MEMORANDUM OPINION AND ORDER re: FORFEITURE
STEPHEN P. FRIOT, UNITED STATES DISTRICT JUDGETable of Contents
I. Introduction... 881
A. The Criminal Charges.. .882
B. The Trials and Convictions... 883
C. The Moving Papers and Evidentiary Hearing.. .884
II. Authority for Forfeiture... 886
A. Forfeiture Procedures... 886
B. General Principles.. .887
C. Forfeitures Based on RICO Conspiracy Convictions... 891
D. Forfeitures Based on Convictions for Operation of Illegal Gambling Business ...894
E. Forfeitures Based on Money Laundering Conspiracy Convictions... 895
F. Forfeiture Based on a Facilitation Theory.. .897
III. The Excessive Fines Clause... 899
IV. Burden of Proof... 905
A. The Government’s Burden.. .905
B. The Defendants’ Burden... 907
V. The Exclusionary Rule... 908
VI. Tracing, Tainting, Commingling and Extrapolation... 910
A. Tracing, Tainting and Commingling. . .910
B. Extrapolation... 915
VII. Notice of the Government’s Intent to Seek Money Judgment.. .916
VIII. Overview of the Government’s Forfeiture Case... 917
A. The Government’s Forfeiture Brief (doc. no. 1999).. .917
1. King... 918
2. Koralewski.. .919
3. Tucker... 919
4. Moran... 920
5. Robles... 920
6. Bramley.. .920
7. Diebner.. .920
B. The Government’s Presentation at the Forfeiture Trial... 920
IX. Findings of Fact... 922
A. Introduction... 922
B. Viability of the Government’s Characterization of Funds Allegedly Laundered ...924
1. Preliminary Matters... 924
*8812. Sources of Deposits and Nature of Activities Generating the Funds... 927
3. Money Laundering Total... 948
4. A Brief Digression... 949
C.Viability of the Government’s Proposed Direct Attributions and Extrapolations ...951
1. General Nature of the Government’s Proposed -Direct Attributions and Extrapolations ...951
2. Extrapolation Based on Legendz Payouts to Bettors through Global Data Payment Services... 954
3. The Government’s Proposed Extrapolations and Direct Attributions re: Proceeds Attributable to Individual Agents and Runners... 955
a. Terry Campbell.. .955
b. Ralph Hernandez.. .957
c. Kelly Diebner... 960
d. Christopher Tanner ...965
e. Michael Lawhorn... 966
f. Joseph Barry.. .970
g-Paul Tucker.. .970
h. Bruce Middlebrook. . .972
i. Rodger Bramley.. .974
j-Paul Wilson.. .976
k. Leon Moran.. .977
1. Joseph McFadden.. .981
m. David Ross... 983
n. Kory Koralewski... 983
0. Luis Robles... 993
4.Conclusion as to the Government’s Proposed Extrapolations and Direct Attributions ...994
D.Viability of the Government’s Contention that all Forfeiture Defendants, Regardless of their Respective Degrees of Participation in the Legendz Enterprise, Should be Subjected to a Forfeiture Money Judgment in the Same Amount.. .995
E. Findings Specific to King...996
1. Forfeiture of Specific Assets... 996
2. Forfeiture Money Judgment... 999
F. Findings Specific to Koralew-ski.. .999
1. Forfeiture of Specific Assets... 999
2. Forfeiture Money Judgment.. .1001
G. Findings Specific to Tucker.. .1001
1. Forfeiture of Specific Assets... 1001
2. Forfeiture Money Judgment.. .1001
H. Findings Specific to Moran... 1006
I. Forfeiture of Specific Assets... 1006
2. Forfeiture Money Judgment.... 1006
I. Findings Specific to Robles... 1008
1. Forfeiture of Specific Assets... 1008
2. Forfeiture Money Judgment.. .1008
J. Findings Specific to Bram-ley... 1009
1. Forfeiture of Specific Assets... 1009
2. Forfeiture Money Judgment.. .1012
K. Findings Specific to Dieber.. .1013
1. Forfeiture of Specific Assets... 1013
2. Forfeiture Money Judgment.. .1013
L. Summary of Forfeiture-Related Findings... 1014
X. Forfeiture-Related Sanctions Against the United States... 1015
XI. Conclusion... 1022
I. Introduction
The government moves for preliminary orders of forfeiture against the defendants Bartice A. “Luke” King,1 Kory Elwin Ko-ralewski, Paul Francis Tucker, Leon Mark Moran, Luis Robles, Zapt Electrical Sales Inc., Rodger Vanpelt Bramley and Kelley *882Ward Diebner. This order refers to these parties, other than Zapt,2 as “the forfeiture defendants.”3
The government seeks forfeiture of specific assets as well as a forfeiture money judgment. See generally, Rule 32.2(b), Fed.R.Crim.P. (procedures for entering a preliminary order of forfeiture). As for the forfeiture money judgment, the government now seeks a judgment in the amount of $231,432,686.73 against the forfeiture defendants. (In the last of the four jury trials that were held in this case, an FBI witness maintained, repeatedly, that the government actually seeks $1 billion by way of forfeiture. Oct. Tr. 2074, 2077, 2114-14.4)
These forfeiture proceedings are, to put it mildly, contested.
The findings of fact in this order are based on the evidence presented at the four jury trials which were held in this case, as well as the James hearings and the nonjury trial which was addressed exclusively to forfeiture issues — all of which generated 11,368 pages of trial transcript.
A. The Criminal Charges
The original Indictment in this case was returned on March 20, 2013. It alleged a ten-year conspiracy period for the two conspiracy counts. Doc. no. 1, at 18, 47.5 As of the date of the Indictment, there was about a nine-year overlap between the ten-year duration of the conspiracy and the duration of the government’s investigation of that conspiracy.
The original Indictment charged fifty-seven defendants, including six of the eight forfeiture defendants now before the court (King, Koralewski, Moran, Robles, Tucker and Zapt). Doc. no. 1. The Indictment charged: Count 1, a racketeering (RICO) conspiracy in violation of 18 U.S.C. § 1962(d); Count 2, operation of an illegal gambling business in violation of 18 U.S.C. § 1955; and Count 3, conspiring in violation of 18 U.S.C. § 1956(h) to commit money laundering under 18 U.S.C. §§ 1956 and 1957. The Indictment included forfeiture allegations stating the government’s intent to seek criminal forfeiture of at least $1,000,000,000.00 as well as specific items of real and personal property.
*883On August 16, 2013, the United States filed a Bill of Particulars for Forfeiture of Property, giving notice to defendants King, Tucker and Moran (and other defendants not involved in these forfeiture proceedings), that the United States was seeking forfeiture of additional specific property listed in the Bill of Particulars. Doc. no. 346.
On August 21, 2013, a grand jury sitting in the Western District of Oklahoma returned a Superseding Indictment against fifty-nine defendants, including the eight forfeiture defendants before the court in these proceedings (King, Koralewski, Tucker, Moran, Robles, Zapt, Bramley and Diebner). Doc. no. 354. The Superseding Indictment included modified versions of the same three counts. The Superseding Indictment included forfeiture allegations stating the government’s intent to seek criminal forfeiture of at least $1,000,000,000.00 and specific items of real and personal property. Each of the forfeiture defendants was convicted, by a jury, on one or more of the three counts.
On October 15, 2013, a Second Bill of Particulars for Forfeiture of Property was filed providing notice (to the extent material here) to King, Bramley and Diebner of the government’s intent to pursue forfeiture of additional items of real and personal property. Doc. no. 465.
B. The Trials and Convictions
The sheer number of defendants necessitated multiple trials. Accordingly, the forfeiture defendants, along with other defendants, were grouped in four sets of defendants to be tried together in four separate trials. All of the defendants who have been tried have agreed to have forfeiture determined by the court rather than by a jury.
The first trial began on February 12, 2015, and a jury verdict was returned on March 3, 2015. Forfeiture defendants Tucker, Robles and Zapt were tried in this first set of defendants. Tucker was found guilty on all three counts. Robles was found guilty on all three counts. Zapt was not a defendant with respect to Count 2, and was found guilty on Count 1 and Count 3. The court denied their post-trial motions. United States v. Robles, 2015 WL 12852050 (W.D. Okla. July 1, 2015).
The second trial began on April 14, 2015, and a jury verdict was returned on April 30, 2015. In the second trial, King (the lead defendant) was found not guilty on Count 1 (by any standard, the flagship count), and guilty on Count 2 and Count 3. The court denied King’s post-trial motion. United States v. King, 2015 WL 12852049 (W.D. Okla. July 10, 2015).
The third trial began on May 4, 2015, and a jury verdict was returned on May 22, 2015. Forfeiture defendants Moran, Bramley and Diebner were tried in this third set of defendants. All three of those defendants were found guilty on all three counts. The court denied their post-trial motions. United States v. Bramley, 2015 WL 12852048 (W.D. Okla. Sept. 16, 2015).
The fourth trial began on October 19, 2015, and a jury verdict was returned on November 4, 2015. Forfeiture defendant Koralewski was in this fourth set of defendants. Koralewski was found guilty on Count 1. He was found not guilty on counts two and three. The court denied Koralewski’s post-trial motion. United States v. Dorn, 2016 WL 7388658 (W.D. Okla. Mar. 31, 2016).
Below is a count-by-count summary of the verdicts relevant to these proceedings.
*884[[Image here]]
C. The Moving Papers and Evidentiary Hearing
The issues addressed in this order are those raised by the moving papers considered as a whole, which are extensive.
The government moved for preliminary orders of forfeiture against defendants King (doc. no. 1493), Koralewski (doc. no. 1975), Tucker (doc. no. 1512, amended motion), Moran (doc. no. 1482), Robles (doc. no. 1887), Zapt (doc. no. 1494), Bramley (doc. no. 1481) and Diebner (doc. no. 1480). Setting defendants Tucker and Zapt aside for the moment (because the government’s motions against these two defendants stand on a slightly different procedural footing, as discussed next), defendants King, Koralewski, Moran, Bramley and Diebner dispute the government’s entitlement to forfeiture of specific property identified in each of these motions. These defendants, plus Robles,6 also dispute the government’s entitlement to a money judgment against them.
The government’s motions for preliminary orders of forfeiture against Tucker and Zapt have already been granted by agreement, to the following extent. The government’s amended motion for a preliminary order of forfeiture against Tucker was granted (doc. no. 1797), as was the government’s motion for a preliminary order of forfeiture against Zapt. Doc. no. 1798. These motions were granted with respect to the specific property identified in the motions, as to which there was no dispute; however, the government’s entitlement to a money judgment against Tucker remains in dispute, as does forfeiture of one of Tucker’s bank accounts, as explained in Part VIII(A)(3), below. The government has not explicitly sought a money judgment against Zapt. See, doc. no. 1999-7, at 3-4 (describing money judgment sought against Tucker, but stating nothing about a money judgment from Zapt). Furthermore, it does not appear that the government seeks to forfeit any assets of Zapt which have not already been preliminarily forfeited. See, doc. no. 1798 (preliminary order of forfeiture by Zapt of contents of account X0897, in approximate amount of $1,657.94, located at TD Bank, *885Mount Laurel, NJ, in the name of ZAPT Electrical Sales). Thus, although Zapt was given permission to participate in the forfeiture proceedings along with Tucker, nothing remains to be decided at this stage with respect to Zapt.
After the government filed its motions for preliminary orders of forfeiture, the forfeiture defendants filed response briefs. See, King (doc. no. 1530), Koralewski (doc. no. 1983), Tucker (doc. no. 1548), Moran (doc. no. 1513), Robles (doc. no. 1897), Zapt (doc. no. 1548), Bramley (doc. no. 1523) and Diebner (doc. no. 1525).
Viewed in light of the complexity of these proceedings, the government’s motions were somewhat perfunctory. For example, the motions listed forfeiture statutes from the Superseding Indictment, but they did not specify which particular forfeiture statutes the government invoked in support of particular forfeitures. Notably, the government’s motions seeking forfeitures from King, Koralewski, Tucker, Bramley, Zapt and Diebner did not mention a personal money judgment against these defendants.
On January 19, 2016 (about twelve years after the Legendz investigation started and nearly three years after the April, 2013 takedown in which the government seized hundreds of thousands of records from the defendants), the government filed its opening brief in support of its forfeiture claims. That brief (herein: Forfeiture Brief) was the government’s definitive statement of its forfeiture case. It was filed pursuant to Case Management Order No. 7, doc. no. 1879, which required the government to:
(i) state the statutory basis for the government’s forfeiture claims with respect to each forfeiture defendant, (ii) specifically identify the property sought to be forfeited, as to each forfeiture defendant, (iii) state with reasonable specificity the factual basis on which the government asserts that there is a nexus between the property and the offense(s) of conviction, (iv) specify the amount of any forfeiture money judgment sought against any forfeiture defendant, with a reasonably detailed explanation of the factual basis for the government’s determination of the amount sought, and (iv) address any other relevant matters as set forth in Part F of the Chambers Procedures of the undersigned judge, as found at the court’s website.”
Id. at 1-2.
In its Forfeiture Brief, the government set out, in separate exhibits for each of the forfeiture defendants, the government’s position as it stood heading into the evi-dentiary hearing, regarding the property, and the size of personal money judgments, sought in these proceedings. See, doc. no. 1999-1 (King); doc. no. 1999-5 (Koralewski); doc. no. 1999-7 (Tucker); doc. no. 1999-3 (Moran); doc. no. 1999-8 (Robles); doc. no. 1999-7 (Zapt); doc. no. 1999-2 (Bramley); doc. no. 1999-4 (Diebner).
As already stated, Case Management Order No. 7 also required the government to specify the statutory basis invoked by the government in support of each forfeiture. Per the government’s Forfeiture Brief, the sole statutory basis pressed by the government in support of forfeitures based on RICO conspiracy convictions was 18 U.S.C. § 1963(a). The sole statutory basis pressed by the government in support of forfeitures based on convictions for operating an illegal gambling business was 18 U.S.C. § 1955(d). And the sole statutory basis pressed by the government in support of forfeitures based on money laundering conspiracy convictions was 18 U.S.C. § 982(a)(1). Doc. no. 1999, pp. 11-*88612.7
All of the forfeiture defendants filed briefs in response to the government’s Forfeiture Brief. See, King (doc. no. 2074); Koralewski (doc. no. 2071); Tucker (doc. no. 2077); Moran (doc. no. 2067); Robles (doc. no. 2078); Zapt (doc. no. 2077); Bramley (doc. no. 2075); Diebner (doc. no. 2070).
As required by the court’s order of March 7, 2016, (doc. no. 2082), the government also filed a reply brief to address arguments made by some defendants in their response briefs, contending the government had failed to give adequate notice it intended to seek a personal money judgment. Doc. no. 2101.
An exclusionary rule argument was raised by King in his motion, seeking to exclude the documents commonly referred to as the Karlo Stewart documents. Doc. no. 2169. The government filed a response brief addressing that issue. Doc. no. 2180. (King had previously filed a successful motion to suppress those documents, and they were not admitted into evidence at his jury trial. United States v. King, 212 F.Supp.3d 1113 (W.D. Okla. 2015)).
The government filed pre-hearing proposed findings and conclusions. Doc. no. 2130. The government’s forfeiture claims against the forfeiture defendants now before the court were tried to the court in three days of hearings in late April, 2016. Doc. nos. 2174-76.
Following the hearing, proposed findings of fact and conclusions of law were submitted by all parties. See, doc. no. 2242 (government); doc. no. 2275 (King); doc. no. 2273 (Koralewski); doc. no. 2276 (Tucker); doc. no. 2270 (Moran); doc. no. 2278 (Robles); doc. no. 2276 (Zapt); doc. no. 2315 (Bramley); doc. no. 2274 (Dieb-ner). In addition, the court has permitted the forfeiture defendants to adopt proposed findings and conclusions of the other forfeiture defendants. See adoption orders, doc. no. 2288 (King);8 doc. no. 2285 (Kora-lewski); doc. no. 2284 (Tucker); doc. no. 2289 (Moran); doc. no. 2286 (Robles); doc. no. 2284 (Zapt); doc. no. 2287 (Diebner). (Bramley did not ask to adopt the other forfeiture defendants’ proposed findings and conclusions, but the court gives him the benefit of the others’ proposed findings and conclusions where applicable to his situation.) At previous stages of the briefing, the forfeiture defendants have also, at times, incorporated each others’ arguments by reference.
II. Authority for Forfeiture
A. Forfeiture Procedures
“As soon as practical after a verdict ... of guilty ... on any count in an indictment ... regarding which criminal forfeiture is sought, the court must determine what property is subject to forfeiture under the applicable statute.” Rule 32.2(b)(1)(A), Fed. R. Crim. P. “If the government seeks forfeiture of specific property, the court must determine whether the government has established the requisite nexus be*887tween the property and the offense. If the government seeks a personal money judgment, the court must determine the amount of money that the defendant will be ordered to pay.” Id.
“The court’s determination may be based on evidence already in the record, ... and on any additional evidence or information submitted by the parties and accepted by the court as relevant and reliable.” Rule 32.2(b)(1)(B), Fed. R. Crim. P. “If the forfeiture is contested, ... the court must conduct a hearing after the verdict ... of guilty.” Id.
“If the court finds that property is subject to forfeiture, it must promptly enter a preliminary order of forfeiture setting forth the amount of any money judgment, directing the forfeiture of specific property, and directing the forfeiture of any substitute property if the government has met the statutory criteria.” Rule 32.2(b)(2)(A). “The court must enter the order without regard to any third party’s interest in the property. Determining whether a third party has such an interest must be deferred until any third party files a claim in an ancillary proceeding under Rule 32.2(c).” Id. “Thus, the ancillary proceeding has become the forum for determining the extent of the defendant’s forfeitable interest in the property. This allows the court to conduct a proceeding in which all third-party claimants can participate and which ensures that the property forfeited actually belongs to the defendant.” Advisory Committee Notes, 2000 Adoption, Subdivision (b).
“Unless doing so is impractical, the court must enter the preliminary order sufficiently in advance of sentencing to allow the parties to suggest revisions or modifications before the order becomes final as to the defendant under Rule 32.2(b)(4).” Rule 32.2(b)(2)(B), Fed. R. Crim. P.
“At sentencing — -or at any time before sentencing if the defendant consents — the preliminary forfeiture order becomes final as to the defendant.” Rule 32.2(b)(4)(A), Fed. R. Crim. P. “If the order directs the defendant to forfeit specific property, it remains preliminary as to third parties until the ancillary proceeding is concluded under Rule 32.2(c).” Id.
B. General Principles
These forfeitures are criminal forfeitures. Criminal forfeiture is part of the sentence — punishment—imposed on a defendant who has been convicted in a criminal case. See, Libretti v. United States, 516 U.S. 29, 38-39, 116 S.Ct. 356, 133 L.Ed.2d 271 (1995) (forfeiture is an element of the sentence imposed following conviction). A criminal forfeiture order is an in personam proceeding against a defendant in a criminal case, as distinguished from civil forfeiture which is an in rem proceeding brought against the thing. United States v. $39,000 in Canadian Currency, 801 F.2d 1210, 1218 (10th Cir. 1986); United States v. Vampire Nation, 451 F.3d 189, 202 (3d Cir. 2006).
Criminal forfeitures fall into two categories, forfeiture of property, and forfeiture in the form of a money judgment. See, Rule 32.2(b)(1), Fed. R. Crim. P., and Advisory Committee Notes, 2000 Adoption, Subdivision (b) (“Subdivision (b)(1) recognizes that there are different kinds of forfeiture judgments in criminal cases. One type is a personal judgment for a sum of money; another is a judgment forfeiting a specific asset.”).
The burden rests on the government to establish the nexus between the property forfeited and an offense of conviction that authorizes forfeiture. United States v. Bader, 678 F.3d 858, 894-95 (10th Cir. 2012). This aspect of the matter will *888receive more attention in Part IV(A), below.
As for money judgments, although the criminal forfeiture statutes in issue in this case do not explicitly refer to money judgments, courts uniformly recognize that money judgments representing unlawful proceeds are appropriate. See, United States v. McGinty, 610 F.3d 1242, 1246 (10th Cir. 2010) (involving a forfeiture of the defendant’s house and proceeds from the sale of his boat and its motor, following defendant’s conviction of misapplication of bank funds; the forfeiture statute involved was 18 U.S.C. § 982(a)(2), which refers to property consisting of or derived from proceeds obtained directly or indirection as a result of the violation).
There are two primary reasons for permitting forfeiture money judgments. First, criminal forfeiture is a sanction against the individual defendant rather than a judgment against the property itself. McGinty, 610 F.3d at 1246, quoting United States v. Hall, 434 F.3d 42 (1st Cir. 2006). Because the sanction follows the defendant as part of the penalty, the government need not prove that the defendant actually has the forfeited proceeds in his possession at the time of conviction. Id. Second, permitting a money judgment as part of a forfeiture order prevents a defendant from ridding himself of his ill-gotten gains to avoid the forfeiture sanction. Id.
The nature of a particular forfeiture order will depend on the relevant forfeiture statute as well as the facts of a given case. Id. at 1248. Hybrid orders which forfeit specific property of a defendant, and which also enter a money judgment against a defendant, may be appropriate and are permitted. Id. at 1248-49.
As will be seen in the portions of this order which address the individual forfeiture statutes relied on by the government in this case, those statutes permit forfeiture of proceeds.9 Moreover, there is universal agreement that at least the portion of property that derives from proceeds of illegal activity is subject to forfeiture. See, United States v. Rudaj, 2006 WL 1876664, *4, and n.6 (S.D.N.Y. 2006) (stating, in discussion of bank fraud forfeiture under 18 U.S.C. § 982, that “there is universal agreement that at least the portion of the property that derives from proceeds of illegal activity is subject to forfeiture”; and citing United States v. Genova, 333 F.3d 750, 762 (7th Cir. 2003), for the proposition, in a RICO case, that ill-gotten gain is forfeited but value added independently by accused is not forfeitable gain.)
With the exception to be discussed immediately below, the general rule is that joint and several liability applies to forfeitures sought in a multi-defendant case. See, United States v. Contorinis, 692 F.3d 136, 146-47 (2d Cir. 2012) (government sought forfeiture of over twelve million dollars following conviction of conspiracy to commit securities fraud and insider trading; court noted that criminal forfeiture focuses on the disgorgement by a defendant of his ill-gotten gains, citing McGinty for the proposition that the calculation of a forfeiture amount in criminal *889cases is usually based on the defendant’s actual gain, at 146; court noted this general rule is somewhat modified by the principle that a court may order a defendant to forfeit proceeds received by others who participated jointly in the crime, provided the actions generating those proceeds were reasonably foreseeable to the defendant).10
For purposes of this case, the exception to the general rule of joint and several liability is that the doctrine of joint and several liability does not permit the government to forfeit racketeering proceeds under the RICO forfeiture statute, 18 U.S.C. § 1963(a)(3), where the proceeds in question were never “obtained, directly or indirectly” by the particular forfeiture defendant from whom forfeiture is sought. This exception to joint and several liability is addressed later in this order, in connection with the RICO forfeiture statute.
Similarly, it is generally (but not inevitably) true that a court may order a defendant to forfeit proceeds received by others who participated jointly with the defendant in the crime. In these proceedings, however, the government only argues for joint and several liability with respect to the conspiracy convictions.11
Moreover, to obtain a money judgment based on joint and several liability for proceeds received by others, at least in the circumstances of this case, a foreseeability requirement must be met so that the government can get a judgment for only so much of the proceeds as were foreseeable to that defendant.12 Long before the forfei*890ture trial, the government was on notice that foreseeability was very much in issue. See, Tucker’s forfeiture brief, doc. no. 2077, at 5-6 (filed on February 29, 2016, nearly two months before the forfeiture trial). After the forfeiture trial, in a filing that was adopted by the other forfeiture defendants,13 Tucker again argued foreseeability, as a legal matter and as a factual matter. Doc. no. 2276, at 23-25, ¶¶ 88-92. But neither the government’s forfeiture brief nor its proposed findings and conclusions address the issue of foreseeability— an issue that is undeniably, and prominently, in the mix as to forfeiture based on money laundering,14 given the far-flung nature of the Legendz betting operation and the geographic and functional differences among the defendants. The government’s filings make no reference, to say nothing of providing any developed argument, on that issue or as to the burden of proof on that issue (discussed in Part IV, below). Any argument that foreseeability is not required as a prerequisite to joint and several forfeiture liability is therefore waived. However, aside from the government’s waiver, the court concludes that each forfeiture defendant’s exposure to money laundering forfeiture is limited to those transactions that the government proves were foreseeable to that defendant.
As discussed later in this order, at least in the context of this case, joint and several liability does not eliminate eighth amendment concerns; rather, at least arguably, it heightens them. For example, United States v. Jalaram, Inc., 599 F.3d 347 (4th Cir. 2010), involved forfeiture based on convictions for violating, and conspiring to violate, the Mann Act; money laundering; and conspiracy to commit money laundering. In Jalaram, the Fourth Circuit recognized that in single-offender cases, it would be very difficult and perhaps impossible for the defendant to show that the forfeiture of proceeds was disproportionate to the gravity of his offense, which the Fourth Circuit speculated might explain the desire of some of its sister circuits to simplify the analysis by holding such forfeitures exempt from constitutional scrutiny in the first place. Id. at 355. Jalaram states, however, that such a “proposed shortcut may work a grave injustice in cases involving joint and several liability.” Id. The court explained that “[i]n such cases, some defendants inevitably disgorge more money than they received from the conspiracy, thus forfeiting property they obtained lawfully in order to satisfy the forfeiture judgment. In a case where a defendant played a truly minor role in a conspiracy that generated vast proceeds, joint and several liability for those proceeds might result in a forfeiture grossly disproportionate to the individual defendant’s offense.” Id.
Additional legal principles of general application are addressed later, under separate headings. Those topics (including the Excessive Fines Clause, burden of proof, the exclusionary rule, tracing, commingling, extrapolation and double counting, and the sufficiency of the government’s notice of its intention to seek money judgments) address issues raised in the forfeiture defendants’ papers. Before addressing those topics, it is useful to address the *891specific forfeiture statutes relied on by the government in these proceedings.
C. Forfeitures Based on RICO Conspiracy Convictions
As the statutory authority for forfeitures based on a RICO conspiracy in violation of 18 U.S.C. § 1962(d), the government relies on 18 U.S.C. § 1963(a). Doc. no. 1999, p. 11 (in which government was required to identify all forfeiture statutes invoked). Section 1963(a) provides as follows.
Whoever violates any provision of section 1962 of this chapter ... shall forfeit to the United States, irrespective of any provision of State law—
(1) any interest the person has acquired or maintained in violation of section 1962;
(2) any—
(A) interest in;
(B) security of;
(C) claim against;
(D) property or contractual right of any kind affording a source of influence over;
any enterprise which the person has established operated, controlled, conducted, or participated in the conduct of, in violation of section 1962; and
(3) any property constituting, or derived from, any proceeds which the person obtained, directly or indirectly, from racketeering activity ... in violation of section 1962.
The government’s proposed findings and conclusions cite subsection (3). See, e.g., doc. no. 2242, pp. 57-58, ¶ 237, citing § 1963(a)(3) (the only subsection of the statute expressly cited or quoted in the government’s proposed findings and conclusions). Subsections (1) and (2) (which are quoted in the government’s Forfeiture Brief) provide that a defendant’s conviction under RICO subjects all of his interests in the enterprise to forfeiture. See, United States v. Cauble, 706 F.2d 1322, 1349 (5th Cir. 1983) (punishment imposed by a RICO forfeiture “deprives that defendant of all of the assets that allow him to maintain an interest in a RICO enterprise”; jury may not find that less than the full amount of the defendant’s interest in an enterprise is subject to forfeiture under § 1963). The government’s present exclusive reliance on subsection (3) is understandable. The evidence does not show that there is (or, for a long time, has been) anything left of the Legendz gambling enterprise other than such assets as may be within the reach of subsection (3).
The RICO forfeiture provision is broadly drafted and has long been liberally construed. United States v. Fruchter, 411 F.3d 377, 384 (2d Cir. 2005), cert. denied, 546 U.S. 1076, 126 S.Ct. 840, 163 L.Ed.2d 708 (2005).
Proceeds forfeitable based on a conspiracy to violate RICO potentially include gross proceeds of the RICO conspiracy, not just net profits after expenses. See, United States v. Christensen, 828 F.3d 763, 822-23 (9th Cir. 2016), cert. denied, — U.S. —, 137 S.Ct. 628, 196 L.Ed.2d 517 (2017) (concluding in a RICO forfeiture that “proceeds” means gross receipts, but noting some circuits have held otherwise; affirming the district court’s conclusion that the gross receipts of the investigative agency which was at the center of the criminal enterprise rather than the agency’s profits, were subject to forfeiture under § 1963(a)(3)). Christensen reasoned that legislative history of RICO indicates that “proceeds” was used in lieu of the term “profits” to alleviate the unreasonable burden on the government of proving net profits). Bader, 678 F.3d 858 (10th Cir. 2012), involved a forfeiture *892based on a conviction for various drug-related crimes involving human growth hormone. Defendant argued that the forfeiture verdict, based on calculation of proceeds from gross drug sales instead of net sales, was improper under United States v. Santos, 553 U.S. 507, 128 S.Ct. 2020, 170 L.Ed.2d 912 (2008). Id. at 892-94. The Court of Appeals held the defendant had waived this argument, but went on to state that Santos should be “confined to its factual setting,” concluding that “ ‘proceeds’ means ‘profits’ for the purpose of the money laundering statute only where an illegal gambling operation is involved.” Id. at 894, quoting United States v. Fishman, 645 F.3d 1175, 1193-94 (10th Cir. 2011), cert. denied, 565 U.S. 1115, 132 S.Ct. 1046, 181 L.Ed.2d 740 (2012). In Bader, the court stated that defendant’s reliance on Santos was misplaced because his convictions involved neither the federal money laundering statute nor an illegal gambling operation. Bader, 678 F.3d at 894.15 And see, United States v. Miller, 2009 WL 2949784 (D. Kan. 2009), in which the defendant conceded, and the court agreed, that “it is gross receipts that are considered forfeita-ble under federal forfeiture statutes invoking the term ‘proceeds.’ ” Id. at *4. Defendant had argued that “proceeds” under 18 U.S.C. § 982(a)(2) must be interpreted as “profits” under Santos. Id. Rejecting that contention, Miller discussed the legislative history of the Crime Control Act of 1984, pointing out that Congress explained that it used “proceeds” in the RICO forfeiture statute in lieu of the term “profits” in order to alleviate the unreasonable burden on the government of proving net profits. Id. at *5. (Santos is a bit more problematic in the money laundering forfeiture context, and, on that score, will get further discussion in Part E, below.)
As for the issue of joint and several liability for racketeering proceeds, this order has already indicated that although, as a general proposition, joint and several liability (limited by principles of foreseeability and the Eighth Amendment) applies to forfeitures which are based on conspiracy convictions, joint and several liability does not apply to permit forfeiture of racketeering proceeds from a particular forfeiture defendant who never, in any sense of the word, “obtained” those racketeering proceeds. In the court’s view, the “obtained” language in the RICO forfeiture statute, 18 U.S.C. § 1963(a)(3), allows no other conclusion.
In this regard, the undersigned’s reading of the RICO forfeiture statute is the same as the D.C. Court of Appeals’ reading of almost identical language in 21 U.S.C. § 853(a), a drug forfeiture statute, as explained with clarity in United States v. Cano-Flores, 796 F.3d 83 (D.C. Cir. 2015), cert. denied, — U.S. —, 136 S.Ct. 1688, 194 L.Ed.2d 790 (2016). Title 21 U.S.C. § 853(a), like the RICO forfeiture *893statute, authorizes forfeiture of “proceeds the person obtained, directly or indirectly” as a result of the violation. Cano-Flores is at odds with the majority view, as Cano-Flores recognizes. See, 796 F.3d at p. 91 (discussing the contrary view in other circuits). That said, the undersigned is convinced that to join the majority of courts— which hold that the doctrine of joint and several liability permits forfeiture of racketeering proceeds which were obtained by others, although those proceeds were never obtained by the particular forfeiture defendant in question16 — would be to read the “obtained, directly or indirectly” language entirely out of the RICO forfeiture statute. In an analysis that is hard to argue with, Judge Williams’s plain language persuasively explains the statute’s equally plain language. Id. at 91-95. And see, United States v. Honeycutt, 816 F.3d 362, 379-80 (6th Cir. 2016) (involving forfeiture of drug proceeds; court follows the majority of courts, and applies joint and several liability among conspirators to forfeiture sought under 21 U.S.C. § 853(a)(1), although noting that Cano-Flores is contra), cert. granted, — U.S. —, 137 S.Ct. 588, 196 L.Ed.2d 462 (2016).
The court’s agreement with Cano-Flores affects the result in these forfeiture proceedings with respect to only one of the forfeiture defendants — Koralewski. This is because the other forfeiture defendants were convicted on the money laundering conspiracy charge, subjecting them to the broader reach of the money laundering forfeiture statutes, none of which include the “obtained” language found in the RICO forfeiture statute. Koralewski, however, was convicted solely on the RICO conspiracy charge.
Eighth amendment limitations are addressed later, in Part III. At this point it is enough to observe that courts have applied an excessive fines analysis to forfeitures that are based on RICO convictions. See, e.g., Rudaj, 2006 WL 1876664 at *8 (under gross disproportionality test of United States v. Bajakajian, 524 U.S. 321, 332, 118 S.Ct. 2028, 141 L.Ed.2d 314 (1998), the forfeiture sought by the government, consisting of over five million dollars and three properties, was not grossly disproportionate to the gravity of the defendants’ offense). Although the “shall” language in the RICO forfeiture statute suggests that forfeitures which meet the statutory requirements are mandatory, this language does not trump eighth amendment limits. See, United States v. Corrado, 227 F.3d 543 at 552 (noting mandatory “shall forfeit” language in RICO forfeiture statute, but stating that although the statute appears to require total forfeiture of illegal proceeds, courts can reduce the forfeiture to make it pro*894portional to the seriousness of the offense so as not to violate the eighth amendment prohibition against excessive fines).
The government need not provide a precise calculation of the proceeds from a RICO enterprise; however, estimates must be conservative, and overly speculative evidence about what constitute proceeds will not support a RICO forfeiture.
[B]ecause it is intended to be a potent means of punishment, the Government need not provide a precise calculation of the proceeds. See United States v. Lizza Industries, Inc., 775 F.2d 492, 498 (2d Cir. 1985) (“RICO does not require the prosecution to prove or the trial court to resolve complex computations, so as to ensure that a convicted racketeer is not deprived of a single farthing more than his criminal acts produced.”). A total amount derived from “conservatively estimating” the revenue regularly collected or received will suffice where the evidence establishes the approximate amounts. Corrado, 227 F.3d at 555. But evidence regarding proceeds that is overly speculative will be insufficient to support a forfeiture award. Id. at 557-58.
Rudaj, 2006 WL 1876664 at *3.
Rudaj strikes the right balance. The court’s reckoning should be conservative in the sense that the evidence should be assessed with some semblance of intellectual rigor, lest loose talk, even under oath (of which there has been no shortage in this case), be allowed to render a defendant a pauper for life. But to-the-penny precision — within the general confines of a result produced by a reasonably rigorous analysis of the evidence — is another issue altogether. That degree of precision is rarely possible and, if required, would hollow out much of the forfeiture legislation.
D. Forfeitures Based on Convictions for Operation of Illegal Gambling Business
As the statutory authority for forfeitures based on operation of an illegal gambling business in violation of 18 U.S.C. § 1955, the government relies on 18 U.S.C. § 1955(d). Doc. no. 1999, p. 12. Section 1955(d) provides that: “Any property, including money, used in violation of the provisions of this section may be seized and forfeited to the United States.”
In addition, although not mentioned in the government’s Forfeiture Brief as required by the court, the government’s proposed findings and conclusions relating to forfeitures based on operation of an illegal gambling business (doc. no. 2242, ¶ 205 at p. 51, ¶ 249 at p. 60), rely on 18 U.S.C. § 981(a)(1)(C).17 That statute provides that the following property is subject to forfeiture to the United States:
Any property, real or personal, which constitutes or is derived from proceeds traceable to a violation of ... any offense constituting “specified unlawful activity” (as defined in section 1956(c)(7) of this title), or a conspiracy to commit such offense.
Continuing the chain of cross references, § 1956(c)(7) incorporates offenses listed in § 1961(1) (RICO) and § 1961(1) includes, among other things, violations of § 1955 (illegal gambling business).
Proceeds which the government seeks to forfeit based on illegal gambling convictions means gross receipts, not net profits. See, e.g., 18 U.S.C. § 981(a)(2)(A) (“For purposes of paragraph (1) [§ 981(a)(1)] ... [i]n cases involving ... illegal services [or] *895unlawful activities, ... the term ‘proceeds’ means property of any kind obtained directly or indirectly, as the result of the commission of the offense giving rise to forfeiture, and any property traceable thereto, and is not limited to the net gain or profit realized from the offense.”)
The forfeiture defendants argue that, unlike the forfeiture statutes which the government relies on to support forfeiture based on RICO and money laundering conspiracy convictions, 18 U.S.C. § 1955(d) uses the permissive “may” rather than the mandatory “shall,” so that forfeitures under this statute are not mandatory. For example, Diebner’s response brief (doc. no. 2070, p. 10, responding to the government’s Forfeiture Brief), cites United States v. Premises Known as 318 South Third Street, 988 F.2d 822, 827 (8th Cir. 1998), for the contention that a forfeiture under § 1955(d) is permissive and allows for judicial discretion. That decision held that under the permissive language of § 1955(d), “a court can refuse a forfeiture if it seems to work a disproportionate penalty in light of the peculiar facts of a particular case.” Id. at 828. Thus, arguments regarding the “may” language of § 1955(d) are taken into account in Part III, below (discussion of the Excessive Fines Clause, where disproportionality is considered).
E. Forfeitures Based on Money Laundering Conspiracy Convictions
As the statutory authority for forfeitures based on conspiring to launder money under 18 U.S.C. §§ 1956 and 1957 in violation of § 1956(h), the government relies on 18 U.S.C. § 982(a)(1). Doc. no. 1999, p. 12. Section 982(a)(1) provides as follows:
The court, in imposing sentence on a person convicted of an offense in violation of section 1956, 1957, or 1960 of this title, shall order that the person forfeit to the United States any property, real or personal, involved in such offense, or any property traceable to such property.
As stated in United States v. Bornfield, 145 F.3d 1123 (10th Cir. 1998), cert. denied, 528 U.S. 1139, 120 S.Ct. 986, 145 L.Ed.2d 935 (2000), a case involving forfeiture of funds in the defendant’s bank account following a money laundering conviction, “[t]he key to whether property is forfeitable [under § 982(a)(1)] is whether it was ‘involved in’ or ‘traceable to’ the offense.” Id. at 1135.18 More about this in Part VI, below.
The forfeiture defendants argue that when the government, seeking forfeiture, relies on transactions it says constituted laundering of illegal gambling proceeds, the transactions are subject to the same Santos date limitation which applied at trial. The court disagrees.
Santos, and the legislation passed soon after that decision to address the definition of “proceeds” as used in 18 U.S.C. §§ 1956 and 1957, created issues in the jury trials in this case concerning which transactions the government could use to prove money laundering or money laundering conspiracy. See generally, doc. nos. 921 and 1247 (Memorandum and Order with Respect to Santos Issues). The government did not prove, either in the jury trials or in forfeiture proceedings, any net profits in the sense required for money laundering transactions involving illegal gambling proceeds prior to May 20, 2009 (that date being the effective date of the legislation, 123 Stat. 1617, § 2, enacted to overrule Santos). Accordingly, in each of the four jury trials in this case, the jury was instructed that, for purposes of Count 3 (and *896with one exception), it could not consider transactions to be money laundering transactions if those transactions occurred before May 20, 2009. The exception was that the date limitation did not apply to transactions shown to be within the scope of § 1956(a)(2)(A). (Because that statute does not use the term “proceeds,” a conspiracy to violate it does not depend on the Santos definition of “proceeds” as net profits. See, e.g., doc. no. 1206, pp. 85-86, jury instructions used in trial of Robles, Tucker and Zapt.)
In Bader, 678 F.3d 858, the Court of Appeals upheld a $4.8 million proceeds forfeiture following convictions for drug crimes involving human growth hormone. The court rejected defendant’s argument that the calculated forfeiture amount was improperly based on aggregated sales rather than on net sales. Id. at 893-94. Although the court deemed the Santos issue waived, it stated that defendant’s “reliance on Santos is ... misplaced, as his convictions pertain neither to the federal money-laundering statute nor to an illegal gambling operation.” Id. at 894. By contrast, this case involves forfeiture under federal money laundering statutes and the government seeks forfeiture of laundered funds on the basis that they are proceeds of an illegal gambling operation. Therefore, unlike the situation in Bader, Santos remains potentially relevant to the forfeiture issues in this case, but that relevance is probably only theoretical, for this reason: In Santos, the Court was construing the term “proceeds” as found in the relevant money laundering statute, specifically 18 U.S.C. § 1956(a)(1)(a)(i). And, as has been stated, the Court held that “proceeds” meant net profits, not gross receipts, thus causing much confusion in the lower courts, as the Tenth Circuit noted in Bader. Id. at 894. But, in this case, at this juncture, the issue is forfeiture, not criminal liability for money laundering, and the operative words, for the purpose of determining the scope of the forfeiture to which the government is entitled are to be found in the relevant forfeiture statute, specifically 18 U.S.C. § 982(a)(1). Those operative words render forfeitable any property “involved in [the offense of conviction], or any property traceable to such property.” Id. (emphasis added). Money or other property can surely be “involved in” or “traceable to” a money laundering offense without being, in an accounting sense, net profits of a specified unlawful activity. See Cassella, Stefan D., Asset Forfeiture Law in the United States, § 25-4, p. 918 at nn. 63-65 (2d ed. 2013), stating that Santos has no application to civil or criminal forfeiture, presumably meaning that money laundering forfeitures are not limited to net profits. The offense of conviction in Bader was not money laundering and it did not involve gambling. Consequently, the most that can be said about Bader’s take on Santos from the perspective of the defendants in this case is that the Tenth Circuit may have assumed — but clearly did not decide — that Santos would apply to money laundering forfeiture in a gambling case. Furthermore, as a practical matter, the impact of Santos and the date limitation that results from Santos and the legislation overruling that ease is lessened in this case because the government seeks the same funds under other forfeiture statutes.
Per the general rule (which this court has held is not applicable to forfeitures of racketeering proceeds) that joint and several liability applies in the context of forfeitures based on conspiracy convictions (limited by foreseeability and the Eighth Amendment), it is only necessary to note here that courts have applied joint and several liability in the context of forfeitures based on money laundering convictions. See, United States v. Seher, 562 *897F.3d 1344, 1372-73 (11th Cir. 2009) (it was reasonable to hold defendant liable for all proceeds that were a reasonably foreseeable result of the money laundering conspiracy, regardless of whether he still possessed them; citing United States v. Caporale, 806 F.2d 1487, 1506-09 (11th Cir. 1986), which affirmed joint and several liability for RICO conspirators).
Courts have applied an excessive fines analysis to forfeitures that are based on money laundering conspiracy convictions. See, e.g., United States v. Viloski, 814 F.3d 104 (2d Cir. 2016). Eighth amendment limitations are discussed in the next part of this order. The “shall” language in the statutes which authorize forfeiture based on convictions for conspiring to launder money does not alleviate eighth amendment concerns. See, e.g., United States v. Warshak, 2008 WL 2705044 at **3-5 (S.D. Ohio 2008) (government argued forfeiture of money laundering proceeds was mandatory and not disproportionate, at *3; court found joint and several liability was required where defendants were guilty of conspiracy to launder money, and that money judgment forfeiture did not qualify as excessive given the sophistication of the offenses, the scope and duration of the scheme, and the length of potential prison terms and statutory fines, so that Eighth Amendment was not triggered in that ease, at *5).
F. Forfeiture Based on a Facilitation Theory
The government seeks forfeiture of Bramley’s house on a facilitation theory, arguing that Bramley maintained an office in his residence which he used to conduct illegal gambling activity and to conduct financial transactions involving criminal proceeds. See, Forfeiture Brief, doc. no. 1999-2, p. 6 of 17. The government’s brief contends that the Bramley house was used to “carry out his crime, and clearly used to conceal proceeds of his crime and to conduct financial transactions involving the criminal proceeds.” Id.
At a minimum,19 a facilitation theory of forfeiture must meet the “substantial connection” test. Cassella, Stefan D., Asset Forfeiture Law in the United States, (2d ed. 2013) § 26-1, p. 938. Courts unanimously hold that “involved in” should be read broadly to include any property used to facilitate the money laundering offense. Id., § 27-7, p. 978. A “facilitating” theory can have a long reach. It is not limited to property that is integral, essential or indispensable to the offense. It can reach property that makes the prohibited conduct less difficult or more or less free from obstruction or hindrance. Id., § 26-3, p. 942. But proving that property made the crime easier to commit or harder to detect is only part of the government’s burden, because many things may make a crime easier to commit in some trivial or incidental way, which is not enough. Id., § 26-4, p. 951. To prevail under the facilitation theory, the government must show that the degree to which the property facilitated the crime was substantial. This is known as the “substantial connection test.” Id.20
*898Thus, to forfeit the Bramley residence on the basis that it was facilitating property with respect to any of the crimes for which Bramley was convicted, there has to be more than an incidental or fortuitous connection between the property and the crime in question; moreover, the connection has to be substantial.
Obviously, whether the proof supports forfeiture of a residence on a facilitation theory of forfeiture is a case-by-case determination based on the evidence before the court, as a review of some cases will show.
In United States v. Iacaboni, 221 F.Supp.2d 104, 115 (D. Mass. 2002) (rev’d in part on other grounds, 363 F.3d 1 (1st Cir. 2004)), the court stated that evidence showing the defendant used a house to promote his gambling business was sufficient to permit forfeiture under § 1955(d). There was evidence that envelopes with money and documents were dropped off and picked up in the garage of the house; that faxes and phone calls related to the enterprise came to and went from the residence; that football tickets were “corrected” on the premises; and that meetings were held in the house. 221 F.Supp.2d at 115. The court said: “Given these facts, the Government would have a strong case for forfeiture of the house under the statute permitting forfeiture of ‘property ... used in violation’ of the gambling statute, 18 U.S.C. § 1955(d), if this provision had been cited in the formal charge.” Id. The court denied forfeiture of the house, however, because the government had not sought forfeiture under that statute, and had, instead, sought forfeiture under 18 U.S.C. § 982(a)(1), the money laundering forfeiture statute. The court held there was no evidence that the house was involved in money laundering. Iacaboni, 221 F.Supp.2d 104, 115-116.
In United States v. Nicolo, 597 F.Supp.2d 342 (W.D.N.Y. 2009), the government sought property which it claimed Nicolo used to facilitate the receipt of illegal payments and to conduct financial transactions with the same funds. Id. at 356. The government asserted that the properties helped maintain an overall appearance of a legitimate business to further the illegal money-laundering activity. Id. The government contended that Nicolo generated invoices using the addresses of the properties as the address of his business, that he faxed invoices from those properties, and that checks were sent to those addresses in payment of those invoices. Id.
The court held that the evidence relied on by the government was more tenuous than the type of 'evidence which courts have used to find that real property is involved in money laundering, citing a Second Circuit decision in which the evidence showed the defendant had deposited the proceeds of his fraud into the business operating accounts of the companies which he ran at the subject premises, and that the premises “served as a conduit for the proceeds of the illegal transactions.” Id., citing States v. Schlesinger, 261 Fed.Appx. 355, 361 (2d Cir. 2008). In Nicolo, the court held that defendant’s acts such as faxing invoices, and use of the properties as a *899mailing address for his business, were activities which were merely incidental or fortuitous in relation to Nicolo’s offenses. Nicolo, 597 F.Supp.2d at 357. The court stated that even if Nicolo physically wrote checks while at the properties to Roeder, his wife and co-defendant, (which the court noted there was no evidence he did), it would not show that the properties themselves had anything more than a fortuitous, transitory and tangential relationship to the money laundering. Id. at 359. That was not enough to justify forfeiting the properties under § 982(a)(1). Id.
The application of these principles to the proposed forfeiture of Bramley’s house will be addressed in Part IX(J), below.
III. The Excessive Fines Clause
At the outset, it is important to note that, in places, the government’s filings (such as in its Forfeiture Brief,21 and in its proposed findings and conclusions22), acknowledge that the Eighth Amendment’s prohibition against excessive fines limits these forfeitures.23
In addition, and alternatively, the court concludes that eighth amendment limitations apply for the reasons stated below.
Criminal forfeitures which proceed against a defendant (as contrasted with civil in rem forfeitures) are designed to punish the offender. United States v. Bajakajian, 524 U.S. 321, 332, 118 S.Ct. 2028, 141 L.Ed.2d 314 (1998). Although the government argued in Bajakajian that the forfeiture of defendant’s currency was constitutional because it involved an instrumentality of the crime, the Court rejected that argument, stating that whether defendant’s currency was an instrumentality was irrelevant; “the forfeiture is punitive, and the test for the excessiveness of a punitive forfeiture involves solely a proportionality determination.” Id. at 333-34, 118 S.Ct. 2028. The Court continued: “The touchstone, of the constitutional inquiry under the Excessive Fines Clause is the principle of proportionality: The amount of the forfeiture must bear some relationship to the gravity of the offense it is designed to punish.... We now hold that a punitive forfeiture violates the Excessive Fines Clause if it is grossly disproportional to the gravity of a defendant’s offense.” Id. at 334, 118 S.Ct. 2028.
Some courts and commentators describe the Tenth Circuit as holding that forfeitures of criminal proceeds are not subject to an excessive fines clause analysis.24 *900Based on Bajakajian and the authorities cited below, the undersigned disagrees with those descriptions. A case often cited as indicating the Tenth Circuit does not apply the Excessive Fines Clause to criminal forfeitures is United States v. One Parcel of Real Property Described as Lot 41, Berryhill Farm Estates, 128 F.3d 1386 (10th Cir. 1997), a civil forfeiture case involving drug proceeds. In Berryhill, the government sought forfeiture of, among other things, personal property located in the home of the defendant.25 According to the government, the defendant’s property had been obtained with proceeds traceable to exchanges of controlled substances and was therefore forfeitable under 21 U.S.C. § 881(a)(6). Id. at 1389. Berryhill is a pre-Bajakajian decision.26 Berryhill holds “as a matter of law that forfeiture of drug proceeds pursuant to § 881(a)(6) can never be constitutionally excessive.” Id. at 1395. Berryhill reasons as follows. “Because the amount of proceeds produced by an individual drug trafficker is always roughly equivalent to the costs that drug trafficker has imposed on society, the forfeiture of those proceeds can never be constitutionally excessive.” Id. at 1395-96. “Therefore, the forfeiture is proportionate as a matter of law and [defendant’s] Eighth Amendment objection is rejected.” Id. at 1396. Thus, Berryhill simply held, in a pre-Baja-kajian context, and in a case in which the government sought personal property obtained by the defendant with drug proceeds, that eighth amendment concerns were satisfied as a matter of law. Berryhill does not preclude an excessive fines clause analysis in these post-Bajakajian proceedings, either by its terms or under its rationale.
Moreover, civil forfeiture cases decided post-Bajakajian suggest the Tenth Circuit Court of Appeals would recognize that eighth amendment arguments apply.27
The undersigned concludes that the Eighth Amendment applies.
The next question is; what criteria should this court use to determine whether a particular proposed forfeiture is grossly disproportionate to the gravity of the defendant’s offense? Some general observations, as well as some specific criteria set out in Bajakajian, help answer that question.
*901First, it is important to note that the test is applied in an individualized manner, because it requires comparing the amount of the forfeiture to the gravity of the defendant’s offense. See, Bajakajian, 524 U.S. at 336-37, 118 S.Ct. 2028 (“If the amount of the forfeiture is grossly disproportional to the gravity of the defendant’s offense, it is unconstitutional.”). Id. at 336-37, 118 S.Ct. 2028. Especially in a multi-defendant case like this one, this focus on the particular defendant’s offense is of more than passing importance. For example, as the court has previously noted, “the federal criminal offense of conspiracy to launder money does not require much more than bad thoughts.” United States v. Bramley, 2015 WL 12852048, at *8 (W.D. Okla. Sept. 16, 2015). For purposes of evaluating the gravity of criminal conduct, the statutory offense of conviction is the beginning, not the end, of the analysis.
Second, Bajakajian rejected the government’s argument that forfeiture of the entire amount of cash which had not been reported when it was transported out of the country, was a “perfectly calibrated” forfeiture. The Court rejected this “inherent proportionality” argument. Id. at 339, 118 S.Ct. 2028. This is relevant to some of the defendants in this case — particularly the runners who transported cash but didn’t generate those proceeds and derived no exceptional financial benefit from the betting activity that produced the proceeds for the Legendz enterprise.
Third, with regard to specific factors weighed by a court, Bajakajian sets out several factors considered in the dispro-portionality analysis. An important part in the analysis is played by the legislature’s enactments regarding the types and limits of punishments, such as potential sentences and fines; the Court also considered whether the defendant falls into the class of persons at which the criminal statute was principally directed; and the Court considered the harm which the defendant caused. Id. at 336-39, 118 S.Ct. 2028. In addition, the Court looked at the relationship of defendant’s offense to other illegal activity, of which it found there was none in that case. Id.
Applying these Bajakajian factors in a civil forfeiture case (involving residential property which the government alleged had been used to commit or facilitate drug offenses), the Tenth Circuit explained as follows. “One of the most important [factors from Bajakajian] was Congress’s judgment about the appropriate punishment for the owner’s offense. Maximum statutory fines provide guidance on the legislative view of the seriousness of the offense.” United States v. Wagoner County Real Estate, 278 F.3d 1091, 1100 (10th Cir. 2002). The fines set out in the sentencing guidelines are a way of translating the gravity of a crime into monetary terms. Id. Additional factors for consideration with regard to the gravity of the offense include the extent of the criminal activity, related illegal activities, and the harm caused to other parties. Id, citing Bajakajian. Wagoner County also recognizes that a proportionality analysis is factually intensive, so that a catalog of factors is not necessarily exclusive. Id. at 1101. Thus, in addition to the Bajakajian factors, factors such as “the general use of the forfeited property, any previously imposed federal sanctions, the benefit to the claimant, the value of seized contraband, and the property’s connection with the offense,” are relevant considerations. Id.28
*902Bearing in mind that, as has been discussed, the proportionality analysis is necessarily undertaken on an individualized basis, it is worth noting that in hierarchical organizations in general (and in large gambling organizations in particular), there will be substantial differences among the defendants in terms of their roles in the criminal activity. Thus, in United States v. Lyons, 870 F.Supp.2d 281, at 292 (D. Mass. 2012), a RICO prosecution involving a large gambling enterprise, Judge Saris expressly recognized the propriety of a Baja-kajian-driven proportionality analysis. In affirming the district court opinion (which it described as “detailed, thoughtful, and well-researched,” United States v. Lyons, 740 F.3d 702, 733 (1st Cir. 2014)), the First Circuit also noted that an individual defendant’s role in the organization is relevant to the proportionality analysis. Id.
A non-exclusive list of factors pertinent to a disproportionality analysis includes the following.
(i) Congress’s judgment about the appropriate punishment for the defendant’s offense, as reflected in the maximum statutory fíne for the counts of conviction and the fines set out in the Sentencing Guidelines;
(ii) whether the defendant falls into the class of persons at which the criminal statute is principally directed;
(iii) the gravity of the defendant’s offense and the extent of his criminal activity;
(iv) any related illegal activities;
(v) the harm caused to other parties;
(vi) the general use of the forfeited property;
(vii) any previously imposed federal sanctions;
.(viii) the benefit reaped by the defendant as a result of his criminal conduct;
(ix) the value of the property; and
(x) the property’s connection to the offense.
Aside from these considerations, at least two circuits have expressly held that deprivation of a defendant’s future ability to earn a living should be considered.29 See, United States v. Viloski, 814 F.3d 104 (2d Cir. 2016); United States v. Levesque, 546 F.3d 78 (1st Cir. 2008). As noted in Viloski, Bajakajian emphasizes that the Excessive Fines Clause grew out of the English constitutional tradition, including the Magna Carta, which required that a fíne “should not deprive a wrongdoer of his livelihood.” Viloski at 111, quoting Bajakajian, 524 U.S. at 335, 118 S.Ct. 2028.30 Viloski reasons that in light of its “strong constitutional pedigree, it seems unlikely that the Bajakajian Court meant to preclude courts from considering whether a forfeiture would deprive an offender of his liveli*903hood.” Viloski at 111. Levesque also cites passages from Bajakajian, and concludes that the notion that a forfeiture should not be so great as to deprive a wrongdoer of his livelihood is “deeply rooted” in the history of the Eighth Amendment. Levesque, 546 F.3d 78, 83-84. For example, Levesque states that “a court should consider a defendant’s argument that a forfeiture is excessive under the Eighth Amendment when it effectively would deprive the defendant of his or her livelihood. Such ruinous monetary punishments are exactly the sort that motivated the 1689 [English] Bill of Rights and, consequently, the Excessive Fines Clause.” Id. at 84-85.
Viloski and Levesque part ways, however, when it comes to exactly how concerns over a defendant’s ability to provide a future livelihood should be folded into the eighth amendment mix. Viloski holds that deprivation of the future ability to earn a living is “a component of the proportionality analysis, not a separate inquiry.” Viloski, 814 F.3d at 111-12. Thus, for the Second Circuit, a forfeiture which destroys a defendant’s livelihood may nevertheless be constitutional, depending on defendant’s culpability and other circumstances. Id. at 112. Under Viloski, the impact of a money judgment on a defendant’s future ability to earn a living is not a free-standing cap on the government’s ability to obtain forfeiture, rather, it is one consideration in the disproportionality assessment. Levesque, on the other hand, regards deprivation of future livelihood as a question which is “beyond,” and “is separate from,” the mul-ti-part “test for gross disproportionality....” Id. at 83-85. Levesque concludes by stating that “it is not inconceivable that a forfeiture could be so onerous as to deprive a defendant of his or her future ability to earn a living, thus implicating the historical concerns underlying the Excessive Fines Clause.” Id. at 85. Thus, Levesque conceives of the deprivation of livelihood inquiry as a cap which stands apart from the various factors looked at under the gross disproportionality analysis.
Both Viloski and Levesque give a reasonable reading to Bajakajian and the history of the Excessive Fines Clause, saying that deprivation of future livelihood plays a role under the Eighth Amendment. The court adopts the Levesque approach, because the treatment of the livelihood criterion as a free-standing reference point is more consonant with the holding, reasoning, and historical antecedents articulated by the Court in Bajakajian. The Court’s historical discussion in Bajakajian is illuminating:
[T]he [Excessive Fines] Clause was taken verbatim from the English Bill of Rights of 1689. That document’s prohibition against excessive fines was a reaction to the abuses of the King’s judges during the reigns of the Stuarts, but the fines that those judges imposed were described contemporaneously only in the most general terms. See Earl of Devonshire’s Case, 11 State Tr. 1367, 1372 (H.L. 1689) (fine of £30,000 excessive and exorbitant, against Magna Charta, the common right of the subject, and the law of the land). Similarly, Magna Char-ta — which the Stuart judges were accused of subverting — required only that amercements (the medieval predecessors of fines) should be proportioned to the offense and that they should not deprive a wrongdoer of his livelihood:
‘A Free-man shall not be amerced for a small fault, but after the manner of the fault; and for a great fault after the greatness thereof, saving to him his contenement; (2) and a Merchant likewise, saving to him his merchandise; (3) and any other’s villain than ours shall be likewise amerced, saving his wainage.’ Magna Charta, 9 Hen. *904III, ch. 14 (1225), 1 Stat. at Large 6-7 (1762 ed.).
Bajakajian, 524 U.S. 321, 335-36, 118 S.Ct. 2028.
The Court’s very evident concern about deprivation of livelihood, and the Magna Carta’s concern about the same thing— “wainage” being tools and implements of husbandry used to earn a livelihood— strongly suggest that the livelihood criterion is not to be folded into the mix of other excessive fines criteria, to make a difference, or not, depending on how the other criteria shake out.
Although the court, for the reasons stated, adopts the Levesque approach to the livelihood issue, the results stated in this order would be the same whether determined under the Levesque or the Viloski approach. Pursuant to Levesque, the fact-findings portion of this order applies two separate screens (not to be confused with the initial screen for foreseeability) to determine how the Eighth Amendment impacts a particular forfeiture sought by the government. First, the court will screen out (disallow) a requested forfeiture to the extent that the forfeiture would be grossly disproportionate to the gravity of the defendant’s offense as determined under the list of factors described above. Second, the court will screen out (disallow) a requested forfeiture to the extent that the forfeiture would be so onerous as to be potentially ruinous to the defendant’s ability to earn a living.
Finally in this regard, the court emphasizes that a defendant’s present inability to satisfy a forfeiture, in and of itself, is not at all sufficient to render a forfeiture unconstitutional, and is also not the correct inquiry. Levesque, 546 F.3d at 85. This is so because the court cannot consider, as a discrete factor, a defendant’s personal circumstances, such as his age, health or present financial condition. Viloski, 814 F.3d at 112. Of course, it is possible a defendant’s health or financial condition might bear on his ability to earn a future living and thus be indirectly relevant. Id. at 113. But that is a different thing from considering a defendant’s present financial circumstances as a discrete factor, which is not permitted. Id. at 113, 115-16.31 Thus, looking at the day of sentencing as Day 1 of the rest of the defendant’s financial life, forfeiture may render the defendant a pauper on Day 1. But as for Day 2 and thereafter, the court clearly has the authority to consider the impact of the forfeiture sought by the government on the defendant’s ability to make a living.
Measured by the joint and several $231,432,686.73 forfeiture money judgment it seeks, the government’s objective, if realized, would sentence the forfeiture defendants life terms of impoverishment or living off the books, or both. The text, the historical antecedents, and the judicial treatment of the Excessive Fines Clause, as expounded in Bajakajian and other relevant precedents, tell us that the statutory forfeiture tools wielded by the government in this case, powerful though they are, do not trump the limitations imposed by the Eighth Amendment.
To summarize, the Eighth Amendment’s Excessive Fines Clause requires the court to consider, first, the Bajakajian factors (supplemented by other factors which may *905be relevant) to determine whether the forfeiture in question is grossly disproportionate to the defendant’s offense. Second, the court considers whether the forfeiture in question would be so ruinous to the defendant’s future ability to provide a livelihood as to be unconstitutional for that reason. Accordingly, the forfeitures the government seeks will be tested for foreseeability, disproportionality, and deprivation of the ability to earn a livelihood.
IV. Burden of Proof
A. The Government’s Burden
The government has the ultimate burden of establishing the forfeitures by a preponderance of the evidence. United States v. Smith, 770 F.3d 628, 637 (7th Cir. 2014). Thus, the burden is on the government to prove, by a preponderance of the evidence, the requisite nexus between the property sought to be forfeited and the underlying crime of conviction. United States v. Susaeta, 2009 WL 928657, *2 (D. Utah 2009) (“The government’s burden at a forfeiture hearing is to prove, by a preponderance of the evidence, the nexus between the property and the offense.”); Rule 32.2(b)(1)(A) (“the court must determine whether the government has established the requisite nexus between the property and the offense”).
The preponderance of the evidence standard also applies to the government’s burden to establish the amount of the criminal forfeiture money judgment. United States v. Executive Recycling, Inc., 953 F.Supp.2d 1138, 1158 (D. Colo. 2013), citing Bader, 678 F.3d 858, 893, for the proposition that a forfeiture money judgment must be supported by a preponderance of the evidence.
“[O]nce the defendant has contended, with some evidentiary support, that at least some of the value in a given asset came from lawful, non-forfeitable sources, then the prosecutor must demonstrate how much is forfeitable.” United States v. Genova, 333 F.3d 750, 762 (7th Cir. 2003). And see, Bornfield, 145 F.3d 1123, 1135 (involving money-laundering-based forfeitures; “forfeiture of legitimate and illegitimate funds commingled in an account is proper as long as the government demonstrates that the defendant pooled the funds to facilitate, ie., disguise the nature and source of, his scheme,”); Rudaj, 2006 WL 1876664 at *4 (noting “universal agreement that at least the portion of property that derives from proceeds of illegal activity is subject to forfeiture”).
Defendants object to the preponderance of the evidence standard when applied to forfeitures based on RICO convictions, citing the Third Circuit’s rule that the government must establish entitlement to RICO forfeitures beyond a reasonable doubt. See, United States v. Pelullo, 14 F.3d 881, 906 (3d Cir. 1994). Most courts addressing the issue have held otherwise, relying primarily on statements in Libretti v. United States, 516 U.S. 29, 116 S.Ct. 356, 133 L.Ed.2d 271 (1995), to the effect that forfeitures based on RICO convictions are a part of sentencing. See, e.g., United States v. Fruchter, 411 F.3d 377, 383 (2d Cir. 2005) (“Libretti remains the determinative decision. Accordingly, the district court did not err when it applied a preponderance standard to the determination of Braun’s forfeiture,” in determining criminal forfeiture based on RICO conviction); United States v. DeFries, 129 F.3d 1293, 1312 (D.C. Cir. 1997) (government must prove forfeiture allegations under 18 U.S.C. § 1963(a) by a preponderance of the evidence, citing Libretti for the proposition that criminal forfeiture is an aspect of sentencing). This court concludes that the preponderance of the evidence standard is correct. That makes a difference in *906this case. Much of the forfeiture relief granted in this order would be unavailable if the government’s forfeiture case were measured by any standard more demanding than the preponderance standard.
The government also has the burden of proof on the issue of foreseeability as a prerequisite to joint and several liability for money laundering forfeiture. Because United States v. Lyons, 870 F.Supp.2d 281, at 292 (D. Mass. 2012), was a gambling case, Judge Saris’s treatment of the foreseeability issue in that case is illuminating.
In Lyons, as in the case at bar, the sports betting operation was geographically far-flung. The betting operation (“Sports Offshore,” a/k/a SOS) was based in Antigua. It operated “in Massachusetts, Florida, South Carolina and other parts of the United States from 1996 through 2010.” Id. at 284. As a result, some of the defendants didn’t know much about what the others were doing. Thus:
[T]he government did not introduce evidence that Thomas informed Lyons that SOS was operating in Florida and did not meet its burden to prove ... that the proceeds SOS generated in Florida were reasonably foreseeable to Lyons. Accordingly, the checks totaling $416,770 in SOS gambling proceeds that Eremian deposited into his personal bank account were not reasonably foreseeable to Lyons.
Id. at 289 (emphasis added).
The quoted passage goes to the question of what the government has the burden of proving was foreseeable. The answer, per Judge Saris, is that the government must prove that the generation of proceeds, arguably in a given locale, by a given individual was reasonably foreseeable to a defendant who would be, for forfeiture purposes, chargeable with those proceeds. This court agrees. The next question is: How does this work in a multi-defendant gambling case with defendants operating in various places, with varying roles and varying levels of involvement. In Lyons, that issue was parsed as follows:
In the circumstances of this case, the government has proven that proceeds generated by SOS agents and bettors known to a defendant were reasonably foreseeable to that specific defendant. On the other hand, I find that the government has not met its burden with respect to agents not known to the defendants, even though the existence of such agents would be conceivable.
Id. at 292 (emphasis added).
The court agrees with this approach as far as it goes. If a given agent, say, Campbell, was at least generally known to one of these forfeiture defendants as an active Legendz agent, then it is highly likely that the betting proceeds generated by that agent (i.e., money “involved in” the offense of conviction for purposes of Count 3 — see, 18 U.S.C. § 982(a)(1) and discussion in Part 11(E), above) were reasonably foreseeable to that defendant. This court does not exclude the possibility, however, that there are situations in which illegal betting proceeds generated by a particular agent may have been reasonably foreseeable to a particular forfeiture defendant even if that defendant was not specifically aware of that agent or of the details of his activities. This is, of necessity, a fact-bound determination.32 Finally, on this topic, it should be *907noted that the court’s assignment of the burden of proof on foreseeability to the government, as the plaintiff, for forfeiture purposes, is consistent with judicial treatment of that issue in other contexts.33
B. The Defendants’ Burden
The defendants other than King assert that the forfeiture relief available to the government is limited by the Excessive Fines Clause, as is discussed in Part III, above. If facts relevant to the application of the Excessive Fines Clause are actually in controversy, the burden of going forward with the evidence — sometimes called the burden of production— rests with the defendant, as the proponent of the limitation.34 But on these issues, it should be borne in mind that, in terms of proof of facts, the burden of going forward with the evidence comes into play and potentially makes a difference only where there actually is a fact issue. Thus, for instance, where the evidence or other reliable information shows without contradiction that it would not take much of a forfeiture judgment to destroy a defendant’s future ability to earn a livelihood, the burden of going forward (in default of which the defendant loses on that issue) is wholly academic — it does not come into play as a tie breaker because there is no tie to break. That said (and regardless of whether burden of going forward makes a difference as to factual matters), the ultimate burden of persuasion as to eighth amendment limitations remains with the defendants. This is so both as to the application of the Bajakajian factors (supplemented by other factors which may be *908relevant) and as to the livelihood issue. On the livelihood issue, the ultimate burden rests with the defendant to persuade the court that it should apply a limitation on forfeiture due to deprivation of future ability to earn or provide a livelihood. See, Viloski, 814 F.3d 104, 115; and see, Fogg, 666 F.3d 13, 18-19 (district court incorrectly put the burden on the government to prove that the defendant could pay, overlooking the fact that defendant had made no eighth amendment argument).35
V. The Exclusionary Rule
During King’s trial, the court excluded documents which Karlo Stewart surreptitiously took, with the encouragement of the government, from the offices of Data Support Services. (For present purposes, DSS may be considered to have been the equivalent of Legendz.) As noted in Part 1(C), above, those documents were obtained in violation of King’s fourth amendment rights. United States v. King, 212 F.Supp.3d 1113 (W.D. Okla. 2015). King argues that ruling means the Stewart documents are also excluded from consideration in these criminal forfeiture proceedings against him. (Some of the other forfeiture defendants make similar arguments, even if not explicitly premised on the exclusionary rule. For example, Koralewski’s proposed findings concern the fact that the government’s “agent list” was excluded in his trial; he argues it lacks reliability. Doc. no. 2273, pp. 1-2.) King’s argument has much appeal. The government seeks to render him a pauper based, to significant degree, on documents that are in the possession of the government only because of a search that clearly violated King’s fourth amendment rights. But the court concludes, reluctantly, that King’s reliance on the exclusionary rule is, under authority by which this court is bound, misplaced.
Criminal forfeiture is a component of the sentence imposed following a defendant’s conviction. Libretti, 516 U.S. 29, 38-39, 116 S.Ct. 356. And see, United States v. Sanders, 743 F.3d 471, 472-73, 475 (7th Cir. 2014) (Easterbrook, J., characterizing 18 U.S.C. § 3661 as a statute “which provides that all evidence is admissible at sentencing”; rejecting a proposed exception in that case for egregious violations of the Fourth Amendment). And the Tenth Circuit has held that it was not error for a court, at sentencing, to consider evidence obtained during an illegal search, so long as there was no evidence that the officers’ unconstitutional actions were taken with the intent to secure an increased sentence (or, presumably, an increased forfeiture as a part of sentencing). United States v. Ryan, 236 F.3d 1268, 1272-73 (10th Cir. 2001). (King has not argued, nor is there evidence to show, that in obtaining the Stewart documents law enforcement sought to secure an increased sentence or an increased forfeiture amount. Accordingly, that possible exception to the general rule need not be further considered.)
Citing One 1958 Plymouth Sedan v. Commonwealth of Pennsylvania, 380 U.S. 693, 700-02, 85 S.Ct. 1246, 14 L.Ed.2d 170 (1965), King contends that the primary rationale for applying the exclusionary rule in civil forfeiture proceedings is that such proceedings are quasi-criminal, and he therefore argues that the rule, perforce, *909must apply in proceedings which are actually criminal in nature as opposed to qua-sf-crimmal.36 There are differences, however, between civil and criminal forfeiture proceedings which, in addition to the fact that criminal forfeiture occurs as a part of sentencing, may help explain the different treatment afforded to the exclusionary rule in these two contexts. Criminal forfeiture proceedings may only ensue after the evidence, filtered through all rules and protections which apply in a criminal trial, has been found sufficient to convict a defendant beyond a reasonable doubt. Civil forfeitures, on the other hand, do not have an equivalent preliminary stage for testing the prosecution’s evidence; in civil forfeiture, forfeitability is determined by a preponderance of the evidence and neither the property owner, nor anyone else for that matter, need be convicted of the crime giving rise to the forfeiture.
Furthermore, the exclusionary rule is a prudential doctrine. Davis v. United States, 564 U.S. 229, 236, 131 S.Ct. 2419, 180 L.Ed.2d 285 (2011), citations and quotations omitted. Exclusion is not a personal constitutional right, nor is it designed to redress the injury occasioned by an unconstitutional search. Id. The rule’s sole purpose is to deter fourth amendment violations. Id. at 236-37, 131 S.Ct. 2419. Its operation is limited to situations in which this purpose is efficaciously served. Id. at 237, 131 S.Ct. 2419. Where suppression fails to yield appreciable deterrence, exclusion is clearly unwarranted. Id. The Tenth Circuit recognizes that applying the exclusionary rule at the sentencing stage of a criminal trial would, in the ordinary case, have a minimal deterrent effect on law enforcement’s conduct. Ryan, 236 F.3d 1268, 1271. There is nothing about the facts of this case which would take it out of this general rule, debatable though that general rule may be.
Thus, despite the inadmissibility of the Stewart documents during King’s trial, the documents are not excluded in these criminal forfeiture proceedings against King. This ruling makes the Stewart documents admissible. It does not suggest what weight, if any, the court will ultimately give the documents, but the court will note, first, that significant segments of the government’s forfeiture case are rooted in data extracted from the spreadsheets taken by Karlo Stewart at the instance of the government, in violation of King’s fourth amendment rights (see, e.g., Tr. 720-21), and secondly, that, throughout the forfeiture trial, King was careful to preserve his objection to the use of the documents that were taken in violation of his fourth amendment rights. E.g., Tr. 70, 71, 437, 440. The court excused King’s counsel from having to reassert *910that objection throughout the forfeiture trial. Tr. 440.
VI. Tracing, Tainting, Commingling and Extrapolation
The jury’s guilty verdict may turn a prince into a frog, but the frog, even though possibly headed for prison, doesn’t forfeit anything until the government proves a forfeiture case. And because of the natural import of the statutory language defining a defendant’s liability to forfeiture, that forfeiture case may present issues and problems of proof — for both sides — that differ markedly from the issues and problems of proof that arose in the underlying trial of the issue of guilt or innocence. For that reason, the discussion returns to the relevant statutory language. The forfeiture statutes that are relevant in this case use various words to express the concept that there must be some connection between the offense that triggers forfeiture and the assets sought to be forfeited — what Rule 32.2 refers to generically as the “nexus.” Those statutory provisions are discussed in detail in Part II, above, but brief repetition is necessary here.
For present purposes:37
On Count 1, forfeiture reaches property “constituting” or “derived from” any “proceeds” which the defendant “obtained, directly or indirectly, from racketeering activity.”38
On Count 2, forfeiture reaches property that was “used in violation of’ § 1955 or that “constitutes” or is “derived from” proceeds “traceable” to a violation of § 1955.39
On Count 3, forfeiture reaches property “involved in” the offense of conviction, or property “traceable to” such property.40
As will be seen, the judicial decisions do pay attention to these words defining the connection the government must prove.
A. Tracing, Tainting and Commingling
Given the nature of most of the criminal activity (especially including organized criminal activity, as in the case at bar) the federal courts see, the property that is ultimately found and forfeited is often not the property that was used to commit the offense (the “corpus”) or that was directly generated by the offense. Concealment and the mere passage of time see to that. Consequently, the forfeiture statutes, or at least those that are relevant here, permit the government to cast a wide net. It falls to the courts, addressing concrete cases, to determine the reach of the statutory language in those concrete cases. The Tenth Circuit and the other Courts of Appeals have provided much valuable guidance.
The Tenth Circuit’s 1998 decision in Bornfield, 145 F.3d 1123, already discussed in this order in other contexts — a decision that was bracketed in point of time by two equally illuminating Third Circuit decisions, United States v. Voigt, 89 F.3d 1050 (3rd Cir. 1996), cert. denied, 519 U.S. 1047, 117 S.Ct. 623, 136 L.Ed.2d 546 (1996) and United States v. Stewart, 185 F.3d 112 (3rd Cir. 1999), cert. denied, 528 U.S. 1063, 120 S.Ct. 618, 145 L.Ed.2d 512 (1999)— provides a good starting point.
*911Bornfield was a money laundering ease. Accordingly, as the Court of Appeals recognized, the key to the determination of forfeitability was the question of whether the property in question was “involved in” or “traceable to” the offense. Id. at 1135. The case involved two bank accounts, one of which (a personal account) had been a conduit for tainted money and one of which (a business account) had not. But, as the court described it, there was “mass confusion between [the] two accounts.” Id. at 1137. On appeal (and apparently in the district court), reliable differentiation between the two accounts was made impossible because the jury instructions at the forfeiture stage associated the personal account (through which tainted funds had passed) with the name of the business. Id. Consequently, the court noted that “we cannot tell with certainty the proper identifications of the accounts.” Id. Moreover, at the time of the indictment, the tainted funds were no longer in his personal account, with the result that “any money contained in the personal account would be subject to forfeiture solely under a facilitation theory.” Id. at 1138. After the forfeiture jury trial, the district court apparently attempted to scab over these problems by ordering “forfeiture based on the substitute assets provision of the same asset alleged to be forfeitable.” Id. at 1139.
In reversing the forfeiture order, the court noted, as an initial matter, that, although, in the indictment, the government sought forfeiture of a specific account, “the account itself is not ‘property’ subject to forfeiture, but merely a container to hold the ‘property’ or money subject to forfeiture. In other words, the account number is merely used for identification purposes and is not itself a forfeitable item.” Id. at 1137 n. 7 (citing Judge Easterbrook’s opinion for the Seventh Circuit in United States v. $448,342.85, 969 F.2d 474, 476 (7th Cir. 1992)).41 Thus, the court’s analysis centered on “the funds in Bornfield’s personal account” (id. at 1138), and not on the account itself as forfeitable property.
Addressing the fact that the tainted funds were no longer in the account at the time of the indictment, the court concluded that the only remaining basis for forfeiture would be a facilitation theory, rendering forfeitable “all of the money contained in the account at the time of the offense,” id. (emphasis added), but — because of the conceptual basis for the facilitation theory — not limited to the amount of the corpus (the corpus being the laundered finds that had passed through the account). Id. at 1138. It was “clear error” to measure the forfeiture by the amount of the corpus because “the corpus was no longer in the account.” Id. And as for facilitation— reaching tainted and untainted funds commingled in an account — the government was obliged to “demonstrate[ ] that the defendant pooled the funds to facilitate, i.e., disguise the nature and source of, his scheme.” Id. at 1135. Citing, among other cases, the Third Circuit’s decision in Voigt, the Court of Appeals also explained that the “traceable to” prong of § 982(a)(1) reaches “property where the acquisition is attributable to the money laundering scheme rather than money obtained from untainted sources.” Id. at 1135.
The district court’s resort to a substitute assets theory did not save the forfeiture order in Bornfield:
An asset cannot logically be both forfei-table and a substitute asset. To allow *912such an anomaly would render the substitute assets provision meaningless. Assets involved in or traceable to the offense are forfeitable once the requisite nexus is established. See 18 U.S.C. § 982(a). The substitute assets provision allows the forfeiture of other assets not already forfeitable when the forfeitable asset is unavailable due to some act or omission of the defendant. [Citations omitted.] Here, the district court ordered forfeiture based on the substitute assets provision of the same asset alleged to be forfeitable. However, the court could not logically order the forfeiture of § 982(a)(1) forfeitable assets pursuant to § 982(b) and § 853(p), the substitute assets provisions.
Bornfield, at 1139.
Accordingly, Bornfield establishes that:
• Accounts, as such, are not forfeita-ble, even on a facilitation theory. What the government must go after is property — money or other property, but not an account as such — that is either (i) the corpus of the offense or (ii) acquired with tainted funds, or (iii) used to facilitate the offense.
• The “mere pooling or commingling of tainted and untainted funds in an account does not, without more, render the entire contents of the account subject to forfeiture.” Id. at 1135. (Or, as the court stated in United States v. $448,342.85: “Once we distinguish the money from its container, it also follows that the presence of one illegal dollar in an account does not taint the rest — as if the dollar obtained from fraud were like a drop of ink falling into a glass of water.” 969 F.2d at 476.)42
• Because of the fungible nature of money, if “the corpus of the money laundering offense” is not shown to have been in the account “at the time of the indictment,” the money in the account is forfeitable, if at all, only on a facilitation theory. Id. at 1138.
• If the government satisfies the prerequisites of the facilitation theory, forfeiture “of legitimate and illegitimate funds commingled in an account is proper,” and not limited by the amount of the corpus. Id. at 1135.
• But the money that is forfeitable on a facilitation theory is limited to “what was in the account at the time of the offense.” Id. at 1138 n. 11.
• The substitute asset legislation permits forfeiture of assets not already forfeitable when the defendant has put the forfeitable asset beyond reach.
In Bornfield, the Tenth Circuit’s focus was on fungible property — money—in bank accounts. The Third Circuit’s decision in Voigt, cited with approval in Bornfield, dealt with tangible property (jewelry) purchased with funds in an account that had served as a conduit for criminally laundered funds. After the tainted funds were deposited, there were “numerous intervening deposits and withdrawals.” Voigt, 89 F.3d at 1082. The question was whether “the money used to purchase the jewelry in question was ‘traceable to’ money laundering proceeds.” Id. at 1084. Because of the commingling problem, the government *913sought to reach the jewelry via the substitute assets provision, § 982(b), that was added to § 982 in 1988. Id. at 1085.43 The Third Circuit rejected this argument, as did the Tenth Circuit two years later in Bornfíeld, but the Third Circuit’s follow-on reasoning is instructive because of the government’s arguments, in that case, relating to the tangible property. The Third Circuit started out with the basics:
We hold that the term “traceable to” means exactly what it says, [footnote omitted] In light of our holding on the burden of proof, this means that the government must prove by a preponderance of the evidence that the property it seeks under § 982(a)(1) in satisfaction of the amount of criminal forfeiture to which it is entitled has some nexus to the property “involved in” the money laundering offense.
[[Image here]]
Where the property involved in a money laundering transaction is commingled in an account with untainted property, however, the government’s burden of showing that money in the account or an item purchased with cash withdrawn therefrom is “traceable to” money laundering activity will be difficult, if not impossible, to satisfy.
Id. at 1087.
Because the jewelry had been “purchased with funds from an account into which money laundering proceeds had been commingled with other funds, and after numerous intervening deposits and withdrawals,” the court concluded that the jewelry could not be found to be “traceable to” money laundering activity. Id. at 1088. This reasoning would apply a fortiori where the government’s evidence does not show the immediate source of the funds used to purchase a tangible asset.
As has been seen, in 1992, the Seventh Circuit opined that “the presence of one illegal dollar in an account does not taint the rest.” United States v. $448,342.85, 969 F.2d 474, 477 (7th Cir. 1992). In an opinion that limited an expansive reading of its decision in Voigt, the Third Circuit held that the converse is also true: The commingling of a smidgeon of clean money with a heap of dirty money will not stand in the way of forfeiture of the dirty money. In United States v. Stewart, 185 F.3d 112 (3rd Cir. 1999), cert. denied, 528 U.S. 1063, 120 S.Ct. 618, 145 L.Ed.2d 512 (1999), the Third Circuit distinguished Voigt and granted forfeiture of $2.6 million that remained in an account into which the defendant had deposited $160,000 of untainted money and $3 million of tainted money. Id. at 129-30. There had not been numerous deposits and withdrawals, and the account was frozen “almost immediately” after the $3 million in dirty money was deposited. Id. (One $600,000 withdrawal had been made by agreement, deemed to have included the untainted funds. Id. at 130.)
The question before the court in Stewart was “whether the government may forfeit directly tainted funds from an account that has been frozen from the time of the illegal transfer but that also contains untainted money.” Id. Pointing out that Voigt was not a case in which “one readily could separate out the amount subject to forfeiture,” id. (citation omitted), the court noted that: “Here, the government clearly *914traced laundered funds forfeited by the jury to Stewart’s account. Stewart does not contest this tracing, which in any event the government clearly established.” Id. at 129-30. Thus, Voigt’s treatment of commingling did not dictate the result in Stewart because, as the Seventh Circuit has noted (discussing Voigt and Stewart in tandem), Voigt was “a very simple case involving few individual deposits.” In re Rothstein, Rosenfeldt, Adler, P.A., 717 F.3d 1205, 1213 (11th Cir. 2013). In Roth-stein, the commingling occurred over a long period of time and involved numerous transactions. Consequently, the Seventh Circuit concluded that: “The sheer volume of financial information available and required to separate tainted from untainted monies in this case leads us to the conclusion that it is far more appropriate to apply the Third Circuit’s rule in Voigt than the exception to that rule it lays out in Stewart.” Rothstein, 717 F.3d at 1213.44
As for hard assets, if a hard asset was acquired (or the mortgage on it serviced) partly with untainted funds, the record must “establish how much of the value came from [untainted sources] and how much from [tainted sources].” United States v. Genova, 333 F.3d 750, 762 (7th Cir. 2003). Cf., United States v. One 1980 Rolls Royce, 905 F.2d 89, 90 (5th Cir. 1990) (rejecting government’s contention that if one dollar of drug money was used to purchase an asset, the entire asset is for-feitable). And, as has been noted (in Part IV(A), above), once the defendant has contended, with some evidentiary support, that at least some of the value in a given asset did come from lawful, non-forfeitable sources, the burden of showing how much is forfeitable rests with the government. This evidentiary burden, resting on the government, is especially significant where, from the nature of the case, the government’s forfeiture case encompasses both (i) funds (or a judgment representing those funds) forfeitable because they are tainted and (ii) assets acquired at least in part with those funds: “[T]he value may be forfeited once, as cash or as an interest in the property, but not twice.” Genova, id.
The short of the matter is that commingling will not stand in the way of forfeiture if the account in question was used exclusively, or nearly so, for illicit purposes (or, by the same token, if the hard asset in question was acquired or financed exclusively, or nearly so, with illicit money). But if the commingling problem is not cured by proof of facilitation, significant commingling can present a significant obstacle to forfeiture.
The case at bar presents a tangle of issues as to tracing,45 commingling and *915related matters. The application of Bornfield,46 Voigt and Stewart (and cases of similar import, like Genova and Rothstein) to the government’s forfeiture claims that are now before the court is discussed below.
B. Extrapolation
As-has been discussed and will be further discussed in Part VII, below, the government is entitled to seek forfeiture by way of money judgments against the defendants. Forfeiture will not be limited to identifiable assets. But the need to determine, on a reasoned basis, the amount of a forfeiture money judgment raises the issue of extrapolation. Or, more precisely, the limits of extrapolation.
Extrapolation is necessary, and permitted by forfeiture law, in the absence of records that would directly provide, for the relevant time periods, the information necessary to enable the court to do the arithmetic. The law of extrapolation (to the extent that law, as opposed to the routine task of looking for plausible factual inferences, has any bearing on it) is that if the government proposes an extrapolation for the purpose of arriving at the amount of a forfeiture money judgment, that proposed extrapolation must be supported by evidence — measured by the preponderance standard — that provides a beginning point reliable enough to lead to the proposed end point (the amount to be included in the judgment) without unacceptable speculation and conjecture.
On one hand, it is quite obvious that “[t]he calculation of forfeiture amounts is not an exact science. The court need not establish the loss with precision but rather need only make a reasonable estimate of the loss, given the available information. A court is permitted to use general points of reference as a starting point for calculating the losses or gains from fraudulent transactions and may make reasonable extrapolations from the evidence established by a preponderance of the evidence at the sentencing proceeding.” United States v. Treacy, 639 F.3d 32, 48 (2d Cir. 2011) (internal citations omitted). On the other hand, if there are simply not enough reliable reference points (or if those reference points, though intrinsically reliable, are only marginally probative for other reasons, such as timing), the proposed extrapolation will cross the line into the realm of impermissible speculation and guesswork. See, e.g., Corrado, 227 F.3d 543, 557-58 (6th Cir. 2000) (thirty years of weekly draws of illegal gambling *916proceeds too speculative); Lyons, 870 F.Supp.2d 281, 287 (D. Mass. 2012) (rejecting government’s efforts to forfeit cash gaming proceeds which it argued represented an “extrapolation estimate” for the years in question; four-year time period over which the government sought to extrapolate was not similar to the original eight-year time period for various reasons, including law enforcement’s disruption of operations, and changed economic conditions such as the financial crisis in 2008)47; and United States v. Basciano, 2007 WL 29439, at *2 (E.D.N.Y. Jan. 4, 2007) (total amount of forfeited assets may be determined by “conservatively estimating” the revenue regularly collected or received, but evidence of such revenue may not be overly speculative).
VII. Notice of the Government’s Intent to Seek Money Judgment
Defendants argue that lack of notice, or lack of adequate or timely notice, precludes or estops the forfeitures which the government seeks in these proceedings. The court disagrees.
Rule 32.2(a) provides that “A court must not enter a judgment of forfeiture in a criminal proceeding unless the indictment or information contains notice to the defendant that the government will seek the forfeiture of property as a part of any sentence in accordance with the applicable statute.” The rule further provides that “The indictment or information need not identify the property subject to forfeiture or specify the amount of any forfeiture money judgment that the government seeks.” Id.
The Advisory Committee Notes to the 2009 Amendments, Subdivision (a), state: “The amendment ... makes it clear that the indictment or information need only provide general notice that the government is seeking forfeiture, without identifying the specific property being sought. This is consistent with the 2000 Committee Note, as well as many lower court decisions.”
The Superseding Indictment states that “notice is hereby given to the defendants that the United States will seek forfeiture as part of any sentence” for a conviction under counts one, two or three. Doc. no. 354, ¶ 110 (forfeiture allegation 1, pertaining to racketeering conspiracy offense); ¶ 115 (forfeiture allegation 2, pertaining to illegal gambling business offense), ¶ 119 (forfeiture allegation 3, pertaining to money laundering offense; stating that upon conviction the government will seek forfeiture of any property involved in such offenses or traceable to such property). The RICO forfeiture allegation of the Superseding Indictment lists specific interests subject to forfeiture, stating that “The interests of the defendants subject to forfeiture ... include but are not limited to at least $1,000,000,000.00 and all interests and proceeds traceable thereto.... ” Doc. no. 354, ¶ 112 (RICO). And see, ¶ 116 (the illegal gambling business forfeiture allegation, stating that “property of the defendants subject to forfeiture ... include but are not limited to at least $1,000,000,000.00 and all interests and proceeds traceable thereto, including but not limited to the following assets....”); ¶ 19 (money laundering forfeiture allegation, stating that “property of the defendants subject to forfeiture ... include but are not limited to at least $1,000,000,000.00 and all interests and proceeds traceable thereto, including *917but not limited to the following assets .... ”)•
Courts which have addressed the issue have found that the government is not required to provide notice of the possibility of a money judgment. See, e.g., United States v. Odom, 2007 WL 2433957, *4 (S.D. Miss. 2007) (defendant complained that the bill of particulars was silent as to any money judgment the government was seeking; court noted defendant had cited no law requiring the government to provide notice of the possibility of a money judgment and that courts which have addressed the issue have found no notice required, citing cases). As previously stated, the rationale which underlies permitting money judgments in the amount of unlawful proceeds is that the defendant may not have the proceeds in his possession at the time of conviction and should not be able to avoid forfeiture by ridding himself of ill-gotten gains. McGinty, 610 F.3d 1242, 1246. As this rationale suggests, facts which may prompt the need for a money judgment as part of a criminal forfeiture, or which may determine the amount of any money judgment eventually sought via forfeiture, are not necessarily facts which are fixed at the time of the indictment, or even at other stages.
Finally, there is no evidence that the timing of the government’s announcements regarding its intent to seek personal money judgments has prejudiced the forfeiture defendants.48
The forfeiture defendants’ notice arguments are rejected as a basis for concluding that the government is precluded or estopped from seeking money judgments.
VIII. Overview of the Government’s Forfeiture Case
An understanding of the outcome of the forfeiture proceedings will be aided by a review of the evolution of the government’s factual contentions in support of its forfeiture claims. The court’s determinations as to the merits of those factual contentions are not discussed here. They are addressed in Part IX, below.
A. The Government’s Forfeiture Brief (doc, no. 1999)
Although the broad outlines of the government’s forfeiture claims may be discerned from the motions for preliminary *918order of forfeiture, most of which were filed in mid-2015, the government’s Forfeiture Brief, doc. no. 1999, provided (as specifically required by the court — see Part 1(C), above) the first detailed statement of the basis for its forfeiture claims.
The government does not assert, either in its Forfeiture Brief or in its lengthy post-trial proposed findings and conclusions (doc. no. 2242), that any of the specific items of real or personal property it seeks to forfeit are sought as substitute assets. Consistent with that, the government does not cite, or otherwise purport to invoke, the legislation49 authorizing it, in some situations, to go after substitute assets.50 Consequently, the government’s forfeiture case presents no occasion for consideration of forfeiture of substitute assets.
Before going further, one other matter should be noted. The government’s Forfeiture Brief did, in general terms, acknowledge, as a legal matter, that the government is not entitled to run up the forfeiture score by double counting. Doc. no. 1999, at 20. But nowhere did the government attempt to address or explain the instances of double counting that are embedded in its forfeiture case, as is discussed in several places in this order. To the contrary (and remarkably), the government would have the court sift the record and do the math: “After determining which assets are subject to forfeiture and the appropriate theories for forfeiture, and after determining the full amount of the forfeiture judgments, the United States respectfully requests that the Court adjust the amount of the judgments and forfeiture of assets accordingly to ensure that there is no double counting.” Id. at 21 (emphasis added).
1. King
As to King, the government seeks forfeiture of specific amounts of funds (e.g., $978.18 in one instance) .from eighteen specifically identified bank and brokerage accounts (doc. no. 1999-1), as well as three automobiles, $10,236.86 in currency, 58 pieces of jewelry, 88 pairs of shoes (including, apparently, all of his 20 pairs of Jimmy Choo shoes, doc. no. 653-3), 75 purses, a jukebox, sports memorabilia, $303,527.46 as a substitute res for a residence in Texas, and five parcels of real property. Id.
As set forth in its Forfeiture Brief, the government also sought a money judgment against King in the amount of $222,358,480.00. Id. at 13. That total is substantially based on a series of extrapolations. Those proposed extrapolations are discussed in Part IX(C), below.
In support of this wide-ranging forfeiture, the government asserts that “[b]e-tween 2003 and 2013, Mr. King only worked for Legendz Sports [transcript citation omitted]. Consequently, his only source of income derived from his illegal gambling operation with Legendz Sports.” *919Id. at 3. The government’s citation is to the direct examination of Joseph McFadden, in which he testified that “to [his] knowledge,” Mr. King did not have “any job other than with Legendz” between 2002 and 2013. Oct. Tr. 607. (McFadden is a businessman and former Legendz agent who, at all times relevant to this case, resided in Florida, not Panama. Apr. Tr. 844-849, 862.) The government concluded that “Because Mr. King’s only source of income from 2003 through 2013 derives from his operation of the Legendz Sports enterprise, the United States contends that all property seized and listed for forfeiture constitutes proceeds of Mr. King’s criminal activity.” Doc. no. 1999-1, at 3-4. This assertion deserves, and will get, more discussion later in this order.
Eleven of the bank accounts listed for forfeiture are accounts in the name of Serena King and various children of Mr. and Mrs. King. Id. at 5-8. As for Mr. King’s seventy-five purses, the government states that “Mr. King’s only source of income was from the Legendz Sports illegal gambling enterprise. The approximately seventy-five miscellaneous purses constitute proceeds of his illegal gambling operations associated with Legendz Sports.” Id. at 11.
2. Koralewski
As to Koralewski, the government seeks to forfeit, by way of identifiable property, a gas lease covering the Koralewski residence in Parker, Colorado, as well as specific amounts found in two Wells Fargo bank accounts, totaling about $1,500.00. In addition, the government sought a “Substitute res” in the amount of $172,039.65, representing the proceeds of the sale of the Koralewski residence. Doc. no. 1999-5, at 1. The government asserted, by way of basis for the nexus between the property and Koralewski’s offense of conviction (Count 1 only) that, during the relevant period, “Koralewski’s only source of income derived from his association with Legendz Sports.” Id. The government elaborated by asserting that: “Koralewski failed to earn legitimate income from 2003 through 2013.” Id. at 2.
In its Forfeiture Brief, the government also sought a money judgment against Ko-ralewski for an amount slightly in excess of $1.2 million. Id. at 3. (As will be discussed, the government now seeks a forfeiture money judgment against Koralewski and the other forfeiture defendants in an amount in excess of $230 million.)
3. Tucker
As to Paul Tucker, the government sought, in its Forfeiture Brief, slightly less than $9,000.00 in currency which had been seized, as well- as the contents of four bank accounts. (Item 7 on the list on p. 1 of doc. no. 1999-7 is an exact duplicate, including dollar amount, of the description of the account listed as item 3 on that page, so the government’s Forfeiture Brief only implicates four accounts even though accounts are listed in five paragraphs.) The government’s Forfeiture Brief, filed in January, 2016, overlooks the fact that the cúrrency and three of the four bank accounts (those listed in paragraphs 3, 5 and 6 in doc. no. 1999-7, at 1) had been forfeited by agreement as set forth in the Preliminary Order of Forfeiture filed on November 13, 2015 (doc. no. 1797). Consequently the only account remaining in play is the one listed in paragraph 4 of doc. no. 1999-7, p. 1, an account at “Wells Fargo/Wachovia Bank” in the amount of $3,490.64.
The government also sought, in its Forfeiture Brief, a money judgment against Tucker in an amount in excess of $8.8 million “for his participation in the Le-*920gendz Sports enterprise.” Doc. no. 1999-7, at 3.
4. Moran
As to Moran, the government sought, in its Forfeiture Brief, approximately $60,000.00 in currency, as well as a money counter, a watch and a diamond ring. Doc. no. 1999-3, at 1. The government also stated that it sought a money judgment against Moran in an amount slightly less than $3 million, arrived at substantially by way of estimating the amount of money Moran picked up and delivered as a runner for the Legendz organization. See, doc. no. 1999-3, at 4-6.
5. Robles
In its Forfeiture Brief, the government sought a money judgment for slightly more than $815,000.00 against Robles, on the basis of the funds he picked up and delivered as a runner, plus the salary paid to him while he served as a runner. Doc. no. 1999-8, at 1-2.
6. Bramley
As to Bramley’s hard assets and financial assets, the government sought, in its Forfeiture Brief, forfeiture of his house in Plano, Texas, the contents of five investment accounts and bank accounts (in unspecified amounts), a Mercedes Benz automobile, slightly more than $23,000.00 in currency and slightly more than $1,100.00 in “collectible” currency. Doe. no. 1999-2, at 1. By way of nexus, the government asserted, among other things, that “Bram-ley’s only source of income (excluding Social Security benefits), between 2003 and 2013, was income he received from his participation in the Legendz Sports conspiracy.” Id. at 2. As to the residential property in Plano, Texas, the government also asserted a facilitation theory, alleging that “Bramley maintained an office in the residence that he used to conduct illegal gambling activity.” Id. at 5-6.
In addition to the hard assets and financial assets, the government also sought a money judgment against Bramley in an amount slightly in excess of $5.1 million.
7.Diebner
As to Diebner, by way of hard assets and financial assets, the government sought, in its Forfeiture Brief, forfeiture of a residence in Houston, Texas, the contents of three bank accounts (unspecified as to amount), approximately $8,400.00 in U.S. currency and miscellaneous items of jewelry. Doc.-no. 1999-4, at 1. In support of forfeiture of the currency, the government stated, in its Forfeiture Brief, that: “Over the course of his involvement with the Legendz Sports enterprise, [Diebner] is responsible for approximately $18,567,345.60 in proceeds.” Doc. no. 1999-4, at 3. In the very next paragraph, in support of forfeiture of the jewelry, the government said: “Over the course of his involvement with the Legendz Sports enterprise, [Diebner] is responsible for approximately $3,679,585.80 in proceeds.” Id. No explanation is offered for the difference between these two figures.
As to Diebner, the government also sought a money judgment in the amount of $18,567,345.00. Id. at 4.
B. The government’s presentation at the forfeiture trial
The government called two witnesses at the forfeiture trial. The first witness was Bart Lowrance, an asset forfeiture investigator employed by the Federal Bureau of Investigation. Tr. 8. The government’s other witness was David Jansen, also a forfeiture investigator, apparently on a contract basis. Tr. 427. Jansen’s description of his status was: “I do work for the FBI.” Tr. *921427.51 The defendants called no witnesses. (This case has had two case agents, one FBI agent and one agent of the Internal Revenue Service Criminal Investigation Division, from a very early stage. Neither of those case agents testified at the forfeiture trial or, for that matter, at any of the four jury trials.) The government introduced 228 exhibits at the forfeiture trial, of which fifty-eight were charts and summaries of voluminous documents. (In this order, the court will also refer, as needed, to exhibits introduced at one or more of the four jury trials.) The defendants introduced four exhibits (two by Moran and one each by Bramley and Diebner).
For reasons that will become apparent, some discussion of the extent of the knowledge of the government’s forfeiture witnesses is appropriate.
At the forfeiture trial, Lowrance made a discernible, and mostly successful, effort to stay within the limits of what he knew. And the main thing he knew was that his arithmetic was correct. For the most part, he was right about that. Lowrance, as the sponsor of the government’s updated recap of its forfeiture figures (GX C89),52 acknowledged that he did not examine any of the underlying evidence, testimony or documents backing up the extrapolation leading to the attribution of $1,055,112.00 to Koralewski. Tr. 387-88. Nor, at least as to Koralewski, did he look at any of the underlying summaries that are consolidated onto GX C89 (App. 1 hereto), a summary that Lowrance agreed was “an overall look at the financial forfeiture case.” Tr. 383, 388. He believes the numbers in GX C89 to be correct, but he does not know them to be correct, because he “did nothing independently of [Jansen] giving you [Low-rance] a number and you putting it on the document.” Tr. 389-90. Consequently, he “wouldn’t know whether any of the information Jansen relied on was problematic, inaccurate, disproved, debunked, or anything else.” Tr. 390. The same is true as to Bramley and Moran. Tr. 381, 391. Likewise, as to Tucker, Lowrance simply did a multiplication (for extrapolation purposes, as discussed below) of a number supplied by Jansen. Tr. 392.
The government’s second (and last) witness at the forfeiture trial was Jansen. A former agent with the Internal Revenue Service Criminal Investigation Division, Jansen is employed by an organization that contracts with the Department of Justice to do forfeiture and money laundering-related work for the FBI. In his words: “I do work for the FBI.” Tr. 427. He got involved in this case in 2011. His main activity has consisted of examining financial records for the purpose of “tracing *922money to assets and then trying to determine the source funds that bought them.” Tr. 432. Although Jansen acknowledged that the source documents he had to work with were incomplete (Tr. 432), he was, in comparison to Lowrance, clearly more familiar with those documents.
Although the testimony of Lowrance and Jansen was, in some ways, informative, it is worthy of note, before going further, that they added nothing to the record in this case in terms of first-hand knowledge imparted by testimony from percipient witnesses.
IX. Findings of Fact
A. Introduction
Broadly speaking, the government’s forfeiture case breaks down into two categories: (i) forfeiture of specifically-identified assets (partly financial assets and partly tangible assets such as houses, cars, shoes and purses) and (ii) forfeiture in the form of money judgments against the forfeiture defendants. As stated in Part I, above, the government originally contended that the forfeiture in this case would amount to a billion dollars. Regardless of what the merits of that number might be, or might ever have been, the government’s forfeiture contentions have been very much a moving target. This is so despite, or perhaps because of, the fact that the court has repeatedly required the government to provide specifics — first in terms of allegations and later in terms of proof — of just what it seeks to forfeit, both by way of specific assets and by way of forfeiture money judgments.
As for forfeiture money judgments, the government, when first required to clearly state what it sought (Case Management Order No. 7, doc. no. 1879), specified the amounts set forth at length in Part VIII(A), above, in its Forfeiture Brief. This, to put it mildly, has changed, as indicated in the following table.
Table 1
Government's Positions as to Amount Sought by way of Forfeiture Money Judgment
(All amounts in dollars)
Defendant Amount sought in Forfeiture Brief, doc. no. 199953 Amount sought In GX C89, without govt’s updated figures54 Amount sought in GX C89, with govt's updated figures55 Amount sought in Proposed FoF and CoL, doc. no. 2242
King 222,358,480.00 209,293,468.38 227,508,341.16 231,432,686.73
Koralewskl 1,222,856.00 209,293,468.38 227,508,341.16 231,432,686.73
Tucker 8,842,560.00 209,293,4.68.38 227,508,341.16 231,432,686.73
Moran 2,993,660.00 209,293,468.38 227,508,341.16 231,432,686.73
Robles 815,197.50 209,293,468.38 227,508,341.16 231,432,686.73
Bramley 5,102,789.00 209,293,468.38 227,508,341.16 231,432,686.73
Dlebner 18,567,345.60 209,293,468.38 227,508,341.16 231,432,686.73
*923The total amount of the forfeiture money judgment the government now seeks against each of the forfeiture defendants is the sum of:
• Legendz payouts to bettors based on Global Data Payment transactions. This is the top section on the government’s summary, GX C89 (App. 1), amounting to more than $25 million. This item is based on an extrapolation.
• The betting proceeds the government attributes, by way of a series of direct attributions and (in some instances) extrapolations, to the individual Legendz agents and runners. This is the middle section (“Agents”) on the first page of the government’s summary, GX C89, amounting to more than $142 million.
• The total of the funds the government asserts were illegally laundered through the various bank accounts associated with the Legendz enterprise. This is the section under the heading of “Money Laundering Accounts/Items” on the first two pages of government’s summary, GX C89, amounting to more than $43 million. This item is not based on extrapolations.
In the second category listed above (betting proceeds attributable to individual Le-gendz agents), the government attributes illicit proceeds to a total of fifteen agents and runners. Id. Of those fifteen, six are forfeiture defendants now before the court (Koralewski, Tucker, Moran, Robles, Bramley and Diebner). Id.
The government seeks a forfeiture money judgment against all of the forfeiture defendants, including King, for the same grand total: $231,432,686.73. Doc. no. 2242, at 48, ¶ 189. Thus, given the government’s final factual approach to determining the amount of the forfeiture money judgments, the viability of the government’s proposed extrapolations (the first two categories listed above) presents an issue common to all of the forfeiture defendants — because it directly affects all of them. For instance, the viability of the government’s proposed extrapolation as to Tucker, a Florida bookie, affects Moran, a California runner. Likewise, but as matters distinct from the merits of the government’s proposed extrapolations, there are common issues as to the viability of the government’s charac*924terization of funds allegedly laundered (the third category listed above), and as to the viability of the government’s contention that there should be no differentiation among the defendants as to the amount of the forfeiture money judgment to be entered. Those common issues are treated separately, below.
B. Viability of the government’s characterization of funds allegedly laundered
By way of factual findings, the court will start with the third category listed above — the government’s request for inclusion in its forfeiture money judgment of more than $43 million attributable to funds the government asserts were illegally laundered through the various bank accounts associated with the Legendz enterprise. As has been noted, the government seeks inclusion of this amount in a forfeiture money judgment against all of the forfeiture defendants.
The government’s case under this heading depends on two things: (i) the facts as to the source of the funds deposited, and (ii) the facts as to the nature of the activities generating the funds in the hands of that source. These two concepts — the source of the funds and the nature of the activities that generated the funds — are related but, for analytical purposes, are not identical.
1. Preliminary matters
One important thing to bear in mind about the figures shown under the heading of “Money Laundering Accounts/Items,” GX C89 (App. 1, Lines 1-16) is that these items are not extrapolations. As will be discussed, these items depend on some assumptions, but they are not extrapolations.56
Before going further, one of those assumptions deserves detailed treatment, because it has a very substantial effect on the factual merit of the government’s contention that the $43 million shown on GX C89 under the Money Laundering heading should be included in the forfeiture money judgment calculation.
When required to definitively state the factual basis for its forfeiture case, the United States Government told the court eleven times, in these words or in substance, that: “King lacked a legitimate source of income between 2003 and 2013,” doc. no. 1999-1, at 3.57 More specifically, the government represented that “his only source of income derived from his illegal gambling operation with Legendz Sports.” Id. (emphasis added). King’s asserted lack of a legitimate source of income during the conspiracy period is an essential underpinning of significant elements of the government’s forfeiture case against King and the other forfeiture defendants. The government’s contention necessitates a review *925of the evidence — from the government’s witnesses — at the four jury trials.
During the period relevant to this case, King had interests in at least four businesses other than the Legendz sports betting enterprise. They are: Grupo Legendz, Vaporama Internacional, S.A., Magna Tours and Platinum Vacations.
Grupo Legendz. Grupo Legendz was knowledgeably described by the government’s witnesses at the four jury trials. Through Grupo Legendz, King operated a brick and mortar bookmaking business in Panama City. Feb. Tr. 242. “Grupo Le-gendz was the company that owned the SuperBook, the actual physical sportsbook across from the Marriott.” Apr. Tr. 1277; Oct. Tr. 1124. (It had started out in the Marriott in Panama City and then moved to the location near the Marriott. Oct. Tr. 1085.) Consequently, the SuperBook’s bank account was in the name of Grupo Legendz. Feb. Tr. 900; May Tr. 431. King owned Grupo Legendz and thus was the owner of the SuperBook. May Tr. 417; Oct. Tr. 792, 1084. It was “a separate business ¿from the Internet sports-betting operation” (Oct. Tr. 1085) and was a legal, licensed sportsbook, holding a license from the Panamanian government (Oct. Tr. 1301), or as per another government witness: “a perfectly legal business, a walk-in casino in Panama City.” Oct. Tr. 434. It was a place “where you could bet the dogs, horses, different things.” May. Tr. 260. In addition to providing sportsbook services, the licensed casino had “a lounge and a restaurant and a nightclub.” Id. at 989 (and see the discussion of Vaporama, below). Patrons could “go in and sit down and have drinks and place bets on sports, horse wagering.” Apr. Tr. 601.
Although Grupo Legendz accounts apparently did receive funds generated by the Legendz gambling enterprise that is the subject of this case (Oct. Tr. 803), Lowrance acknowledged that if King’s characterization of Grupo Legendz and Va-porama was correct (that being the characterization that was based on testimony from government witnesses at the four jury trials — testimony that was neither contradicted nor disputed in any other way by any party), then it “might be possible” that the $3,746,816.83 in allegedly laundered funds shown in Line 1 on GX C89 “represents both lawful and unlawful proceeds being counted into that account.” Tr. 187. In light of the uncontradicted testimony of the government witnesses at the jury trials, this was a remarkably tepid concession by Lowrance. The evidence before the court will permit no finding other than that Grupo Legendz was a lawful business situated in Panama City, generating lawful revenues for Bartice King.
Vaporama Internacional S.A. The uncon-tradicted testimony (from a government witness) at the April trial established to the satisfaction of the court that Vaporama was a company, affiliated with King, that held the liquor license for the sportsbook that was operated by Grupo Legendz in Panama City. Apr. Tr. 1277.58 There was no evidence to the contrary, nor was there any suggestion, from any party, to the contrary. As is discussed below, Vaporama deposited $1,934,515.94 into an account at Banco General, S.A. in Panama between 2006 and 2013. GX C64A (summary).59 And as has been discussed, the government *926seeks to include this $1.93 million in the forfeiture money judgment. GX C89, p. 2. But, bearing in mind the government’s repeated representations to the court that King had no legitimate source of income during the conspiracy period, the Vapora-ma account is significant for other reasons. As Lowrance acknowledged, application of simple arithmetic to the government’s summaries leads to the conclusion (“I presume so.” Tr. 264) that Vaporama made a profit of about $780,000 during the period covered by the government’s summaries (GX C64A and C64B). There is no basis in the record for any inference that the income generated by Vaporama, as the bar and restaurant co-located with the licensed sportsbook operated by Grupo Legendz in Panama City, was anything other than legitimate. This, as will be seen, affects the viability of the government’s contentions as to the forfeitability of the hard assets that the government seeks to forfeit on the premise that King had no legitimate source of income during the conspiracy period.
Magna Tours and Platinum Vacations. At the four jury trials, government witnesses also testified about King’s interests in two other businesses: Magna Tours and Platinum Vacations. Magna and Platinum were separate entities. Feb. Tr. 823.
Magna Tours was a travel agency — an IATA-certified travel agency (Feb. Tr. 842; Apr. Tr. 1078), otherwise described as a “walk-in retail travel agency.” Feb. Tr. 719. “You can go in and book a cruise, book airline tickets.” Id. As a “legitimate travel agency” (Oct. Tr. 1035-36), Magna had its own, separately-keyed, leased space in the building that also housed the Legendz call center. Feb. Tr. 844; May. Tr. 344.
King was the founder, and, for a time, the sole owner of Magna Tours. Feb. Tr. 720; Apr. Tr. 657. Later, he was a partner in this business with Reggie Berthiaume and Michael Lawhorn (Feb. Tr. 328-29), both of whom (as government witnesses) testified in detail about that business. Ber-thiaume was in this business with King from 2004 to 2010 — very much within the conspiracy period. Feb. Tr. 720; Apr. Tr. 596. In addition to serving the needs of the Legendz organization (Oct. Tr. 1086), Mag-na Tours had corporate accounts in Panama. Feb. Tr. 720. With Berthiaume’s help, they were “picking up a lot of corporate accounts” (id. 721), so the business was “pretty successful” (Apr. Tr. 596), concisely described by Berthiaume as “a profitable business.” Apr. Tr. 657. It was “growing at about 20 percent annually in sales.” Apr. Tr. 596.60 Magna Tours was also “the beneficiary of any travel that was booked through the sale of’ vacations by Platinum Vacations. Feb. Tr. 722. Initially, King and Berthiaume split the profits 50/50. Apr. Tr. 596. Although King owned Magna Tours at least until 2010 (May Tr. 430), the company was ultimately purchased by Antonio Jenkins-Lara. Feb. Tr. 842. At that time, it was still “an operating business and a going concern.” Apr. Tr. 659.61
Beginning in approximately 2006, King, Berthiaume and Robles were partners, one-third each, in Platinum Vacations. *927Feb. Tr. 722; Apr. Tr. 597.62 King supplied the start-up capital and equipment. Feb. Tr. 723; Apr. Tr. 597. Platinum Vacations sold vacation packages — ultimately needing “maybe 20 to 30 operators” at its call center in an office “around the corner from Legendz.” Feb. Tr. 722, 830-31. To patronize Platinum, “you can call and plan your vacation through them and they would give you the best package that they could for your vacation.” Oct. Tr. 1088. Platinum had nothing to do with Legendz or with gambling. Feb. Tr. 831, 1168. Its offices were separate from the Legendz facility. Feb. Tr. 1167. Although the business was ultimately unsuccessful, Platinum’s revenues grew, at one point, to $10,000 a day. Feb. Tr. 722.
This, in a nutshell, is where the record stood after the government had called various combinations of four witnesses at four jury trials who testified without contradiction — or even an intimation of doubt from the government or anyone else— about the particulars of King’s outside businesses during the period in which, as the government put it after those four trials had been completed: “his only source of income derived from his illegal gambling operation with Legendz Sports.” Doc. no. 1999-1, at 3. Finally, on this subject (for now), it should be noted that the government’s forfeiture witnesses offered no estimation of the ratio of King’s legitimate income vs. his illegal gambling income. The government made no attempt to determine this. Tr. 649. Thus, Jansen had no conclusion at all as to “how much money attributable to Mr. King is tainted up versus clean or legitimate.” Tr. 629. This is consequential for forfeiture purposes.63
2. Sources of deposits and nature of activities generating the funds.
As stated above, the government’s case for inclusion of $43 million in the forfeiture money judgment based on deposits under the heading of Money Laundering on GX C89 requires resolution of two separate (but in some instances related) factual issues as to each of the sixteen line items in the money laundering section of GX C89: (1) What was the source of the funds deposited? (2) What was the nature of the activities that generated those funds in the hands of that source? The court will now address these issues as to each of the sixteen line items shown under the Money Laundering heading on GX C89.64
*928GX C89, Line 1: Source of funds and nature of activities. The forfeiture money judgment the government seeks would include $3,746,816.83, representing funds deposited into a Legendz account at Banco Nacional de Panama (BNP). See, GX C89, Line 1; doc. no. 2242, at 21, ¶¶ 83, et seq.; Tr. 171. The relevant entry on GX C89 takes us to GX46, 46A, C58A and B and 1264. Three of those exhibits, GX 46 and 46A, and 1264 are irrelevant for present purposes — they consist of documents establishing the fact of the formation of “Le-gendz Gamming [sic] Corp.,” evidencing the opening of an account at BNP, and evidencing transactions in that account in 2007-2012. Government Exhibit C58A, á summary of deposits into this account, supplies the dollar amount — $3,746,-816.83 — that the government would include in the forfeiture money judgment. But of the 233 deposit transactions summarized on GX C58A, the source of 163 is unknown, as shown on the face of the exhibit and as acknowledged by FBI witness Lowrance:
Q. [by Mr. Mays] And with respect to those payments, you can’t — that are reflected for those unknown payors, you can’t testify that those represent lawful or unlawful proceeds being deposited into that account, can you?
A. No, I can’t.
Q. It’s impossible for you to know since the payor is unknown?
A. Correct.
Tr. 184-85.
In addition, as to 46 of those 233 deposits, the source is shown to be Grupo Le-gendz. GX C58A, p. 1-2. Grupo Legendz, as discussed in Part IX(B)(1), above, was a source of legitimate income for King. In fact, there is no basis in the record for any suggestion that Grupo Legendz was a source of anything other than legitimate income for King.65 After being confronted *929with the facts as to Grupo Legendz, Low-rance maintained that the Banco Nacional account was nevertheless “a tainted account” (Tr. 182), apparently justifying the government’s attempt to include in its forfeiture money judgment all of the $3.75 million in Banco Nacional deposits shown in Line 1 of GX C89. But of those 233 deposits, the source of 163 is unknown, and 46 were from Grupo Legendz, for a total of 209 out of 233 deposits that are devoid of significance for forfeiture purposes — or at least significance favorable to the government. The twenty-four deposits from identified sources other than Grupo Legendz account for about fifteen percent of the $3.75 million sought by the government in Line 1. Those 24 deposits provide little, if any, help to the government. One, in the amount of $250, is from Alfred Ly, a cooperating witness. Mr. Ly began cooperating with the government, acting for the government as a subagent under Bruce Mid-dlebrook, in 2006. Oct. Tr. 1570, 1604-07. The only check attributed to Ly in GX C58A is dated in March of 2009, when Ly was a cooperator. Three other checks, totaling $8,700, are from FBI agents, acting as bettors. GX C58A, p. 2 (Hash and Lyon). Another one of the 24 checks is from Yaporama, a legitimate business operated by King as is discussed above.
Unsurprisingly, when confronted with the facts establishing the deficiencies in GX C58A, Lowrance was unable to venture a guess as to the percentage of the account activity shown in GX C58A that actually consists of “deposits from bettors or reflects earnings into this account from unlawful sports betting.” Tr. 186-87. Facts notwithstanding, the government has blithely asserted that: “Records from [this Banco Nacional account] show that from October 2005 through July 2012, Legendz Gamming Corp., an entity used by Le-gendz Sports and Mr. King to collect funding from its post-up bettors, collected approximately $3,746,816.83 in deposits for illegal gambling.” Doc. no. 1999-1, at 19. In its post-trial proposed findings, the government seeks to include the entire $3.75 million in its forfeiture money judgment, without so much as a nod in the direction of the uncontroverted evidence that thoroughly undermines its factual contentions as to Line 1. Doc. no. 2242, at 21, ¶ 83.
Save for the $250 from Alfred Ly and the $8,700 from the FBI agents posing as bettors, there is a failure of proof as to both the source of the $3,746,816.83 comprising Line 1 and as to the nature of the activities generating those funds. The court finds that Line 1 contributes $8,950 to the forfeiture money judgment calculation.
GX C89, Line 2: Source of funds and nature of activities. The allegedly laundered funds in this line on GX C89 are shown on the government’s recap to total $7,167,005.65. But, at the forfeiture trial, the government stated that this number should be $7,152,574.65. Tr. 172. This matches up with the government’s supporting summary of the deposits, GX C49A.
Before going further with respect to Line 2, it should be noted that Line 2 is in a different category than the'other fifteen lines under the heading of Money Laundering on GX C89. The reason is that the deposits into this account — $7.15 million— are the basis for the government’s proposed extrapolation from twenty-four months to 108 months, to attribute $25 *930million to Legendz payouts to bettors, as shown in the top box on GX C89. Tr. 201. The government does not propose any extrapolations from any of the other fifteen accounts the government asserts were used for money laundering purposes. Thus, aside from any other problems there may be with this extrapolation (see Part IX(C)(2), below), any deficiencies in the government’s reckoning as to the deposits into this account are magnified by a factor of more than four for purposes of the government’s forfeiture money judgment case.
The summary the government uses to back up Line 2, GX C49A, is based on deposits into a bank account of Global Data Payments Services S.A. (GDPS). In turn, the source document for this summary is GX 1258, an account ledger from HSBC Bank in Panama, which has entries for the deposits summarized on GX C49A and the payments summarized on GX C49B. The deposits shown by GX C49A total $7,152,574.65, consisting of 136 deposit transactions, all of which were in 2007 or 2008. The source of the deposited funds is shown on the exhibit to be unknown for 104 of the 136 deposits. Those 104 deposits, for which the source of the funds is unknown, account for approximately $5.43 million of the $7,152,574.65 in total deposits, yet .the government seeks to include this entire amount in the forfeiture money judgment.66
The thirty-two deposits for which the source is known are from fourteen persons or entities. Of those fourteen, Lowrance recognized only four (International Gold-store, International Data Processing, Elite Pay Systems and Agathe Services, Ltd.) as Legendz affiliates that “receive bettor payments from the United States.” Tr. 79.
International Goldstore and Elite Pay Systems need not detain the court long. The evidence from the four jury trials established quite clearly that those entities were used for receipt and payment of illegal betting proceeds. Those two entities account for $61,604.40 on GX C49A.
Agathe Services is another matter. Agathe was not mentioned at any of the four jury trials. Agathe is not mentioned in the government’s Forfeiture Brief (doc. no. 1999) or in its proposed findings of fact. Doc. no. 2242. At the forfeiture trial, Low-rance could not cite any document that would implicate Agathe as a recipient of (and, in turn, as a source of) betting proceeds. Tr. 189. All he could refer to was “[djiscussions with other investigators in this case” (id.), presumably the case agents. Thus, no percipient witness has told the court, in the jury trials or in the forfeiture trial, that Agathe was a recipient of anything, let alone illegal betting proceeds. And, not to put too fine a point on it,67 the court declines to include deposits *931originating with Agathe in the forfeiture money judgment calculation on the basis of what the case agents (who did not testify in any of the jury trials or in the forfeiture trial) may have told Lowrance.
Likewise, Lowrance had no personal knowledge as to International Data Processing, which was the source of one deposit into the GDPS account. Tr. 193-94. It was his understanding that International Data was “affiliated with the Legendz enterprise” (id. at 193), but he did not know in what way it was affiliated. Id. at 194. He had no “idea what that payment of $39,975 represents.” Id. The testimony at the jury trials provides no real help as to International Data. That company was mentioned once, and only once, at each of the four jury trials. Feb. Tr. 912; Apr. Tr. 1186; May Tr. 433 and Oct. Tr. 972. In the first three jury trials, the singular mention was to the effect, in the testimony of Karlo Stewart (the Chief Financial Officer of the Legendz organization, Feb. Tr. 891) that he opened an account at a Costa Rican bank for International Data Processing Services. There was no mention of the uses to which that account was put, other than an unelaborated statement by Stewart in the May trial that International Data Processing was “a company that was set up to receive funds.” May Tr. 433. There was no mention of whose funds were to be received, which is no small point in light of the fact that, as has been seen, King had several sources of legitimate cash flow during the conspiracy period. Stewart’s lack of knowledge as to the use of the International Data Processing account was demonstrated at the October trial, where he testified that, for lack of knowledge, he had been unable to respond to an inquiry from a bank as to the provenance of some funds that were being sent to an International Data Processing account from Mexico. Oct. Tr. 972.68
Another curious thing about the government’s approach in Line 2 is that the transactions summarized on exhibit C49A — deposits into an account associated with Legendz — are used as the basis for an extrapolation (from $7.15 million to $25 million) arriving at a dollar amount for payouts, under the heading of “Legendz Payouts to Bettors.” GX C89, p. 1; Tr. 205. As shown by GX C49B, the amount of the payouts for a virtually identical period was known (and, by the way, added up to a significantly lower dollar amount), but the government chose to estimate payouts from GDPS by adding up the deposits into its bank account. The illogie of estimating payouts by adding up deposits when the actual amount of the payouts was known was acknowledged by Lowrance on cross examination:
Q. [by Mr. Mays] Well, let me ask it this way: If you wanted to find out how much money somebody spent in a month, would you look at how much they put into their account or how much they. took out of their account?
A. Took out of their account.
Q. All right. And in this particular instance, you looked at how much money *932went into the account and then extrapolated from there, correct?
A. Yes.
Tr. 205.
As will be seen in Part IX(C)(2), below, the court rejects the $25 million extrapolation (in the top box on GX C89) that the government has based on Line 2. So, even though the government does not, as a fallback position, ask the court to treat Line 2 as a money laundering item (as it does with the other fifteen lines under the Money Laundering heading), the court will do so. Measured by the preponderance of the evidence, Line 2 is good for $61,604.40 of dirty money, fair game for the forfeiture money judgment. Thus, the beginning point for the government’s proposed 108-month extrapolation in the top box on GX C89 is $61,604.40, not $7,152,574.65. Other aspects of that proposed extrapolation are discussed in Part IX(C)(2), below.
GX C89, Line 3: Source of funds and nature of activities. In Line 3, the government seeks to include $4,179,355.00 in the forfeiture money judgment. The backup summary for Line 3 is GX C42D, which shows 32 funds transfer transactions. That summary is, in turn, backed up by bank records, mostly evidencing wire transfers, from Bank of America. GX 1405A.
The evidence does reflect what the immediate source of the funds was for each of these 32 transactions. So, strictly speaking, the government’s proof adequately establishes the source of the funds comprising Line 3. In this respect, the quality of the government’s evidence in support of Line 3 differs noticeably from the quality of that evidence in support of other money laundering items on GX C89.
The transactions comprising Line 3 occurred between April 8, 2008 and January 12, 2009. They fall into three categories:
• One transfer from Global Data Payment Services to Platinum Advantage, Ltd. ($117,530.00)
• Six transfers from Seven Capital Business, Inc. to Platinum Advantage, Ltd., totaling $1,531,125.00.
• Twenty-five transfers from Platinum Advantage, Ltd. to Virtual Automated Technologies, totaling $2,530,700.00.
Thus, per GX C42D and the underlying source documents, Platinum Advantage took in approximately $1.65 million and transferred $2.53 million to Virtual Automated Technologies. Notably, the $4.18 million comprising Line 3 is the sum of the transfers into and out of Platinum Advantage. Plainly double counting, a matter that is not mentioned, much less explained, by the government.
In support of this item, citing its summary (GX C42D), the government says that: “Platinum Advantage records show that between August 2008 and January 2009 Virtual Automated Technologies, a payment processor Legendz used to remit funds to bettors, received approximately $4,179,355.00 in illegal proceeds from Platinum Advantage and Seven Capital — entities created to move bettor funds.” Doc. no. 2242, at 15, ¶ 58. The government’s statement that Virtual Automated Technologies received this amount is demonstrably wrong — by more than $1.6 million. As shown above, in seven instances, involving more than $1.6 million, the recipient was not Virtual Automated Technologies. In fact, in those seven instances, the payee was Platinum Advantage, not Virtual Automated Technologies, a fact that is also borne out by a careful review of the source documents, GX 1405A.
The government cites only two passages from trial testimony to support this item. Doc. no. 2242, at 15. The first passage is from the forfeiture trial, specifically, Tr. *93377-78. In that testimony, Lowrance did nothing other than recite the (erroneous) assertion that GX C42D shows $4,179,355 in funds received by Virtual Automated Technologies. The second passage cited by the government is from the February, 2015 trial — Feb. Tr. 742-50. But that passage makes no reference to any payments at all involving Virtual Automated Technologies, nor does it refer to GX C42D or to the source documents, GX 1405A. The only flow of funds suggested by that passage from the February 2015 trial is that money flowed from Seven Capital to Platinum Advantage to a company called NxSystems — an entity to which the government makes no reference at all in its proposed findings. Nor does the government explain why the court would find comfort in the cited testimony as clearing up the government’s demonstrably erroneous assertions as to these payments.
Neither Lowrance nor Jansen could shed light on the nature of the activities generating the funds involved in the transactions summarized on GX C42D. Jansen did not address the matter at all, and Lowrance had only this to say:
Q. Can you testify with respect to these dollar amounts that are here on this chart as to the source of those funds?
A. No.
Q. And do you have any documentation to support those funds representing proceeds from illegal betting activity, bank records, financial records?
A. No.
Tr. 208.
So, the government’s effort, in Line 3, to include $4.18 million in the forfeiture money judgment is infected by: (i) bad math, (ii) double counting, (iii) a misrepresentation by the government as to the flow of funds (Doc. no. 2242 at 15, ¶ 58 — perhaps excusable because it simply parrots Low-rance’s erroneous testimony), and (iv) inability of the government’s lead forfeiture witness to say whether any of this money represented the proceeds of illegal betting activity.69 Moreover, the relevant portions of the source documents consist only of paperwork for wire transfers. There are no bank statements, as such, which would give an overview of account activity, running balances, and sources of funds. The absence of evidence giving the big picture as to the sources of funds going into the Bank of America account is no small matter — especially so in light of the fact that the source documents, such as they are, show just under a million more dollars going out of the account than into the account. But the government would have the court presume, sans evidence, that every dime of the money going out of the account arrived in the account as dirty money.
Now, to be fair, the court can say quite readily, but at a high level of generality, that these transactions are suspicious. There was considerable testimony in the four jury trials about Platinum Advantage and Virtual Automated Technologies, Ltd. being involved in receiving and paying out gambling proceeds. But, as has been not*934ed, there is not a shred of testimony from the jury trials or the forfeiture trial that would support the government’s asserted characterization of these particular payments. Nor, it should be noted, does the government’s Forfeiture Brief (doc. no. 1999) make any reference at all to the transactions comprising Line 3, or to the bank account involved in Line 3. The only reference to Platinum Advantage in the government’s Forfeiture Brief is on p. 14 of doc. no. 1999-1, where the government refers to Platinum Advantage in an entirely different context, involving a different time period,'a different transaction total, and a radically different approach — specifically an extrapolation to $36.2 million.
Suspicion is not enough. The government’s contentions in support of Line 3 do not withstand scrutiny, either as a matter of simple math or in terms of evidentiary support. Line 3 is rejected.
GX C89, Line 4: Source of funds and nature of activities. By way of Line 4, the government would include $4,630,377.83 in the forfeiture money judgment, as representing bettor losses paid to Legendz. Line 4 is supported by the summary that was admitted as GX C52A, which, in turn is supported by numerous checks, and related documents, as source documents. GX 879 A-E. These exhibits establish that the source of the deposits is as asserted by the government.
The government has shown to the satisfaction of the- court that the $4,630,377.83 deposited into the Investment Consulting Services account at Creditcorp Bank, as shown on GX C52A, was proceeds of illegal betting activity within the scope of the Indictment. The government’s summary, GX C52A, associates virtually every transaction with a specific agent or bookie, for instance, “Bramley” or “Tucker.” “The government [is] not required to prove that individual [deposits] were illegitimate when the evidence created a reasonable, indeed an overwhelmingly strong inference that all [deposits] ... were illegitimate.” United States v. Elder, 682 F.3d 1065, 1073 (8th Cir. 2012). The quality of the evidence as to the nature of the transactions comprising Line 4 is much more satisfactory than, say, the quality of the evidence in support of Line 5, discussed immediately below.
The court finds that Line 4 contributes $4,630,377.83 to the forfeiture money judgment calculation.
GX C89, Line 5: Source of funds and nature of activities. Line 5, by which the government seeks to include $3,613,897.27 in the forfeiture money judgment, aptly illustrates why the court, in its analysis of the government’s money laundering-related forfeiture case, undertakes separate analyses of the evidence as to (i) identification of the sources of the funds deposited and (ii) the nature of the activity that generated the funds coming from the identified sources. But, before getting to that, another aspect of Line 5 must be addressed.
In its post-trial proposed findings of fact, the government, citing the relevant summary (GX C53A), tells the court that “[f|rom June 2005 through June 2012, International Goldstore received a total of approximately $3,613,897.27 in deposits on behalf of Mr. King and Legendz Sports.” Doc. no. 2242, at 23. That $3.6 million number is also what the government’s forfeiture recap, GX C89, shows in Line 5 for deposits to an International Goldstore account. But, as Lowrance acknowledged at the forfeiture trial (Tr. 224), this number is clearly wrong. Lowrance’s explanation: “Just didn’t double-check the numbers.” Tr. 224. It is off by $1,887,101.27, a fact the government inexplicably ignores in its post-trial proposed findings. See, doc. no. 2242, at 23, ¶ 88 (using the $3,613,897.27 *935figure). The correct number, $1,726,796.00, is less than half of the number proposed by the government.
This item also presents problems aside from bad math. Of the ninety-eight entries on the relevant summary, GX C53A, the government noted that the payor was “known” to be a bettor as to only fifty-nine. As to the rest, that column is either blank or says: “suspected.” The government has neither contended nor shown that those unproven deposits (ie., deposits that did not consist of funds that were discernibly generated by illegal betting activity) then became “part of the corpus of the money laundering conspiracy,” United States v. Huber, 404 F.3d 1047, 1059 (8th Cir. 2005) or “part of a transaction that also involved proceeds of a specified unlawful activity,” id. at 1061 n. 11, so as to become forfeitable (and thus fair game for inclusion in a forfeiture money judgment) on that basis. Nor has the government invoked a facilitation theory as to this account (or, for that matter, as to any other financial account involved in this case). The government has offered no alternative calculation to account for the defects in GX C53A, and the court declines, for a variety of reasons, to embark upon the task of coming up with a more supportable dollar amount.
Line 5 contributes nothing to the forfeiture money judgment.
GX C89, Line 6: Source of funds and nature of activities. In Line 6, the government seeks to add $5,425,621.68 to the forfeiture money judgment. This represents funds deposited into a Data Support Services (DSS) account at Banco General Panama between May, 2004 and May, 2013. GX C47A; Tr. 173-74. The sum total of the government’s explanation of the basis for including this $5.43 million in the forfeiture money judgment is:
• From May, 2004 through May, 2013,. Data Support Services received these funds “on behalf of the Le-gendz Sports enterprise.” Doc. no. 2242, at 23, ¶ 90.
• “DSS was a Legendz Sports entity; specifically, it was the Legendz’s physical call center that received calls from bettors in the United States.” Id.
The summary supporting this $5.43 million item is GX C47A. That summary exhibit shows 309 deposits into the DSS account at Banco General. Of those 309 deposits, the source of fifty-eight is shown to be unknown. The immediate source of 238 is shown to be Banvivienda, but the origin of the funds is unknown. And the source of seven is shown to be Alegui Services S.A. DSS is an entity that functioned as an integral component of the Legendz enterprise. Banvivienda is a bank in Panama. Tr. 14. The name of the actual counterparty remitting from an account at Banvivienda is not shown by GX C47A or by the source documents, GX 1255. Thus, all that is known as to these deposits is that they came from some unknown person or entity having an account at that bank. (And, to be sure, the Legendz organization had accounts at that bank, although the government ventures no suggestion as to why funds would be moved from a Legendz-affiliated account at Banvivienda to a Legendz-affil-iated account at Banco General before being remitted to an ultimate recipient, such as King.)
The testimony of Lowrance at the forfeiture trial established that the government was unable to determine the source of the deposits shown to be “unknown” on GX C47A, and the government has no way to identify that source. Tr. 229. Lowrance could not testify that those deposits had any relation to sports betting activity. Id. *936The government was unable to determine the identity of the actual source of the funds that came from Banvivienda. Tr. 230, 232, 288 (all the funds from Banvivienda are “from unknown depositors”). Aside from pointing out the uncontested fact that the Legendz organization had accounts at that bank, Lowrance could not associate those deposits with King or Legendz, nor could he say that those deposits had anything to do with illegal sports betting. Tr. 230, 232. (For the reasons that 'are discussed above with respect to Grupo Le-gendz, inquiry as to whether any particular deposit was generated by illegal sports betting was, throughout the forfeiture trial, a fair question.)
To be sure, given the fact that DSS was, to put it mildly, an integral component of the Legendz sports betting enterprise, there is reason to suspect that the funds coming into the DSS account may have been illegal gambling proceeds. But the evidence in this case precludes any finding, with a wave of the hand, that all funds coming into a DSS account were dirty money. The government supposes this to be the case (Doc. no. 2242, at 23, ¶¶ 89-90), but this supposition rests on (among other things) the government’s altogether baseless assertion that King had no legal source of income during the conspiracy period. Nowhere in the government’s presentation as to Line 6 did the government provide any evidence that any of the $5.43 million in deposits encompassed by that item (almost all of which were from unidentified sources) originated with illegal sports betting. Although this may seem to be a remarkable conclusion as to funds coming into an entity like DSS, it is a fact that DSS was a multi-purpose umbrella entity. In that respect, DSS did not serve the narrow (and, for all the record shows, exclusive) purpose of receiving illegal betting proceeds that was served by, say, Investment Consulting Services (discussed above with respect to Line 4). Thus, the government’s proof with respect to DSS simply will not support the permissive inference that the court easily indulged, with the guidance of United States v. Elder, 682 F.3d at 1073, in its analysis of Line 4.
Line 6 is rejected.
GXC89, Line 7: Source of funds and nature of activities. In Line 7, the government seeks to include in the forfeiture money judgment $3,911,702.84, representing funds deposited into a Grupo Legendz account at Banco Nacional de Costa Rica. The backup summary for Line 7 is GX C50A and the source documents are in GX 1259.
In its Forfeiture Brief, the government told the court: “Bank records from Grupo Legendz bank account number XXXXX0921 held with Banco Nacional show that from April 2008 through April 2013, approximately $3,911,702.84 in illegal gambling proceeds was deposited into the account.” Doc. no. 1999-1 at 20 (emphasis added). This contention by the government died aborning, because the source of 300 of the 324 deposits into the Grupo Legendz account, comprising Line 7, is unknown. This is shown by the backup summary (GX C50A) as well as the source documents (GX 1259) and was confirmed by Lowrance:
THE COURT: Wait just a minute. On most of the transactions shown by C50A show the payor to be unknown. Is that because they were cash deposits or what?
THE WITNESS [Lowrance]: Your Honor, we didn’t get complete records, so I think the statements show some of these entity names. I think at some point in the process the bank down there changed its format and was able to include the name of the deposits on the statements, but there is a period of time *937when we didn’t get — we don’t know who deposits — what the source of the deposit was. We got no deposit tickets. We didn’t get any items.
[[Image here]]
Q. [Cross examination by Mr. Mays] So out of the nine pages — out of the ten pages, we have over nine pages of unknown depositors, correct?
A. Yes.
Q. Can you tell the Court whether or not those deposits represent legal or illegal sports betting proceeds for the unknown entries?
A. We don’t have the underlying documents, so I couldn’t speak to them.
Q. All right. You have included them in your total on the forfeiture overview regardless of that fact, correct?
A. Yes.
Tr. 93-94, 241.
That alone is sufficient to render those transactions meaningless for forfeiture purposes, even aside from the evidence in the four jury trials establishing without contradiction that, in terms of its own actual operations, Grupo Legendz generated nothing but lawful revenues as a brick and mortar sportsbook in Panama City — more or less like the Ladbrokes shops that dot the United Kingdom.
In its proposed findings, the government backs off the patently unsupportable representation quoted above — but only slightly (and not nearly enough).70 In its proposed findings, the government would have the court find that “[t]hose deposits are, in part, from entities that received or processed illegal gambling proceeds on behalf of Mr. King and Legendz Sports.” Doc. no. 2242, at 24 (emphasis added). But the government still includes every penny of this $3.9 million in its request for a forfeiture money judgment. Doc. no. 2242, at 47, ¶ 187.71
The twenty-four deposits included in Line 7 (out of the total of 324 deposits) for which the source is known include eleven sources. They are (in the order in which they appear in the back up summary, GX C50A):
Absolute Privilege for the World S.A.
Data Support Services
Entiant International LTD
Inversiones Altamonte S.A.
Ivan Xavier ALV
Legendz Gamming [sic]
NYTO Trade, Inc.
Pentraci Corporation LTD
Premiere Data Management Corp.
UDS International Software S.A.
Yaporama Internacional
These eleven known sources account for $567,339.65 of the $3.9 million the government includes in Line 7. But the government goes no farther than to assert that the money from four of those eleven sources was illegal gambling proceeds. *938Those are: Absolute Privilege (doc. no. 2242, at 24; Tr. 91); Pentraci Corporation (id.); Entiant International (Tr. 92); and UDS International (Tr. 92). The court will add Legendz Gamming [sic] and Data Support Services to the list, because they hardly needed to be mentioned by the government. Those two entities clearly did receive and pay out illegal gambling proceeds. That makes a total of six entities implicated as sources of dirty money deposited into Banco Nacional and included in Line 7.
As to three of those six implicated entities, extended discussion is not necessary. The evidence from percipient witnesses at the jury trials clearly established that UDS International, as well as Legendz Gamming and Data Support Services (as mentioned above), processed illegal gambling proceeds. Those three account for $234,482.50 in Line 7. The other three, Absolute Privilege, Pentraci and Entiant require more discussion.
As for Absolute Privilege, Lowrance did testify that “Absolute Privilege is a payment processor. They show up on the Le-gendz spreadsheets.” Tr. 91. The court does not doubt Lowrance’s bona fides in saying that, but the court is unpersuaded. For one thing, Absolute Privilege was not mentioned at all at any of the four jury trials. In the jury trials, the government called several witnesses with first-hand knowledge of the internal operations of the Legendz enterprise. Most of those with first-hand knowledge were, for reasons of self-interest, keenly interested in pleasing the government. But in jury trials that generated more than 8,500 pages of transcript, none of them mentioned Absolute Privilege. Secondly, the only spreadsheets cited by the government (GX 65, 66 and 67) do not, in terms or by implication, show Absolute Privilege to be a processor of illegal gambling proceeds. Those three spreadsheets show a variety of entities (such as Western Union) for which there are entries in the “processor” column. They also show activity involving businesses like the SuperBook and Vaporama, that were operating legally in Panama. Without some explication from a witness with actual knowledge, the court declines to find, even by a preponderance of the evidence, that Absolute Privilege was a processor of illegal gambling proceeds.
Pentraci Corporation was mentioned in the May and October jury trials, by Karlo Stewart in each instance. For reasons that have been discussed at greater length in other orders72 (and will be discussed at greater length in this order), the court is unwilling to give Karlo Stewart a blank check in terms of believability. His multiple state and federal felony convictions for serious crimes of dishonesty see to that. But Stewart’s testimony to the effect that Pentraci was used by King as a processor of illegal gambling proceeds was specific enough and detailed enough to pass muster on a preponderance standard. The court will credit that testimony.
Entiant is easier, mainly because the court doesn’t have to rely solely on the testimony of Karlo Stewart. Entiant was discussed in all four of the jury trials, in a fair amount of detail. Michael Lawhorn, a defendant and cooperating witness, credibly identified Entiant as an entity used to receive illegal gambling proceeds, a proposition that is also borne out by Legendz records. The Line 7 funds that came from Entiant will be included in the forfeiture money judgment calculation.
To summarize, as to Line 7, the deposits attributable to Legendz Gamming, Data Support Services, UDS International, Pen-*939traci and Entiant will be included, for a total of $364,754.92 out of the $3.9 million sought by the government.
GX C89, Line 8: Source of funds and nature of activities. In Line 8, the government would have the court include $1,380,865.13 in the forfeiture money judgment, representing funds deposited into an account in the name of Fundación Los Olivos (FLO) at Credicorp Bank in Panama. The backup summary for this item is GX C48, which, in turn, summarizes the source documents, GX 1256.
To take the facts favorable to the government first, some bettor checks clearly were made payable to FLO. May Tr. 767-68 ($45,000 check from Cabrera, a/k/a Los Cucos, in February, 2012 — deposited into Banco Nacional, not into Credicorp Bank. See GX 887I.).73
Cutting in the other direction, the government points the court to no bettor funds going to FLO later than October, 2012. GX 887P (check into Banvivenda, not Credicorp).74 The Credicorp account that is the subject of Line 8 was not even opened until late June, 2013, more than two months after the takedown in this case and nearly two months after King voluntarily returned to the United States to answer the charges in this case. GX 1256; Tr. 109, 253. FLO’s Credicorp Bank account was funded by two cashier’s checks from Ban-vivienda on June 25, 2013. GX C48 (summary); GX 1256 (source documents). The origin of those funds from Banvivienda is not indicated in the source documents or, understandably, in the summary. Low-rance did not know where those funds came from. Tr. 255. He acknowledged that the funds going into the account “are untraceable.” Tr. 256. They cannot be connected to illegal sports betting by way of any documentation or witness statements. Id. Aside from the opening deposits on June 25, 2013, the only funds going into that account were the bank’s payments of interest on the funds in the account. Tr. 255.
Suspicion about the origin of the funds in the FLO account at Credicorp Bank is understandable. But even so, if those funds originated with the Legendz organization, they could have had either lawful or unlawful provenance. In any event, suspicion does not amount to a preponderance of the evidence. Line 8 is rejected in toto.
GX C89, Line 9: Source of funds and nature of activities. The government seeks to include $1,934,515.94 in the forfeiture money judgment on the basis of that dollar amount of deposits into a Banco General account in the name of Vaporama Interna-cional, S.A. GX C89, Line 9; Tr. 96-97. But, of the 2,339 deposits into that account (which include all but about $60,000 of the $1,934,515.94 the government would have the court include in the forfeiture money judgment), the source — ie. gambling of some kind? Let alone illegal gambling? — is unknown as to all but two. As Lowranee acknowledged: “Q: And all of those [2,337 of the 2,339] the depositor or the person putting the money into the account is unknown, correct? A: That’s correct.” Tr. 258. Lowranee was, consequently (and unsurprisingly) unwilling to tell the court that those deposits were “related to illegal sports betting.” Tr. 259.
*940The two deposits into this Banco General account from known sources were: a $45,170.65 deposit from Grupo Legendz and a $15,000 deposit from Data Support Services. GX C64A; Tr. 257-58. As has been seen, Grupo Legendz operated a legal sportsbook in Panama City. Even allowing for the fact that King may also have used Grupo Legendz as a convenient conduit for illegal gambling proceeds,75 the fact remains that the record provides no basis for any overarching assumption that funds in the hands of Grupo Legendz are illegal gambling proceeds. That $45 thousand item is rejected.
As for the $15 thousand that went into Vaporama’s account at Banco General from Data Support Services, it can easily be said, as an initial matter, that the main reason Data Support Services existed was to serve as a host entity and call center for the illegal sports betting enterprise operated by Legendz. But lest this $15,000 be thought of as low hanging fruit for the government, it should be borne in mind that the government’s witnesses confirmed that Data Support Services also served some of King’s other, legal, businesses, including AJAX, a dialer company, Platinum Vacations and Magna Tours. Apr. Tr. 1273-74; May Tr. 1123-24. Karlo Stewart, a key government witness, confirmed that Data Support Services was, for bank regulation reasons, insulated from gambling revenue. Apr. Tr. 1251-52. Moreover, as Lowrance acknowledged, this $15,000 came from the account that is the subject of the government’s Line 6, which is discussed above. Tr. 257-58. As is discussed with respect to Line 6, nowhere in the government’s presentation as to Line 6 did the government provide any evidence that any of the $5.43 million in deposits encompassed by that item originated with illegal sports betting.
And lest it be thought that funds coming out of that account made the $1.9 million fair game for inclusion in a forfeiture money judgment (even though the government’s calculation is based entirely on funds going into that account), Lowrance identified only one individual payee (Anas-tasio Bayne) as having any association with Legendz. Tr. 260. Bayne was the payee on fourteen of the 550 checks written on the account (those fourteen totaling about $34,500), according to the government’s summary. GX C64B. Grupo Legendz was the payee on two of the checks (for just over $9,000), and Legendz Gamming [sic] was the payee on one (for $5,000). As to all of the remaining checks out of that account, the payee is either unknown (346 of the 550 checks) or an individual who is not in any way implicated as having any involvement with the Legendz gambling enterprise. Tr. 260.
Line 9 is rejected.
GX C89, Lines 10 and 11: Source of funds and nature of activities. Lines 10 and 11 will be addressed together because they are substantially interrelated. In Line 10, the government seeks to add $1,119,321.72 to the forfeiture money judgment. In Line 11, the figure is $336,463.38. But nearly seventy percent of the Line 11 money is also included in Line 10.
Lines 10 and 11 both involve accounts owned by Inversiones Altamonte — one account at Banco Panama and one at Credi-corp Bank. Inversiones was not mentioned by any witnesses at any of the four jury trials. Yet, at the forfeiture trial, Lowrance referred to Inversiones (without any elaboration at all, from him, from Jansen, or *941government counsel) as an affiliate of Le-gendz. Tr. 107. Nothing was brought out in the forfeiture trial that would indicate that Lowrance would have had any actual knowledge of an affiliation between Inver-siones and Legendz, but there is no need to dwell on that point because, as will be seen, it does not matter whether Inver-siones is an affiliate of Legendz.
Government Exhibit C54B summarizes deposits into Inversiones’ account at Banco Panama.76 The deposits into that account are from eight identified sources. And there are six deposits from unknown sources. One of those eight identified sources is Legendz Gamming [sic] Corp., which was the source of a $109,994.65 deposit on April 19, 2012. From the perspective of the forfeiture defendants, there is no good explanation for this item. It is not credibly arguable that funds coming directly from Legendz Gamming [sic] Corp. in April, 2012 consisted of anything other than proceeds of illegal gambling activities within the conspiracy period. After that, it gets more complicated.
Probably because of the government’s overall approach to the subject of “tainting,” a matter that is discussed later in this order, the government points, in its testimony or in its proposed findings, to only two of the other seven sources of deposits into the Line 10 Inversiones account as being sources of illegal gambling proceeds. Those are: Paige Overseas, Inc. and KHML Payment Processing, Inc. Tr. Ill, 114, 266-68. See also, the government’s proposed findings, doc. no. 2242, at 26-27, ¶ 98. This requires examination of what Lowrance claimed to know and what he does not know about these two companies.
Paige Overseas. Paige Overseas was not mentioned in any of the four jury trials. On direct examination at the forfeiture trial, Lowrance testified as follows:
Q. [Mr. McGarry] And who is Paige Overseas?
A. They are a payment processor.
Tr. 111.
This flat statement of fact was not garnished with any elaboration or qualification. But then came cross examination:
Q. [Mr. Mays] What business was Paige Overseas in?
A. I believe it was a payment processor.
Q. Okay. Do you know whether or not it was a payment processor?
A. No. I just know it was affiliated with Mr. Rigoberto Nolan Forbes and he has been indicted in this matter.
Tr. 268.77
The court’s firm expectation that any witness — but especially an FBI witness— will testify only to that which he knows to be true does not depend on whether the witness is on direct or cross examination. Sadly, Lowrance’s categorical testimony on direct examination is not the first instance, in this case, of an FBI witness testifying flatly to things he did not know to be true.78 And whether or not Lowrance believed Paige to be a payment processor (a proposition which, for Lowrance’s sake, *942the court will assume to be true, even though, when pressed, he admitted he did not know that to be a fact) is ultimately of no great concern, because he also admitted that he knew nothing about the origin of the funds that came into the Inversiones account at Banco Panama from Paige. Tr. 268. The deposits from Paige Overseas, Inc. are of no consequence for forfeiture purposes.
KHML Payment Processing, Inc. Low-rance also asserted that KHML Payment Processing, Inc. was a payment processor. Tr. 114.79 Specifically, he testified that KHML “is a payment processor that is affiliated with an individual by the name of Jorge Zuniga.” Tr. 266. (KHML was not mentioned in any of the jury trials, nor was Jorge Zuniga.) However, it is reasonably clear that any notions Lowrance has about KHML are based on what he “understand[s] from reading FBI reports.” Tr. 266. The experience of the undersigned in this case teaches that is too slender a reed, particularly when the matter is viewed in light of the fact that Lowrance has no knowledge as to the source of the funds remitted by KHML and could cite no financial records that would show where KHML got the money. Tr. 267, 269. True, the rules of evidence do not apply in this forfeiture proceeding, but the rules of persuasion do. The court is unpersuaded that it should include in the forfeiture money judgment calculation the $624,445.32 that the government attributes to KHML. GX C54B.
Of the $1.2 million sought by the government in Line 10, per GX C54B, $109,994.65 passes muster. Attention now turns to Line 11, which is closely related to Line 10.
Line 11, based on GX C54C and GX C54D, relates to the Inversiones account at Credicorp Bank. The first noteworthy transaction in that Credicorp account is the deposit of the $229,951.78, into that account, of funds that came, in the exact same amount, from the Inversiones account at Banco Panama that is the subject of Line 10. As GX C54C indicates, and as Lowrance acknowledged, that $229,951.78 deposit into the Inversiones/Credicorp account effected the closing of the Inver-siones/Banco Panama account. Tr. 116.80 In other words, at the outset, there is an overlap of nearly $230 thousand between Line 10 and Line 11, which accounts for nearly seventy percent of the $336,463.38 sought by the government in Line 11. The government acknowledges the overlap (doc. no. 2242 at 28, ¶ 102), but offers no explanation as to why that money should be included in the forfeiture money judgment.
The rest of the transactions (twenty-one in all) comprising Line 11 are equally meaningless for forfeiture purposes. Of those twenty-one, fourteen are payments of interest on a certificate of deposit that was bought with funds from the Inver-siones/Banco Panama account (e.g., the account addressed in Line 10). The government does not attribute the other seven *943deposits comprising Line 11 to any questionable source.
The government’s net yield from Lines 10 and 11 is $109,994.65 for the forfeiture money judgment calculation.
GX C89, Line 12: Source of funds and nature of activities. In Line 12, the government seeks to add $3,636,171.24 to the forfeiture money judgment, based on the summary in GX C66A and the source document, GX 1271. This $3.6 million consists of deposits into an account of Woodcastle Foundation at Towerbank, a Panamanian bank. Woodcastle was not mentioned at any of the jury trials. At the forfeiture trial, the government’s witnesses did not shed any particular light on the origins or corporate purpose of Woodcastle, but that was not really necessary. The primary focus, given the government’s deposit-oriented approach to building the forfeiture money judgment, was on the sources of the deposits and the nature of the activities generating those deposits.
Of the twenty-five deposits into the Woodcastle account, adding up to the $3.6 million sought in Line 12, six deposits came from identified sources. Two of those six are from Fundación Los Olivos, three are from Investment Consulting Services (ICS) and one is from Olmos Overseas Ltd. Those six account for $532,500.00 of the $3.6 million. King argues that “the government did not prove that any of the deposits into the Woodcastle account involve proceeds of illegal gambling.” Doc. no. 2275 at 13. The court disagrees. The evidence at the jury trials clearly established that Fundación, ICS and Olmos were used as conduits for illegal gambling proceeds. There was no evidence that any of those entities were used for any other purpose. They may have received some legitimate revenues, but, faced with cogent evidence favoring the government on this score, the court is no more inclined to speculate in favor of the defendants than it is inclined to speculate in favor of the government on other issues.
Lowrance acknowledged on cross examination that he did not know the source of the deposits from ICS (Tr. 274), but the court is unpersuaded by King’s contention that the court should reject this item. The evidence at the jury trials as to the pervasive illicit use of ICS was overwhelming. The deposits from Fundación, ICS and Olmos account for $532,500.00 in Line 12, and this will be included in the forfeiture money judgment calculation. GX C66A.
As for the rest of the $3.6 million in Line 12, the origin of the funds (let alone the nature of the activities generating the funds) is unknown. The government asserts, correctly, that “[t]his account received deposits from entities known to receive and move illegal gambling funds on behalf of the Legendz Sports enterprise” (doc. no. 2242 at 28), but the government refers only to the three entities discussed above in support of that proposed finding. Id. The testimony at the forfeiture trial was to the same effect. Tr. 118-125. Low-rance did try to impugn some of the unknown deposits by saying, in substance, that they at least came from banks which were known to have been used by entities like Fundación (Tr. 271), but that tie is too tenuous, for reasons that have already been discussed in detail.
Accordingly, the court finds that Line 12 contributes $532,500.00 to the forfeiture money judgment calculation.
GX C89, Line 13: Source of funds and nature of activities. In Line 13, the government focuses on another Woodcastle Foundation account. -That account, also, was at Towerbank. GX C65A. The $1,490,912.21 encompassed by Line 13 came into the account by way of fifty-four deposit transactions. Of those, fifty-one are transfers *944from the other Woodcastle account at Tow-erbank (the account that is the subject of Line 12) and three are of unknown origin. Id.; Tr. 121, 123, 277. With two exceptions, every one of the transfers from the Line 12 account to the Line 13 account occurred during the same time period that the funds originally flowed into the Line 12 account. Compare, GX C66A (funds into the Line 12 account) and GX C66B (funds out of the Line 12 account) with GX C65A (funds into the Line 13 account). The two deposits that are exceptions were made within two days after the last of the funds flowed into the Line 12 account. Id. In other words, save for the three Line 13 deposits that are of unknown origin, Line 13 is an exercise in double counting. The Line 13 deposits that are also included in Line 12 are meaningless for forfeiture purposes, as are the three deposits of unknown origin. That accounts for all of Line 13.
In its Forfeiture Brief and in its Proposed Findings of Fact and Conclusions of Law, the government offers no explanation or rationale for this double counting. At the forfeiture trial, Lowrance made one vague reference to 18 U.S.C. § 1957, apparently to support his notion that double counting is permissible where there is a money laundering conviction under that section. Tr. 235. But the government’s proposed findings and conclusions make no reference at all to § 1957, or to any theory, under the applicable money laundering statutes, that would support double counting, or to any other theory on which the government’s proposed double counting might pass muster. On the subject of double counting, the government advances only one proposition: “The United States is not entitled to double counting of proceeds for purposes of forfeiture.” Doc. no. 2242, at 61, ¶ 252. The court agrees. And that makes Line 13 all the more inexplicable in light of the fact that the double counting was made unmistakably clear at the forfeiture trial:
Q. [by Mr. Goldstein] If you compare C66A and C65A, it shows simply that the money that was deposited into line item 13, the second Wood-castle Foundation Towerbank account simply came from the antecedent Woodcastle Foundation, meaning line item 12, correct?
A. Yes.
Q. And so you’ve counted this money towards a forfeiture judgment, this is an example of what I would contend to be your double-counting, meaning you’ve included the same money in both line items 12 and line items 13 simply because it went from one account to a second account, correct?
A. Yes.
Q. You’ve also included a significant amount of money in your C89 chart that came from either lawful sources or unknown sources, correct?
A. Yes.
Q. Meaning there’s — there was no discriminating or limiting principle applied by you in preparing C89. You clearly included in C89 significant amounts of money for which you have no knowledge whether or not it derives from illegal betting, correct?
A. I don’t have specific tracing to show each of those dollars is attributable to illegal gaming.
Q. It’s not simply that you don’t have tracing. I mean, there are entries for unknown. You would have no way of knowing whether or not that significant amount of money had come from betting, legal betting, il*945legal betting, or some other business interest, correct?
A. That’s true.
Tr. 398-99 (Lowrance).
Although not argued in defense of obvious double counting, the government does point out that some funds from the Line 13 Woodcastle account “were distributed to individuals and entities affiliated with the Legendz Sports enterprise.” Doc. no. 2242, at 29, ¶ 105. But, aside from the fact that the government’s case for a forfeiture money judgment is pitched on deposits, not -withdrawals, that cannot close the gaping holes in the proof as to the character of the funds coming into the account, nor can it provide a rationale for double counting.
Line 13 is a nonstarter.
GX C89, Line 14: Source of funds and nature of activities. In Line 14, the government focuses on an account in the name of 6 Monkeys S. de R.L. at The Bank of Nova Scotia, also known as Scotiabank. The overwhelming evidence in the jury trials established that entities with the 6 Monkeys moniker were controlled by Mr. and Mrs. King and existed for the benefit of the King family. Defendants wisely do not assert otherwise.
By way of the transactions comprising Line 14, the government proposes to add $445,866.52 to the forfeiture money judgment. Detail for this entry on GX C89 is provided by the summary at GX C41, for which the source document is GX 1266. This $445,866.52 consists of four deposits. Three of those deposits amount, in total, to $3,500. For those, the source is unknown. The remainder is represented by a $442,366.52 deposit remitted by Starlink Investissements Inc. An October, 2007 email cited by the government clearly shows that Starlink is associated, in some way, with King. GX 214 (received, Oct. Tr. 966). This is supported by the testimony of Karlo Stewart at the jury trials to the effect that Starlink had a Swiss bank account that was on “investment account” for King. Feb. Tr. 913 (investment account); Apr. Tr. 1187; May Tr. 433.81 Testimony at the October trial fortified the connection between King and this account, as well as the proposition that this account was the account referred to in the October, 2007 email. Oct. Tr. 967-69.
At the forfeiture trial, the government had nothing other than the October, 2007 email to support the suggestion that the $445,866.52 deposit into the 6 Monkeys account from Starlink consisted of funds that originated with DSS. Tr. 128, 280 (nothing other than the email). The problem is that the funds came to the 6 Monkeys account from Starlink in April, 2012, about four and a half years after the date of the email that suggests that, in October, 2007, DSS was a source of funds for Star-link. It is not enough that, as the government asserts, Starlink was “an entity which received funds from DSS.” Doc. no. 2242, at 30, ¶ 107. A necessary (but not sufficient — see below) prerequisite to the government’s success as to Line 14 is to produce evidence sufficient to persuade the court by a preponderance of the evidence that these funds — this $445,866.52— originated with DSS. (To be sure, DSS was not the only entity from which illegal gambling proceeds could have come, but the government’s case as to Line 14 is pitched on that assertion.) The court is unpersuaded.
Assuming that the Starlink deposit did come from DSS, it is not a forgone condu*946sion that all funds coming from DSS were the proceeds of illegal gambling, as is discussed with respect to Line 6, above. For that additional reason, the court concludes that the government has not proven Line 14.
GX C89, Line 15: Source of funds and nature of activities. In Line 15, the government seeks to add $153,000.35 to the forfeiture money judgment, consisting of ten deposits into King’s personal account at The Bank of Nova Scotia, as summarized in GX C41. Three of those deposits, from unknown sources, amount to a total of $2,500. Of the others, one, in the amount of $54,755.35, is from the 6 Monkeys account at the same bank. And this $54,755.35 came from the 6 Monkeys account to King’s account not long after the $445 thousand went into the 6 Monkeys account from Starlink, as discussed above. So, aside from all of the other factual defects in the government’s case with respect to the 6 Monkeys account (Line 14, discussed above), that $54,755.35 plainly represents double counting.
The other six deposits comprising Line 15 are from DSS. Four of those are from late 2010 or early 2011 and two are from early 2013 (GX C41), a period easily encompassed by the beginning and ending dates of the flow of funds into DSS as discussed above with respect to Line 6 and as shown by GX C47A. Consequently, the government’s reliance on these deposits amounts to double counting, in addition to the factual insufficiency of the government’s contentions with respect to funds flowing into the DSS account at Banco General, as discussed with respect to Line 6, above.
As is discussed above with respect to Line 13, the government makes no attempt, in its Forfeiture Brief or in its proposed findings and conclusions, to support the repeated instances of double counting that the court has addressed so far. And Lowrance’s vague reference to § 1957 as supporting double counting “every time that money changes an account” (Tr. 235) is unpersuasive. In United States v. Tedder, 2003 WL 23204849 (W.D. Wisc. July 28, 2003), a § 1957 money laundering case, the government argued that “it is entitled to consider as forfeitable the entire amount of money that was ‘involved’ in the money laundering activity, so that it may count the same money more than once if it was involved in more than one transaction.” Id. at *4. In analyzing the government’s contention, the court was careful to note that the government is entitled to take profits earned on laundered money, as well as property used to facilitate the money laundering offense. Id. But:
It is another thing for the government to claim the right to funds calculated to reflect the number of times they were moved around.
If defendant moved $11,000 three times to three different bank accounts, he did not possess or handle a total of $33,000, any more than my transferring $10 from my wallet- to my bank account means that I now possess $20 or that I “gained” an extra $10 through the transaction. The purpose of the forfeiture laws is to deprive the perpetrator of his ill-gotten gains.
Id.82
The United States Department of Justice agrees:
It should also be noted that, with regard to proceeds, “double counting” or “dou*947ble recovery” through forfeiture is not permissible and, therefore, it is improper to forfeit more than the total value of the defendant’s unlawfully-obtained proceeds. For example, if the defendant obtains proceeds from an offense, he may be made to forfeit the total value of those proceeds or ordered to forfeit property traceable to those proceeds, but he cannot be ordered to forfeit the sum of both.
8 Department of Justice Manual, § 9-110.000C, CRIMINAL RICO: 18 U.S.C. §§ 1961-1968; A MANUAL FOR FEDERAL PROSECUTORS (United States Department of Justice, 6th rev. ed., May, 2016).
Line 15 fails on several counts, and is rejected.
GX C89, Line 16: Source of funds and nature of activities. In Line 16 (the last item in the government’s money laundering list), the government seeks to add $245,136.55 to the forfeiture money judgment because: “From March 2008 through September 2012, Bartice and Serena King purchased money orders totaling $245,136.55, which were used to pay credit card bills and mortgages.” Doc. no. 2242, at 31, ¶ 112. Although the government cites some exhibits, that is the sum total of the government’s explanation of this item.
The government’s summary supporting Line 16 is GX C31. That exhibit summarizes 320 money order purchases between March, 2008 and September, 2012. Of those, one is shown to have been purchased by “B A King” (presumably defendant Bartice King) and four are shown to have been purchased by “King,” all of which is supported by the court’s review of the source document (GX 1105). The remaining 315 are either shown to have been purchased by “S King” (presumably defendant Serena King, who was acquitted on all counts) or have no indication of the identity of the purchaser. They total $245,136.55. The money orders were made payable to various credit card issuers. GX 1105. These money orders received only the slightest mention at the forfeiture trial. Tr. 160, 176.
As an initial matter, Line 16, consisting of money that came to the Kings (almost entirely Serena King, so far as the source documents show) by means not shown by the record and was then spent to buy money orders to pay their credit card debt, is a glaring departure from the government’s deposit-oriented, money laundering forfeiture case, a matter discussed in note 64, above. That inescapably raises the specter of double counting, for which the government offers no explanation at all, even though the court invited an explanation:
THE COURT: So from those exhibits, the — in this second category that starts two-thirds of the way down the first page of C89 and goes a little over halfway down the second page, is this the sum total of the dollars into those accounts or out of those accounts or what?
THE WITNESS [Lowrance]: Your Honor, I don’t recall. I think it’s the dollars in and the dollars out.
THE COURT: On what theory do you present to me the sum total of dollars in and dollars out for forfeiture purposes?
THE WITNESS: I’m just — I would have to go back and look at the chart to refresh myself.
THE COURT: Okay. Just to be fair, as we speak — and you’ve got another two-days-plus to clarify. But as we speak, that $43 million number[83] is a bit enig*948matic to me. I’ll just put it to you that way.
[AUSA]: We can look into that, Your Honor, and provide an answer to the Court.
THE COURT: Okay. Very well.
Tr. 159-160 (emphasis added).
No factually and legally supportable answer has been provided. And lest the significance of the exchange quoted above be lost in the give and take: On the first morning of the forfeiture trial (after four jury trials and twelve years after the investigation began), neither the government’s lead witness (and sponsor of the summary that encapsulated the government’s forfeiture case) nor the AUSA were able to tell the court with any assurance whether the government’s money laundering forfeiture summary was pitched on dollars in or dollars out or both. By that afternoon, they were able to allow that they used deposits exclusively. Tr. 234. But that, also, was not true, as Line 16 demonstrates.
The court is disinclined to reach out and find double counting where there is a plausible scenario that would, when measured on a preponderance standard, suggest that double counting is not a problem. But Line 16 is a nullity if those dollars were not illegal gambling proceeds (a matter discussed below). If they were illegal gambling proceeds, then any notion that they are not already accounted for in one or more of the flows of funds already addressed (Lines 1-15) begs explanation. The government offers none.
All of this is aside from the fact (repeated several times so far, and destined to be repeated several more times) that, contrary to the government’s representations to the court, Bang did have legal sources of income during the conspiracy period. Frustrating to the government though it surely is that Serena King walked up to grocery store or post office counters scores of times, cash in hand, to buy money orders, it was incumbent on the government to show (or at least raise a reasonable generalized inference — see the discussion of Elder, above) that cash was dirty. The government has not done so. If, instead of pretending that King had no legal source of income, the government had dealt squarely with that fact and attempted to show why, in the grand scheme of things, that should make no difference, that might be another matter. But the government has not done so.
Line 16 is rejected.
3. Money laundering total.
The result of the court’s findings as to the accounts under the heading of “Money Laundering Accounts/Items” on GX C89 (App. 1 to this order), is:
*949Line 1: $ 8,950.00
Line 2: 61,604.40
Line 3: 0
Line 4: 4,630,377.83
Line 5: 0
Line 6: 0
Line 7: 364,754.92
Line 8: 0
Line 9: 0
Line 10: 109,994.65
Line 11: 0
Line 12: 532,500.00
Line 13: 0
Line 14: 0
Line 15: 0
Line 16: _0
Total: $5,708,181.80
4. A brief digression.
At least in a general sense, the government’s approach to the money laundering accounts, Lines 1-16, sheds light on the government’s approach to other aspects of its forfeiture case, as addressed in this order. For that reason, and given the government’s limited success in making its forfeiture case as to the money laundering accounts, it is appropriate to pause at this point to examine more closely the conceptual basis for the government’s proposed inclusion of more than $43 million of allegedly laundered funds in the forfeiture money judgment that it seeks against every forfeiture defendant.
It is evident that the government’s forfeiture logic as to most of these money laundering accounts is as follows:
Step 1: An account that was maintained in the name of an entity that may or may not have been associated with the Legendz enterprise received some betting proceeds even though the source of the funds or the nature of the activities generating the funds is unknown as to most (and in some instances nearly all) of the deposits.
Step 2: Ergo, the account is tainted and is to be treated as a forfeitable vessel.
Step 3: Because the account is a forfei-table vessel, every deposit into that account during the conspiracy period (or at least those deposits for which records have been obtained) picks up the taint, becoming forfeitable funds, regardless of whether the source of the funds for any given deposit is known.
Step 4: Ergo, the government is entitled to add up all of the deposits into that account and get a forfeiture judgment for that amount (or get forfeiture of the account itself).
This reasoning process was made clear at the forfeiture trial in several ways, not the least of which is that the government was noticeably indifferent to the extent to which its money laundering-based forfeiture judgment is predicated on identification of dirty money:
Q. [by Mr. Goldstein] You’ve also included a significant amount of money in your C89 chart [App. 1] that came from either lawful sources or unknown sources, correct?
*950A. Yes.
Tr. 398 (Lowrance).
Thus, as has been noted, the key to the government’s approach is to taint an account as if it were a vessel that would pass the taint on to all funds in the account, regardless of the source. For example, in the case of Bramley, the government was well aware that his Social Security benefits (lawful income by any standard) were transferred into one of his Scottrade accounts (Tr. 642), but the deposit of bettor funds into his Scottrade accounts “taint[ed] up those accounts from a forfeiture standpoint.”84 Tr. 508. The result of “taint[ing] up” an account, per the government, is that the entire account is forfeited (doc. no. 1999-2, at 1, 7; GX C81) even though the (i) government does not argue facilitation as to that account, and (ii) the government does not even hazard a guess as to the ratio between Bramley’s legitimate income vs. his illegal income. Tr. 629-30. No matter. The government’s “taint up” theory excuses the government from worrying about legitimate income:
Q. [by Mr. Jumes] Okay. You trace to find if there’s gambling checks that ever went into an account—
A. That’s correct, sir.
Q. —but you didn’t do any tracing to see how much money is actually sourced by legitimate funds versus illegitimate funds within any account in this case?
A. Yes, that’s a fair statement.
Tr. 649 (Jansen).
The potency of the government’s “taint up” theory — if it had one iota of legal merit — is perhaps best illustrated by the government’s treatment of the Vaporama account at Banco General. The sources of 2,337 of the 2,339 deposits into that account were unknown, as discussed above with respect to Line 9. So, of the $1.9 million going into that account, a $15,000 deposit from Data Support Services was sufficient to make the entire $1.9 million fair game for inclusion in the forfeiture money judgment. GX C64A; GX C89, Line 9. Thus, Lowrance cited the $15,000 from DSS — less than .008 of the deposits into the Vaporama account — as sufficient to “taint that account.” Tr. 264. This was made quite clear: it was sufficient that “there was a payment into the account from the Data Support Services account.” Id. (The $15,000 payment from DSS was the only deposit identified as being from DSS.)
Q. [by Mr. Mays] So it’s like a drop of oil into a bucket of water, that $15,000 taints the entire account?
A. I would initially look at it like that, yes.
Tr. 264 (Lowrance).
There is nothing “initial” about this approach. This is the theory by which the government still seeks to add $1.9 million to the forfeiture money judgment. Doc. no. 2242, at 26, 47-48; GX C89; GX C64A.
It comes, thus, as no surprise that the government does not cite the Tenth Circuit’s decision in Bornfield in its Forfeiture Brief or in its proposed findings and conclusions: “mere pooling or commingling of *951tainted and untainted funds in an account does not, without more, render the entire contents of the account subject to forfeiture.” Bornfield, 145 F.3d at 1135.85
C. Viability of the government’s proposed direct attributions and extrapolations
Under the heading of “Legendz Payouts to Bettors” on GX C89, the government proposes an extrapolation, to slightly over $25 million, which it seeks to include in the forfeiture money judgment. Then, under the heading of “Agents” on GX C89, the government seeks to attribute more than $159 million (as revised, see the discussion below) in betting proceeds to fifteen individuals — and then to hold the seven forfeiture defendants now before the court jointly and severally liable for that sum, as part of the overall $231 million forfeiture money judgment the government seeks against all of these defendants in its Proposed Findings of Fact and Conclusions of Law. Some of those attributions are based on extrapolations and some are posited simply as arithmetic totals, unaugmented by extrapolation. The court now turns to these items, as found under the first two headings on the first page of GX C89.
1. General nature of the government’s proposed direct attributions and extrapolations
As has been noted, at the forfeiture trial, the government sought (for the first time, despite having been ordered to state its position definitively about three months before the forfeiture trial) a money judgment against each of the forfeiture defendants equal to the sum total of (i) the extrapolated amount the government proposed to attribute to Legendz payouts to bettors, plus (ii) the amount of betting proceeds the government attributes to all of the agents and runners (fifteen in all).
As has also been noted, these two categories, based mostly on extrapolation and totaling just over $184 million, are shown on the first page of App. 1 (GX C89) and may be briefly described as follows:
First, the government would include $25,098,950.69 in the money judgment on the basis of “illegal proceeds processed by Global Data Payment Services.” Doc. no. 2242, at 14. (For reference purposes, this is the extrapolation to $25,084,519.78 shown as the total under the heading of “Legendz Payouts to Bettors” in the top box on GX C89. See, App. 1. As is discussed later, these numbers do not agree because the government changed this figure in its Proposed Findings of Fact and Conclusions of Law, doc. no. 2242.) Global Payment Data Services was a payment processor controlled by King. It functioned essentially as a clearinghouse for the receipt and remittance of betting proceeds. This proposed extrapolation is discussed in Part IX(C)(2), below.
Secondly, the government proposes a series of extrapolations (and, in some instances, direct attributions without extrapolation) of betting proceeds attributable to agents and runners. This would add $159,006,791.24 to the forfeiture money judgment. (For reference purposes, this is the series of items totaling $142,281,247.37 shown under the heading of “Agents” in the first page of GX C89. See, App. 1. The government revised GX C89 to insert the updated numbers shown in red on the first page of that exhibit, but did not update the total at the bottom of the “Agents” box, or, for that matter, the grand total on p. 4 of GX C89. So, the government asked the court, in its calculations, to eliminate *952the numbers immediately above the red numbers and substitute the red numbers in its calculation. Tr. 158. This yields a total, per the revised version, of $159,006,791.24.)
Taking into account the numbers in its Forfeiture Brief, the numbers in the original version of GX C89, the numbers in the revised version of GX C89 and the numbers in its post-trial proposed findings of fact, the government, in preparing for and presenting its forfeiture case, twelve years after the Legendz investigation began, was unable to bracket the amount of its proposed attributions of betting proceeds to agents and runners more tightly than plus or minus approximately $18.3 million, as shown here:
Table 2
Government's Positions as to Amount Sought, for Forfeiture Money Judgment Purposes, by way of Attribution to Agents and Runners
(All amounts in dollars)
Name Amount sought in Forfeiture Brief doc. no. 199 986 Amount sought In original version of GX C89 Amount sought In updated version of GX C89 Amount sought In Proposed FoF and CoL, doc. no. 2242
Campbell 43,160,000.00 26,290,472.00 38,888,787.60 38,888,787.60
Tucker 8,842,560.00 7,958,304.00 7,958,304.00 7,958,304.00
Middle-brook 4,524,520.00 4,449,353.14 4,449,353.14 4,449,353.14
Moran 2,993,660.00 6,491,205.00 6,491,205.00 6,491,205.00
Ross 1,642,813.00 1,265,831.18 5,485,268.45 5,485,268.45
Tanner 17,667,895.00 15,143,910.00 15,143,910.00 15,143,910.00
Barry 9,139,255.20 15,275,612.57 15,275,612.57 15,275,612.57
Hernandez 22,232,393.85 18,121,267.20 18,121,267.20 18,121,267.20
Lawhorn 20,000,000.00 18,000,000.00 18,000,000.00 18,000,000.00
Wilson 3,000,000.00 4,000,000.00 18,000,000.00 18,000,000.00
McFadden 1,656,000.00 1,784,800.00 1,784,80.00 3,907,791.00
Bramley 5,102,789.00 4,953,478.78 4,953,478.78 4,945,633.78
Diebner 18,567,345.60 16,770,000.00 16,770,000.00 17,432,674.00
Koralewski 1,222,856.00 1,055,112.00 1,055,112.00 935,112.00
Robles 815,197.50 721,901.50 721,901.5 727,901.50
Total 160,567,285.15 142,281,247.37 159,006,791.24 '159,350,820.24
As discussed in Part VI(B) of this order, there is no doubt that reasonable extrapo*953lations, based on reliable evidence, are permissible in determining the amount of a forfeiture judgment. An indispensable premise for any financial extrapolation is the proposition that the sample data set proffered as the jumping-off point for the extrapolation consists of transactions known to have the characteristic — in this case, transactions generating funds from betting activity having some connection with the illegal gambling enterprise charged in the Indictment and proven at trial — that the government would have the court attribute to the larger dollar amount resulting from the proposed extrapolation.
Another indispensable premise is the proposition that it is more probably true than not that the level of activity in the base period is representative, and thus a reliable indicator, of the organization’s level of illegal bookmaking activity, on the whole, for the period the government proposes to encompass with its proposed extrapolation. It is true that the government does not have to prove forfeiture amounts with to-the-penny precision (or even anything very close to that). But it is equally true that when the government would have the court extrapolate from a base data set to a final number, there can be no outcome other than rejection of the proposed extrapolation when the base data set is not shown to be an accurate reckoning of illegal activity during that period, when the activity in the base period, even if reliably proven, is not shown to be representative of the period to which it would be extrapolated, or when the difference between the length of the base period and the length of the total period is just too great to provide a basis for confident extrapolation from one to the other.
The government’s proposed extrapolations are, in some instances, from a known number of weeks or months of betting activity to a nine-year period encompassed by the Superseding Indictment. The government’s proposed extrapolations incorporate various combinations of assumptions, depending on which individual is involved. At a minimum, the government’s extrapolations depend on an assumption that the average periodic dollar amount of gambling proceeds generated by a particular bookie, agent or master agent during a known period is representative of the betting activity generated by that individual for the entire period in question, which is nine years87 as to most of the individuals who produced those proceeds. That assumption encompasses a raft of variables, as will be seen.88
Another assumption, well-supported as to some individuals but not as to others, is that the individual in question did, in fact, participate in the Legendz enterprise for the entire period asserted by the government. Yet another assumption is that there is no overlap, or double counting, as between funds remitted to Panama by master agents and funds generated by bookies working under those master agents.
With the government’s proposed extrapolations in mind, the court will now address the merits of those proposed extrap*954olations and direct attributions. As has been noted, the government’s proposed extrapolations fall into two categories: (i) extrapolations based on payouts to bettors through Global Data Payment Services, and (ii) extrapolations proffered to arrive at a dollar amount of gambling proceeds to be attributed to each of the agents and runners (including the forfeiture defendants).
2. Extrapolation based on Legendz payouts to bettors through Global Data Payment Services
Under the heading of “Legendz Payouts to Bettors” at the top of the first page of GX C89 (App. 1), the government extrapolates from one of the money laundering accounts, GX C89, Line 2, to arrive at a 108-month extrapolation in the now revised amount of $25,098,950.69 which it seeks to include in the forfeiture money judgment against all of the forfeiture defendants. This is said to represent “Le-gendz Payouts to Bettors.” GX C89. See also, doc. no. 2242, at 14, ¶ 54. (On the government’s forfeiture recap, GX C89, this figure is $25,084,519.78, but the government now says that this figure should be $25,098,950.69. Doc. no. 2242, id.)
As is discussed above in Part IX(B)(2) with respect to Line 2, even though the source is unknown for the vast majority of the deposits,, representing seventy-six percent of the dollar amount, the government uses $7.15 million, representing transactions that occurred only in 2007 and 2008, as the beginning point for its proposed extrapolation to 108 months (from January 1, 2004 to December 31, 2012), totaling $25,084,519.78. GX C89, p. 1 (top box). This represents, in round numbers, $25 million of the $231 million for which the government seeks a money judgment against each of the forfeiture defendants. But the court has rejected the beginning number ($7.15 million), finding that only $61,604.40 of that is fair game. Accordingly, if the extrapolation proposed in the top box in GX C89 is tenable, the base number for that extrapolation is $61,604.40.
The government’s proposed extrapolation under the heading of “Legendz Payouts to Bettors” (top box on GX C89) is based on a multiplication from 24 months to 108 months. Tr. 200. The court is unwilling to use twenty-four months as the base period for an extrapolation to 108 months. To take but one example of the circumstances cutting strongly against this proposed extrapolation, the base period for the extrapolation to 108 months is entirely in 2007 and 2008, but includes only five transactions (out of 136) dated after July, 2008. GX C49A. The government’s extrapolation thus implicitly assumes that betting activity was the same before and after the financial implosion that occurred beginning in the fall of 2008. As has been noted (Part VI(B), above), in Lyons, 870 F.Supp.2d 281, 287 (D. Mass. 2012) (a RICO sports betting ease), Judge Saris rejected the government’s proposed extrapolation from a ninety-six month period (1997 through 2005) to a fifty-two month period (January, 2006 through April, 2010), partly because the financial crisis of 2008 affected the comparability of the two periods. Id. On the basis of all of the evidence (from four jury trials and the forfeiture trial) as to the secular variations in the betting activity of the Legendz bettors during the conspiracy period, the court is far from satisfied that the twenty-four month period proposed by the government is a reliable jumping off point for its proposed extrapolation to 108 months.
The government has failed to satisfy the court, by a preponderance of the evidence, that its proposed extrapolation for payouts to bettors is sufficiently rehable for inclusion in the forfeiture money judgment. *955Viewed either as an extrapolation from a $7.15 million base or from a $61,604.40 base, the government has not carried its burden of proof. The court rejects this extrapolation.
3. The government’s proposed extrapolations and direct attributions re: proceeds attributable to individual agents and runners
Under the heading of “Agents” on GX C89 (App. 1), the government proposes a series of direct attributions and extrapolations89 to assign $159,006,791.24 in betting proceeds to fifteen Legendz “agents” (six90 of whom are the forfeiture defendants now before the court) and then to add that amount to the forfeiture money judgment.91 To the extent that these items are extrapolations, they consist of a base dollar amount, a base period and an extrapolation period. These must be addressed individually, because there is substantial variation in the factual merit of these proposed extrapolations. Some of the problems are:
• Extrapolating on the basis of as- , sumptions negated by the government’s evidence (or other uncontro-verted evidence) at the jury trials.
• Extrapolating on the basis of assumptions otherwise unsupported by the evidence (including, among others, assumptions as to the similarity of the level of activity in the base period vs. the entire period encompassed by the extrapolation).
• Extrapolating from a base period that is only a small fraction of the entire period proposed for the extrapolation.
• Extrapolating on the basis of an assumption that King did not have a lawful source of cash flow during the conspiracy period.
These, and kindred matters, will be addressed as the government’s proposed extrapolations and direct attributions are considered. The reference point here is the fifteen lines under the heading of “Agents” on GX C89, App. 1 to this order. Those items will be addressed in the sequence in which they appear on GX C89.
a. Terry Campbell
The government proposes a $38,888,787.60 extrapolation as to the activity of Terry Campbell, a Legendz bookie. The government originally said this should be an extrapolation to $26,290,472.00. GX C89 (under the heading of “Agents”). Some more general aspects of this proposed extrapolation should be addressed.first, because they shed light on the court’s concerns about this and other extrapolations proposed by the government under the “Agents” heading.
*956At the forfeiture trial, Lowrance testified that he 'prepared or participated in the preparation of the government’s charts. He vouched for the accuracy of the government’s extrapolations. See, e.g., Tr. 140 (as to Campbell). As to Campbell, Low-rance testified at first that, using GX 496, representing one week of activity, “we” extrapolated on the basis of a beginning number, for the base period, of $54,955. Id. Later, the government came up with additional records, GX 497, supporting a ten-week base period. Id. This resulted in the replacement of the original $25,290,472.00 extrapolation with an extrapolation to $38,888,787.60.
After it was brought to Lowrance’s attention on cross examination that the source document he cited would not provide the base data he said it would, he backed off:
Q. [by Mr. Mays] So is it your testimony, sir, that this extrapolation is based off of a week in April of 2013?
A. I believe that I had about ten weeks’ worth of records and I believed that those were either Exhibit 496 or 497.
Q. Plainly, that’s not the case; would you agree?
A. Yes.
Tr. 326.
Lowrance later asserted that the source document for the extrapolation as to Campbell should be GX 1257 (Tr. 418-19), notwithstanding his earlier testimony, and even though the government’s summary refers to GX 496 and 497 — even going so far as to cite GX 497 for the updated ($38.8 million) figure.
Regardless of what the correct source document for the government’s proposed extrapolation might be, Lowrance was able to confirm that the base data is not based on actual collections. Tr. 330. The base data is “not an accounting of what was actually collected or paid out.” Tr. 331.
Terry Campbell entered a guilty plea in this case in early January, 2016, well before the forfeiture trial, giving the government an opportunity to discuss with Campbell the interpretation of the fairly cryptic underlying records and the reliability of the proposed $38.8 million extrapolation. But Lowrance did not discuss any aspect of this extrapolation with him. Tr. 322. Neither, to Lowrance’s knowledge, did any of the other individuals involved in arriving at this extrapolation discuss either the extrapolation or the source document with Campbell. Tr. 324-25. They didn’t even consider doing that (Tr. 325), although that simple step would obviously have gone a long way to engender confidence in an extrapolation by which the government seeks to add $38.8 million to the forfeiture judgment. (The court can say without hesitation, based on Campbell’s appearances as a witness, that he was intelligent, articulate and cooperative with the government.) Lowrance was unable to tell the court whether the government’s Campbell extrapolation was consistent with Campbell’s trial testimony. Tr. 322.92 Not having discussed the Campbell extrapolation with Campbell, Lowrance was unable to tell the court what Campbell’s arrangement was with King, or whether the level of Campbell’s bookmaking activity varied from month to month. Id.
To understand the significance of the government’s failure to avail itself of a readily available source of confirmation of a $38.8 million extrapolation, it may be *957helpful to step back to look at the forest instead of the trees. First, we are talking about $38 million. In a garden variety civil suit (if it could be called that) for $38.8 million, the idea of not consulting the witness in the best position to give credence to that figure would be beyond comprehension. This is not Monopoly money, as the government’s causal approach might suggest. Second, the records from which the government proposes to extrapolate are far from self-explanatory — and Lowrance disclaimed any special expertise in interpreting them. Tr. 330 (“I can’t say that I understand all those workings like you’re talking about.”).93 Third, as to the cooperators (which includes Michael Lawhorn, Bruce Middlebrook, Paul Wilson, Joseph McFadden and David Ross as well as Campbell), the time and effort required to get at least some minimal confirmation is, in the grand scheme of things in this case of epic duration, minimal. Fourth, no one, anywhere, is in a better position than Campbell to tell the government (and the court) what can reliably be inferred from the records purporting to show Campbell’s own bookmaking activity. Confirmation of a $38.8 million extrapolation from the very person whose activities are the basis for the extrapolation would boost the credibility of the extrapolation by an order of magnitude. These considerations are so self-evident that the court cannot but conclude that the government simply did not want to run the risk of having to deal with inconvenient factual guidance from the individuals who were in the best position to provide it.
The sense the court gets, taking the Campbell extrapolation as only one example, is that as long as the government had in hand documents that could be interpreted to support a nine-year extrapolation to some figure in the tens of millions of dollars, the government was decidedly disinclined to ask any questions that might upend that calculation. Given the magnitude of the dollar amounts the government wants the court to include in the forfeiture money judgment, and given the gravity of the matter for all concerned, the court sees no good explanation for the government’s disinclination to take even simple steps to get verification that would add some credence to the base data it uses in its extrapolations.
The quality of the evidence supporting the government’s proposed $38.8 million extrapolation as to Campbell is unsatisfactory. The court is unpersuaded of the reliability of the extrapolation or of the reliability of the base number. Moreover, this is not a situation in which it might be reasonable to pick a lower number just to be on the safe side. That would be as speculative as the government’s proposed extrapolation is in the first instance.
The Campbell extrapolation is rejected.
b. Ralph Hernandez
A variation on the problem of extrapolations negated by the evidence is the problem of extrapolations that are simply unsupported by the evidence, exemplified by the government’s proposed extrapolation with respect to gambling proceeds attributable to Ralph Hernandez. Ralph Hernandez served the Legendz enterprise in several capacities. He was a bookie, a runner, and a repository of funds generated by others. As the government asserts, he sent funds to Panama on numerous occasions.
*958As shown in the extrapolation overview exhibit (GX C89), the government proposes an extrapolation to $18,121,267.20 on the basis of Hernandez’s activity. That extrapolation is based on summary exhibit GX C51, showing what the government says are Hernandez’s shipments of funds to Panama. That summary, in turn, is based on a source document, GX 198. Thus, the summary exhibit and the source document are building blocks of the government’s overview exhibit (the exhibit which shows the government’s extrapolated figures). Because the source document (GX 198) is the foundation of this extrapolation (source document GX 198 supports summary exhibit GX C51, which supports the extrapolation overview, GX C89), the court begins with the source document.
Government Exhibit 198 is a list of tracking numbers (except for a few entries that say “hand deliver”). For each tracking number, there is a date (but not the year), a dollar amount and a name, of sorts, such as “BIGMIKE” or “GATO” or “SKI.” Many of the names on the list are recognizable as monikers of Legendz bookies or agents. The natural inference to be drawn from GX 198 is that it is a record of shipments of funds generated by the various individuals whose Legendz monikers appear adjacent to the date of the shipment and the tracking number.
This is where this matter gets complicated, because Hernandez himself was a Le-gendz bookie. Feb. Tr. 708; Apr. Tr. 578. Recall that GX 198 is the source document for the summary exhibit GX C51. This summary exhibit is sponsored by Low-rance, who testified (Tr. 69) that he “assist[ed]” in preparing GX C51. The summary exhibit (GX C51) includes the dates, the tracking numbers and the dollar amounts; and, for every item, it shows “GEORGIE” under the heading of “WHO SENT IT.” It may well be true that “Geor-gie” sent the money, as urged by the government per summary exhibit GX C51, and the court will so assume. (“Georgie” was Hernandez’s moniker. Tr. 68-69.) Notably, however, the summary exhibit does not include the names — “BIGMIKE” or “GATO” or “SKI,” etc. — which appear in GX 198, the source document for the summary.
Lowrance explained the summary (GX C51), as follows:
Q. [Mr. McGarry] Could you please identify that [GX C51] for the Court.
A. This is a summary chart built on tracking numbers of packages that were received at the Legendz facility on behalf of Georgie or Ralph Hernandez.
MR. McGARRY: The government would move to—
THE COURT: When you say “on behalf of,” I’m not quite sure what you mean.
THE WITNESS: My understanding is that Mr. Hernandez would owe this money to Legendz for his bettor activity and he would be forwarding this and getting credit for it with Legendz.
Tr. 69 (emphasis added).
Shortly after that, Lowrance answered “yes” when asked whether the $7.5 million total on GX C51 (the summary) reflected “his” (Hernandez’s) “association with the Legendz enterprise.” Tr. 71. Lowrance later testified that the $18 million Hernandez extrapolation is included in GX C89 (the government’s forfeiture overview— App. 1) because “we had betting records for Mr. Hernandez.” Tr. 141 (emphasis added). Lowrance could look at the summary (GX C51) and easily conclude that it reflected nothing but Hernandez’s activity, because the summary didn’t show any of the numerous other monikers listed on the source document.
*959This, quite obviously, raises questions as to whether Lowranee had even looked at the source document (GX 198), because even the briefest glance at that source document — the only one cited by the government in support of this extrapolation— shows that the funds transmitted per that source document were generated by numerous persons other than “GEORGIE.” To be sure, “GEORGIE” is listed as the source of some of the remittances — specifically 142 out of a total of 578. But it is simply not possible to look at the source document and then say that the summary shows Hernandez’s “bettor activity” or Hernandez’s “betting records.”94
This is no small point because, aside from the questions this raises as to Low-rance’s familiarity with the source documents backing the summaries he sponsored, the other names on the source document are names to which the government also proposes to attribute betting proceeds for forfeiture purposes. If Low-rance’s characterization of the activity shown by the summary (and, implicitly, by the source document) were correct (which it is not), there would be no double counting problem here. But, as will be seen, there is.
There are other problems with this extrapolation. The beginning number for the government’s proposed extrapolation as to Hernandez is $7,550,528.00. GX C51. As has been discussed, this summary is based on shipments, apparently of some combination of cash and checks, shown on an internal Legendz source document. GX 198. But this Legendz spreadsheet, although reflecting a day and month for the shipment, does not show the year for any of the shipments. The years are not known. Tr. 338-334. The government’s extrapolation is based on an average for 45 months, as purportedly shown by GX 198 (Tr. 141), but that exhibit actually only covers 39 months. The government then divides the total by 45 and extrapolates to 108 months to arrive at an extrapolated figure of $18,121,267.20. GX C89. (The use of a divisor of 45 months, rather than the 39 months actually shown by the source document, to arrive at the monthly average results in an understatement of this extrapolation by more than $2 million. But that is of no moment because this extrapolation is rejected for other reasons.)
The government, in its proposed findings of fact, posits a monthly dollar average for Hernandez that can only result from using a divisor of 39 (doc. no. 2242, at 18, ¶ 69), but repeats Lowrance’s error by referring to 45 months. Id., ¶ 70. To support the 108-month multiplier, the government cites testimony purportedly from Bruce Middlebrook at the April, 2015 trial of Bartice King. Id. at ¶ 71. But Middle-brook did not testify at that trial, and the subject of Hernandez’s tenure with the Legendz enterprise is not discussed on the page cited by the government. Testimony (not from Middlebrook) at the April, 2015 trial does indicate that Hernandez got started with King before the beginning of the conspiracy period (Apr. Tr. 319), but nothing in the transcript of that trial establishes one way or the other whether Hernandez stayed with the Legendz organization long enough to support a 108-month extrapolation.
Moreover, and of at least equal importance, it is clear from the testimony at the February and April jury trials that Hernandez was a bookie (Feb. Tr. 708; Apr. *960Tr. 578) and a runner (Feb. Tr. 294; Apr. Tr. 277-78, 393-95) and a trusted holder of cash for Bartice King, a point that the government thought important enough to bring out by way of leading questions. Apr. Tr. 578. See also, Feb. Tr. 708 (Hernandez “held money for [King]” in his cellar); Apr. Tr. 578 (same). Thus, for instance, as a runner and cash repository, Hernandez—per the government’s evidence—collected money from Paul Wilson. Apr. Tr. 393; Oct. Tr. 694-95. But the government also seeks to base a $3.9 million, 108-month extrapolation on Wilson’s activity, separate from the Hernandez extrapolation. GX C89. In fact, the source document attributes remittances to five monikers (aside from Wilson) which the government asserts were the Legendz monikers of individuals as to whom the government seeks, separately, to attribute betting proceeds for purposes of getting a forfeiture money judgment.95
The sum of the matter as to the proposed $18 million Hernandez extrapolation is that the base number does not represent what Lowrance says it represents,96 and, tfeating that base number as what it actually is, it is fatally infected with double counting. And the year is not known for any of the activity shown by the source document, which is aside from the fact that the government has cited no basis for a terminal date for Hernandez’s participation in the Legendz operation that would support a 108-month extrapolation.
The $18,121,267.20 extrapolation and the base number, $7,550,528.00, are untenable. The court rejects both.
c. Kelly Diebner
Two things are not in doubt about Dieb-ner. First, he was a Legendz bookie in Texas. Secondly, he generated a substantial amount of illegal betting proceeds for the Legendz enterprise.
By way of extrapolation, the government sought, as shown in GX C89, to attribute $16,770,000.00 to Diebner. As will be seen, this figure was advanced during the testimony of Lowrance at the forfeiture trial. Using an entirely different factual approach, a different extrapolation, to $17,432,674.40, was advanced during Jansen’s testimony at the forfeiture trial. The court will first address the $16.7 million extrapolation advocated by Lowrance.
Neil Myler, a cooperating defendant, is central to both of the government’s alternative extrapolations. Myler was a Le-gendz runner who was active in Texas. He started out as, essentially, a general helper for Bartice King, in activities unrelated to sports betting. After a time, he transitioned into his service as a trusted runner for the Legendz organization and for King, personally. Ultimately, Myler developed qualms about his involvement in the organization and went his own way in late 2011 or early 2012. Apr. Tr. 1671. He testified in three of the four jury trials and his testimony was generally credible. (Every word of his testimony may well have been true, *961but much of what he had to say was uncorroborated and incapable of verification.)
The $16.7 million extrapolation is, in Lowrance’s words, “based on a statement given to the FBI by Mr. Neil Myler.” Tr. 141-42. In response to a question from the court, Lowrance confirmed that this figure was based on an interview with the FBI, and not on trial testimony. Tr. 142. Low-rance was correct about that. Myler’s jury trial testimony, although credibly establishing that he handled large amounts of betting proceeds for the Legendz organization, did not (and did not need to) establish any particular dollar amount as to Dieb-ner.
Per Lowrance’s account of the statement given to the FBI by Myler, “[h]e talked about exchanging money at the amount of $200,000 a month with Mr. Diebner over the period of time.” Lowrance elaborated: “So we took the 215,000 that he told us he dealt with Diebner and his associates, multiplied that out over 78 months, and came up with a $16,770,000 figure.”97 Id. Lowrance had nothing else of substance to say to substantiate the $16.7 million extrapolation. As he put it: “I was given a number [$215,000] to calculate out for the purposes of putting it on Mr. King’s sheet [GX C89]. I have not examined the underlying documents that make up Mr. Diebner’s sheet at all and I would not be able to discuss it with any clarity.” Tr. 373. Accordingly, Lowrance agreed that he had “no idea how much money he’s [Diebner] responsible for, how much his gains are, where they came from, and whether they’re valid.” Tr. 374.
The government has a source document, GX 325 (discussed below), which it cites in support of the $17.4 million extrapolation, but the government does not rely on that or any other source document to support the $16.7 million extrapolation. Tr. 589. So, as to the $16.7 million extrapolation, we are left with Lowrance’s interpretation (and recollection) of an FBI 302 report (Tr. 373) which presumably reflects what the writer of that report understood Myler to be saying about his dealings with Dieb-ner. The writer of that report did not testify and was not identified at the forfeiture trial. (Myler did not testify at the forfeiture trial.)
This is not the stuff of which $16.7 million forfeitures are made. No percipient witness has testified in support of the $215,000 multiplicand. No document supports it. Aside from that, the reliability of the seventy-eight month multiplier is subject to substantial doubt. Tr. 576-78. For reasons which presumably require no elaboration at this point, the court declines to rely on hearsay twice removed from the putative source to arrive at a basis for including this $16.7 million in the $231,432,686.73 forfeiture money judgment the government seeks.98
*962At the forfeiture trial, the court expressed skepticism about the $16.7 million extrapolation. Tr. 142 (“FYI, that’s not going to carry a whole lot of weight.”). Jansen, the government’s second (and last) witness at the forfeiture trial, articulated an alternative extrapolation. This is the $17,432,674.40 extrapolation. This extrapolation does have the benefit of a source document, GX 325. Government Exhibit 325 is some cryptic handwritten notes that were given to Myler by an unknown person. Oct. Tr. 1698-99 (Myler); Tr. 528 (Jansen). Myler referred to these notes very briefly in the October jury trial, but this exhibit was not introduced or otherwise referred to in any of the other jury trials — including Diebner’s trial.
The government characterizes GX 325 as “29 handwritten notes” which “summarize, based on [Myler’s] testimony, payments from Diebner’s betting group that he processed either through Mr. Diebner himself or through Rok, a runner for Mr. Diebner.” Tr. 527 (Jansen). These notes (GX 325) do contain numbers and what appears to be names, or at least monikers. Some of the twenty-nine pages are fairly well organized and some are gibberish. Compare, pages 4 and 5 of GX 325. For the most part, they are not dated. Within their four corners, they do not, individually or collectively, indicate what time period they cover. Many of the entries have, from left to right, a moniker, a number, a delta (A) and then another figure. The first number may be a code number. Oct. Tr. 1699. In a few instances (e.g., p. 9), the first number has a “Z” prefix. Viewed in eon-text, they surely relate to betting of some sort. But they are, beyond question, cryptic in terms of the ability of an outside observer to reach reliable conclusions as to their import.
Myler referred to these notes in only one of the three jury trials in which he testified. In the October trial (in which Myler was not subject to cross examination on behalf of Diebner, because Diebner stood trial in the May trial), Myler testified, believably, that these notes “represent the Saturday pick-up of bets that were lost the prior week.” Oct. Tr. 1699. These are not Myler’s notes. Id. The notes are “listings of people that lost money that would have to pay to Legendz.” Id.
At the forfeiture trial, Lowrance made no reference to GX 325 (because he was advancing the $16.7 million extrapolation, not the $17.4 million one). Jansen was the sponsor of the $17,432,674.40 extrapolation. Tr. 589. As for Jansen, the beginning point is that this $17.4 million extrapolation was neither his product nor was he willing to vouch for it with any particular gusto. He testified that “the FBI analyzed the individual notes [GX 325].” Tr. 529 (emphasis added). In this context, it is clear beyond question that Jansen, a forfeiture contractor, did not, for this purpose, include himself in “the FBI.” See also, Tr. 593. He testified that the “29 notes” equated to twenty-nine weeks. Tr. 529. They arrived at a weekly average. Tr. 530. But neither Jansen nor the government said what that weekly average was, or, for that matter, what total dollar amount is represented by the notes.99 They *963then multiplied that unspecified weekly average times fifty-two to arrive at an annual average. Id. They then assumed a 6.5-year time period {e.g. seventy-eight months).100 Id. This results in the $17,432,674.40 extrapolation. Tr. 531. Jansen expressly disclaimed any ability to come up with the $17,432,674.40 figure on the basis of GX 325, the only source document cited by the government in support of this extrapolation: “That was taken from some analysis done by the FBI.” Tr. 589. Jansen looked at the notes and “accepted the analysis that was done.” Tr. 590. He had no idea what the A symbol meant in the notes. Id. What the court has from Jansen in support of this $17.4 million extrapolation is his version of what an unidentified FBI analyst had to say about notes given to Myler by an unknown person.
Thus, the court does not have before it, either by way of testimony or proposed findings of fact, any combination of numbers leading to an extrapolation to $17,432,674.40,101 let alone a satisfactory basis for a finding as to any weekly or annual average as the base number for a (questionable) seventy-eight month extrapolation. And the $17,432,674.40 — ipse dixit though it is, without evidence as to how a weekly average can be reliably extracted from GX 325 — is not even Jansen’s ipse dixit.
But there is a combination of numbers that leads to an interesting — and perhaps telling — result. When required to do so on cross examination, Jansen expounded on the $17,432,674.40 extrapolation as follows:
Q. [by Mr. Gerson] Okay. For the 17 million, you’re telling the judge that somehow these notes represent $50,000 a week that involve Kelley Diebner. And I’m using “involved” right now, not — don’t go further.
A. Yes. These numbers were extracted from these 29 handwritten notes to arrive at — at a number whieh was roughly $1.4 million divided by 29 weeks to arrive at that 50-plus thousand-dollar average per week times 52 weeks.
Tr. 589.
Although neither Jansen nor any other witness has pointed out (or even claimed to know) how GX 325, the twenty-nine weeks’ worth of notes, supported a $50,000 per week beginning number (or any other beginning number), Jansen, as can be seen, unflinchingly supported that number in support of the FBI’s $17.4 million extrapolation. Well, $50,000 per week for seventy-eight months (using the usual rule of thumb that, on the average, there are 4.3 weeks in a month) comes out exactly to the amount of Lowrance’s extrapolation: $16,770,000.00. Jansen may well have gotten the $50,000 number from “the FBI,” arid he may well have been content to testify to it, but it does not support any math that leads to $17,432,674.40.102 And *964all of this is aside from (i) the fact that the seventy-eight week extrapolation period has not persuasively been supported, and (ii) the fact that the government, as has been noted, has not attempted to explain or account for the inevitable trends and variations within the proposed extrapolation period.
The $17.4 million extrapolation, like the $16.7 million extrapolation, is not the stuff of which forfeiture judgments are made. The court rejects it.
Nothing other than the two extrapolations the court has rejected was offered by the government, at the forfeiture trial, by way of attribution of betting proceeds to Diebner. By some standards, that should be the end of the matter. But the government offers a fallback approach in its proposed findings of fact. Doc. no. 2242, at 40-42. Those matters were not argued at the forfeiture trial and the court is disinclined to work all the way through the government’s fairly elaborate newly-constructed contentions. Considerations of fairness, among other things are in play here. However, there is some low-hanging fruit which the court will pick for the government.
As has been noted, the Cabreras (Sergio Cabrera and his son, Sergio Cabrera Pine-da) were major betting customers of Dieb-ner. Both of the Cabreras testified at the May trial. The Cabreras own eighteen restaurants in the Houston area. May Tr. 735. With Diebner as their Legendz • bookie, they bet on basketball, football and baseball. Id. 736. Sergio would place bets by calling a 1-800 number; his son placed bets online. Id., 737, 739. They would place as many as twenty bets a week. Id. 737. For the most part, they paid their losses by check. Id. 758.
As evidenced by checks received in evidence at the May trial, the Cabreras’ losses totaled $787,100.00. GX 887A-887T. This will be included in the forfeiture money judgment calculation. Diebner argues that his forfeiture money judgment exposure is limited to his “personal financial gain.” Doc. no. 2274, at 7. Some of the other forfeiture defendants also advance this personal financial gain argument, citing United States v. McGinty, 610 F.3d 1242 (10th Cir. 2010), for the proposition that “forfeiture is based on the offender’s gain.” Id. at 1247 (internal quotation and citation omitted). This could make a substantial difference depending on what the bookie’s profit and loss sharing arrangement was with King. But the court does not agree that, in this context, forfeiture is capped by the amount of a defendant’s personal financial gain. McGinty involved a conviction on one count of misapplication of bank funds. The Court of Appeals, in the passage relied upon by the defendants, was comparing restitution concepts (restitution being based on the victim’s loss) with forfeiture concepts (based on the offender’s gain), because the trial judge had treated the two as redundant, resulting in the government’s appeal in that case. In reversing the district court, the Court of Appeals was, in the relevant passage, making the point that restitution and forfeiture have differing legal underpinnings and serve differing purposes. McGinty does not stand, at least in the context of the facts of the case at bar, for the proposition that forfeiture is limited to the offender’s gain. As has been discussed at length in this order, in the case at bar, the search for “proceeds” or for funds “involved in” the offenses of conviction is not limited by any defendant’s personal gain. The entire *965$787,100.00 will be included in the forfeiture money judgment calculation.103
Michael Gaplan, another Diebner betting customer, credibly testified that he lost about $5,000 betting on football and basketball with Diebner, using the 1-800 number Diebner gave to him. May Tr. 799-800. That will be included in the forfeiture money judgment calculation, for a total of $792,100.00 attributable to Diebner.104
d. Christopher Tanner
Tanner was a Legendz bookie in Florida. For Tanner’s activities, implicitly attributable to Tanner’s Legendz-related bookmaking, the government proposes a base dollar amount of $2,528,985.00. Working from that base number, the government proposes an extrapolation to $15,143,910.00. The $15.1 million extrapolation is based on the government’s spreadsheet (GX C71), which, in turn, is based on a source document consisting of handwritten notes (GX 528) seized at Tanner’s home on the day he was arrested. Tr. 336. GX 528 was admitted into evidence in the February trial. Feb. Tr. 1785.
Lowranee did not know who prepared the government’s spreadsheet supporting the Tanner extrapolation. Tr. 336-37. Whoever prepared that spreadsheet, Lowranee knows nothing of his or her ability to interpret Tanner’s handwritten notes. Tr. 339. And Lowranee, himself, has never reviewed the records on which that summary is based. Tr. 336-37. Those notes (GX 528) are hardly self-explanatory. Although the government’s spreadsheet purports to reflect activity in 2009, Lowranee could not say whether or not the spreadsheet reflects money collected in 2009. Tr. 339. He did not know what Tanner’s arrangement was with King, and he was not aware of whether the unidentified person who prepared the government’s spreadsheet ever contacted Tanner to verify the accuracy of the spreadsheet or the extrapolation based on it. Tr. 339. (Tanner entered a guilty plea in this case and was sentenced, on the basis of a Rule 11(c)(1)(C) plea agreement that was approved by the court, to ten months of probation.)
Lowranee could not shed any light on whether the betting activity reflected by Tanner’s handwritten notes was exclusive to the Legendz organization — a fair question, given the fact that the spreadsheet has a column of figures under the heading of “5 Dimes,” an offshore sports betting organization wholly unrelated to (and in competition with) Legendz. Tr. 337; Feb. Tr. 139, 483, 1115, 1245; May Tr. 122. Consequently, Lowranee conceded that, for purposes of this case, a spreadsheet reflecting activity associated with 5 Dimes *966would not be accurate.105 Tr. 338. “5 Dimes” appears in the header on GX C71 along with fourteen other names. Potentially as significant, the header also includes several other names that have been not been mentioned in the five trials in this case as having any association with Le-gendz. The activity shown in the columns under those other names may or may not be associated with Legendz — there is no evidence on that score one way or the other. The lack of any explanation of those names, cheek by jowl with 5 Dimes, does nothing to give the court confidence in the base number supporting the government’s proposed extrapolation of $15 million in Legendz-related activity to Tanner. (Jan-zen made no substantive reference to Tanner in his testimony, and consequently did nothing to rectify these deficits in Low-rance’s testimony about Tanner.)
As for the extrapolation from the $2.52 million base number shown on the government’s spreadsheet, Lowrance cited nothing to support the government’s proposed extrapolation period, let alone account for trends or variations during the extrapolation period that would affect the reliability of the $15.1 million extrapolation.
There is no need to dwell on the extrapolation proposed for Tanner, because the base number has not been substantiated as a reliable reflection of Tanner’s Legendz-related activity, and only that activity, for 2009 or any other period. There is no evidence as to what the numbers on the spreadsheet actually reflect. One of the columns is labeled “CT,” which probably stands for Christopher Tanner himself. There is no explanation as to how the dollar amounts in the column relate to the rest of the spreadsheet. There is no explanation (nor is one conceivable) as to why the base number proposed by the government should be predicated, in part, on activity under the heading of “5 Dimes,” a competitor of Legendz. There is no explanation as to whether any of the other names across the header (e.g., aside from 5 Dimes) are also Legendz competitors. Some of the names, like “Margarita” and “Ringo,” appear to be names of individuals, but that is not true of most of the names. Lowrance acknowledged that all he did was “some math,” and it is clear that the only math he did was to take the base number that was given to him and multiply it by the extrapolation period that was given to him. The government’s proposed findings cite nothing to rectify any of these deficits. Doc. no. 2242, at 17.
The government has failed to establish, even on a preponderance standard, a reasonably reliable dollar amount to attribute to Tanner’s Legendz-related activity.
e. Michael Lawhorn
The government seeks to add exactly $18,000,000.00 to the forfeiture money judgment by way of attribution of betting proceeds to Michael Lawhorn. GX C89. The mathematical building blocks of this extrapolation are exceedingly simple. At the October trial, Lawhorn, a successful master agent and runner in Florida, testified, as relevant here, as follows:
Q. Now, how much money as a bookmaker did you make before you joined Legendz?
A. Hundred, 150 maybe, thousand a year.
*967Q. Okay. And what about after you joined up with Legendz?
A. Some years a million or so, million dollars.
Q. Was that an average per year?
A. Yeah, that’s an average throughout the thing, a million a year, uh-huh.
Oct. Tr. 1135.106
On this basis, Lowrance elaborated as follows at the forfeiture trial:
Q. Looking at Michael Lawhorn’s entry there, can you explain how you arrived at $18 million.
A. Yes. Mr. Lawhom testified that he had a 50 red deal with Luke Bang. Which was explained to me as a 50/50 split except when Mr. Law-horn’s clients won and Mr. King had to pay them and then Mr. King was reimbursed before the split continued. So basically it was a 50/50 split. And Mr. Lawhorn testified that he made a million dollars a year roughly on average during his time involved in the Legendz enterprise. So we just took that — if he was making a million dollars a year, that meant Mr. King was making a million dollars a year. So we multiplied that out by two for the nine-year time period for an $18 million money judgment.
Q. So then that 18 million is also an extrapolation over the nine-year period?
A. Oh, yes.
Tr. 144.
The nine-year extrapolation period is not disputed (and may, in fact, be conservative). The “50 Red” arrangement was explained at all four of the jury trials, and in the most detail at the October trial. Oct. Tr. 278-79. Essentially, it is a 50/50 split, between the local bookie and the house (Legendz) of bookmaking wins (e.g., bettors’ losses) and bookmaking losses (e.g., bettors’ winnings), with a gradual recoupment by Legendz of the money the local bookie owes as his half of the bettors’ winnings. In its proposed findings, the government repeats, without elaboration, the calculation set forth above. Doc. no. 2242, at 19, ¶¶ 73-75.
The government cites no other testimony to support this $18 million figure. Strictly speaking, it is an extrapolation only in the sense that the government, relying on Lawhorn’s laconic estimate (quoted in full above), posits a $1 million dollar a year figure, on a “50 Red” basis, yielding $18 million in proceeds over nine years (108 months).
But the situation with Lawhorn is not as uncomplicated as the government’s approach would suggest.
Lawhorn was indeed a successful master agent107 for Legendz. At one point, he had “around 40” agents working for him. Oct. Tr. 1237. His million-dollar-a-year estimate, and the government’s assumption *968that estimate really is reliable for a nine-year period, glosses over some complications in Lawhorn’s story. These complications, to be discussed, are not adequately clarified in Lawhorn’s jury trial testimony, probably because there was no need, in those trials, to smooth the record out for forfeiture purposes.
The first complication arises from the fact that Lawhorn, in giving his answer, and the government, in relying on that answer, are glossing over the fact that there were significant interruptions in Lawhorn’s day-to-day activities as a local bookie or master agent for Legendz in Florida. In 2003, Lawhorn went to Panama to serve as a trainer for Legendz for “four or five months”. Oct. Tr. 1130, 1177. This may not be within the nine-year extrapolation period (that is not completely clear), but it is an example of a pattern of interruptions in Lawhorn’s activity in Florida as a local bookie or master agent. Lawhorn went to Panama again in 2008. Oct. Tr. 1130. By Lawhorn’s account, he went to Panama “very often.” Id. at 1131. In fact, “I became a resident at one time.” Id. “I moved to Panama.” Oct. Tr. 1154. Among other activities, Lawhorn “was running the bar” at the Super-book (discussed in Part IX(B)(1), above) in Panama, an activity that is not obviously consistent with working as a master agent in Florida. Oct. Tr. 1157. As a trainer in Panama, Lawhorn “taught people how to accept bets” and “helped work on the stage when the bet-tings would come in,” for which he was paid “750 a week.” Oct. Tr. 1168-69.
Lawhorn also served as a runner for Legendz. Oct. Tr. 1134. As defendants Robles and Moran will attest, that is a decidedly less profitable role than service as a bookie or master agent. The impact of that on what we otherwise might assume was a million-dollar-a-year pace as a master agent is not explained.
One mathematical element of the government’s extrapolation as to Lawhorn is the proposition that he made his money (and generated proceeds for Legendz) on a “50 red” basis, as has been discussed. But, for some undetermined period during his association with Legendz as an agent or master agent, Lawhorn made book on a “pay-per-head” basis. Oct. Tr. 1266. In its order denying the post-trial motions of the defendants Moran, Diebner and Bramley, the court explained the pay-per-head system:
As a PPH [pay-per-head] bookmaker, Bramley paid a flat weekly fee (on a per-bettor basis) to Data Support Services (based in Panama, and beneficially owned by defendant Bartice King) for ‘back office’ services. Thus, if Bramley had 1000 bettors in his book of business in a given week, and the agreed PPH fee was $18, Bramley would owe $18,000 to DSS for its services for that week. As a PPH bookmaker, Bramley was in business for himself — he kept the losing bettors’ losses and was on the hook for their winnings. Wagering risk was not shared with, or otherwise borne by, Le-gendz.
United States v. Bramley, 2015 WL 12852048, at *3 (W.D. Okla. Sept. 16, 2015).
Lawhorn’s pay-per-head fee was $18 per bettor. Oct. Tr. 1266. His explanation of his pay-per-head arrangement was exactly the same as quoted above with respect to Bramley: “So if I had 200 active clients and only a hundred played, I would only pay for that hundred clients. So it would be a hundred times $18. So I would pay $1,800 for that week.” Oct. Tr. 1320. On a pay-per-head basis, all Legendz got was the per-head fee. Id. at 1327. Obviously, during the unspecified period that Law-horn operated on a pay-per-head basis, one of the basic premises for Lowrance’s ex*969trapolation (“[A] 50/50 split except when Mr. Lawhorn’s clients won and Mr. King had to pay them and then Mr. King was reimbursed before the split continued”) did not apply. To be sure, the total proceeds (presumably all going to Lawhorn, except for the $18 fee) may have been the same, and that is all the government needs to be concerned with for forfeiture purposes. But it is equally easy to envision Lawhorn operating more conservatively when he was bearing all of the risk. None of this is explained in the testimony.108
Even without these complications, the government’s beginning point — facile testimony from which an equally facile extrapolation is drawn — requires somewhat of a leap of faith. Regardless of whether the court might otherwise be willing to make that leap, the complications arising from Lawhorn’s sojourns in Panama (running the bar, training the staff) and from his ventures into pay-per-head bookmaking on a 100 percent risk-bearing basis, undermine the court’s confidence in the government’s proposed $18 million extrapolation.109 These variations in Lawhorn’s activities while he was associated with the Legendz organization were of very little moment for purposes of the jury trials, but they could make a major difference in any attempt to reliably extrapolate from his testimony in those trials for forfeiture purposes. And as was the case with the Terry Campbell extrapolation, the court is disinclined to adjudicate the proposed Lawhorn extrapolation by picking a reduced extrapolation just to be on the safe side. The record does not provide a reasoned basis, grounded in reliable evidence, for picking a lower number by way of extrapolation, even though it is plain that Lawhorn generated a substantial amount of . illegal gambling proceeds for the Legendz organization. Cf., Lyons, 870 F.Supp.2d 281, 287, where the government’s proposed extrapolation was rejected because of uncertainties that were not adequately explained or accounted for. As has been discussed, the government is not required to prove up a dollar amount for a forfeiture money judgment with to-the-penny precision (or even anything close to that). But the party with the burden of proof is obliged to present a reasoned basis, and the judge is obliged to arrive at a reasoned basis, for a decision. Picking a number somewhere short of 18 million just because the true number surely is in there somewhere is not, by any reasonable measure, a comprehensible method of adjudication. At least for forfeiture purposes, the burden of proof entails more than just proving that some generalized factual concept is more probably true than not.110
The Lawhorn extrapolation is rejected.
*970f. Joseph Barry
As to Joseph Barry, who was neither a defendant nor a witness at any stage of this case, the government’s extrapolation to $15,275,612.57 is based on the $3,960,344.00 shown on GX C55, a government spreadsheet. Tr. 72-73. GX C55 shows a 28-month period of activity attributed to Barry (September, 2007-Decem-ber, 2009). The government divides that $3.9 million by 28 and multiplies it by 108 (months) to arrive at the $15.2 million extrapolation as to Barry. Government Exhibit C55, the government spreadsheet on which the government’s extrapolation is based, was not prepared by either of the government’s forfeiture witnesses. Tr. 344. Lowrance has no knowledge that would support the extrapolation period of 108 months as to Barry. Tr. 345-46.111 (Jansen did not mention Barry in his testimony.) The four jury trials in this case yielded not a single mention of Joseph Barry. The government cites no evidence that would support its 108-month extrapolation period. For reasons which should, by now, be too obvious to require elaboration, the court declines to take the government’s proposed 108-month extrapolation period on faith.
This extrapolation is rejected. (The government proposes no alternative attribution to Barry based on discrete transactions. Doc. no. 2242, at 17-18, ¶¶ 67-68.)
g. Paul Tucker
By way of extrapolation, the government seeks to include $7,958,304.00, attributable to the activity of Paul Tucker, in the forfeiture money judgment against all of the forfeiture defendants. GX C89 (App. 1). This is based on a source document (GX 546), showing Tucker’s activity from January 5, 2009 to March 30, 2009, generating $221,064.00 in betting proceeds. GX C86, p. 7. That calendar quarter is then multiplied by four (to annualize) and then by nine (years) to arrive at the $7.96 million attributed to Tucker, the premise being that “the volume of bets is uniform across the year.” Tr. 415 (Lowrance).
Thus, the government proposes to use that January through March calendar quarter thirty-six times over for purposes of its extrapolation. In support of that approach, the government solemnly represents to the court that: “This time period would not represent prime betting activity.” Doc. no. 1999-1, at 15. In other words, never mind the NFL conference championship playoffs, the Superbowl, the NCAA *971Bowl Championship Series,112 and the NCAA Division I Basketball Tournament commonly called March Madness. To his credit, Jansen acknowledged that “it’s very much a possibility” that the three months in this first calendar quarter, held out by the government as not representing prime betting activity, “are the busiest time periods for a bookmaker.” Tr. 655. The government’s extrapolation period as to Tucker, nine years, is probably correct. And, as is discussed below, the court finds that the $221,064.00 base number is reasonably reliable as a measure of Tucker’s bookmaking activity in the first quarter of 2009. But the extrapolation fails for two reasons. First, the notion (asserted by the government in writing but debunked by Jansen in his testimony) that the first quarter of any year is a typical quarter to use as an exemplar of sporting activity (especially the type of sporting activity that draws wagers) for the entire year is untenable. Second, and more generally, the use of a three-month period as the base period for a 108-month extrapolation is simply a bridge too far. Too many things have to be presumed in favor of the government for that to be a reasonable extrapolation. See, Lyons, 870 F.Supp.2d 281, 287, where a much more conservative extrapolation was rejected for the same reason.
This leaves the question of whether there is a basis for a reliable attribution of betting proceeds to Tucker on the basis of discrete transactions. The government, mindful of the court’s invitation to propose alternative attributions of betting proceeds based on discrete transactions, does proffer an alternative attribution as to Tucker. Doc. no. 2242, at 45-46, ¶¶ 175-183. And Tucker would seem to be a good candidate for an alternative attribution. He was a successful Legendz bookie in Florida for a long time, and the government was able to get some records reflecting some of that activity.
As an initial matter, it is difficult to determine what the total for the government’s proposed alternative attribution is. GX C82 would suggest that it is $2,225,099.20, net of the amount ($183,-042.18) that has been forfeited by agreement. But the government’s proposed findings posit $2,265,238.00, with no explanation (or even acknowledgement) of the difference between that number and the number in the government’s exhibit.
At the forfeiture trial, Jansen was the government witness who was answerable for the alternative attribution, and he deserves credit for his candor.
One element of the government’s alternative attribution is $80,689.00 extracted from “Gooch records of 12-16-07 to 11 — 9— 08.” GX C82. Jansen acknowledged that this amount erroneously includes $2,700 that did not actually represent proceeds because the check by which that amount was purportedly paid bounced. Tr. 663. Jansen also acknowledged that it is possible that other source documents for GX C82 would show that other checks had bounced. Id. In fact, another returned check was pointed out. Tr. 664.
Another component of the alternative attribution is $1,477,561.53 shown to have been derived “From Legends Spreadsheets.” GX C82. Thus, this compilation is from internal Legendz documents— which is significant because the Legendz organization included “Pauls” other than the defendant Paul Tucker (a fact of which Jansen was unaware — Tr. 665). The government spreadsheet from which *972this $1.48 million amount is derived shows that any variation of “Paul” in the Legendz source document resulted in the attribution of the transaction to Paul Tucker, even though there were specific entries labeled “Paul,” “FPAUL” and “Paul (FLA)” in the source documents from Legendz. E.g., GX 62, from which the $1,477,561.53 figure in GX C83 and GX C82 is derived. Jansen acknowledged, after being shown that there was at least one other “Paul” in the bookie network, that there would be no reason to designate a person “FPAUL” or Paul Florida if those were the same person as “Paul.” Tr. 666. This potentially makes a very substantial difference in the attribution of proceeds to Paul Tucker, of Florida. In its proposed findings, filed more than two months after the forfeiture trial (and with the trial transcript in hand), the government ignores these discrepancies altogether. Doc. no. 2242, at 46-47.
Other components of the government’s alternative attribution are outlined on a summary exhibit as to Tucker (GX C82—4th through 8th lines), but were not the subject of any testimony at the forfeiture trial or at any of the jury trials. They have not been tested in any way, mainly because the government has made no attempt to prove them in the first instance.
But there is a residuum of evidence that reasonably supports an attribution of proceeds to Tucker. Even though the court has rejected the government’s proposed extrapolation based on Tucker’s bookmaking activity in the first quarter of 2009, the court is satisfied that base number ($221,-064.00), drawn from records seized from Tucker’s own computer, is reliable for that period. GX 546; Tr. 565.113 The evidence satisfactorily supports the attribution of $221,064.00 in illegal gambling proceeds to Tucker.
h. Bruce Middlebrook
On the basis of GX 1134H as a source document (and no other document — see, doc. no. 2242 at 16, ¶ 60 and GX C89), the government proposes to attribute (by extrapolation) $4,449,353.14 in gambling proceeds to Bruce Middlebrook for inclusion in the forfeiture money judgment against all of the forfeiture defendants. The source document, GX 1134H, was found on Mid-dlebrook’s computer. Oct. Tr. 362.114 That source document drew only two questions on direct examination of Lowrance. Tr. 147-48. (Jansen had nothing to do with this extrapolation, and had never seen the source document before. Tr. 728-29.) Government Exhibit 1134H starts in September (year not specified) and ends in March (no year specified). Consequently, Low-rance testified that “we took this as seven months’ worth of records.” Tr. 347. (In light of the fact that Jansen had never seen the source document, it is not clear who is included in Lowrance’s “we.”) That seven months of activity generated the $288,384 number, but exactly what that represents is not completely clear. That number, divided by seven and multiplied *973by 108, leads to the $4,449.353.14 extrapolation.115 Tr. 146.
Lowrance did not know the years in which the seven-month period shown on GX 1134H began and ended, but he did allow that the September through March period “would be the busy time of year for a sportsbook.” Tr. 348. Although Middle-brook was a cooperating defendant and a witness for the government, neither Low-rance nor (to his knowledge) anyone else did anything to confirm with Middlebrook the accuracy of the figure to which the government proposes to extrapolate from the somewhat cryptic GX 1134H. Tr. 348-49. Consequently, even though Middle-brook lives in this locality, appeared as a witness for the government on three separate days during the October trial (a few weeks before the government’s Forfeiture Brief was due), and has -been readily available to the prosecution team, the government’s $4.4 million extrapolation as to Middlebrook is “uncorroborated by any conversations with Mr. Middlebrook.” Tr. 349. In this instance, as with some others discussed in this order, if this proposed extrapolation were a standalone $4.4 million lawsuit pending in this or any other court, the notion that it is not necessary to talk to the one person who is in the best position to confirm or cast doubt on the $4.4 million number would be unfathomable. This is especially true when the matter is viewed in light of the fact that Lowrance did not even know what year the base numbers were drawn from.
The government’s proposed extrapolation based on the record taken from Mid-dlebrook’s computer is rejected. The difference between the length of the base period and the 108-month extrapolation period is just too much — especially so, given the fact that the actual period (what years?) covered by GX 1134H is not known. The lack of any indication as to the actual period encompassed by the source document adds an element of speculation to the proposed extrapolation, but that is a matter separate from the sheer length of the proposed extrapolation period vs. the base period.
Although the government proposes no alternative attribution to Middlebrook, based on discrete transactions, the court has considered whether the base number for the rejected extrapolation is, itself, a reliable number for dollar-for-dollar attribution as to the period covered by GX 1134H, whatever that period may be. It is a close question, but the court concludes that the base number has been satisfactorily established and is fair game for attribution to Middlebrook. This is a close question because, as has been noted, Jansen had never seen the source document and thus could not support the inference the government would have the court draw from that document. As for what that source document actually represents, all Lowrance said is that Middlebrook’s “records for that seven-month time period totaled $288,384.” Tr. 146. In keeping with his general tendency to stay within the realm of what he actually knew, Lowrance *974did not expand on that statement. Thus, the forfeiture trial did not shed light on what subset of betting-related activity is actually represented by the $288,384 number, or, indeed, on how that number was derived from GX 1134H. At the October trial, Middlebrook described a very similar source document in the same series (specifically, GX 1134A) as, essentially, a routine weekly “running sheet” of his group. Oct. Tr. 363. The import of that document was not challenged at the October trial, and Lowrance’s somewhat fleeting characterization of the very similar GX 1134H was not challenged at the forfeiture trial. By the narrowest of margins, the court finds that the $288,384.00 in betting-related activity shown by GX 1134H should be attributed to Middlebrook for forfeiture purposes.
i. Rodger Bramley
As the court has discussed at length in another order, there are some substantial legal issues afoot as to Bramley’s criminal liability in this case. See, United States v. Bramley, 2015 WL 12852048 at **5-6 (W.D. Okla. Sept. 16, 2015). But one thing that is not in doubt is that Bramley was a successful and productive bookie in the Dallas area.
The government proposes to attribute $4,953,478.78 to Bramley. This is not an extrapolation. This attribution, as proposed by the government, is the total of the dollar amounts gleaned from (i) eleven types of records purportedly relating to Bramley’s activities as a Legendz bookie, and (ii) the trial testimony of five witnesses. See, GX C81 (drawn “from ten years’ worth of documents” — Tr. 621). Lowrance did not substantively address this attribution. The government’s witness on this one was Jansen.
Not all of the sixteen components of this attribution need extended discussion. Most of them are well-supported (and either not challenged at all, or not persuasively challenged, by defendants). The ones that need discussion will now be addressed.
The reckoning as to Bramley begins with $967,847.00 extracted from internal records of Legendz, reflecting checks attributable to “Doc,” which is Bramley, received by Legendz. (Within the Legendz organization, Bramley’s moniker was “Doc.” There is no confusion as to whether that moniker might refer to more than one individual.) At the forfeiture trial, it was suggested, on cross examination of Jansen, that these payments “did not end up in Mr. Bramley’s coffer.” Tr. 613. They did not end up in his “coffer,” else they would not have been received by Legendz in Panama. But whether these payments ended up in Bramley’s coffer is, in any event, of no moment for forfeiture purposes. Bramley also raised the possibility that these payments represent “pay-per-head” fees paid by Bramley to Legendz. Tr. 615. If they are, that also makes no difference — if the court is correct that Bram-ley’s participation in the Legendz operation as a pay-per-head bookie is irrelevant to his criminal liability. United States v. Bramley, 2015 WL 12852048, at *3 (W.D. Okla. Sept. 16, 2015). This $967,847.00 will be included in the forfeiture money judgment calculation.
Based on GX C43.3, and the source documents for that summary, the government includes $117,164.00 in its total as to Bramley. This consists of checks deposited into one of Bramley’s accounts at ViewPoint Bank. GX C81. Jansen did not know “how much of that money within the ViewPoint account comes from legitimate sources versus illegitimate sources.” Tr. 640. That makes no difference for presént purposes because the court finds by a preponderance of the evidence that these checks (in evidence as GX 871) were from bettors. These dollars were betting pro*975ceeds. The fact that various of Bramley’s accounts did receive deposits from lawful sources will be of consequence when the court addresses the issues as to forfeiture of specific assets owned by Bramley. This relates to the government’s “taint up” approach to some aspects of these forfeiture proceedings. But these dollars were betting proceeds and are fair game. This $117,164.00 will be included in the forfeiture money judgment calculation.
The government proposes to include $117,968.00 in the forfeiture money judgment, attributable to Bramley, consisting of bettor checks he used to service the mortgage on his house. GX 871; Tr. 489. This is a matter distinct from the question of whether the house itself is forfeitable (to be discussed later). This $117,968.00 total will be adjusted to $114,882.41 because of a double counting problem (pointed out by the government — Tr. 522-23; 635-36). This $114,882.41 will be included in the forfeiture money judgment calculation.
Citing twelve pages of the testimony of David Pray, a witness at the May trial, the government would add $256,500.00 to the forfeiture money judgment. GX C81. That Pray was a bettor in Bramley’s stable is not in doubt. As would be the case with any witness whose testimony is touted as the basis for calculating a dollar amount of betting proceeds, the building blocks- for determining the amount of the betting proceeds generated by Pray, attributable to Bramley, are: beginning and ending dates, frequency of bets, and dollar amounts of bets. Those building blocks, if present, would yield a raw number. That raw number must then be tested for reliability, bearing in mind that vague and uncertain testimony can undermine the reliability of a calculation based on that testimony even if, as the court presumes with Pray, the witness is doing his best to tell the facts as he recalls them.
By Pray’s account, he began betting with Bramley “sometime in the early 2000, certainly before the fall of 2005.” May Tr. 1054. That is not a good start for a beginning date for a quarter of a million dollar calculation. Pressed to sharpen that up, he said: “this is a guess, but it’s probably in the two-to three-year range” before 2005. Id. 1054-55. When asked about the period of time in which he placed bets with Bram-ley, he would only say: “So assuming that I started gambling in 2003, I bet continuously until recently.” Id. At 1060-61 (emphasis added).
As for an ending date, Pray said that he was directed to another website, called lookitup.net, in August of 2014. Id. at 1059. Legendz used several names for websites at various times, but the court does not recall lookitup.net as a Legendz website. Nor is it clear whether the government’s calculation of $256,500.00 on the basis of Pray’s testimony is based on August, 2014 as an ending date. Pray estimated that he placed bets nineteen weeks of the year, basically the NFL season. Id. At 1062. His estimate for his betting volume was “generally 1,500 to $2,000 a weekend.” Id. At 1063.
At the forfeiture trial, neither of. the government witnesses explained how the government arrives at $256,500.00 as to Pray. The government’s proposed findings shed no light on this. Doc. No. 2242, at 35, ¶ 131. The record before the court simply will not bear the weight of a $256,500.00 calculation on the basis of Pray’s vague and uncertain testimony. This item is rejected.
The government’s calculation of $200,000.00 in betting proceeds attributable to Mark Polan is rejected for similar reasons. True, he estimated that his losses amounted to $200,000. He put it this way: “Probably under $200,000, close to, probably.” May Tr. 791. His uncertainty may *976well have been based on the fact that he could not recall with any certainty when he started betting: “My best guess would be maybe eight years ago.” Id. 785. Even on a preponderance standard, this is too flimsy. This calculation is rejected.
Citing two pages from the testimony of Neil Myler, a Legendz runner, at the May trial, the government proposes including $1,200,000.00 in the attribution of betting proceeds to Bramley. This is based on the number of times that Myler picked money up from Bramley, as the multiplier for the amount of money picked up each time. As for the number of pick-ups, Myler said: “I reckon between 30 and 40 times.” May Tr. 1527. Granting that Myler was not a disinterested witness, his estimate is believable. Myler estimated that the amounts involved in these collections ranged from “40” to “175,” which the court presumes to be thousands. May Tr. 1528. The government’s $1,200,000 calculation, though not explained at the forfeiture trial, appears to be based on the low end of the estimate of the dollar amount ($40,000), multiplied by the low end of the estimate of the number of collections (30). The court finds, by a preponderance of the evidence, that this calculation is sufficiently reliable to be included in the forfeiture money judgment calculation.116 The court is mindful that it has rejected an extrapolation, relating to Diebner, that is based in part on Myler’s dealings with Diebner. But, as is discussed in that section of this order, the court was asked, in that instance, to make a $16.8 million extrapolation on the basis of Low-ranee’s description of what he recalled to be the substance of the relevant portion of an FBI 302 interview report. That left the court with no sound basis for evaluation of Myler’s account of that matter (or even an opportunity to determine with certainty whether Myler actually said what was attributed to him). This $1,200,000.00 will be included in the forfeiture money judgment calculation.
One other instance of likely double counting, involving $4,760, was pointed out by the government at the forfeiture trial-candor which the court appreciates. Tr. 522. The remainder of the items comprising the government’s calculations (GX C81) as to betting proceeds attributable to Bramley are well-supported.117 After deducting the components of the government’s calculation that the court has rejected, the amount of betting proceeds attributable to Bramley is $4,489,133.19.
j. Paul Wilson
The government originally sought, by extrapolation, to attribute $4,000,000.00 in betting proceeds to Paul Wilson, a Le-gendz bookie in California. GX C89. But now, the updated total of proceeds attributable to Wilson is $3,907,791.00. GX C89 (“Updated”); Tr. 149. This is an extrapolation from $1,157,864 attributable to Wilson’s activity as the government says is reflected on an internal Legendz source document, GX 61.
The government asserts that GX 61 shows thirty-two months of activity, from February of 2008 to August of 2010 (Tr. *977150; doc. no. 2242 at 19, ¶ 76), so the government divides the beginning number by thirty-two and multiplies it by 108, to arrive at the $3.9 million extrapolation. Tr. 150, 349-50. Although Wilson, a cooperating defendant, testified as a government witness in each of the four jury trials in this case, Lowrance did not discuss this $3.9 million extrapolation with Wilson and received no confirmation from Wilson as to its accuracy. Tr. 350. This is significant for several reasons, not the least of which is that there is no reading of GX 61 that will square with Lowrance’s testimony that it supports a thirty-two month base period from February of 2008 to August of 2010 (or any other thirty-two month period, or, for that matter, anything close to a thirty-two month period). Lowrance — maintaining that he performed this calculation himself (Tr. 349) — confirmed this thirty-two month base period no less than three times. Tr. 150, 350.
It is tempting to try to repair Low-rance’s calculation (which is repeated, to the dollar, in the government’s proposed findings), but the court declines to do so, for a variety of reasons. The government’s proposed extrapolation as to Wilson is rejected. The government proposes no alternative extrapolation based on discrete transactions.
k. Leon Moran
Leon Moran was a Legendz runner in California. Oct. Tr. 1243. On the basis of an extrapolation from an internal Legendz document, GX 473, the government seeks to attribute $6,491,205.00 in betting proceeds to Moran. GX C89. Alternatively, the government proposes to attribute $3,218,914.00 to Moran by drawing inferences from specific items of evidencé. Both approaches are without merit, for reasons to be discussed.
The proposed $6.5 million extrapolation fails for several reasons.
First, the document on which this extrapolation is based, GX 473, is a spreadsheet reflecting betting activity. Tr. 578. Regardless of whether it relates to Moran, all it reflects is betting activity, not the activities of a runner. Although some participants in the Legendz enterprise worked in both capacities (e.g., Michael Lawhom, as has been discussed), the two functions are decidedly different. The grand jury found probable cause to believe that Moran was a runner and indicted him on that basis — in a Superseding Indictment that specifically recognized the various roles of the participants in the Legendz enterprise, designating “Executive Staff,” “Agents,” “Runners” and “Bookies.” Doc. no. 354 at 4-9. The court has held the government to that designation. See, Order of May 18, 2015, doc. no. 1443, at 1-2 (“For two distinct reasons, one based on the Fifth Amendment and the other based on the Sixth Amendment, Moran may not be prosecuted as an agent in this case.”) The government provides no argument, developed or otherwise, in support of the implicit proposition that the court should, at this point, adjudicate forfeiture as to Moran on a basis materially differing from the basis presented to the grand jury, found by the grand jury, and set forth in the Indictment. There may be an argument for disregarding, for forfeiture purposes, the differentiation that the court enforced at an earlier stage of this case. But the government has not made that argument, if there is one.
Second, the court has not been persuaded that Moran is the person referred to in GX 473, the document on which the $6.5 million extrapolation is based. The moniker used in GX 473 is “Mastiff.” Moran’s moniker as a runner was “Makavelli.” May Tr. 569. Not even Karlo Stewart, a Legendz insider and a highly cooperative government witness, could, when specifically *978asked, associate “Mastiff’ with Moran. May Tr. 569. In Moran’s jury trial (the May trial), no witness associated “Mastiff’ with Moran. No witness in the February or October jury trials associated that moniker with Moran. In the April trial, the only testimony associating Moran with “Mastiff’ came by way of the answer of an FBI witness to a leading question from an AUSA. In that trial, the same witness, extraordinarily compliant with leading questions from government counsel, gave false testimony on a contested factual issue in response to another leading question from an AUSA, as discussed at length in United States v. Bramley, 2015 WL 12852048, at *21 et seq. (W.D. Okla. 2015). (The court found — and continues to believe — that it was quite unlikely that he knowingly gave false testimony, but that is beside the present point.) That testimony from the April jury trial does not persuade the court that Moran is “Mastiff.” The short of the matter is that no witness who was actually in a position to testify on the basis of actual knowledge (and the government has called more than a dozen witnesses who were directly involved in the Legendz operation) has associated Moran with “Mastiff.” The $6.5 million extrapolation fails if Moran was not “Mastiff,” and the government has not persuaded the-court that he was.
Third, it is difficult to determine (and the government offers no help on this) whether GX 473 reflects the activity of a bookie, and not just a single bettor. The appearance of GX 473 is decidedly different from most of the bookie sheets that are in evidence. The only individual it refers to is Mastiff. If that person is Moran, then it is far from clear that this document reflects anything other than Moran’s personal betting activity. (A possibility that Jansen readily acknowledged. Tr. 754.)
For these three independent reasons, the $6.5 million Moran extrapolation .fails.
The government’s alternative approach to attributing betting proceeds to Moran is also without merit, for two reasons. First, as will be seen, the record thoroughly undermines any sense of confidence the court might otherwise have in the reliability of Jansen’s work, as his own work product, on the alternative calculations as to Moran. That alone compels rejection of the alternative calculation, but that also bears directly on the court’s conclusions as to the second reason for which the court rejects the alternative calculation — double counting.
The alternative calculation adds up to $3,218,914.00. GX C74. This total is the sum of thirteen separate items, some relatively complex, such as interpretation of betting sheets and win/loss spreadsheets, and some simple, such as looking at a figure scrawled on the front of an envelope and calling that betting proceeds— $187,200 in betting proceeds. GX 1401; Tr. 747.
On the first basis for rejection of the government’s alternative calculation, a short digression is necessary. Lowrance and Jansen were latecomers to this case. They are not the case agents. Jansen was the government’s witness in support of the government’s alternative calculation as to Moran. His advocacy of that calculation encompassed two indispensable functions: (i) recognizing and reliably evaluating the significance of source documents in the overall context of this case, and (ii) then accurately doing the math after determining that the source documents do mean something that should make a difference for forfeiture purposes. The court is left with the distinct impression that, in many important respects, Jansen’s testimony, necessarily conveying to the court the net result of those two functions, represented *979the intellectual product of persons other than Jansen. In a case as complex as this one, it is difficult to prove anything that way. For instance, some of the components of the government’s alternative calculation required review and understanding of voluminous internal Legendz documents, many of which were not intended to be understood by strangers — tedious and exacting forensic work, by most any standard. To be sure, some of the components of the alternative extrapolation were, on the surface, less complex, such as adding up the amounts of a set of checks. But even at that, the witness who would attempt to persuade the court that those numbers should make a difference in the outcome should, as a beginning point, be a person who is selling the court his ovm intellectual product, based on a thorough understanding of the reasons for which the numbers the government has put on paper should find their way into a forfeiture judgment.
The government’s alternative calculation includes $22,825.00, based on six internal Legendz betting records. GX C74 (referring to GX 58-64). Those internal records are betting spreadsheets. Perhaps those spreadsheets are not as cryptic as some other records in this case (such as individual bookie sheets), but they do leave room for interpretation. For the most part, they don’t indicate the year in which the transactions occurred. In some instances, there is room for interpretation, with the benefit of extrinsic knowledge, as to just what agent or bookie any given entry relates to. But the interpretation and the math for this item were not done by Jansen. That was done by Greg Melzer. Tr. 737. Melzer is a contract forfeiture investigator. Tr. 716. He did not testify at the forfeiture trial.
The alternative calculation includes $1,299,940.00 for “Odes Thompson checks per Legendz spreadsheets.” GX C74. The checks are GX 883A-883S. Four of these checks contain a reference to Paul Wilson. None of them refer to Moran, Makavelli or Mastiff. When asked the source of this calculation, Jansen responded: “Again, I believe that would have come from Greg Melzer.” Tr. 737. So, when Jansen was asked about any connection between one of these checks (for $5,000 — showing Le-gendz agent Paul Wilson, a/k/a Big Dog, as the remitter) and Moran, we had the following exchange:
Q. [by Mr. Corgan]: I’m lost. Does Makavelli, Moran, Mastiff go by Paul Wilson?
A. [Mr.‘ Jansen]: Not that I’m aware of, but I think there’s — and I don’t know this for sure, but I think there’s some connection between the money that Paul Wilson would have paid would have been picked up by Mr. Moran.
Q. Okay. And is this one of those checks?
A. I believe it is, based on it being incorporated in this exhibit.
Q. Oh, so because it’s incorporated, we should accept it as being linked to Mr. Moran?
A. Yes.
Tr. 739 (emphasis added).
This is cold comfort for a finder of fact. When pressed, Jansen ventured that “I think” there is a connection between Wilson and Moran. Id. But “I don’t know the specifics.” Tr. 729-40. Understandably, Jansen was pressed on this point:
Q. [by Mr. Corgan]: Is this $5,000 Big Dog/Robert Paul Wilson money or is it Leon Moran money?
A. Well, it would have been money potentially — I mean, that — without going back to some more information, *980I really don’t know the answer to that.
Tr. 740.
Jansen was asked essentially the same questions as to another check, attributed to Moran, showing Wilson as the remitter. He posited a connection with Moran in that “I believe that Moran picked these checks up....” Tr. 741. When asked by the court how he knew that Jansen replied: “From my working with Greg Mel-zer.” Id. Obviously, Melzer himself had no way to know that as a percipient witness, but the government was satisfied to present Jansen’s account of what he thought Melzer believed to be true from a source not specified at the forfeiture trial or in the government’s proposed findings, doc. no 2242 at 38, ¶ 141.
Jansen gave essentially the same account as to GX 883M, N, O, P and Q. Tr. 743. The best he could say was that he was “pretty sure” that Moran picked the checks up from Wilson (Tr. 743), a proposition that, as far as the record discloses, he could not have espoused with any more assurance than anyone else in the courtroom. “It’s based on my learning about this from Mr. Melzer, as I recall, yes.” Tr. 744. So it boiled down to this:
Q. [by Mr. Corgan]:Okay. Let me ask you — I’ve kind of surmised this, but in regard to C74 [GX C74, the government’s summary as to Moran], am I correct in my understanding that most of the numbers in there were numbers provided to you by Mr. Melzer and then you either accepted or rejected them and you prepared this chart?
A. Yes.
Tr. 760.
On another item in the government’s summary as to Moran, there was a discrepancy as to the amount of alleged betting proceeds evidenced by GX 476, a handwritten sheet of figures. Tr. 760. The difference, amounting to $5,000, was between $45,790 and $40,790. Tr. 761. Jansen had nothing of substance to offer:
Q. [by Mr. Corgan]: Would it — am I going too far, Mr. Jansen, if I say pretty much at this point you’re just guessing about a lot of this?
A. Yes, at this — as it relates to how I — I don’t know for sure how I dropped it off to 40,000.
Tr. 761.
Turning to the substance of some of the items in the government’s alternative calculation (and still ignoring double counting for the time being), the outlook does not improve. One attribution of betting proceeds, amounting to $187,000, was urged because an empty envelope (from a dentist) that was found in Moran’s house when it was searched showed, in handwriting, four figures that added up to $187,100. GX 1401. Neither the envelope nor any other evidence or testimony connects this exhibit with illegal gambling. The fact that it was found in Moran’s house was enough. What transactions did this represent? Jansen: “I have no idea.” Tr. 749. (And, aside from that, there was a discrepancy even in adding up the four numbers. Jansen’s explanation: “That was provided to me by Mr. Melzer’s information.” Tr. 747.)
Another item, perhaps a bit more plausible, was a digital image, yielded by the search of Moran’s computer, showing a stack of paper currency with a note leaning against the stack, suggesting (believably) that the stack amounted to $50,000. Moran wasn’t shown to have ever had the money. He wasn’t even shown ever to have had a physical photograph. But this digital image resulted in the attribution of $50,000 to Moran. Tr. 571.
*981The court could work its way through the rest of the thirteen items comprising the government’s alternative calculation as to Moran, and some of them are indeed better than others. Even so, the court declines to pick its way through the rest of them to determine which might pass muster despite the absence of meaningful supporting testimony.
Wholly aside from the deficits that have been discussed so far as to the government’s proposed alternative calculation as to Moran, there is an unavoidable, and insoluble, problem of double counting. With the possible exception of GX 473, discussed above, the numbers that add up to the government’s alternative calculation as to Moran áre posited because he was a runner. He was a runner in California. He dutifully picked up gambling proceeds from bettors and took them to agents or, more typically, picked the proceeds up from agents and sent them to Legendz. The evidence from the four jury trials established unmistakably that betting proceeds were, as needed, transported all over the United States. E.g., May Tr. 887, 906 (Lawhom delivering cash to Las Vegas and Nebraska). Billy Howell: “Big Lou would pick up from me in Orlando or we would have money picked up in California or Las Vegas, Oklahoma, and Texas.” May Tr. 1469. Although it is true that the court has rejected the government’s proposed extrapolations as to Wilson and Campbell (both of whom operated in California, where Moran was a runner), the record, taken as a whole, leaves the court with no assurance that money the government seeks to attribute to Moran just because he touched it is not also attributable — and attributed — to one of the other agents as to whom the government has made a satisfactory showing. Jansen’s testimony as to Wilson is illustrative, even though the court has rejected the government’s proposed extrapolation as to Wilson: “I think there’s some connection between the money that Paul Wilson would have paid would have been picked up my Mr. Moran.” Tr. 739.
The extrapolation and the proposed alternative calculation as to Moran are rejected.
1. Joseph McFadden
The government’s proposed attribution to Joseph McFadden is not an extrapolation. McFadden was an agent for Legendz in Florida. He was a credible witness for the government in the October trial. (He entered a plea of guilty in this case and has been sentenced.)
McFadden started as a Legendz agent in 2008. Oct. Tr. 592. His book of business grew to about fifty bettors. Id. 593. He stopped bookmaking in April, 2013, after he was arrested in this case. Id. 600.
In its Forfeiture Brief, the government attributed $1,656,000.00 to McFadden. Doc. no. 1999-1, at 18. In the government’s forfeiture overview and at the forfeiture trial, this became $1,784,800.00. GX C89; Tr. 153, 351. In the government proposed findings, this number became $1,588,000.00. Doc. no. 2242, at 21, ¶ 82. And all of these versions were drawn from the same source — McFadden’s testimony at the October trial.
The government’s proposed attribution of $1.6 million to McFadden (per the government’s proposed findings) is in a bit of a different category from most of the items included under the “Agents” heading on GX C89. The difference is that none of this money has been shown, or even suggested, to have been generated by McFadden’s activity as a Legendz agent. Instead, this $1.6 million is the sum of five separate and mostly dissimilar events and transactions. As to some of these events, McFadden could be loosely considered to have *982been a runner, but none of this money has been shown to have been generated by McFadden as an agent. These events and transactions are listed below. (The dollar amounts in italics add up to the $1,588,000 now proposed by the government in its proposed findings.)
On dates not disclosed by the record, McFadden visited King’s home in Houston. Oct. Tr. 602. This included a birthday party at the Kings’ home. Oct. Tr. 603. At the birthday party, McFadden delivered to King a “satchel of money,” amounting to $4-00,000 (pack- , aged in $10,000 bundles). Oct. Tr. 603, 605-06.
• Beginning in 2003, King invested $196,800 in a business venture, unrelated to illegal gambling, with McFadden and one other individual. Oct. Tr. 619, 21-22, 24. King made his investment in several stages by delivering or having a helper (such as Myler or Lawhorn) deliver cash to McFadden. Oct. Tr. 619-20. The investment was liquidated in stages, resulting in the return of $208,000 to King.118 Oct. Tr. 621, 623.
• In 2009 or 2010, King called McFadden and asked that McFadden accept delivery of a package of cash. Oct. Tr. 628. McFadden was, at the time, at a restaurant in Houston, so McFadden stayed there until the cash, $30,000 in “three stacks” of $10,000, was delivered. Oct. Tr. 628. At King’s request, McFadden delivered the cash to King’s wife, Serena.
• In 2007 or 2008, McFadden helped to clean up and recondition (with fabric softener) some $350,000 in cash that had deteriorated because it had been buried in Nebraska. Oct. Tr. 630, 644. The condition of the previously-buried cash was such that McFadden was only able to recover about $175,000 of it. Oct. Tr. 631. All of the reconditioned cash was delivered to King by McFadden or Neil Myler, less a $17,000 fee retained by McFadden. Oct. Tr. 631.
• In 2004 or 2005, McFadden allowed King to use Zima Holdings, a company owned by McFadden and John Reynolds, as a repository for the proceeds of the sale of a house King owned in Florida. Oct. Tr. 617, 624-25, 627. This transaction — resulting in the deposit of “over 600,000”119 into a Zima account— occurred at a time when King, McFadden and Reynolds contemplated that King would be buying Zima from McFadden and Reynolds. Oct. Tr. 625-26. The closing on the house occurred the day before the transfer of Zima to King was to have been completed, but the transfer of Zima fell through, so the proceeds of the sale of the house were transferred from Zima to King. Oct. Tr. 626.
Giving the government the full benefit of the preponderance standard, and drawing reasonable inferences from all of the evidence in this case, the court concludes that it is more probably true than not that the cash transactions to which McFadden testified — described above in the first four paragraphs — were effected with funds that “constituted” or were “derived from” proceeds of the unlawful activities of which the defendants have been convicted, or *983were “traceable to” property “involved in” those offenses. As to those, the nexus requirement has been satisfied. But as to the $600,000 in proceeds of the sale of the Florida house, it is too much of a stretch. In its proposed findings, the government does not even go so far as to assert that this $600,000 constituted, or was traceable to, criminal proceeds, and the court would not, in any event, so find. Doc. no. 2242, at 21, ¶ 81. All the government does is recite the fact of the transfer, perhaps under the influence of its overarching — and completely untenable — factual assertion that King had no cash flow from lawful sources during the conspiracy period.
The McFadden transactions net $988,000 for the forfeiture money judgment.
m. David Ross
For a number of years, David Ross was an active bookie and runner for Legendz in the Oklahoma City area. At all times relevant to this case, he has lived in central Oklahoma. Oct. Tr. 298, 321, 450. He entered a plea of guilty in this case and testified for the government in all four of the jury trials. In its Forfeiture Brief, the government went through the math step by step and came up with $1,642,813.00 in betting proceeds attributable to Ross by way of extrapolation. Doc. no. 1999-1, at 16. In the first version of GX C89, this extrapolation shrank to $1,265,831.18. In his testimony at the forfeiture trial, Low-rance went through the math step by step — purportedly using the same methodology — and came up with an extrapolation to $5,485,268.45. Tr. 154. (Jansen did not address the Ross extrapolation.) The following matters are noteworthy with respect to this proposed $5.5 million extrapolation:
• Lowrance, the government’s only witness in support of this extrapolation, did not read any of Ross’s trial testimony. Tr. 352.
• Neither Lowrance nor any other agent conferred with Ross (a highly cooperative and easily available government witness) about this extrapolation. Tr. 353.
• Lowrance has no idea whether the government’s proposed extrapolation as to Ross is consistent with Ross’s testimony. Tr. 353.
• Lowrance does not “recall exactly where the records came from” that support this extrapolation. Tr. 353.
• Consequently, as Lowrance put it: “I don’t know anything about the underlying activity.”120 Tr. 353.
In other words, the FBI’s approach as to Ross was to look just far enough to get to the desired result, as a purely mathematical exercise, and then look no further, lest inconvenient facts, perhaps imparted by a person (Ross) with firsthand knowledge of those facts, get in the way. There is no need for the court even to assess the relationship between the base period and the proposed 108-month extrapolation period. This extrapolation is rejected. The government proposes no alternative calculation as to Ross. Doc. no. 2242, at 16-17, ¶¶ 63-64.
n. Kory Koralewski
Koralewski was convicted on Count 1 (racketeering conspiracy) and acquitted on Count 2 (illegal gambling business) and Count 3 (money laundering conspiracy). For purposes of the forfeiture money judgment it seeks, the government would attribute $935,112 in illegal betting proceeds *984to Koralewski. GX C73. (As shown by GX C73, the government’s attribution to Kora-lewski started out at $1,055,112. But the government no longer advocates the $120,000 in proceeds it based on the testimony of Michael Schory, so the total is now $935,112. Doc. no. 2242, at 44 n. 3.)
A fair evaluation of the government’s forfeiture case against Koralewski requires, for one thing, an understanding that Koralewski and King have been good friends for “many years,” Oct. Tr. 537. In fact, Legendz chief financial officer Karlo Stewart knew Koralewski to be King’s “good life-long childhood friend.” Oct. Tr. 997.121 Although Koralewski had legitimate employment in the U.S., and spent the vast majority of his time in the U.S., he was no stranger to the Legendz headquarters in Panama. He clearly participated, at least sporadically, in the operations of Le-gendz in Panama and he was at least sporadically involved in some Legendz-re-lated activities in the U.S.
Consistent with his talents and his friendship with King, Koralewski got involved in training Legendz employees in Panama and with recruiting prospective bookies to affiliate with Legendz. Koralew-ski was close enough to the Legendz organization to find it convenient to list “DSS” as his employer on a loan application, even when he lived in Colorado. GX 398C (employer listed as: “Skis BBQ/ DSS” of El Dorado Hills, California).122 Thus, the situation with respect to Kora-lewski is unmistakably complicated by the fact that Bartice King’s relationship with Koralewski was very much a combination of personal friendship and business. Of the fifty-nine indicted defendants in this case, there are three — King’s wife, Serena, Edward Buonanno, and Koralewski — who, more than any others, had a close personal relationship with King. And in the case of Koralewski, there is evidence plausibly suggesting that even the personal relationship had financial aspects that were separate from the financial aspects of the business relationship. That muddles the forfeiture case against Koralewski to some extent.
Another complication with respect to Koralewski (not as problematic, but still in the mix) lies in the fact that Koralewski was indicted and prosecuted in this case as a member of the “Executive Staff’ of Le-gendz. Doc. no. 354, at 4-5. The grand jury did not label him as an “Agent,” a “Bookie,” or a “Runner” (id. at 6-8), and the government has disclaimed any intent to characterize him as an agent or bookie. Doc. no. 1576, at 1. But, in the October jury trial, the government still wanted to prove that Koralewski was a runner. In an order entered at the outset of that trial, the court addressed the fact that the Indictment, even though characterizing Ko-ralewski as a member of the executive staff, did allege some acts that amounted essentially to service as a runner:
Although this latter allegation is consistent with service as a runner, the government will be permitted to prove the truth of that allegation. Moreover, the government will not be precluded from proving other instances of conduct consistent with what a runner might also *985do. Sometimes executives do things that underlings also do. But the court concludes that Koralewski may not be characterized in any broader sense, in arguments or by way of evidence, as a runner. The court has no basis for finding that the grand jury considered any such evidence, and the Superseding Indictment suggests no such overall role on Koralewski’s part.
Doc. no. 1727, at 4-5.
At the forfeiture trial, Koralewski renewed his objection. Tr. 154. The court received evidence as to Koralewski subject to that objection. Tr. 155. The government’s proposed findings and conclusions do not address this issue. Of the $935,112 the government attributes to Koralewski for purposes of its forfeiture money judgment, only $10,790 could be regarded as attributable to Koralewski solely by virtue of functioning as an agent. Those dollars will be excluded from the forfeiture money judgment calculation for other reasons, but, in any event, they would be excluded on the basis of the court’s determination that Koralewski may not be prosecuted in this case as an agent. The remainder of the $935,112 is sought by the government on a basis consistent with a fair reading of the allegations in the Indictment.
One other preliminary matter relating to Koralewski should be addressed. Koralew-ski was convicted only on Count 1, racketeering conspiracy, and the court has concluded, consistent with the D.C. Circuit’s decision in United States v. Cano-Flores, 796 F.3d 83 (D.C. Cir. 2015), that, for forfeiture purposes, Koralewski is not on the hook for racketeering proceeds which were “obtained” by others (all as is discussed in Part 11(C), above). That will shield Koralewski from vicarious liability for proceeds received, say, by Bramley, as a result of bookmaking by Bramley in Texas that Koralewski had nothing to do with. The court concludes, however, that the government’s factual theory — if meritorious — as to the direct attributions it proposes with respect to Koralewski does qualify the money involved in the relevant transactions, discussed immediately below, as money “obtained” by Koralewski within the meaning of 18 U.S.C. § 1963(a)(3).123
With this background in mind, the court turns to the specific items the government seeks to attribute to Koralewski for forfeiture money judgment purposes. The $935,112 the government seeks to attribute to Koralewski consists of four items, per GX C73 and the government’s proposed findings (doc. no. 2242, at 43-44):
• $10,790 which the government says is betting proceeds evidenced by Le-gendz spreadsheets.
• $711,812 which the government says is proven by “shipping tracking records.”
• $202,510 which the government characterizes as “Legendz Sports Payments.”
• $10,000 which the government bases on the testimony of David Ross.
The $10,790 in betting proceeds.
The first item, $10,790 in betting proceeds, relates to the betting activity of *986Ashish Patel ($4,800), Robert Hurst ($2,561), and Craig Hardy ($3,429).
The government’s Forfeiture Brief says nothing about this proposed attribution of betting proceeds. See, doc. no. 1999-5, where the government addresses “Money Judgment with Factual Basis” as to Kora-lewski. But the government introduced this item at the forfeiture trial. GX C87; Tr. 538. In its post-trial proposed findings, the government’s entire assertion with respect to this item is:
Legendz financial spreadsheets attribute approximately $10,790.00 in illegal proceeds to Mr. Koralewski. (Exhibit 59); (Exhibit 208); (Exhibit C87) (summarizing Legendz spreadsheets with regard to proceeds attributable to Mr. Koralew-ski); (Tr. Apr. 28, 2016, Vol. III, at 532-534).
Doc. No. 2242 at 43-44, ¶ 164.
Ashish Patel ($4,800). Called as a witness by Koralewski, Ashish Patel testified at the October jury trial. He is an orthopedic technician with Kaiser Permanente in Sacramento, California. Oct. Tr. 2187. “[A]t some point in years past,” he bet on sports. Id. at 2188.
As relevant to the matters now before the court, the gist of Patel’s testimony at the October trial was:
• He never used a bookmaker or agent to place his bets. Id.
• He did his sports betting by using a web site and a $100 prepaid debit card. Id.
• He never met Koralewski until the day before he testified at the October trial. Id. at 2190.
• He never bet the $4,800 the government attributes to him. Id. at 2190.
• He never sent any money to Legendz. Id. at 2192.
• In February, 2015, eight months before the October trial, he was interviewed by the FBI. Id. at 2191. In that interview, he gave the FBI “the same information you’ve given the jury today.” Id. He told the FBI that he “played for a hundred dollars and lost it and that was the end of it.” Id.
• He engaged in no betting activity after he lost the hundred dollars. Id.
Patel’s testimony at the October trial was credible. The government’s perfunctory cross examination cast no shadow on Patel’s account of his betting activity and his lack of any contact with Koralewski or Legendz. Oct. Tr. at 2192-95.
At the forfeiture trial, by way of answers to three questions (and no more) on direct examination, Jansen attributed the $4,800 to Patel, and to Koralewski, because Patel’s name showed up on a Legendz spreadsheet in the same line as “SKI.” (GX 59). Tr. 532.124
On cross examination of Jansen, it got more interesting:
Q. [by Mr. Autry] Were you aware that Mr. Patel told the FBI, I never bet $4,800 with Legendz Sports or anybody else; were you aware of that?
A. No, sir.
Q. Were you aware Mr. Patel said, The only time I bet I did it on online with a prepaid debit card for a hundred dollars, and when I lost the hundred dollars, I quit betting?
A. No, sir, I’m not aware.
Q. Were you aware that Mr. Patel testified in the criminal trial involving Mr. Koralewski?
*987A. No, sir, I’m not.
Q. Were you aware that Mr. Patel, when he testified, said he did not know Mr. Koralewski at all?
A. No, sir, I did not.
Q. Were you aware that Mr. Patel, when he testified, said he had absolutely no betting or gambling dealings whatsoever at any time in any fashion with Kory Koralewski, Ski, or whatever moniker you want to use; were you aware of that?
A. No, sir.
Q. Okay. Were you aware of the fact that Mr. Patel stated that he never wrote a $4,800 check for a gambling debt?
A. No, sir.
Q. Okay. Were you aware of the fact that the FBI or the government never subpoenaed Mr. Patel’s bank records to either confirm or deny that there was any $4,800 check?
A. No, sir.
Tr. 678-79.
Thus, when he took the stand at the forfeiture trial — six months after Patel went under oath to give his account of the matter at the October jury trial — the fact that Patel had testified in that trial, and had given the testimony recounted above, was news to Jansen.
Robert Hurst ($2,561). Hurst lives in the Sacramento area. Oct. Tr. 2197. He is retired after working for forty-nine years for a power equipment distribution company (thirty-four years as president of the company). Id. Hurst, like Patel, testified at the instance of Koralewski at the October trial. His testimony was credible.
Hurst has placed bets on sporting events. Id. He did that “mostly online.” Id. at 2198. He was shown a Legendz spreadsheet that showed his name on the same line with “Ski,” indicating that the “amount deposited” was $2,561. Id. at 2198-99; (GX 208). In his betting activities, he never dealt, that he knew of, with a person named Ski. Id. at 2199. Hurst does not know “Ski.” Id. at 2201. Hurst had never met Koralewski until the night before he, Hurst, testified at the October trial. Id. Koralewski has never picked any money up from Hurst. Id. When the FBI came to Hurst’s house to interview him, he told the FBI the same thing he told the jury. Id. at 2202.
At the forfeiture trial, Jansen’s testimony linking Hurst and Koralewski consisted, in its entirety, of one answer: “There’s two transactions, the fourth and fifth line down [on GX 208], attributed to Ski. The ‘Received on’ date is March 2nd; the pending amount for the first transaction is $3,429; and the second transaction is $2,561; for a total package of $5,990. The first transaction comes from the sender Craig T. Hardy and the second one comes from Robert Hurst.” Tr. 533-34.
As for Jansen’s familiarity with Hurst’s testimony at the October trial, Jansen had this to say:
Q. [by Mr. Autry] Okay. Were you aware that Mr. Hurst was interviewed by the FBI?
A. No, sir, I was not.
Q. So I guess you wouldn’t be aware that Mr. Hurst told the FBI that he did not know Mr. Koralewski, never placed a bet with him, that Mr. Koralewski was not his bookie and that he was not involved in any way, shape or form with Mr. Koralewski in laying any bets?
A. No, not aware of that.
Q. Were you aware that Mr. Hurst, like Mr. Durborough and Mr. Patel, testified in the criminal trial involving Mr. Koralewski?
*988A. No.
Q. And that Mr. Hurst testified just like these other two guys, that he did not know, had never met and never dealt with in a gambling context with Ski or Kory Koralewski?
A. No, I did not.
Tr. 681.
Craig Hardy ($3,429). The situation with Craig Hardy is a bit more complicated because Hardy is dead. Oct. Tr. at 2204. Consequently, the jury at the October trial heard from Hardy’s bookie as to Hardy’s betting activity. That bookie, called as a witness by Koralewski, was William Dur-borough. Durborough lives in the Sacramento area and is retired after thirty years with the Sara Lee Baking Company. Id. at 2203. Hardy worked for Wonder Bread, and Durborough knew Hardy for fifteen or twenty years. Id. at 2204.
Durborough and Hardy both bet on sports, and they did so through an internet sportsbook called ZMVP. Id. (Ample evidence in this case establishes that ZMVP was part of the Legendz sports betting enterprise.) In his capacity as a bookie, Durborough’s moniker was “Apple.” Id. at 2206. At the October trial, Durborough was shown GX 59, which, on three separate lines, lists Apple as the Legendz agent with respect to Hardy’s betting activity. (Durborough saw that document for the first time the day before he testified at the October trial. Id. at 2205. He had been interviewed by the FBI about six weeks before that. Id.) Durborough confirmed that he is the “Apple” associated with Hardy’s betting activity on GX 59. Id. at 2206.
Durborough was then shown GX 208. Government Exhibit 208 shows “Ski” on the same line as Craig Hardy, associating both of them with the $3,429 in betting activity at issue here. This led to a description by Durborough of his betting-related activity, Hardy’s betting-related activity, and (importantly for present purposes), Durborough’s familiarity with Hardy’s betting. As far as sports betting was concerned, Durborough’s account was that “Craig [Hardy] answered to me [Durbor-ough].” Id. at 2208. On the question of how Hardy could possibly know Koralewski, Durborough said: “I collected from him [Hardy] when he would lose. I just had a few guys, friends, that wanted to play, so Craig was one of those guys. And I got him into MVP [ie., Legendz]. So I don’t know how he would know this gentleman [Koralewski] or anybody else from MVP.” Id.
Durborough testified that he never had any contact with Koralewski during the time that he and Hardy were active in sports betting. Id. at 2207. Durborough told the FBI that he did not know anybody named Ski. Id. Durborough never exchanged money with Koralewski. Id. at 2208. He never reported to Koralewski. Id. Durborough testified firmly and persuasively that, to his knowledge, Koralewski never had “anything whatsoever to do with Mr. Hardy and Mr. Hardy’s betting.” Id. at 2208. See also, id. at 2210 (“Craig would have told me if he did.”). Although it would not be possible for Durborough to know for a fact that Hardy never dealt with Koralewski, Durborough’s account of his relationship with Hardy leaves the court well-satisfied that it is improbable that Hardy dealt with Legendz (or anyone associated with Legendz) other than through Durborough.
Durborough’s testimony was believable. Six or seven weeks before the October trial, Durborough told the FBI the same thing he told the jury in the October trial. Oct. Tr. at 2208.
As for Jansen:
Q. [by Mr. Autry] Well, as a matter of fact, were you aware that a gentle*989man named William Durborough, a/k/a Apple, testified as a defense witness—
A. No, sir, I’m not.
Q. —in Mr. Koralewski’s trial?
A. I’m not.
Q. And that he indicated that he handled Mr. Hardy, who is a personal friend of his for many years’ duration in the gambling business?
A. Okay.
Q. Were you aware of that?
A. No, I wasn’t.
Q. And that he didn’t know, talking about Mr. Durborough, Mr. Kora-lewski at all; were you aware of that?
A. No, sir.
Q. And that he had no gambling dealings with Mr. Koralewski at all; were you aware of that?
A. No, sir.
Q. And that Mr. Hardy had no gambling dealings with Mr. Koralewski at all because Apple, Mr. Durbor-ough, was his bookie or agent, were you aware of that?
A. No, sir.
Tr. 674-75.
Jansen’s account of what he did not know is also believable. Jansen may be forgiven for having a tone of exasperation when he summed the matter up: “I didn’t know that these guys had testified.” Tr. 682.
The court is far from satisfied that this $10,790 is chargeable to Koralewski for forfeiture money judgment purposes. The foregoing summary speaks for itself, and that alone is sufficient to undermine the court’s confidence in the government’s proposed attribution of these betting proceeds to Koralewski. (In its proposed findings, the government makes no attempt at all to explain the testimony of Patel, Hurst and Durborough at the October trial, or Jansen’s testimony, quoted above, at the forfeiture trial.) But there is more.
The two internal Legendz spreadsheets (GX 59 and GX 208) relied upon by the government to attribute $10,790 in betting proceeds to Koralewski, are two of the internal Legendz spreadsheets surreptitiously taken by Karlo Stewart from the Legendz headquarters at the behest of the government. Tr. 437 (GX 59); Oct. Tr. 958 (GX 208). In United States v. Bramley, 2015 WL 12852048, at *14 (W.D. Okla. Sept. 16, 2015), the court noted the substantial baggage Stewart carries as a witness. He is a repeat fraudster with state and federal convictions for serious crimes of dishonesty, including (especially noteworthy in the present context) forgery. Aside from that, as a deportee from the United States (Feb. Tr. 869) who also left Panama one jump ahead of the Panamanian authorities (id. 1104), his immigration-related aspirations exponentially increase his motivation to please the government. Id. at 1105. As a general proposition, the court has, in this and other orders in this case, relied on testimony of Stewart favorable to the government only when that testimony has been corroborated by other evidence.
Internal financial and betting-related spreadsheets were the stuff of Stewart’s day-to-day life at Legendz. Feb. Tr. 972, 996. Stewart could, in the blink of an eye, doctor up a spreadsheet in any way he thought expedient for the purpose of pleasing the government.125 Stewart cooperated with the government for a considerable *990period of time while he remained employed by Legendz and was producing the various tranches of internal Legendz documents to the government. See, generally, Feb. Tr. at 1050-59. He had good reason to know what persons the government was curious about. The court views spreadsheets produced by Stewart with the same gimlet eye with which it views his testimony. '
The court rejects the government’s proposed attribution of $10,790 in betting proceeds to Koralewski.
The $711,812 based on the “shipping tracking records.”
The government asserts that:
Legendz Sports’ shipment records show that Mr. Koralewski sent approximately 70 packages to Legendz Sports, containing approximately $711,812.00 in illegal proceeds. (Exhibit 198); (Exhibit C62) (summarizing Mr. Koralewski’s shipments of proceeds to Legendz); (Tr. Apr. 28, 2016, Vol. Ill, at 534-539).
Doc. no. 2242, at 43, ¶ 163.
That is all the government’s proposed findings have to say about this item. The quoted assertion, if true, is fair game for forfeiture purposes because, even though Koralewski was acquitted of money laundering conspiracy, this kind of activity could reasonably be brought within the scope of the “Executive Staff’ label the grand jury attached to Koralewski. As the court noted in its order at doc. no. 1727, sometimes executives do things that underlings also do.
The source document for this proposed $711,812 attribution is GX 198, an internal Legendz spreadsheet taken by Karlo Stewart from Legendz. Tr. 534. That spreadsheet is a list of tracking numbers, apparently reflecting shipments of cash or other forms of payment to Legendz. As described by Karlo Stewart, GX 198 is a “monthly reconciliation of packages that would have arrived.” Oct. Tr. 916. Each entry on GX 198 consists of a date (no year specified, and no year known), a tracking number, a dollar amount, a name under the heading of “WHO SENT IT,” and, where applicable, an indication (“ARR”) that the package did arrive. Seventy of these entries show “WHO SENT IT” to be “SKI.” They total $711,812. GX C62; Tr. 683.
As noted, GX 198, the internal Legendz spreadsheet showing tracking numbers and related information, is the source document for this proposed attribution. But that source document has its own underlying source documents. The source documents for GX 198 are GX 68-108, forty exhibits in all. Oct. Tr. 1024. Those forty exhibits are smaller spreadsheets by which Edward Buonanno, the general manager of Legendz, “recorded the incoming packages” consisting of “[cjhecks coming from the United States.” Oct. Tr. 933. Each one of those smaller spreadsheets covers two or three days. E.g., GX 100 (two days); GX 106 (three days). And, as Jansen acknowledged at the forfeiture trial, not a single one of the forty source documents for GX 198 refers to “Ski.” Tr. 690. This leaves GX 198, and any inferences reasonably to be drawn therefrom, twisting in the wind, as Jansen acknowledged:
Q. [by Mr. Autry] So with respect to the tracking numbers, there is no evidence that would indicate they are connected in any way independently of this document [GX 198] to Mr. Koralewski, right?
A. That’s correct, sir.
Q. In other words, the tracking numbers have not been verified in connection with him outside of this document, right?
*991A. That’s correct, sir.
Tr. 686-87.
Neither Jansen nor, to his knowledge, anyone else on the prosecution team, ever investigated the tracking numbers by which the government seeks to hold Kora-lewski accountable for $711,812. Tr. 686. Moreover, Jansen was unaware that Karlo Stewart had testified that GX 68-108 were the source documents for GX 198. Tr. 689-90. This is apparently because, as Jansen acknowledged, he has not “read any of Karlo Stewart’s testimony from any of these proceedings.” Tr. 691.
In evaluating this proposed $711,812 attribution of betting proceeds to Koralew-ski, the court will readily grant that a spreadsheet showing seventy shipments from the United States to Panama (assuming that is what it actually reflects — probably a fair assumption) with “SKI” listed in the “WHO SENT IT” column is enough to give rise to a suspicion that Koralewski sent these packages. There is not much that Koralewski would not have done for Bartice King. As a management-level employee of some large companies, or as an entrepreneur in various small businesses (more about those matters in Part IX(F), below), it would not have been much of an inconvenience for Koralewski to pick up a package of checks or cash from a runner and take it to a UPS or FedEx office. But the evidence leaves the court short of concluding with reasonable confidence (even on a preponderance standard) that Kora-lewski is chargeable with this $711,812. For reasons that have already been discussed, a document produced by Karlo Stewart does not come before the court with much intrinsic reliability. Thus, GX 198 does not have a starting point any better than the simple fact that it is held out by Karlo Stewart to be a document he produced from within Legendz. The probative value of GX 198 can go up or down from there. It goes down because “Ski” is not mentioned in any of the forty source documents for GX 198 even though those forty source documents are studded with names that are very familiar to the court in this case.
The court is unpersuaded by this proposed $711,812 attribution. It is rejected. The $202,510 the government characterizes as “Legendz Sports Payments.”
The government’s proposed finding as to this item is:
Mr. Koralewski received checks valued at approximately $202,510.00 consisting of illegal proceeds from the Legendz enterprise. (Exhibit 915); (Exhibit C29); (Tr. Apr. 28, 2016, Vol. Ill, at 589-541).
Doc. no. 2242, at 43, ¶ 160.
The most damning thing about these forty-five checks, from the perspective of a forfeiture defendant who seeks to avoid being held accountable for them for forfeiture purposes, is that they came from Central America, where Legendz was based. The checks are all what we in the U.S. would call cashier’s checks. They are all payable to Kory Koralewski. They were issued by Banco de Costa Rica in San Jose, Costa Rica — not by a Panamanian bank or from a Panamanian office of any other bank. They are not drafts against the account of any account holder, nor is the name of any remitter or account holder shown on any of the checks. Jansen acknowledged that the checks show no link to a bank customer. Tr. 698.
Each of the checks is for 4,500 U.S. dollars, save one in the amount of $4,510. They were issued essentially monthly (there are three exceptions), beginning in early 2005 and ending in late 2008. GX C29. Many of the checks were payable through Citibank in New York, apparently as a correspondent bank. Others were payable through Bank of America or the Mia*992mi office of Banco Internacional de Costa Rica. GX 915, GX 1119, GX 1120.
At the forfeiture trial, Jansen’s attention was called to many of these $4,500 checks, and his typical explanation was that they were “considered to be payments from the Legendz enterprise.” Tr. 540. Otherwise stated: “It was funds from Banco de Costa Rica that we considered associated with Legendz enterprise.” Tr. 543. Jansen also acknowledged that these checks “formed the basis of the government’s allegation in Count 3 against Koralewski for money laundering conspiracy,” of which Koralew-ski was acquitted. Tr. 701. (The court does not consider that to be fatal to the government’s position as to these payments, because, if, as the government asserts, they were betting proceeds, then they would also be within the permissible ambit of Count 1.)
As to the source of the funds, Jansen acknowledged that:
Q. [by Mr. Autry] You can’t trace a single dime of any of these checks to illegal sports-betting money generated by Legendz Sports, ZMVP or any subsidiary' — -I never can pronounce that — subsidiary company involved with Legendz, true?
A. Yes, sir.
Q. Okay. So you don’t know what these checks are for, right?
A. No, I don’t. I would want to look at probably the books and records of what I believe the source of the funds is Legendz Sports.
Q. But you don’t have anything to confirm that, correct?
A. No, sir, I don’t.
Q. You don’t have any documents that would show a dime of any of these checks came from dirty, illegal sports-betting money controlled by Legendz or any of its related companies, true?
A. That’s correct.
Tr. 699-700.
In its proposed findings, the government offers nothing beyond a bare description of the checks, saying, essentially, that they are what they are. Doc. no. 2242 at 43, ¶ 160.
The government is not required to produce direct evidence that any given payment or series of payments consists of dirty money. That approach would largely reduce the right to get a forfeiture money judgment to a nullity. The required nexus may certainly be proven by circumstantial evidence, and circumstantial evidence can in many instances be compelling. Bearing that in mind, the questions the court asks about these payments are: (1) Where did the money come from? and (2) What do the payments represent? There is no evidence as to the source of the funds for these forty-five checks. The court certainly suspects that the money came from Bar-tice King or an entity controlled by him. But the difference between suspicion and the preponderance of the evidence is evidence. The government has failed to satisfy the court, by a preponderance of the evidence, that this money came from the Legendz enterprise.
Inferences favorable to the government about what these payments represent are also hard to draw. These payments have no resemblance to the way credits and payments between Legendz and its bookies and agents were handled. The fact that these forty-five payments came every month (with three exceptions) and were in identical amounts (with one ten-dollar exception) is air-tight proof that these were not routine transactions between the house and an agent or bookie. The fact that these payments may have come from Bartice King (which will be assumed for the mo*993ment), taken together with the fact that there is no way that they were routine house-bookie payments, brings to the fore two other facts that are problematic for the government. First, King and Koralew-ski were extraordinarily close Mends. Kar-lo Stewart recalled King mentioning that he, King, was a godfather to one of Kora-lewski’s children. Oct. Tr. 998. Second, we have the fact that, contrary to the government’s repeated groundless statements to the court, King clearly did have lawful sources of cash flow during the conspiracy period.
As to this $202,510 item, the government’s burden (having obtained a conviction of Koralewski only on Count 1) is to satisfy the court that the money was racketeering proceeds paid by Legendz to Ko-ralewski. The government has not carried that burden.
The $10,000 that the government bases on the testimony of David Ross.
The government’s contention as to this item is: “Mr. Koralewski delivered approximately $10,000.00 in illegal gambling proceeds to David Ross. (Tr. Oct. 20, 2015, Vol. II, at 502).” Doc. No. 2242, at 43, ¶ 161.
David Ross, a Legendz bookie, testified in the October jury trial that he drove from Oklahoma to Colorado, where he met Koralewski and received $10,000 from Ko-ralewski. From all the evidence in this case, the court can conceive of no characterization of these funds as anything other than betting proceeds. Ross was a runner as well as a Legendz bookie, and he clearly went to Colorado to pick up betting proceeds. There is no reason for Koralewski to have been giving $10,000 to Ross if that money was not betting proceeds. The question is whether this money represented Koralewski’s personal betting losses or whether it was garden-variety betting proceeds for which Koralewski was the conduit. At the October trial, Ross agreed that this money “could have been” Koralewski’s personal betting losses. Oct. Tr. 540. The court thinks not. There is no evidence that Koralewski was a Legendz bettor. And there is no reason for the court to be concerned about the provenance of this money in anything like the sense that the court has discussed above with respect to the monthly $4,500 payments. The government has satisfied the court, by a preponderance of the evidence, that this $10,000 was betting proceeds. The money was directly handled by Koralewski, so there are no issues as to vicarious liability. The money, having been given by Koralewski to Ross, was clearly “obtained” by Koralew-ski from someone. 18 U.S.C. § 1963(a)(3). It is enough that it was betting proceeds, as the court has found.
In sum, the government has proven illegal betting proceeds attributable to Kora-lewski amounting to $10,000.
o. Luis Robles
Luis Robles was a Legendz runner from Orlando, Florida. Oct. Tr. 1243. The government would attribute $721,901.50 to Robles under the “Agent” heading on GX C89. At the forfeiture trial, the government’s calculation was corrected to $727,901.50. Tr. 171. Doc. No. 2242 at 44, ¶ 171.
This proposed attribution consists of six items. GX C90. They will be addressed in the sequence in which they appear on GX C90.
The first item is $46,197.50, extracted from a Legendz spreadsheet, GX 1273. The relevant page of GX 1273 refers to “Big Lou” (which, beyond question, is Robles) under the heading of “Agent” and shows $46,197.50 under the heading of “Amount.” This item is rejected, for two reasons. First, Robles is prosecuted in this case as a runner, not as an agent. Doc. No. *994354, at 8. Regardless of whether some flexibility might be warranted in a less complex case, the court is convinced that considerations of fundamental fairness (to say nothing of the law under the Fifth Amendment and the Sixth Amendment) dictate that the court hold the government to the roles assigned in the indictment. The court does so with respect to Robles, as it has done with respect to Koralewski and Moran. (The court noted Robles’ objection on this point at the forfeiture trial. Tr. 156.) Second, and more substantively, Lowrance was not able to satisfactorily substantiate the basis for this $46,197.50 item. He had nothing to offer (based on other documents, trial testimony or otherwise) as to what this $46,197.50 was all about. Tr. 358, 369.
The second item comprising the proposed $727,901.50 attribution to Robles is $160,459.00, based on “Derek Hewitt T III intercepts.” GX C90. In his proposed findings, Robles complains that Hewett’s testimony in the February trial did not substantiate all of the contacts that were recorded by the Title III intercepts. Doc. No. 2278, at 5. That is correct (see Feb. Tr. at 1206, et seq.), but the intercepts (GX 810 and 813) are much more reliable than the long after the fact recollection of a witness at the February trial. The recordings are also more reliable than the government’s version of what they prove. GX 810 refers only to $50,000 (“50 dimes,” to use the terminology Robles used in the recorded conversation). GX 813 refers to $60,459 (“60 dimes', 4-5-9” in the recording). This second item has been substantiated to that extent — $110,459 (not the $160,459 asserted by the government).
The third item ($269,000) has not been substantiated. The consensual recordings cited as evidence (GX 946 and 947) contain no references to any dollar amounts, and the only other exhibit cited by the government (GX “1C39”) is not in evidence. Low-rance had nothing of substance to offer on this item. Tr. 157, 362.
The fourth item is $189,245, attributed to the trial testimony of Alan Gould. The testimony cited by the government (pages 1451 and 1453 from the February trial) does not come close to supporting the government’s contention, and Lowrance’s account of this item shed no light on this one. Tr. 355-56.
The fifth item is $50,000, consisting of $2,000 per month in salary allegedly paid to Robles. This would, obviously, consist of twenty-five months. The cited testimony from the February trial (Feb. Tr., 353 and 356) does not support the government’s version of this item. Nothing in the cited testimony establishes the fact of the $2,000 monthly payments, the regúlarity with which they were made, or the duration of that arrangement, if it existed at all. Low-rance could add nothing on this one. Tr. 366-67.
The sixth, and last, item as to Robles is $13,000, asserted to be based on “Reggie/Kihg T III intercept.” The cited exhibit (“819” — see GX C90) is not in evidence. Lowrance’s account of his understanding of the recorded conversation (the one that is not in evidence) suggests that the conversation, if it were in evidence, would be inconclusive at best. Tr. 367.
The evidence satisfactorily supports the attribution of $110,459 in betting proceeds to Robles.
4. Conclusion as to the government’s proposed extrapolations and direct attributions
The result of the court’s findings as to the government’s proposed extrapolations and direct attributions under the heading of “Agents” on GX C89 (App. 1 to this order), is:
*995[[Image here]]
The foregoing summary, viewed alone or in combination with the summary in Part IX(B)(3), above (summarizing the results of the court’s analysis of the money laundering entries on GX C89), might suggest the possibility of capricious results with respect to the forfeiture money judgment sought by the government. As will be seen, the ultimate result will be evened out considerably by the application of the law with respect to (i) joint and several liability, (ii) foreseeability, and (iii) the Excessive Fines Clause.
D. Viability of the government’s contention that all forfeiture defendants, regardless of their respective degrees of participation in the Legendz enterprise, should be subjected to a forfeiture money judgment in the same amount.
The government seeks a forfeiture money judgment against the forfeiture defendants in the amount of $231,432,686.73. As has been seen, the total amount of any forfeiture money judgment will be considerably less than that. But the issue now to be addressed is whether the government’s argument for joint and several liability for the full amount of a forfeiture money judgment has any merit. It does not.
As is discussed in Part (II)(B), above, the forfeiture defendants other than Kora-lewski are subject to joint and several liability for the forfeiture money judgment to be entered in this case. That, at this point, is a given. Yet to be determined is the extent of that joint and several liability. The nominal joint and several liability exposure of the forfeiture defendants other than Koralewski is $12,607,321.99, the sum of $6,899,140.19 by way of attribution of betting proceeds to agents (Part IX(C)(4), above) and $5,708,181.80 by way of money laundering liability (Part IX(B)(3), above).
This nominal joint and several liability is subject to three screens: (i) foreseeability (discussed in Part (II)(B), above), (ii) dis-proportionality, governed by the Excessive Fines Clause (discussed in Part III, above), and (iii) deprivation of the ability to earn a living, also governed by the Excessive Fines Clause (discussed in Part III, above).
As is discussed in more detail in Part 11(B), above, the government, although unmistakably on notice that foreseeability is very much in issue as a limitation on joint and several liability, advanced no legal or *996factual arguments at all on that issue in its Forfeiture Brief (doc. no. 1999), in its pretrial proposed findings and conclusions (doc. no. 2130), in its post-trial proposed findings and conclusions (doc. no. 2242) or by way of any other form of advocacy. As discussed in Part IV(A), above, the burden of proof on foreseeability rests with the government. The court will address the issue of foreseeability as best as may be possible in the absence of any advocacy from the government, as the party with the burden of proof. But, of necessity— and, under the circumstances, as a matter of fundamental fairness — doubts will be resolved against the government.
The second and third screens (dispro-portionality and deprivation of the ability to earn a living) will be addressed separately as to each of the forfeiture defendants. Suffice it to say for present purposes that the net effect of the application of all three screens is significant.
Bearing all of these matters in mind, the court now turns to the remaining substantive tasks with respect to forfeiture: (i) determining the extent of forfeiture of specific assets (ie., tangible property and financial assets), and (ii) determining the extent of forfeiture money judgment liability. The forfeiture defendants will be addressed in the same sequence in which they are discussed in the Forfeiture Overview (Part VIII, above). The three screens — foreseeability, disproportionality, and livelihood — serve separate purposes; all three are important.
E. Findings specific to King
1. Forfeiture of specific assets
The government has cast a wide net in terms of the assets belonging to Bartice King that it seeks to forfeit. As discussed in Part VIII(A)(1), above, this includes bank and brokerage accounts, three automobiles, jewelry, eighty-eight pairs of shoes, seventy-five purses, and five parcels of real property.
The factual linchpin of the government’s proposed forfeiture of King’s hard assets and financial assets is its oft-repeated and palpably untrue statement that King’s only source of income during the conspiracy period was the illegal sports betting operation. In support of forfeiture of specific properties, the government puts it as follows in its Forfeiture Brief: “[K]ing’s only source of income derived from his illegal gambling operation with Legendz Sports.” Doc. No. 1999-1, at 3. This disregards the uncontroverted testimony of the government’s witnesses — testimony the government never questioned or attempted to challenge — establishing King’s history of lawful entrepreneurial activities during the conspiracy period. Those matters are discussed at length in Part IX(B)(1), above, but, to briefly review, those activities included:
• Grupo Legendz a legal, licensed sports-book operating as a walk-in casino in Panama City.
• Vaporama, the holder of the liquor license for the sportsbook, generating $1.9 million in deposits into Banco General.
• Magna Tours, the IATA-eertified travel agency, “a pretty successful” business with corporate accounts, “growing at about 20 percent annually in sales.”
• Platinum Vacations, ultimately unsuccessful but generating $10,000 in daily revenues at one point.
Importantly, with one exception, the government does not trace gambling proceeds into the King assets it seeks to forfeit. Taking the bank accounts as an example, the government cites ten pages of testimony at the forfeiture trial, together with several account documents and transaction records, to support forfeiture *997of five accounts at Amegy Bank in Texas. This trail began with two checks from Banco General in Panama — a bank which received $1.9 million from Vaporama, aside from whatever it may have received from Legendz. GX 1217A. Those are cashier’s checks, not checks drawn on a transaction account. The payee is Bartice King, remit-ter not indicated. The proceeds went into Compass Bank in Texas in April, 2007, six years before the takedown in this case. As for the source of these funds, Lowranee could do no better than to allow that “I believe these funds came from the Banco General DSS account in Panama.” Tr. 21. The exhibits the government cites do not associate the transactions with DSS or any other Legendz entity. Doc. no. 2242, at 8-9, ¶¶ 31-35. See also, Tr. 296-97. Low-rance’s “belief’ may be right, but it is not enough to support a finding by the court because the government is steadfast in its resolve to ignore King’s sources of lawful cash flow during the relevant period. Moreover, assuming that the dollars that ultimately found their way into the Amegy Bank accounts were illegal gambling proceeds (a proposition that has not been established), the money went into the accounts in 2011, and that was only after the balance in the intermediate source account (Compass Bank) dropped to as low as $14,400 in April, 2007.126 The government does not argue, in terms, that this is tracing, which is a good thing, because this is certainly not tracing. Cf. United States v. Certain Accounts, 795 F.Supp. 391, at 398 (S.D. Fla. 1992) (“Like a contagious disease, each direct account [here, Compass] could contaminate any account that had dealings with it. The indirect accounts [here, Amegy] could then conceivably pass on the infection to other accounts, and so forth ad infinitum. The outer limits of this theory would be bounded only by Plaintiffs imagination.”)127
King’s Invesco IRA account is an exception to the government’s general failure to establish a nexus with respect to King’s financial accounts and hard assets. Government’s Exhibits 1218A-D (particularly 1218B) quite satisfactorily show $45,000 coming from DSS into the IRA account. The money in that account, consisting of a remaining balance of $9,962.06, will be forfeited. (The government points to no evidence showing the activity in that IRA account since the $45,000 was put into it in 2010. That, ordinarily, would suggest a tracing problem: What happened in that account between 2010 and 2013? But given the nature of that account and the fact that the balance at issue is a fraction of the illegal gambling proceeds that went into the account in 2010, the court is not inclined to take an exacting approach to this item. Cf. Stewart, 185 F.3d 112, 129.)
As for the parcels of real property,128 the government’s proof fails for several reasons. First, although ownership issues are generally deferred to ancillary proceedings, the court simply observes here that *998the court has not been made aware of evidence establishing Bartice King’s ownership interest in most of them.129 Second, and more importantly at this stage, the government has not shown that these properties were acquired with funds having an illicit origin.130 The funding sources go back to Panama in some instances, but the instruments getting the funds out of Panama are cashier’s checks that do not disclose the source of the funds. See, e.g. GX 853, 1203D, 1203E, 1203F, 1203H, 1204C, 1213C, 1213D, 1213E, 1214A.131 Taking 30 Nestlewood Place as an example: “Q: [by Mr. Mays] Do you know what account at Banvivienda this money came out of? A: I don’t know.” Tr. 314. As has been seen, because there can be no categorical characterization of funds coming into the hands of the Kings in Panama, there can be no categorical characterization of funds coming out of Panama to the Kings. The government’s overarching theory that King never had any lawful cash flow will not save the day because that assertion is simply not true. Another problem with the government’s case against the real properties is that the government’s evidence as to the money used either for down payments or debt service is very much a mélange. Some of the money did come from Panama (illicit source not shown, other than as to 206 N. Vesper Bend) and some of it came from accounts maintained by Bartice or Serena King in various domestic U.S. banks — accounts to which the government would have the court apply an overall taint, which the court, as a matter of forfeiture law, cannot do, as is discussed in Part VI(A), above. See, e.g. GX C37.1: “Mortgage Payments Made from Tainted Sources,” citing only an exhibit summarizing payments from Amegy Bank, Wachovia Bank, Bank of America and money orders.) As has been discussed, if a hard asset was acquired (or the mortgage on it serviced) partly with untainted funds, the record must “establish how much of the value came from [untainted sources] and how much from [tainted sources].” United States v. Genova, 333 F.3d 750, 762 (7th Cir. 2003). The government’s evidence falls far short on this score.
In sum, and reduced to its essentials, the government’s approach with respect to the hard assets and financial assets is that the voids in its evidence with respect to the required nexus with illegal proceeds can be filled by presuming, categorically, that everything King bought or put into savings during the conspiracy period originated with the illegal gambling operation: “Because Mr. King’s only source of income from 2003 through 2013 derives from his operation of the Legendz Sports enterprise, the United States contends that all property seized and listed for forfeiture constitutes proceeds of Mr. King’s criminal activity.” Forfeiture Brief, doc. no. 1999-1, *999at 3-4. For the government, as the party with the burden of proof, the problem is that if the government’s theory is not categorically right, then the government is at risk of being categorically wrong.
As the court put it in Genova: “[0]nce the defendant has contended, with some evidentiary support, that at least some of the value in a given asset came from lawful, non-forfeitable sources, then the prosecutor must demonstrate how much is for-feitable.” United States v. Genova, 333 F.3d 750, 763 (7th Cir. 2003). Where, as here, there is substantial uncontroverted evidence establishing significant on-going lawful entrepreneurial activities during the conspiracy period, the conclusion follows quite easily that there is “some evidentiary support” for the proposition “that at least some of the value” in the assets the government seeks here “came from lawful, nonforfeitable sources.” The court so finds. It fell to the government to “demonstrate how much is forfeitable.” Id. By ignoring King’s lawful entrepreneurial activities and hanging its hat on the proposition that King never had any lawful cash flow during the conspiracy period,132 the government effectively relinquished the opportunity to marshal the proof necessary to salvage some forfeitable value out of assets that may well have been acquired at least in part with dirty money. The government, thus, effectively took an all or nothing approach. The outcome: virtually nothing. King will forfeit none of the specific assets sought by the government other than the Invesco IRA account.133
2. Forfeiture money .judgment
King’s exposure to a forfeiture money judgment in this ease is $12,607,321.99, and that is what it will be. See, Part IX(C)(4), above. As for foreseeability, he has a serious problem. He was at the top of the organization and ran the show. The evidence establishes quite clearly that King was not a passive proprietor. He had trusted subordinates in management, a network of master agents and runners, and “a hundred or so [employees working in the call center] during the busy season.” Feb. Tr. 337-38. But he kept abreast on a day-to-day basis with the operations of his sports betting business. As noted in Part 1(C), above, King has expressly disclaimed reliance on the Excessive Fines Clause, so no test for dispro-portionality (which would probably be a tall order for King), or for deprivation of the ability to earn a living, applies to the court’s consideration of the government’s case for a forfeiture money judgment against King.
The net result of the government’s forfeiture case against King is: no forfeiture of property other than the IRA account and a money judgment against King in the amount of $12,607,321.99.
F. Findings specific to Koralewski
1. Forfeiture of specific assets
As has been noted, the government seeks to forfeit the proceeds of the sale of Koralewski’s house as a substitute res, as well as a gas lease and about $1,500 found in two bank accounts.
In support of the asset forfeiture it seeks, the government told the court, in its Forfeiture Brief, that: “Mr. Koralewski’s only source of income derived from his *1000association with Legendz Sports.” Doc. No. 1999-5, at 1. This was no fluke. The government repeated this representation to the court three months later: “Mr. Kora-lewski’s only source of income derived from his association with Legendz Sports.” Doc. no. 2130 at 21, ¶ 144.
These representations to the court by the government, which are important underpinnings of the asset forfeiture the government seeks against Koralewski, require an examination of the evidence from the October jury trial.
Testifying as a government witness at the October trial, Bruce Middlebrook, a Legendz bookie in central Oklahoma, told the jury that “he always knew” that Kora-lewski had a legitimate job. Oct. Tr. 410. This included jobs at Home Depot and Burlington Coat Factory. Id.
Jansen testified at the October trial that, as part of his investigation, he never looked into Koralewski’s employment history. Oct. Tr. 2110. So, he acknowledged that he “wouldn’t have any idea of what [Koralewski’s] employment history actually is.” Id.
Later in the October trial, three months before the government filed its Forfeiture Brief telling the court that Koralewski’s only source of income derived from his association with Legendz Sports, John Meoni, a homicide detective with the Denver Police Department, was called as a witness by Koralewski. Meoni testified that in about 2000, Koralewski moved to California and was employed by Home Depot there. Oct. Tr. 2215. Meoni stayed in contact with Koralewski. Id. Consequently, Meoni knew, and was able to testify, that, in California, Koralewski continued to work for Home Depot, after which Kora-lewski and his wife started “a sub shop or a deli shop.” Id. At 2216. Around 2004, per Meoni, Koralewski moved back to Colorado. Meoni and Koralewski started a construction company, but that lasted only about two months. Id. At 2217. Koralewski then started working for his brother installing windows and doors, after which he took a job with Burlington Coat Factory in loss prevention. Id. This would have been in the heart of the conspiracy period alleged in the Indictment, during which, according to the government, Koralewski had no legitimate income.
Meoni’s testimony at the October trial was specific and credible. The government did not take issue with his testimony, other than by softening its impact a bit by pointing to some exhibits indicating (correctly) that, during the relevant'period, Koralew-ski also had a relationship with DSS and UDS International Software, both of which were Legendz entities. Oct. Tr. 2220-21.
As stated above, this testimony was given in the October, 2015 jury trial. Three months later, with nothing in the record to cast the slightest doubt on the testimony of Detective Meoni, to say nothing of the uncontradicted testimony of Middlebrook, a government witness, the government made its repeated representations to the court that Koralewski’s only source of income derived from his association with Legendz. (The government’s Forfeiture Brief, and its later filing repeating the same averment, doc. no. 2130, at 21, ¶ 144, cited no supporting evidence, mainly because none existed.)
At the forfeiture trial, after acknowledging most of the particulars described above as to Koralewski’s history of legitimate employment during the conspiracy period (Tr. 672-73), Jansen acknowledged that he had not even been aware that Meoni had given the testimony described above. Tr. 673-74. Understandably, and to his credit, Jansen availed himself of the opportunity to clear the air on cross examination:
*1001Q. You’re not here to tell the Court, as the government has stated from the beginning of this case, that Mr. Ko-ralewski had no legitimate sources of income for the period of time that the indictment spans, correct?
A. No, sir.
Tr. 674.
The government’s general theory in support of forfeiture of KoralewsM’s assets is that he had no legitimate employment during the conspiracy period. The government’s specific theory is that the house in Parker, Colorado was paid for with those monthly $4,500 checks which the court has found have not been shown to have represented illegal proceeds. Doc. no. 2242, at 43, ¶ 162 (referring to deposits from the Costa Rican bank and citing GX 1120— $4,500 checks and deposits thereof). See also, Tr. 708-09 (Jansen confirming government’s theory).
Both theories are without merit. The proceeds of the sale of Koralewski’s house will not be forfeited. (And this is aside from the fact that the government could not, in any event, get both the proceeds of the sale of the house and a money judgment for the money used to pay for the house. Cf., Genova, 333 F.3d 750, at 762: “Forfeiting the whole amount of the bribes he received, and then forfeiting $60,000 of the equity in his home, would be excessive” because to forfeiture money once as cash and a second time as building materials would be double counting).
Now for the gas lease. The government’s singular statement in its Forfeiture Brief in support of forfeiture of the gas lease is: “Because Mr. Koralewski failed to earn legitimate income from 2003 through 2013, the gas lease is proceeds of illegal gambling.” Doc. no. 1999-5, at 2. No further discussion is necessary. The gas lease will not be forfeited. Nor, for the same reasons, will the $1,500 in the bank accounts.
2. Forfeiture money judgment
The forfeiture money judgment against Koralewski (who the court has previously found not to be subject to joint and several liability) will be in the amount of $10,000. This is the amount previously found by the court in its conclusions regarding the government’s proposed extrapolations and direct attributions. See, Part IX(C)(4), above. There is no issue as to foreseeability, because this $10,000 forfeiture is not imposed as a result of vicarious liability. As for analysis under the Excessive Fines Clause, the court is well-satisfied that the $10,000 forfeiture money judgment will be neither disproportionate under the Bajakajian factors nor of a magnitude that would jeopardize Koralewski’s ability to earn a livelihood. (As will be seen, considerations of foreseeability, disproportionality and deprivation of the ability to earn a living will be much more prominent in the analysis of forfeiture as to the three agents — Tucker, Bramley and Diebner — and the two runners, Moran and Robles.)
G. Findings specific to Tucker
1. Forfeiture of specific assets
At the forfeiture trial and in his briefing, Tucker devotes himself almost exclusively to his substantially successful arguments against the government’s proposed extrapolation of betting proceeds attributable to Tucker. As for asset forfeiture, the government has, prima facie, substantiated its case for forfeiture of the account listed in paragraph 4 on p. 1 of doc. no. 1999-7 (the $3,490.64 bank account). As is discussed in Part VIII(A)(3) above, that account is the only specific asset remaining in play as to Tucker.
2. Forfeiture money judgment
As stated above, the beginning point for determination of the amount of *1002the forfeiture money judgment against Tucker (and for the other forfeiture defendants who are potentially subject to joint and several liability) is Tucker’s nominal exposure in the amount of $12,607,321.99 (the sum of the agent attributions and the money laundering calculations).
This brings the forfeiture case to a point that is pivotal for the government and for the three agents and two runners whose forfeitures have yet to be individually addressed: the application of the governing principles with respect to foreseeability, proportionality, and deprivation of the ability to earn a livelihood. These principles have been discussed at length in this order (in Parts 11(B) and IV(A) as to foreseeability and in Part III as to proportionality and deprivation of livelihood). But it is appropriate to pause here to recap concisely the governing principles:
Foreseeability
• There is no issue as to foreseeability for purposes of RICO forfeiture, because there is no vicarious liability for “proceeds” forfeiture under RICO in the first place.134
• The court looks for illegal betting proceeds foreseeable to the defendant in question, although, in this case, that probably does not differ much, if at all, from looking for foreseeable criminal conduct.135
• In this case, if a given agent or bookie was at least generally known to the defendant in question, then the illegal betting proceeds generated or laundered by him were foreseeable.136
Proportionality
• As the Supreme Court put it in Bajaka-jian, the question is whether the forfeiture sought by the government would be “grossly disproportional to the gravity of a defendant’s offense.”
• This is an individualized determination.
• In hierarchical organizations, there will be substantial differences among the defendants in terms of their roles in the criminal activity, and hence, in the gravity of their offenses.
• The sheer amount of money involved in the offense is not necessarily a “perfectly calibrated” measure for purposes of determining proportionality.
• The maximum statutory fine is one useful reference point in determining proportionality.
• Other factors include: the extent of the criminal activity, related illegal activities, and the harm caused to other parties.
*1003Deprivation of Future Ability to Earn a Living
• The court must determine whether the forfeiture sought by the government would “be so onerous as to deprive a defendant of his or her future ability to earn a living.”
• Present inability to satisfy a forfeiture is not sufficient to render a forfeiture unconstitutional. It is permissible for forfeiture to render a defendant a pauper on Day 1. The question is whether he should be deprived of his ability to make a living beginning on Day 2.
• When a court considers constitutional imitations on forfeiture orders due to deprivation of a defendant’s future ability to earn a living, a court is not permitted to consider, as a discrete factor, a defendant’s personal circumstances, such as a defendant’s age, health or present financial condition. This is because forfeitures are not concerned with what a defendant actually has; stated differently, a defendant’s ability to satisfy a forfeiture at the time of conviction, in and of itself, is not at all sufficient to render a forfeiture unconstitutional, nor is it the correct inquiry. It is possible, however, that personal circumstances such as a person’s health or financial condition may bear on a defendant’s ability to make a living in the future and thus be indirectly relevant to this particular inquiry. Vilo-sM, 814 F.3d 104, 112 and n.15 (2d Cir. 2016). It is only in this indirect sense that this order considers any of the forfeiture defendants’ personal, current circumstances.
Foreseeability. With these governing principles in mind, the court turns to the issue of foreseeability as to Tucker. As has been noted, the court does not have the benefit of any advocacy from the government on the issue of foreseeability, an issue on which the government has the burden of proof. The question of foreseeability answers itself as to King, but it is not so simple as to the other forfeiture defendants.
As an initial matter, the court concludes that, with respect to the forfeiture defendants now before the court, the dollar amount of the individual attribution (for instance, $221,064.00 as to Tucker) provides little guidance because, for several reasons, that quantification of proven individual activity is a poor proxy for what was foreseeable, to any one defendant, by way of illegal activity on the part of his co-conspirators. Deriving much guidance from that number would lead either to an arbitrarily high reckoning of foreseeability or to an arbitrarily low reckoning. In either case, it would add an element of arbitrariness to an already decidedly imprecise process.
A much more rational beginning point is, instead, the $12,607,321.99 figure representing the nominal joint liability exposure of each of these forfeiture defendants. With this figure in hand, the court next examines the respective roles of these defendants. What did they do that would give them knowledge of what their co-conspirators were doing? Tucker was a well-established Legendz agent in Florida. The evidence presented in three long jury trials and several other days of evidentiary hearings persuades the court that the three main geographic areas for the Legendz sports betting operation were Florida, Texas and California. Oklahoma and other smaller states were in the mix, but the evidence points fairly strongly to Florida, Texas and California as the predominant markets.
Tucker operated in Florida, but he had good reason to have working knowledge of *1004the overall scope of the operations of Le-gendz. He was active enough in Florida to have an excellent knowledge of what was going on in that state. From his travels to the Legendz headquarters in Panama, he gained a good understanding of the overall scope of the Legendz sports betting operation. At least for a time, he had an apartment in Panama City. He knew, and spent time with, Legendz agents from other parts of the United States.
Based on Tucker’s in-depth knowledge of Legendz activities in Florida and his knowledge, both general and specific, of the domestic U.S. activities of his co-conspirators, the court finds that betting activity giving rise to no less than137 forty percent of the total amount of the nominal joint exposure was foreseeable to Tucker. Forty percent of $12,607,321.99 is $5,042,928.76. The court is content with that number based on all of the evidence before it. To the government, that resulting number may seem breathtakingly low. It may well be. But the court will act only on the basis of what the evidence reliably demonstrates. In short, this is where the deficits in the government’s forfeiture presentation come home to roost. Where the government’s proof falls short, either because of its untenability or because it was not presented at all, the court cannot bridge the gap by imagining what must have really happened and then turning that figment into a forfeiture money judgment.
As reduced for foreseeability, $5,042,928.76 is the amount of Tucker’s forfeiture money judgment exposure. The next step is consideration of the effect of the Excessive Fines Clause.
Proportionality. The court must make an individualized determination as to whether Tucker’s total forfeiture money judgment exposure — $5,042,928.76 at this point — would be grossly disproportional to the gravity of his offense. In a multi-defen-dant case, the court does not assess proportionality on an absolute scale. Granted, if every defendant in a conspiracy committed a perfectly heinous crime in furtherance of the objectives of the conspiracy, then there might be no room, in a proportionality analysis, for distinctions based on role in the organization. The facts would force everyone to the top of the scale, as sometimes happens under the federal sentencing guidelines. But as the First Circuit recognized in its decision in the Lyons case (affirming Judge Saris’s order), a larger forfeiture judgment can be proportionate if a defendant plays a larger role in the organization. Lyons, 740 F.3d at 733. Obviously, the converse is also true.
One aspect of the proportionality analysis that cuts in favor of the government is the fact that, to the extent that the reduction for foreseeability reflects a defendant’s limited role in the organization’s criminal activity, it would be double counting, of sorts, to also give that defendant the benefit of that limited role by way of a proportionality analysis. But the principles of foreseeability and proportionality do bring separate, and important, concepts into play, so the court will not treat the two as being completely overlapping.
As for Tucker’s role in the Legendz organization, he was lower than King but higher than the runners. He was higher than his subagents and local bookies. He was a valued member of the Legendz organization. And in this business, the pri*1005mary driver of value was the ability to reliably generate a substantial and steady flow of betting proceeds. Separating out the benefit Tucker has already received as a result of the court’s consideration of the issue of foreseeability, but taking into account separate factors such as societal harm, the benefit reaped by Tucker from his criminal conduct and the fine range established by Congress for his offenses, the court concludes that considerations of proportionality, scrubbed for double counting as discussed above, require a ten percent reduction in the amount ($5,042,-928.76) otherwise forfeitable by way of a forfeiture money judgment. The resulting number is $4,538,635.88. (This part of the analysis is arguably hocus-pocus, but the undersigned has been living with this case for four years, through four long jury trials, to say nothing of many other hearings, and is reasonably comfortable evaluating the relative culpability of the forfeiture defendants, as measured by the Bajakajian factors and other relevant considerations.)
Ability to earn a living. The last step is consideration of deprivation of future ability to earn a living. The government, although acknowledging the applicability of the Excessive Fines Clause (see Part III, above), and although on notice well before the forfeiture trial that the issue of livelihood was in play (see, n. 29, above) offers no suggestions at all as to the application of the Excessive Fines Clause to these defendants. The government argues for a $231,432,686.73 forfeiture money judgment against all of the forfeiture defendants and leaves it at that. The lack of advocacy from the government on this subject suggests either that the government is indifferent to whether the defendants are rendered paupers for the remainder of their lives or that the government will be content with the result the court may reach, left to its own devices. To borrow the words of the Court of Appeals as written in a different (but not altogether dissimilar) context: There is not “any ready template for conducting the constitutional analysis required under” the Excessive Fines Clause.” Lompe v. Sunridge Partners, LLC, 818 F.3d 1041, 1067 (10th Cir. 2016) (analyzing the constitutionality of a punitive damage award under the Due Process Clause of the Fourteenth Amendment). And, for that reason, reasonable jurists could certainly differ on these issues, as witness the analyses in the reasoned majority opinion and in the reasoned dissent in Lompe.
The facts on which the court relies in addressing the issue of deprivation of future ability to earn a living are uncontro-verted, which obviates any concern about allocation of the burden of proof, as discussed in Part IV(B), above. The relevant facts may be ascertained from Tucker’s presentence report.138
Tucker has more than $560,000 in liabilities. He has a significant negative net worth even if he is credited with the entire equity value of his house (i.e., even if the court disregards any spousal interest). Doc. no. 1821, at 29 (presentence report). His debts appear to be, at least in part, a legacy of a lifestyle that was augmented by his work as a Legendz agent. His liquid *1006assets are virtually nil and his retirement nest egg is negligible — less than $2,000. Comparing his present income with his legitimate income in past years, his earned income is a fraction of what it was. Id. A $200,000 forfeiture money judgment will put Tucker, a fifty-one year old man with diminished earning capacity, significantly under water on Day 1. But he has the ability and skills needed to work himself out of the hole beginning on Day 2. A forfeiture money judgment in that amount will be sufficient to put Tucker in the service of the government for a time, but not forever. A forfeiture money judgment in that amount will be entered against Tucker.139
H. Findings specific to Moran
1. Forfeiture of specific assets
As to Moran, the government seeks to forfeit $60,000 in currency, a money counter, a watch and a ring. Doc. no. 1999-3, at I.
The court is satisfied that the $60,000 is forfeitable. From all the evidence, the court concludes quite readily that this was cash that was in Moran’s possession only because of his service as a trusted Le-gendz runner. The nexus requirement has been satisfied.
The court has no evidence that would tend, at least directly, to establish a nexus as to the money counter. And the government does not proceed against the money counter on a facilitation theory. But the circumstantial evidence is rather compelling, in its totality, that this tool of the trade of a Legendz runner was either the property of the Legendz enterprise or was Moran’s property, acquired with illegal gambling proceeds. Either way, the nexus has been established to the satisfaction of the court.
The nexus has not been established with respect to the watch and the ring. There is no evidence as to how long Moran had them or how he got them. They will not be forfeited.
Accordingly, the asset forfeiture as to Moran will consist of items 1 and 2 on doc. no. 1999-3, p. 1.
2. Forfeiture money judgment
In determining the amount of the forfeiture money judgment to be entered against Moran, the court takes the approach and applies the standards discussed above with respect to Tucker.
Foreseeability. Moran was a Le-gendz runner in California. In its forfeiture papers, the government has made no argument and pointed to no evidence suggesting that the scope of Moran’s knowledge of the Legendz organization and its operations exceeded that’which he would gain as a result of his service as a runner in California. That said, it is also true that California was a major market for. Le-gendz. Moran provided runner service for Alan Campbell and other major Legendz agents in California. The beginning number as to Moran is, at this point, $12,607,321.99, as it is for all of the forfeiture defendants (other than Koralewski) before the court. Based on Moran’s thorough understanding of Legendz operations in California, gained mostly by virtue of his day-to-day contacts with his co-conspirators as a Legendz runner, the court finds that betting activity giving rise to no less than fifteen percent of the total amount of the nominal joint exposure was foreseeable to Moran on the basis of his knowledge of *1007the activities of the master agents, agents, bookies and bettors in California. Fifteen percent of $12,607,321.99 is $1,891,098.28.
Proportionality. On the issue of proportionality, Moran, as a runner (and Robles, as a runner — see below) gets significant benefit for reasons separate from those that are in play under the heading of foreseeability. The money Moran handled as a runner was not his. Having been convicted on the money laundering count, Moran is subject to forfeiture measured, as a beginning point, by the money “involved in” that offense, but it is nevertheless relevant that he did not generate the proceeds, he did not direct their ultimate use, and he never had a beneficial interest in those funds. He was not dealing for his own account, and he did not stand to profit as an agent or bookie would profit from the operation of the sports betting enterprise. As the Fourth Circuit recognized in United States v. Jalaram, 599 F.3d 347 (4th Cir. 2010), there is the potential for “a grave injustice in cases involving joint and several liability” unless the court pays attention to whether a defendant “played a truly minor role in a conspiracy that generated vast proceeds.” Id. at 355. Moran’s culpability stems not so much from any intrinsic evil in what he was physically doing (which did not differ much from what a commercial courier might innocently do) as from what he knew he was helping to accomplish as he did it.
Taking into account those factors that are relevant in a reckoning of proportionality (but separate from foreseeability and antecedent to consideration of deprivation of future ability to earn a living), the court concludes that Moran’s $1,891,098.28 exposure should be reduced by sixty percent (compare the corresponding ten percent reduction for Tucker). That leaves $756,439.31 as the court proceeds to consider the issue as to deprivation of future ability to earn a living.
Ability to earn a living. Moran has a high school education plus two years of college, studying social work. He had to drop out for financial reasons. He can weld, operate a forklift, and help in an office. Doc. no. 1632, at 23 (presentence report). He has also worked as a mechanic, but that is questionable now, as a result of a work-related injury that resulted in permanent partial disability. Id. And see, Defendant’s Exhibit 2 (Moran), California workers’ compensation award, indicating a 34% permanent disability.
Moran’s assets consist of two cars valued at a total of $34,000. He has no debt. He is a fifty-six year old man with noticeably limited earning capacity. Compared to Tucker, Moran has a higher net worth, at $34,000, but less auspicious earning prospects. (Tucker has a diminished earning capacity, as the court has noted, but that capacity is “diminished” from a much higher level.)
Moran has, and supports, a minor child. Doc. no. 1632, at 21-22. The court has given careful consideration to the question of whether, in assessing the impact of a forfeiture on Moran’s “future ability to earn a living,” Levesque, 546 F.3d 78, 85, the court should take the financial responsibilities inherent in parenthood into account. The answer, in this court’s view is an emphatic yes, at least where, as here, the defendant (unlike many who appear before this court) has willingly assumed the responsibilities of parenthood.140
*1008For Moran, a forfeiture money judgment for more than $35,000 would amount to a “ruinous monetary punishment[ ],” Levesque, 546 F.3d 78, 84, so onerous as to deprive a defendant of his future ability to earn a living. The forfeiture money judgment against Moran will be for $35,000.
I. Findings specific to Robles
1. Forfeiture of specific assets
The government does not seek forfeiture of any assets belonging to Luis Robles.
2. Forfeiture money judgment
Foreseeability. Robles was a runner in Florida. That suggests a limited role in the Legendz enterprise, as was the case with Moran. The evidence indicates, however, that Robles’ horizons within the Legendz organization were wider than Moran’s. In 2010, Robles made a trip to the Legendz headquarters in Panama. Feb. Tr. 415. He “stayed at Luke’s [defendant Bartice King’s] house for about a month.” Id. That certainly would have given Robles knowledge of the Legendz sports betting operation considerably beyond the knowledge he would glean as a runner in Florida. As has been noted, Robles was also a. partner with King in Platinum Vacations. That additional contact with King would have deepened Robles’ knowledge of the Legendz operation and its activities outside of Florida.
For purposes of the court’s evaluation of what was foreseeable to Robles, it is clear that even though Robles’ function within the Legendz organization was that of a runner, his familiarity with the operations of the enterprise was more like Tucker’s than it was like Moran’s. As the court noted in Part IV(A), above, in the Lyons case, Judge Saris found that betting proceeds generated by agents and bettors known to a defendant were reasonably foreseeable to that specific defendant. She went on to find that the government had not met its burden with respect to agents not known to a specific defendant even if the existence of such agents would be conceivable. That distinction, sensible though it clearly was in Lyons, would not, in the case at bar, give the government the full benefit of the scope of Robles’ activities within (or in proximity to) the Legendz organization.
The court concludes that betting activity giving rise to no less than forty percent of the total amount of the nominal joint exposure was foreseeable to Robles. In short, the unlawful proceeds-generating activity Robles was in a position to foresee put him on a par with Tucker. Forty percent of $12,607,321.99 is $5,042,928.76.
Proportionality. As for proportionality, Robles benefits (by way of comparison with Tucker and Diebner) for the same reason Moran does. Robles’ culpability is measured principally by the fact that he was only a runner. It is not influenced by the fact that he was a partner with King in a lawful business venture (Platinum Vacations) or by the fact that he knew King well and spent a considerable amount of time with him. In his capacity as a runner, Robles was a small cog in a big machine. An evaluation of Robles’ role in the sports betting operation as a runner in Florida, separate from considerations of foreseeability, leads the court to conclude that Robles’ $5,042,928.76 exposure should be reduced by sixty percent, as was the case with Moran. That leaves $2,017,171.50 as the court proceeds to consider the question of Robles’ future ability to earn a living.
*1009Ability to earn a living. Robles, who recently had his fifty-first birthday, has a high school education (in Puerto Rico) and a Florida real estate sales associate’s license, which he has held for more than fifteen years. Doc. no. 1869, at 24-25 (pre-sentence report). He has been employed on a commission basis in the real estate and time share business for several years. He has significant earning capacity, but, given the nature of his work, that might be better described as earning potential. He owns no substantial assets other than his house, two cars and a joint personal checking account. His net worth is less than $10,000, lower than Moran’s but within the same order of magnitude. The difference between Robles and Moran is the outlook for earned income. Taking into account Robles’ significant but somewhat speculative earnings outlook, together with his low net worth, the court concludes that a forfeiture judgment for more than $100,000 would amount to “ruinous monetary punishment.” That will be the amount of the forfeiture money judgment against Robles.
J. Findings specific to Bramley
1. Forfeiture of specific assets
As stated in Part VIII(A)(6), the government seeks, by way of forfeiture of specific assets, Bramley’s house, five investment accounts, his car, and slightly over $24,000 in U.S. currency, some of which is described as “collectible U.S. currency.” Doc. no. 1999-2, at 1.
Bramley’s house. The government seeks the house on two theories. Neither has merit. The first theory is what Jansen, in effect, called the “taint up the property” theory. Tr. 515. This is discussed in detail in Part IX(B)(4), above. The idea is that if bettor funds were used to make some mortgage payments, the property is gone. Jansen’s more genteel way of saying it was that if checks from “known and suspected bettors” were used to make mortgage payments, then “the property becomes evaluated for forfeiture,” (ie., gone). Tr. 514-15.
Bettor checks used to make mortgage payments amount to $114,882.41 (originally $117,968, but corrected to the lower figure. Tr. 522.). One problem is that these dollars are already, at the behest of the government, in the forfeiture money judgment calculation. See, Part IX(C)(3)(i), above. Another problem is that the government has made no pretense of separating the value attributable to illicit proceeds from the value that is not (doc. no. 1999-2 at 2-5; doc. no. 2242, at 31-32, ¶ 15) because the government argues, in essence, that it needn’t show anything more than the application of bettor checks to the mortgage (which clearly did happen). That will not suffice. See, Genova, 333 F.3d 750, 762.
The government’s other theory is facilitation, the argument being that Bramley used the house to facilitate his bookmaking. Doc. no. 2242, at 32, ¶¶ 116-17. The government cites photographs taken when Bramley’s house was searched. Tr. 516; doc. no. 2242, at 32, ¶ 116. They show: mortgage-related documents sitting on Bramley’s desk; FedEx shipping supplies; cards (apparently debit-type cards) from legitimate casinos such as Bally’s; a Scot-trade account card; an adding machine; a camera; Texas driver’s licenses and cash. GX 1284-1287; Tr. 515-21.
On the basis of these photos, Jansen ventured that Bramley’s use of his office in his house “possibly facilitated the gambling enterprise.” Tr. 521. The items in the photos do not amount to a smoking gun, individually or collectively. Indeed, it is of more than passing interest that a thorough search of the house of a prolific pay-per-head bookie did not turn up more damning papers than the ones described above. Searches of the homes of other bookies in *1010this case turned up reams of betting sheets and kindred records. But it is at least plausible that Bramley did use his house to facilitate his bookmaking. The court will so assume.
Bearing in mind that the government seeks to forfeit Bramley’s entire house, not just his home office, the question is whether Bramley’s use of his house to facilitate his bookmaking was anything more than incidental, as is discussed at length in Part 11(F), above.141 The government presented precious little evidence bearing in any meaningful way on the extent of that use. On this point, the testimony of Neil Myler, Bong’s most trusted runner in Texas, is noteworthy. In June, 2005, Myler made a trip to Dallas to collect money from Bram-ley. They met at a Wal-Mart. May. Tr. 1525. Myler estimated that he met Bram-ley between thirty and forty times. Id. 1527. Significantly, the only other place Myler met Bramley in these thirty or forty instances was Gecko’s Restaurant, north of Dallas. Id. 1527.
The court will not repeat its earlier discussion of the legal framework of the facilitation theory here. Suffice it to say that facilitation, as applied to a decidedly mixed-use asset, can be a slippery concept. Assuming, as seems reasonable, that Bramley did take care of some Legendz-related business from his home, it is clear that the Bramley house was not used to facilitate criminal activity in the way that a drug dealer would acquire and use a stash house to store illegal drugs or the cash generated by the sale of the drugs, or as a place to meet for the purpose of receiving and delivering the drugs.
As is discussed in Part 11(F), the government’s burden on the issue of facilitation is to show that the degree to which the property was used to facilitate the crime was substantial. Sporadic use of the property to facilitate criminal activity may, in those sporadic instances, make the prohibited conduct less difficult, but it is not enough that the property made the crime easier to commit in some incidental way. In that respect, there is a relevant comparison between the case at bar and the cases mentioned in Part 11(F). In Iacaboni, 221 F.Supp.2d 104, 115, a gambling case, the evidence did establish that envelopes with money and documents were dropped off and picked up in the garage of the defendant’s house. And it is instructive that, in Nicolo, 597 F.Supp.2d 342, it was not enough that the defendant used the property as a mailing address for criminal purposes and to fax invoices. The court said that this did not establish anything more than a fortuitous, transitory and tangential relationship to the criminal activity. What little evidence there is as to the locations to which Bramley had betting proceeds sent would suggest that he avoided using his house as a mail drop. May Tr. 1064 (Bramley directing that cash be sent to 2701C West 15th St. in Plano, Texas — not his house at 3120 Crooked Stick.)
Once again, the court’s evaluation must be based on what the evidence fairly shows, and not on what the court might guess was really the extent to which *1011Bramley used his house to facilitate his bookmaking activities. Measured by the law defining the concept of incidental use, the evidence in this case falls short of establishing the forfeitability of Bramley’s house. It will not be forfeited.
The five investment accounts. For the reasons discussed in Parts VI(A) and IX(B)(4), above, the investment accounts will not be forfeited.
Jansen had no idea how much money attributed to Bramley was “tainted up versus clean or legitimate.” Tr. 629-30. He had no idea how much of the money in the Scottrade accounts was dirty. Tr. 643. He “didn’t do any tracing to see how much money is actually sourced by legitimate funds versus illegitimate funds within any account in this case.” Tr. 649. The cases discussed in Part VI(A), above, especially the Tenth Circuit’s decision in Bornfield, 145 F.3d 1123, make it clear that the government’s “taint up” approach cannot satisfy its burden of proof with respect to commingled assets.
Bramley’s car. The government originally sought to forfeit a 2009 Mercedes Benz S500 it asserted Bramley owned.142 Doc. no. 1999-2 (Forfeiture Brief), at 1. As noted in part IX(B)(2), above (n. 64), it does not appear that the government still seeks to forfeit the Mercedes on a facilitation theory. But a brief review of the evidence may be appropriate in ease the court is mistaken about that.
The government originally asserted that Bramley drove a Mercedes Benz to meet Neil Myler. Doc. no. 1999-2, at 8. He may well have done that, but there is no evidence (and the government has cited none) to that effect. In Bramley’s jury trial, Mercedes Benz ears were attributed to a truly remarkable number of individuals who owned or drove them, but not to Bramley. See, e.g., May Tr. at 242, 380, 446, 454, 909, 1125, 1154, 1297 and 1641. In his testimony, Myler did not associate Bramley with any particular car.
Surely, Bramley used some car to go meet Myler at Gecko’s. The problem is that there is no evidence that he used the car the government alleges he owned. Given the property-specific nature of facilitation forfeiture, that is a problem. It would not have taken much for Myler to persuade the court that Bramley used a Mercedes to meet Myler at the restaurant, but he at least had to say that. At that point the court would have had to consider whether Bramley’s use of the car for purposes related to the offenses of which he stands convicted was incidental — which might have been a somewhat easier argument for the government than is the case with respect to Bramley’s house.
In its Forfeiture Brief, the government asserted that Bramley used funds, from his Viewpoint Bank account to make payments on a Mercedes. Doc. no. 1999-2, at 8. For the reasons discussed above with respect to Bramley’s financial accounts, that is of no moment. The government also asserted, originally, that a search of a Mercedes the government associates with Bramley yielded “a large sum of cash believed to be illegal gambling proceeds.” Id. No evidence relating to any such cash was presented at Bramley’s jury trial or at the forfeiture trial.
The Mercedes Benz the government attributes to Bramley will not be forfeited.
Cash. The government seeks to forfeit slightly over $24,000 in cash. At the forfeiture trial, the only testimony about Bram-*1012ley’s cash was the testimony about the cash found in his home.
Jansen was vague about where in the Bramley house some of the cash was found when it was seized. Tr. 518, 520-21. Some of the cash was found in a desk drawer with a stapler. Tr. 520. Jansen could not connect the cash to “any suspected or known gambler in this case.” Tr. 638. Jansen acknowledged that, as to the cash found in Bramley’s house, “there’s really no way to connect or disconnect that money from this case.” Tr. 639. On the basis of sheer amount, the cash found in Bramley’s house is suspicious. But the problem here — and, bearing in mind the burden of proof, this is the government’s problem — is that this is cash, not kryptonite. It is true that the court has found the satchels of cash that were delivered to King by McFadden to be dirty money, but a stash of cash sitting in the house of a 74-year old man is not freighted with the same connotations. No one would be shocked to find out that this cash in fact had something to do with illegal gambling. But that possibility is not the equivalent of proof, even on a preponderance standard. The cash will not be forfeited.
2. Forfeiture money .judgment
Foreseeability. Bramley had ample reason to be very familiar with the scope of the operations of the Legendz organization, which obviously bears directly on what was foreseeable to Bramley. Although he was a Dallas bookie, Bramley traveled to Panama from time to time. Paul Wilson, a Legendz bookie from California, saw Bramley in Panama, at Luke King’s casino. May Tr. 260. Karlo Stewart testified that he saw Bramley in Panama three or four times a year. May. Tr. 484. This included trips in February and March, 2010. May. Tr. 535, 536. Bramley also attended a security meeting in Panama on Super Bowl weekend in 2006. May Tr. 890,1005-06.
Bramley’s bookmaking business was brisk enough that he had “a numbers guy down in Panama.” May Tr. 853. This was Steve Polak. Id. Polak worked for Bramley in Panama, at least during the time before DSS relocated to its new quarters. May Tr. 491.
Bramley’s knowledge of the Legendz organization and its operations put him on a par with Tucker. Based on his extensive knowledge of Legendz activities in Texas and his knowledge, both general and specific, of the domestic U.S. activities of his co-conspirators (gleaned from what he saw and heard on his trips to Panama, if from no other source), the court finds that betting activity giving rise to no less than forty percent of the total amount of the nominal joint exposure was foreseeable to Bramley. Forty percent of $12,607,321.99 is $5,042,928.76. Accordingly, as reduced for foreseeability, $5,042,928.76 is the amount of Bramley’s forfeiture money judgment exposure. The next step is consideration of the effect of the Excessive Fines Clause.
Proportionality. As a pay-per-head bookie, Bramley’s role in the Legendz organization was qualitatively different, in some ways, from Tucker’s. But in those respects that would have a real bearing on proportionality, Bramley’s role did not materially differ. He was a valued Legendz bookie in Texas, like Tucker was in Florida. As a pay-per-head bookie, Bramley made book for his own account, but he paid substantial pay-per-head fees to Le-gendz, which made Bramley’s affiliation with Legendz profitable for King. By any measure, Bramley was a valued member of the Legendz organization (if, as a pay-per-head bookie, he can be considered to have been a member at all, a matter addressed by the court in United States v. Bramley, *10132015 WL 12852048 (W.D. Okla. Sept. 16, 2015)).
The court concludes — eliminating the benefit Bramley has already received by way of consideration of foreseeability— that considerations of proportionality require a ten percent reduction in the amount otherwise forfeitable by way of a forfeiture money judgment. The resulting number is $4,538,635.88.
Ability to earn a living. Allowing for a spousal interest in his house (which he is not forfeiting), Bramley has roughly the same net worth as Tucker, but much poorer earnings prospects. Bramley is out of business as a bookie, now working for not much more than minimum wage as a 74-year old helper in his son-in-law’s construction business. Bramley collects Social Security retirement benefits and does not earn enough to run afoul of the Social Security earnings limits. For thirty years ending in 2007, Bramley made a good living as the owner of a small construction company. His monthly income now is less than $3,500, which is less than his monthly expenses. A $200,000 forfeiture money judgment would likely impoverish Bramley for the rest of his life, or at least for a long time, even if he raised some money by moving to a modest house. That said, the court is not obligated, in assessing what would deprive a defendant of his livelihood, to protect a defendant’s ability to live in the style to which he has become accustomed, if, in so doing, the court would be reducing the forfeiture recovery to virtually nil. The court concludes that a forfeiture money judgment in the amount of $100,000 would put Bramley well under water on Day 1, but would leave him the ability to earn a living going forward. A forfeiture money judgment in that amount will be entered against Bramley.
K. Findings specific to Diebner
1. Forfeiture of specific assets
The government originally sought to forfeit Diebner’s house, three financial accounts, $8,401 in currency, and miscellaneous items of jewelry. Doc. no. 1999-4, at 1. None of these items were mentioned at the forfeiture trial or in the government’s Proposed Findings of Fact or Conclusions of Law. They will not be forfeited.143
2. Forfeiture money judgment'
Foreseeability. In assessing what Legendz-related sports betting (and resultant money laundering) operations were foreseeable to Diebner, it is natural to compare him with Tucker and Bramley, the other two bookies among the forfeiture defendants. The short version is that Dieb-ner’s involvement in activities not directly related to his own bookmaking was noticeably limited. Compared to Tucker and Bramley, Diebner had much less reason to have a working familiarity with the overall scope of the sports betting activities of the Legendz organization. This directly affects the extent to which Diebner can be said to have had knowledge that would have made betting or bookmaking activity involving persons other than himself foreseeable.
Antonio Jenkins-Lara, who spent a fair amount of time at or near the Legendz headquarters in Panama had no recollection of ever having seen Diebner. May Tr. 368. More tellingly, Karlo Stewart had only “heard the name.” Id. 528. Stewart *1014had never met Diebner or seen him at the Legendz headquarters. Id. 615. Michael Lawhorn, who was no stranger to the Le-gendz headquarters in Panama, had never met Diebner, and had never seen him at various events in Panama that were referred to throughout the May jury trial. May Tr. 954, 966, 976. On the other hand, Diebner did go to Panama for three days in early February, 2010 — specifically February 5th-8th, which coincided with the Super Bowl. May Tr. 530. Neil Myler, who met Diebner in 2005 and was certainly in a position to know who (besides himself) would be holding money for Legendz in Texas, had no recollection of Diebner ever hiding money for King. May Tr. 1653, 1721.
Based on Diebner’s limited knowledge of the scope of Legendz-related activities other than his own activities in Texas, the court finds that betting activity giving rise to no less than ten percent of the total amount of the nominal joint exposure was foreseeable to Diebner. Ten percent of $12,607,321.99 is $1,260,732.19.
Proportionality. Diebner’s role in the overall Legendz sports betting and money laundering conspiracy was limited — as witness the fact that, due to considerations of foreseeability, the court has just reduced Diebner’s exposure to a forfeiture money judgment by ninety percent. To be sure, Diebner’s sports betting business was good enough to require, at least for a time, a runner working for Diebner. May Tr. 1530. But, functionally, Diebner was not in the same league with Tucker or Koralewski as a confidant of King. The evidence does not establish that Diebner helped the Legendz organization achieve its goals in significant ways aside from the benefit naturally flowing from his activities as a successful Le-gendz agent. Taking into account Dieb-ner’s actual role as a local, but apparently successful, Legendz agent in Houston, and assessing that role on the basis of considerations aside from those that made a difference with respect to foreseeability, the court concludes that an additional ten percent reduction in the amount ($1,260,-732.19) otherwise forfeitable is required. The resulting number is $1,134,658.97.
Ability to earn a living. Diebner, a forty-four year old man with a history of profitable employment in fields other than bookmaking, has better prospects than Bramley. His present earning power from legitimate pursuits is in the same general range as Tucker’s. Setting aside from Diebner’s assets a spousal interest in his house, Diebner is left with a net worth of less than $30,000. As is the case with Tucker, Diebner’s liquid assets are virtually nil and he has no retirement savings. Doc. no. 1656 at 28 (presentence report).
Unlike Tucker and Bramley, Diebner has minor children — two of them, both younger than eleven. He is their provider. Doc. no. 1656 at 21. As is discussed above with respect to Moran, this is relevant to the present analysis. Taking into account Diebner’s net worth and earning power, viewed in light of the responsibilities that bear on the question of just what constitutes “a living” in his case, the court concludes that Diebner should be rendered liable for $75,000 by way of a forfeiture money judgment.
L. Summary of forfeiture-related findings
*1015Table 3
Results re: Forfeiture Money Judgment
(All amounts in dollars)
Defendant Individual Attribution ("Agents'1 on GX C89) Joint Exposure (individual attribution total + money laundering total) As reduced: Foreseeability As reduce.d: Excessive Fines
King (See subtotal, below) 12,607,321.99 No reduction No reduction144
Koralewski 10,000.00 N/A N/A No reduction
Tucker 221,064.00 12,607,321.99 5,042,928.76 200,000.00
Moran 0 12,607,321.99 1,891,098.28 35,000.00
Robles 110,459.00 12,607,321.99 5,042,928.76 100,000.00
Bramley 4,489,133.19 12,607,321.99 5,042,928.76 100,000.00
Diebner 792,100.00 12,607,321.99 1,260,732.19 75,000.00
Middlebrook 288,384.00
McFadden 988,000.00
Subtotal for Attributions 6,899,140.19
Add: Money Laundering: 5,708,181.80
Total: 12,607,321.99
The sum total due the government by way of the forfeiture money judgment is a unitary amount, as can be discerned from GX C89 and from the government’s proposed findings and conclusions. Doc. no. 2242 at 48, ¶ 189. That is inherent in the government’s theory of its forfeiture case: the government asserts that all of the forfeiture defendants are liable for all of the amount adjudicated to be due. But that unitary — i.e., joint — liability is moderated by the foreseeability requirement and the Excessive Fines Clause, to the extent set forth in this order with respect to some of the forfeiture defendants. The forfeiture money judgment liability of the forfeiture defendants is joint and several (except as to Koralewski, for reasons explained earlier), but no forfeiture defendant is liable for more than the amount stated in this order with respect to his individual forfeiture money judgment obligation. The liability is also joint in the sense that collection from any forfeiture defendant will be credited to all forfeiture defendants.
X. Forfeiture-Related Sanctions Against the United States
Diebner, joined by the other forfeiture defendants, asks the court to reduce the amount of a forfeiture money judgment otherwise appropriate “in light of the prosecutorial misconduct that has permeated these proceedings.” Doc. no. 2274, at 10. Diebner cites Caplin & Drysdale, Chartered v. United States, 491 U.S. 617, 109 S.Ct. 2646, 105 L.Ed.2d 528 *1016(1989), as confirming the unremarkable proposition that the lower courts do have the authority to deal with “particular abuses” in forfeiture cases “when and if any such cases arise.” Id. at 635, 109 S.Ct. 2646. See also, Libretti v. United States, 516 U.S. 29, 43, 116 S.Ct. 356, 133 L.Ed.2d 271 (1995) (same, citing Caplin & Drysdale).
The question is whether this is such a case.
Diebner complains, in particular, of: “[0]bjectively incorrect assertions in the Government’s forfeiture briefs.” Doc. no. 2274, at 11 (referring to colloquy between court and government counsel at the conclusion of the forfeiture trial), and government attempts to “abuse and overreach with its money judgment calculations [and] extrapolations,” elaborating that the proposed extrapolations are not “grounded in reality.” Id.
These complaints are not frivolous, and will be addressed, but, as an initial matter the court should make clear what it will and will not consider with respect to the complaints advanced by Diebner and (by adoption) the other forfeiture defendants. At an earlier stage of this case, the court commented pointedly with respect to some aspects of the conduct of two of the government lawyers in this case. United States v. Bramley, 2015 WL 12852048, at **16-22 (W.D. Okla. Sept. 16, 2015). Those lawyers were not the ones who presented the government’s forfeiture case. The matters described in the order just cited did not have a direct bearing on the forfeiture proceedings in this case. Moreover, as the court has previously indicated (doc. no. 1880), those matters will be addressed in open court at an appropriate time. Consideration of those matters is beyond the scope of the present discussion.
What the court will consider are those matters that arose in the course of these forfeiture proceedings. The court’s inquiry is two-fold: (i) What, exactly, occurred that could be considered relevant to the court’s consideration of Diebner’s complaint and request for sanctions, and (ii) What effect, if any, did those matters have on the outcome of these forfeiture proceedings.
Diebner is correct — the government has made significant objectively incorrect factual assertions in these forfeiture proceedings. That complaint is, of course significant in itself, but it also has some impact on the “abuse and overreach” of which Diebner complains. The material misstatements of fact that the government has made in writing are in some respects compounded by, and in some respects separate from, a good many vexing (and avoidable) factual problems inherent in the government’s presentation at the forfeiture trial. Because of the interrelationships between the government’s written statements and its presentation at the forfeiture trial, all of those matters will be addressed here.
The record shows (with references to the relevant parts of this order):
• Repeated, demonstrably untrue assertions that King had no legitimate income during the conspiracy period. Part IX(A)(1).
• A positive and unqualified factual statement by an FBI witness as to a matter he later conceded was not within his knowledge. Part IX(B)(2) (re: Paige Overseas).
• Inclusion of significant amounts of deposits in the forfeiture summary with, admittedly, “no way of knowing” whether the money came from illegal sources. Part IX(B)(2) (re: GX C89, line 13).
• An FBI witness purporting to have assisted in preparing a summary exhib*1017it (and thus to be knowledgeable about its substance) when, in fact, he was unable to give meaningful testimony shedding light on its reliability and had no significant role in preparing it. Part IX(C)(3)(f) (re: GX C55).
• An FBI witness attributing betting proceeds to Anish Patel, Robert Hurst and Craig Hardy without even knowing (Part IX(C)(3)(n)):
• That Patel had himself testified at Koralewski’s jury trial, six months before the forfeiture trial.
• That Hurst himself had testified at Koralewski’s jury trial.
• That William Durborough had testified at Koralewski’s jury trial as to the betting activity of the deceased Craig Hardy. (As to all three witnesses: “I didn’t know that these guys had testified.”)
• That the trial testimony of Patel, Hurst and Durborough thoroughly undermined the government’s factual contentions as to attribution of betting proceeds to Patel, Hurst and Hardy.
• Repeated written assertions — contrary to overwhelming and uncontradicted jury trial testimony of which the FBI witness and the AUSAs presenting the forfeiture case were unaware — that Koralewski’s only source of income derived from his association with Le-gendz. Part IX(F)(1).
The matters summarized above, are, of course, described in detail in the cited sections of Part IX of this order. Not addressed there, but appropriate to address here, is some of the backstory.
The government’s repeated untrue assertion that King had no lawful income during the conspiracy period has been noted several times in this order. That assertion, if true, would have led to a significantly more favorable result, for the government, than the result reached in this order with respect to forfeiture by Bartice King. Moreover, if the government attorneys and forfeiture investigators had been aware of the facts as to King’s legitimate income, perhaps they could have dealt with that by marshaling evidence to establish that, in the grand scheme of things, that lawful income was so paltry as to be negligible in comparison to the proceeds generated by the illegal sports betting operation. The government, however, made no attempt to determine the ratio of legitimate funds to illegitimate funds going into the accounts it scrutinized in this case. Tr. 629, 649. For that reason, the following exchange during cross examination of Lowrance is noteworthy:145
Q. [by Mr. Mays] Are you aware that there were portions of the Legendz organization that were involved in legal sports betting in Panama as well as legal sports betting in other parts of the world?
A. I’m not aware of that, no.
Q. You never discussed with [FBI] Special Agent Bowles or [IRS] Special Agent Battershell [now Bowles] or Investigator Melzer or anyone else the fact that Legendz had a legal sportsbook in Panama with a license and they did business with other countries?
*1018A. No.
Tr. 239-40.
One or the other (and usually both) of the case agents were present in the courtroom during the jury trials in which it was made unmistakably clear, from government witnesses, that King did, in fact, have lawful entrepreneurial activities that generated lawful cash flow during the conspiracy period.
The forfeiture trial brought out, in cross examination of Jansen, similar concerns with respect to the representations made to the court about Koralewski:
Q. [by Mr. Hudson] I’m talking about relevant information. Would you agree that if they’ve [the case agents] handed you a spreadsheet that says that Mr. Koralewski is responsible for a bet from a gentleman who has told the FBI that he never made a bet with Mr. Koralew-ski and he’s testified before this judge, after having been sworn, under the penalties of perjury in front of a jury, and said he’s never made a bet with Mr. Koralewski, that’s relevant information to your determination about whether or not those funds should be attributed to Mr. Koralewski in a forfeiture judgment?
A. Yes.
Q. Now, we can safely assume, given that answer, you were not provided all of the relevant information in this case?
A. Yes.
Q. Can you tell me, other than the blind faith that you’ve placed in these people, that the information you were given as to any of the other defendants whose forfeiture judgment amounts you’ve worked on that you received all of the relevant information?
A. I don’t know.
Q. You can’t tell me that, can you?
A. No.
Q. Okay. And because you can’t tell me that, you can’t tell the Court that any of the figures that you’ve arrived at are, in fact, reliable: Yes or no?
A. I feel like the ones I’ve put together are reliable, yes.
Q. Because you want to believe you got all of the relevant information, but you’ve now admitted you don’t know that?
A. That’s correct.
Q. Okay. They’re reliable if and only if you were, in fact, provided all of the necessary and relevant information to make the determinations you’ve made?
A. Yes.
Tr. 721-23.
At the end of the forfeiture trial, the court considered it appropriate to direct a limited inquiry to government counsel for the purpose of shedding more light on the matters that are addressed here:
[THE COURT:] Now, I have one or two questions for Mr. McGarry. And I’m not going to dwell on this, but as an example, I’m looking at Appendix 5 to the government’s opening brief for forfeiture. And you don’t have that in front of you, so I’m just going to give you an exact quote.
“Furthermore, Mr. Koralewski’s only source of income derived from his association with Legendz Sports.” Next page, “Because Mr. Koralewski failed to earn legitimate income from 2003 through 2013, the gas lease is proceeds of illegal gambling.”
*1019That was rather eye-catching. Based on the government’s own evidence[146] last year, that was — those were two rather eye-catching representations to the Court. They were unqualified. “Mr. Ko-ralewski’s only source of income derived from his association with Legendz Sports.” Another unqualified statement, “He failed to earn legitimate income from 2003 to 2013.”
Who did you rely on for that, for those— are those yours? Did you come up with that? Or who did you rely on for those representations to the Court?
MR. McGARRY [AUSA]: Your Honor, when reviewing some of the documents and reports, that was the conclusion that was reached.
THE COURT: Okay. That was a conclusion who reached?
MR. McGARRY: I’d have to look at the specific record that I reviewed, but as far as I’m aware, it would have been from — -that’s Mr. Koralewski and forfeit — it probably would have been a forfeiture brief written by either an agent or investigator from the—
THE COURT: Mr. Bowles?
MR. McGARRY: As far as I know — I can’t recall right now, but—
THE COURT: I mean, the government’s own testimony in a previous trial covered his outside employment.
MR. McGARRY: Yes, Your Honor.
THE COURT: So who gave you that information?
MR. McGARRY: I mean, my communications were with the FBI.
THE COURT: Okay. I don’t want corporate references to the FBI. What person?
MR. McGARRY: Probably would have been Investigator Melzer.
Tr. 771-73.
This brings the court to one other aspect of the matter that is both relevant to the present inquiry and, aside from that, ought to be mentioned as a matter of fairness. The government has had a total of eight counsel of record in this case, in various combinations and at various times, only one of whom has been in the case from the beginning. The AUSAs who presented the forfeiture case were brought into the case at a late stage. The court is quite confident that neither those AUSAs nor any other attorneys in the U.S. Attorney’s office for this district knowingly made misrepresentations of fact to the court. Having been assigned to the case about twelve years after the investigation began, more than two years after the original Indictment was returned, and after four jury trials had been completed, the AUSAs who were assigned to the case at the forfeiture stage were at the mercy of others on the prosecution team, and this was true to an extent far greater than would ordinarily be true, even in relatively complex cases. As for those on the prosecution team other than the newly-assigned AUSAs, the court will refrain, also as a matter of fairness, from making any extended evaluative comment on an individual basis. But it does appear to the court that, as far as some others on the prosecution team were concerned, the witnesses at the jury trials who gave testimony that was potentially inconvenient to the government for forfeiture purposes *1020were damnatio memoriae. As has been seen, the net effect of this eompartmentali-zation of information was that the government’s factual contentions in support of its forfeiture case ranged from the dead-on to the surreal, leavened here and there by the merely untenable.
Before determining whether this is “such a case,” as contemplated by the Court in Caplin & Drysdale, one other matter should be addressed.
Diebner, joined by the other forfeiture defendants, also argues that the court should reduce the amount of forfeiture because the government failed to give adequate or timely notice to the defendants as to the government’s intent to seek a joint and several judgment against all the forfeiture defendants in the full amount the government sought at the forfeiture trial— more than $200 million (as reduced from $1 billion). See, Table 1 in Part IX(A), above. While the court has determined that the notice was not so insufficient as to preclude entry of a forfeiture money judgment, the court has not addressed defendants’ notice argument as a possible ground for other remedial action.
The court recognizes that the amount of a forfeiture money judgment sought by the government may be a moving target as the evidence develops. That said, the government’s stated intentions as to the size of the money judgments it has sought from each of the forfeiture defendants at different stages of this case, have been all over the lot. Here are some examples:
• The Superseding Indictment refers to forfeiture of one billion dollars. This target number was unequivocally reiterated by an FBI witness as recently as the October trial (the last of the four jury trials).
• However, when the government moved for preliminary orders of forfeiture from King, Koralewski, Tucker, Bram-ley, Zapt, and Diebner, its motions mentioned nothing about seeking a personal money judgment against these defendants.
• Some of the defendants’ briefs point out that the first time the government announced its intention to seek a money judgment against them was in the government’s Forfeiture Brief, filed January 19, 2016, doc. no. 1999. At that time the government attached exhibits which stated what it was seeking from each defendant. That Forfeiture Brief was required by the court’s management order to be very specific. The size of the money judgment sought against each forfeiture defendant, as stated in the exhibits to the government’s Forfeiture Brief, varied from defendant to defendant, as shown in Table 1, above.
Then, despite its detailed representations in its Forfeiture Brief as to the relief it sought, the government asked, in its post-hearing proposing findings and conclusions, (doc. no. 2242, p. 48, ¶¶ 189, 190) for a money judgment against each of the forfeiture defendants (referring to “Mr. King and the defendants convicted of the RICO counts,” which covers all of them) in the amount of $231,432,686.73 (or $98,911,830.31 if no extrapolation). As stated in Part VII of this order, the law allows the government a reasonable degree of flexibility in shaping its forfeiture case. The problem here is that given the number of defendants which the government chose to prosecute, the court put case management orders in place to govern these proceedings, with which the government complied in its Forfeiture Brief but then ignored in its subsequent filings by seeking a money judgment of over $200 million from everybody involved in these proceedings.
*1021But for the effect of the application of the principles embodied in the Excessive Fines Clause (e.g. the Bajakajian factors and, separately, the livelihood issue), the court would, without hesitation, and as a matter of fundamental fairness, hold the government to the forfeiture money judgment figures it posited in its Forfeiture Brief.
Two separate (but perhaps related) restraints on the court’s discretionary power to impose a sanction on a litigant are in play here, and the court has concluded that these restraints are fatal to the forfeiture defendants’ request for sanctions.
First, the request for imposition of sanctions on the government amounts, in this instance, to an appeal to the exercise of the court’s inherent powers. In that vein, the court is mindful that “[bjecause of their very potency, inherent powers must be exercised with restraint and discretion.” Chambers v. NASCO, Inc., 501 U.S. 32, 44, 111 S.Ct. 2123, 115 L.Ed.2d 27 (1991). The need for restraint is heightened where, as here, the aggrieved parties’ argument amounts to a request that one co-equal branch of government rebuke another one, in the process denying the latter judicial relief to which it would otherwise be entitled. Cf. Degen v. United States, 517 U.S. 820, 823, 116 S.Ct. 1777, 135 L.Ed.2d 102 (1996) (“The extent of these [inherent] powers must be delimited with care, for there is a danger of overreaching when one branch of the Government, without benefit of cooperation or correction from the others, undertakes to define its own authority.”).
Second, and equally significant here, it is clear from the Supreme Court’s decisions in Caplin & Drysdale and Libret-ti that the reason for depriving the government, because of its conduct, of some measure of the forfeiture relief to which it would otherwise be entitled, must be remedial, not punitive.
Both the need for restraint and the limitations on the purpose for which a sanction might, in these circumstances, be imposed, counsel denial of the forfeiture defendants’ request for sanctions. To take the second factor first, the court will assure the forfeiture defendants that it has taken care to eliminate from its adjudication of the government’s forfeiture claims any advantage the government might have gained on account of the matters addressed in this part of this order. In short, remedial action by way of imposition of sanctions on the government is not necessary to put the forfeiture defendants in as good a position as they would have been in had the government conducted itself punctiliously throughout these forfeiture proceedings. That alone requires denial of sanctions. But if it were a close question, the considerations of restraint discussed above would carry the day.
The forfeiture defendants’ request for sanctions against the government is DENIED.
This is as good a place as any to add one final note on this distasteful subject. Although extended evaluative commentary on the government’s performance in this case might, by some lights, be appropriate, the court chooses not to go down that road. There is more that could be said, but the court has not, in this order, gone down that road any farther than necessary to fairly adjudicate the government’s forfeiture claims against these forfeiture defendants, with, in some instances, an exposition of the reasons for which the record undermines the confidence the court might otherwise have in the reliability of the government’s factual assertions. However, the court cannot but get the impression *1022that, from the standpoint of the FBI, the dollar amounts being bandied about in this forfeiture case are, essentially, play money. What other explanation could there be for touting more than $70 million in extrapolations based on the bookmaking activity of five highly cooperative witnesses147 without so much as picking up the phone to ask any of them whether the extrapolated numbers have any relationship to reality? No good explanation readily comes to mind for the flimsy extrapolations and the superficial — quit looking when you think you have enough in hand to run the number way up — investigative approach of the FBI’s forfeiture witnesses.
The chain of events that brought this case to its present pass is long and complex. The court does not have all (and probably not nearly all) of the facts that would need to be on the table as a prerequisite to any more evaluative commentary than is already in this order. So the court will let the facts plainly shown by the record speak for themselves. The reader can decide for himself whether the government has approached the forfeiture case with any real sense of intellectual rigor. The reader can decide for himself whether the government’s approach to this case bespeaks anything more plainly than a desire to run up the score — in this case, into the hundreds of millions. But one thing should be made clear. The court has not set out to embarrass the government in this order. To the extent, if any, that the government is embarrassed by any of the matters that are set out at length in this order, the government has embarrassed itself.
XI. Conclusion
In accordance with Rule 32.2(b)(2)(B), Fed.R.Crim.P., preliminary orders of forfeiture preliminarily granting the relief to which the government has been found to be entitled will be entered on this date. Rule 32.2 states that “Unless doing so is impractical, the court must enter the preliminary order sufficiently in advance of sentencing to allow the parties to suggest revisions or modifications before the order becomes final as to the defendant under Rule 32.2(b)(4),” ie. at the time of sentencing. Given the scope of this order, the court does not intend to revisit its adjudications of contested legal or factual issues. But if, for example, the court has misidentified property in this Memorandum Opinion and Order Re: Forfeiture, or in the individual preliminary orders of forfeiture to be entered separately later today, or if any party believes there is a mathematical error which should be corrected, that type of specific matter may be brought to the court’s attention by motion filed pursuant to Rule 32.2(b)(2). If the parties agree regarding such matters, the motion shall so state. Any such motion is DUE within fourteen days of the date of this order. If, after that, further briefing is required, the court will set a schedule.
Sentencing proceedings for the forfeiture defendants will be scheduled as will be set forth in Case Management Order No. 8, to be entered on this date.
EXHIBIT C89
*1023[[Image here]]
*1024[[Image here]]
*1025[[Image here]]
*1026[[Image here]]
. When this order refers to "King,” it refers to Bartice A. "Luke” King. Bartice King’s wife, Serena King, was also a defendant in this case. She was acquitted on all counts.
.With regard to a potential money judgment against Zapt, the government has made general statements like those in doc. no. 2242, p. 48, ¶¶ 189 and 190 (government’s proposed findings and conclusions), in which the government submits that "Mr. King and the defendants convicted of RICO counts” are subject to a money judgment in the amount of $231,432,686.73, a set of defendants which would include Zapt. And the government’s Forfeiture Brief (doc. no. 1999, p. 17) includes Zapt among the defendants as against whom it seeks a money judgment. However, this is the very brief which, in the attached exhibits, tailored to the individual defendants, specifies no money judgment sought against Zapt. Doc. no. 1999-7 (identifying money judgment as sought only against Tucker). Consistent with their apparent understanding that no money judgment is now sought against Zapt, Tucker and Zapt’s jointly-filed proposed findings and conclusions focus on challenging the money judgment sought against Tucker. Doc. no. 2276.
. The government’s papers also reference defendant Kelly James Dorn, but no forfeiture issues remain as to Dorn. See, doc. no. 2138 (agreed upon Final Order of Forfeiture).
. In this order, the transcript of the forfeiture trial, held after the four jury trials were completed, is cited as: "Tr.” The four jury trials were all held in 2015. The transcript of the February, 2015 trial is cited in this order as “Feb. Tr.” The transcripts of the April, May and October trials will be cited consistently with that format.
. In this order, page references to filed documents (other than transcripts) are to the pages assigned by the EOF system.
. The government does not seek forfeiture of particular property from Robles, but it does seek a personal money judgment against him.
. Despite the court's clear order requiring the government to identify in its Forfeiture Brief all statutory bases pressed in support of these forfeitures, the government’s proposed findings and conclusions, filed just three months later, expanded the list of forfeiture statutes relied on by the government by citing 18 U.S.C. § 981(a)(1)(C) in support of forfeitures based on Count 2 convictions. Doc. no. 2130, ¶ 228, p. 36. The government's post-hearing proposed findings and conclusions also cite that statute. Doc. no. 2242, ¶ 249, p. 60.
. King expressly did not adopt arguments based on the Eighth Amendment. See, doc. nos. 2283, 2288. King also stated that by adopting others’ arguments, he did not waive his position that the Karlo Stewart documents and their fruits must be excluded under the Fourth Amendment.
. The statute relied on by the government to support money laundering forfeiture in these proceedings does not use the term "proceeds.” But its broad language encompasses proceeds generated by the money laundering conspiracy, as well as any tainted money laundered (proceeds of predicate specified unlawful activity), and money used to disguise the tainted funds. See, United States v. Bornfield, 145 F.3d 1123, 1135 (10th Cir. 1998) (property "involved in” an offense includes commissions or fees paid to the launderer, money being laundered i.e. the corpus, and any property used to facilitate the laundering offense such as money used to disguise the nature and source of the laundered money).
. United States v. Wilson, 244 F.3d 1208, 1212 (10th Cir. 2001), notes that the Tenth Circuit has yet to rule on whether the proper standard for forfeiture (under 18 U.S.C. § 853, in that case) encompasses joint and several liability for gross proceeds.
. The government’s Forfeiture Brief (doc. no. 1999), and its proposed findings and conclusions (doc. no. 2242), assert joint and several liability only as to forfeiture predicated on the conspiracy counts. See, e.g., doc. no. 1999 at 13, 18; doc. no. 2241, ¶¶ 208 (all co-conspirators), 211 (RICO conspirators), 238 (RICO conspirators).
. The foreseeability requirement has been articulated in different ways, potentially leading to divergent results. Does the court look for foreseeable proceeds or for foreseeable criminal conduct? See, e.g., United States v. Christensen, 828 F.3d 763, 824 (9th Cir. 2016), cert. denied, — U.S. —, 137 S.Ct. 628, 196 L.Ed.2d 517 (2017) (discussing distinction between foreseeable proceeds of and foreseeable conduct, and adopting a foreseeable criminal conduct standard). In another part of this order, this court comes down on the side of those courts which describe the foreseeability requirement as contemplating proof by the government that betting proceeds or funds laundered were reasonably foreseeable to the forfeiture defendant in question. See, Part IV(A), below. Regardless, in this case, because of the nature of the betting enterprise, there is, as to any one defendant, at least a rough parity — a roughly linear relationship — between the foreseeable criminal conduct of any one agent or bookie and the proceeds foreseeably generated by that conduct or laundered as a result of it. For that reason, the court finds that under either version of the foreseeability standard, its findings with respect to each individual forfeiture defendant would be the same. It should be noted, also, that not all courts impose a foreseeability requirement. See, e.g., United States v. Yass, 636 F.Supp.2d 1177, 1186 (D. Kan. 2009) (noting that a number of courts hold that co-conspirators in a criminal scheme are jointly and severally liable for all proceeds generated under the scheme, collecting cases at n.51; while other courts hold that a defendant is only required to forfeit so much of the proceeds of the fraud, to the extent they were not received by him, as were foreseeable to him, collecting cases at n.52). Given the far-flung nature of the conspiracies of which certain defendants were convicted in this case, and given the scope of the proceeds sought via money judgments, it is appropriate to require proof of foreseeability before entering judgments for sums which were generated or laundered by others.
. See, motions and orders at doc. nos. 2277, 2279, 2280, 2281, 2282, 2283, 2284, 2285, 2286, 2287, 2288, and 2289.
. As noted above, foreseeability is not in issue as to the Count 2 convictions because the government only asserts joint and several forfeiture liability as to the conspiracy counts. Foreseeability is not in issue as to RICO forfeiture, because, as is discussed in Part C, below, the court has concluded as a matter of law that there is no joint and several (i.e., vicarious) liability for forfeiture under RICO.
. As further support for the conclusion that ''proceeds” in this context does not mean net profits, see United States v. Keeling, 235 F.3d 533 (10th Cir. 2000), cert. denied, 533 U.S. 940, 121 S.Ct. 2575, 150 L.Ed.2d 738 (2001), a pre-Santos decision involving forfeiture under 21 U.S.C. § 853 following drug convictions. In Keeling, the defendant argued he should only be responsible in forfeiture for the amount of the profits rather than the gross proceeds of his narcotics enterprise. The Court of Appeals described this argument as "utterly without merit.” Id. at 537. Keeling quotes United States v. McHan, 101 F.3d 1027, 1042 (4th Cir. 1996), as follows.
Were we to read proceeds in § 853 to mean only profits, ... we would create perverse incentives for criminals to employ complicated accounting measures to shelter the profits of their illegal enterprises. The purpose of forfeiture is to remove property facilitating crime or property produced by crime- — all of which is tainted by the illegal activity.
235 F.3d at 537, quoting 101 F.3d at 1042 (ellipsis in original).
. The majority includes the following cases and rationales. United States v. Corrado, 227 F.3d 543, 553 (6th Cir. 2000), cert. denied, 537 U.S. 1238, 123 S.Ct. 1366, 155 L.Ed.2d 207 (2003), states co-defendants are properly held jointly and severally liable for the proceeds of a RICO enterprise; government is not required to prove the specific portion of proceeds for which each defendant is responsible; such a requirement would allow defendants to mask the allocation of the proceeds to avoid forfeiting them altogether; citing United States v. Simmons, 154 F.3d 765, 769-70 (8th Cir. 1998). Corrado also cites United States v. Caporale, 806 F.2d 1487, 1507 (11th Cir. 1986), cert. denied, 483 U.S. 1021, 107 S.Ct. 3265, 97 L.Ed.2d 763 (1987), for the proposition that "joint and several liability is not only consistent with the statutory scheme [of RICO], but in some cases will be necessary to achieve the aims of the legislation.” Corrado, 227 F.3d at 553. As stated in Simmons, this type of joint and several liability for proceeds "is in accord with the traditional rules with respect to criminal conspiracy, under which all members of a conspiracy are responsible for the foreseeable acts of co-conspirators taken in furtherance of the conspiracy.” Simmons, 154 F.3d 765, 770.
. Title 18 U.S.C. § 981, by its title, pertains to civil forfeitures. But 28 U.S.C. § 2461(c) authorizes criminal forfeitures as part of the sentence, where civil forfeiture is authorized.
. The money laundering crime charged in Bornfield was a violation of 18 U.S.C. § 1957 (engaging in a monetary transaction in criminally derived property). Id. at 1127.
. Different forfeiture statutes use different language to authorize the "facilitation” theory of forfeiture. The forfeiture statutes relied on by the government in our case do not use the word “facilitate.” But § 1955(d) authorizes forfeiture of property "used in” the gambling business offense, and § 982(a)(1) authorizes forfeiture of any property "involved in” the money laundering offense. The "used in” test is akin to the "instrumentalities” test; the "involved in” test is somewhat less rigorous than the "used in” test. Cassella, § 26-1, p. 941.
. Of note, the substantial connection test has been incorporated in the civil forfeiture stat*898ute, CAFRA, 18 U.S.C. § 983(c)(3), which states as follows.
(3) If the Government's theory of forfeiture is that the property was used to commit or facilitate the commission of a criminal offense, or was involved in the commission of a criminal offense, the Government shall establish that there was a substantial connection between the property and the offense.
Cassella, § 26-4, pp. 952-53, states the test was already recognized in a majority of the 'circuits based on forfeiture statutes for drug offenses, and that although the test is codified in the civil forfeiture statute, it applies in criminal forfeiture cases as well.
. In its Forfeiture Brief, the government states: "A court may limit a forfeiture so as not to violate the Eighth Amendment’s prohibitions against excessive fines and unusual punishment. Otherwise, all illegal proceeds are forfeitable.” Doc. no. 1999, p. 16.
. The government proposes the following conclusions of law. "While all illegal proceeds are forfeitable, a court may limit a forfeiture so as not to violate the Eighth Amendment’s prohibitions against excessive fines and unusual punishment.” Doc. no. 2242, p. 56, ¶ 227. "A forfeiture is unconstitutionally excessive if it is grossly disproportional to the gravity of a defendant’s offense.” Id. at ¶ 228.
. All of the forfeiture defendants except King make eighth amendment arguments. As previously mentioned, King excludes these arguments from his adoption of the other defendants’ filings.
. For example, United States v. Lyons, 870 F.Supp.2d 281, 292 (D. Mass. 2012), states there is disagreement within the federal appellate courts regarding when the forfeiture of criminal proceeds triggers eighth amendment protections. Lyons groups the Tenth Circuit with those courts which supposedly hold that in such instances the Eighth Amendment does not apply. Id., n.5. It does so based on Berry-hill, discussed next in the text. The other circuits which Lyons puts in that group are the Fifth, based on statements in a 2005 decision regarding drug proceeds (the 2005 case, in turn, cites pre-Bajakajian cases); and the Seventh and Eighth (based on pre-Bajakajian decisions).
. Defendant also argued the forfeiture of his real property (a house, which the government contended had been obtained with drug proceeds) was invalid. For reasons not material here, Berryhill did not reach the merits of that argument. Berryhill, 128 F.3d 1386, 1392.
. Research has also disclosed two unpublished, post-Bajakajian Tenth Circuit cases, both of which are civil cases involving drug proceeds, which cite Berryhill for the proposition that forfeiture of drug proceeds does not violate the Eighth Amendment: United States v. $189,825.00, 216 F.3d 1089, *7 (10th Cir. 2000); and United States v. One Parcel of Real Property Known as: 16614 Cayuga Road, 69 Fed.Appx. 915, 919-20 (10th Cir. 2003), unpublished.
. At least two Tenth Circuit, post-Bajakajian, civil forfeiture cases address factors used in the proportionality analysis. United States v. Wagoner County Real Estate, 278 F.3d 1091, 1100-01 (10th Cir. 2002), notes the importance of Congress's judgment about the appropriate punishment for the owner’s offense and sets out other factors from Bajakajian. The court noted that Bajakajian addressed a criminal forfeiture which (unlike the civil forfeiture before the court in Wagoner County), “imposes punishment upon an individual who has committed a specific crime.'' Id. at 1100. Lot Numbered One (1) of the Lavaland Annex, 256 F.3d 949, 958 (10th Cir. 2001), is another civil forfeiture case which applies a disproportionality analysis. There, the Court of Appeals stated that in Bajakajian, "the Court concluded that a criminal forfeiture that was grossly disproportionate to the gravity of the defendant's offense would violate the Excessive Fines Clause.” Id. at 958.
. Wagoner County was remanded for additional fact-findings relevant to the Excessive Fines Clause. Id. at 1102.
. In a notice filed well before the forfeiture trial, the court notified all counsel that the livelihood issue was potentially relevant to forfeiture. Doc. no. 2089. The government raised no objection to the court's consideration of the deprivation of livelihood issue. To the contrary, the government proposes the following conclusion of law. “To the extent the Court considers whether the forfeiture would deprive the defendant of his livelihood, this is different from considering as a discrete [Bajakajian] factor a defendant’s present personal circumstances, including age, health, and financial situation,” which should not be considered. Doc. no. 2242, p. 56, ¶¶ 229-30, citing Viloski. The court agrees with the government’s differentiation.
. Bajakajian, 524 U.S. 321, 340, n.15, 118 S.Ct. 2028, notes the respondent did not argue that his wealth or income are relevant to the proportionality determination, or that full forfeiture would deprive him of his livelihood.
. And see, United States v. Fogg, 666 F.3d 13, 18-20 (1st Cir. 2011) (finding the district court had incorrectly based its decision on the defendant's inability to pay, which is not the proper inquiry; the question is not the defendant’s inability to pay but whether defendant's post-incarceration livelihood would be imperiled by the forfeiture; the district court had also incorrectly handled the burden of proof, as noted later in this order, requiring the government to prove that Fogg could pay a forfeiture judgment).
. In Lyons, these foreseeability issues were discussed in the context of RICO forfeiture. See, 870 F.Supp.2d at 290. This court has concluded, contra to Lyons, that, under a straightforward reading of the applicable statutory language, there is no joint and several forfeiture liability under RICO. See Part 11(C), above. But the foreseeability analysis in Lyons *907applies equally to the joint and several {i.e. vicarious) liability issues under Count 3.
. For examples of other types of situations in which the courts place the burden of proving foreseeability on the plaintiff, see: United States v. Thompson, 286 F.3d 950, 960 (7th Cir. 2002) (in the context of "relevant conduct” for sentencing: "we ... require the government to prove that the conspiracy's actions were foreseeable to each defendant to whom it seeks to impute relevant conduct”); and Anchor Savings Bank, FSB v. United States, 81 Fed.Cl. 1, 81 (Ct. Fed. Claims 2008)-(analyzing damages for breaches of contracts by the government, the court considered: "What evidence will carry plaintiff’s burden of proving that lost profits damages should have been foreseen as the • consequence of defendant’s breach of contract?”).
. The difference between burden of production and burden of persuasion (e.g., burden of proof) has been explained as follows by the Supreme Court:
In the two decades after Hill [v. Smith, 260 U.S. 592, 43 S.Ct. 219, 67 L.Ed. 419 (1923)], our opinions consistently distinguished between burden of proof, which we defined as burden of persuasion, and an alternative concept, which we increasingly referred to as the burden of production or the burden of going forward with the evidence. See, e.g., Brosnan v. Brosnan, 263 U.S. 345, 349, 44 S.Ct. 117, 118, 68 L.Ed. 332 (1923) (imposition of burden of proof imposes the burden of persuasion, not simply the burden of establishing a prima facie case); Radio Corp. of America v. Radio Engineering Laboratories, Inc., 293 U.S. 1, 7-8, 55 S.Ct. 928, 930-931, 79 L.Ed. 163 (1934) (party who bears the burden of proof "bears a heavy burden of persuasion”); Commercial Molasses Corp. v. New York Tank Barge Corp., 314 U.S. 104, 111, 62 S.Ct. 156, 161, 86 L.Ed. 89 (1941) (party with the burden of proof bears the “burden of persuasion,” though the opposing party may bear a burden to “go forward with evidence”); Webre Steib Co. v. Commissioner, 324 U.S. 164, 171, 65 S.Ct. 578, 582, 89 L.Ed. 819 (1945) (claimant bears a "burden of going forward with evidence ... as well as the burden of proof”) (emphasis added). During this period the Courts of Appeals also limited the meaning of burden of proof to burden of persuasion, and explicitly distinguished this concept from the burden of production.
Director, Office of Workers’ Compensation Programs, Dept. of Labor v. Greenwich Collieries, 512 U.S. 267, 274-75, 114 S.Ct. 2251, 129 L.Ed.2d 221 (1994).
. Since this criminal forfeiture proceeding is a component of sentencing, the relevant facts on the excessive fines issues may be gleaned from "information,” accepted by the court as "relevant and reliable” (Rule 32.2(b)(1)(B), Fed. R. Crim. P.), even if that information is not admissible under the rules of evidence. Rule 1101(d), Fed.R.Evid. That "information” would include the relevant and uncontested factual portions of a presentence report, as is discussed in Part IX(G), below.
. King also cites United States v. $149,-442.43, 965 F.2d 868 (10th Cir. 1992), a civil forfeiture case which states that “Although forfeiture proceedings are inherently civil in nature, they are not to be effectuated in derogation of one's constitutional rights,” and that “any defects in process used to secure the possession of defendant property may defeat the government’s right to possession, inasmuch as the government will be barred from introducing evidence illegally seized in violation of the fourth amendment to prove a claim of forfeiture.” Id. at 872. These statements are understood as limited to the civil forfeiture context, in which the broad discretion afforded to a judge at the sentencing stage does not apply. At least one court has noted the application of the exclusionary rule in civil but not criminal forfeitures. See, United States v. $511,780, 847 F.Supp. 908, 915, n.3 (M.D. Ala. 1994) (civil forfeiture case in which the exclusionary rule was potentially applicable; court noted that in contrast to civil forfeiture, criminal forfeiture proceedings are considered part of the sentencing process and that, as a general proposition, the exclusionary rule does not apply to sentencing).
.It should, of course, be borne in mind that not all defendants were convicted on all three counts. Thus, as has been noted, the lead defendant, King was acquitted on the flagship count, racketeering conspiracy, but convicted on Count 2 (illegal gambling business) and Count 3 (money laundering conspiracy). In contrast, Koralewski was convicted on the RICO count but acquitted on Counts 2 and 3.
. See Part II (C), above.
. See Part II (D), above.
. See Part II(E), above.
. In United States v. $448,342.85, Judge Easterbrook wrote for the court that: "It makes no sense to confiscate whatever balance happens to be in an account bearing a particular number, just because proceeds of crime once passed through that account.” 969 F.2d at 476.
. After the decision in the $448,342.85 case, Congress enacted 18 U.S.C. § 984, which would have “considerably eased” the government's task in the $448,342.85 case. See, United States v. United States Currency Deposited in Account No. 1115000763247, 1998 WL 299420 (N.D. Ill. May 21, 1998), at *3. But that legislation applies only to civil forfeiture. See, 18 U.S.C. § 984(a)(1).
. Section 982(b) incorporated the substitute assets provision from 21 U.S.C. § 853(p) which was quoted in Voigt at 1085, and provides that: "[i]f any of the property [forfeita-ble under subsection (a)], as a result of any act or omission of the defendant ... has been commingled with other property which cannot be divided without difficulty^] ... the court shall order the forfeiture of any other property of the defendant up to the value of any property described in subparagraphs (A) through (E) of paragraph (1), as applicable.”
. Reading Voigt, Stewart and Rothstein together, it comes as no surprise that, as the court concluded in Rothstein at 1214: "In sum, if ever there was a case where commingled proceeds 'c[ould not] be divided without difficulty' and that therefore required the Government to seek forfeiture pursuant to the statutes’ substitute property provisions, §§ 1963(m) and 853(p), this is that case. For us to conclude otherwise would 'render the substitute asset provision a nullity,' Voigt, 89 F.3d at 1087.” This conclusion is in harmony with the Tenth Circuit's treatment of the same subject in Bornfield. See, Bornfield, 145 F.3d at 1139.
. In most forfeiture cases, the concept of "tracing” is in play in a broad sense: There is a search for assets "traceable to” the criminal activity. See, Parts 11(D) and (E), above. But some forfeiture cases involve "tracing” in a more technical sense. In these cases, "tracing” is the application of a set of accounting conventions to a set of known transactions in an account for the purpose of arriving at a dollar amount that can, with some semblance of rationality, be treated as illicit and therefore forfeitable. See, e.g., Rothstein, 717 F.3d at 1214 (RICO forfeiture; referring to the "lowest intermediate balance” method of tracing). See also, United States v. Henshaw, *915388 F.3d 738 (10th Cir. 2004), a civil tax collection case in which the Tenth Circuit described the tracing principles as being "used by courts in many different areas of law to identify and segregate property that has been mingled with other property in such a manner that it has lost its identity.” Id. at 740. (In an apparently related civil tax case, United States v. Lowrance, 2002 WL 31987131 (N.D. Okla. Oct. 17, 2002), Judge Kern considered four tracing methods: "the lowest intermediate balance approach; the last-in-first-out (LIFO) approach; the proata approach; and the first-in-first-out approach.” Id. at *3.) In the case at bar, the government has not sought to employ any tracing methods. Instead, as will be seen, the government seeks, by and large, to taint otherwise clean assets with at least a minuscule amount of dirty money.
. The Tenth Circuit's decision in Bornfield covers a lot of ground, much of it unfavorable to the government, at least in the context of the facts of the case at bar. The Tenth Circuit's analysis in Bornfield is expressed in definitive terms. As its citation history demonstrates, Bornfield is very much the law of this circuit. The government’s Forfeiture Brief, and its proposed findings of fact and conclusions of law, make no reference at all to Bornfield. But this court is bound by Born-field, whether or not the government chooses to deal with it.
. It is worth noting that, in Lyons, the district judge, after rejecting the government's proposed extrapolation, found the evidence sufficient to support entry of judgment on the basis of specific, identifiable transactions that occurred during the relevant period, yielding a “known total” for purposes of entering a forfeiture judgment. 870 F.Supp.2d at 287.
. Not all of the government's motions for preliminary orders of forfeiture state that the government seeks a money judgment against the defendant in question. This is a surprising failure, and the court cannot guess whether it was the result of an oversight, or whether the government was, at some point, disinclined to seek a money judgment from certain defendants and later decided otherwise. Either way, this failure did not deprive the affected defendants of adequate notice, and it does not limit the government’s ability to obtain a forfeiture money judgment at this stage, for several reasons. The indictment is sufficient to put defendants on notice of their full forfeiture exposure. *See, United States v. Peters, 257 F.R.D. 377, 383 (W.D.N.Y. 2009) (government’s pre-trial representations regarding the maximum forfeiture amount did not nullify, modify or limit the notice contained in the indictment, which put defendant on notice of his full forfeiture exposure). Rule 32.2 does not require the government to file any type of motion for a preliminary order of forfeiture; the rule merely sets out procedures to be followed when the government "seeks” a personal money judgment, as the government eventually made clear that it did here by other pre-(and post-) forfeiture hearing filings. Finally, there is no evidence that the government's late clarification of its intent to pursue money judgments from all of the forfeiture defendants prejudiced a defendant, for example, by limiting his ability to meet the government's arguments for a money judgment once those arguments were fleshed out. (As will be seen, this question of notice of intent to seek a forfeiture money judgment is a matter entirely distinct from issues, such as foreseeability — to take only one example — that will affect the amount of any forfeiture money judgment to be entered.)
. RICO’s substitute asset provision appears at 18 U.S.C. § 1963(m). The more general substitute asset provision is found at 21 U.S.C. § 853(p), which is incorporated, for purposes of Title 18 forfeiture, by 18 U.S.C. § 982(b).
. In its Forfeiture Brief (doc. no. 1999), and at the forfeiture trial (see, GX C89), the government sought forfeiture of "[alpproximately $303,527.46 as substitute res regarding" certain residential real property in Texas on which King made mortgage payments. Doc. no. 1999-1, p. 11. But the government seeks nothing other than a money judgment against King for this $303,527.46, so this does not amount to a claim for forfeiture of a substitute asset in the sense contemplated by § 1963(m) or § 853(p). The same is true of the government’s request for a money judgment in the amount of $172,039.65 as a “substitute res” in consequence of Koralewski's sale of certain real property in Colorado. Doc. no. 1999-5, p. 1.
. Although the precise status of Lowrance and Jansen was not completely clear, the court, without hesitation, regards their activities in this case as being, in every respect, the activities of the FBI.
. "GX” refers to government exhibits. The government exhibits that are charts or tables (i.esummary exhibits admitted under Rule 1006) have the "C” prefix. This order refers repeatedly to the government's financial recap of its forfeiture case, GX C89, because that exhibit provides a reasonably concise overview of the government’s forfeiture case, except as to specific financial or tangible assets identified by the government for forfeiture by defendants other than King. For ease of reference, that exhibit is attached hereto as Appendix 1. GX C89 has been modified by the court in two respects: The account numbers have been deleted, except for the last three digits, and, for clarity of reference, numbers have been entered (1 through 49) to the left of the boxes under the headings of "Money Laundering Accounts/Items” and "Item Seized” on GX C89. The court added these numbers at the forfeiture trial. Tr. 220. Those numbers were used for reference at the forfeiture trial, e.g., Tr. 419. Accordingly, in this order, those numbered boxes are referred to as "Line 1,” etc.
. The government’s Forfeiture Brief made passing references to joint and several liability (e.g., doc. no. 1999, at 13, 18, but where the government specified, as required by the court, exactly what it sought by way of forfeiture money judgment against each of the forfeiture defendants, it provided the numbers set forth in this column. These were unequivo*923cal statements of the amount the government sought at that point. Taking Bramley as an example — with verbiage identical to that used with respect to all the other forfeiture defendants — the government said: "The United States seeks a money judgment in the amount of $5,102,789.00 against Mr. Bramley.” Doc. no. 1999-2, at 9. The government’s original differentiation among defendants was unsurprising, in light of matters that could potentially make a difference in forfeiture exposure among the forfeiture defendants, such as the acquittal of King on Count 1 and the acquittal of Koralewski on Counts 2 and 3, issues as to foreseeability, and issues as to application of the Excessive Fines Clause, among other things.
. This is the amount shown for "Money Judgment Grand Total” in GX C89 (App. 1 hereto). There is no combination of subtotals on GX C89 that adds up to this amount.
. Although the dollar amount shown in this column does not appear anywhere in GX C89, this is the sum of the subtotals comprising the items for which the government seeks a forfeiture money judgment, per GX C89, using the government's updated figures for its proposed extrapolations with respect to proceeds attributable to agents and runners. The government did not seek judgment for this amount before the forfeiture trial (doc. no. 1999 [Forfeiture Brief]; doc. no. 2130 [pre-trial proposed findings of fact and conclusions of law]) or after the forfeiture trial (doc. no. 2242 [post-trial proposed findings of fact and conclusions of law], at 48, ¶ 189), but this is the sum of the subtotals on the final version of GX C89.
. One of these items, Line 2, is used as the basis for the Global Data Payment extrapolation (to $25.1 million) in the top box on GX C89, as is discussed below, but the $7.1 million figure in Line 2 is not itself an extrapolation.
. E.g., doc. no. 1999-1 at 3 ("King's only source of income from 2003 through 2013 derives from his operation of the Legendz Sports enterprise.”); id. at 8 ("Mr. King's only source of income was from the Legendz Sports illegal gambling.”); 9 (same); 10 ¶ 24 (“Mr. King’s only source of income was from the Legendz Sports illegal gambling enterprise.”); 10 ¶ 25 (same); 11 ¶ 26 ("Mr. King's only source of income was from the Legendz Sports illegal gambling enterprise.”); 11 ¶ 27 (same); 11 ¶ 28 (same); 11 ¶ 29 ("Mr. King's only source of income was from the Legendz Sports illegal gambling.”). Three months later (and eight days before the start of the forfeiture trial), the government made the same representation to the court: "Mr. King's only source of income was from the Legendz Sports illegal gambling.” Doc. no. 2130, at 5.
. Lowrance did not know what Vaporama’s purpose was, what type of business it conducted, or whether the business it conducted was lawful or unlawful. Tr. 241-42.
. The source document for the government’s summary is GX 1267, a 548-page exhibit apparently consisting of account origination documents, account statements for ''Vaporama Internacional, S.A.,” and transaction documents—all in Spanish.
. Lowrance acknowledged at the forfeiture trial that he had not read Berthiaume's testimony. Tr. 82. Possibly, other members of the prosecution team thought it unnecessary— perhaps even inconvenient — for Lowrance to be aware of that testimony. This aspect of the matter is discussed elsewhere in this order.
. Magna was named as a defendant in the original Indictment in this case but not in the Superseding Indictment.
. Platinum Vacations is not to be confused with Platinum Advantage, an unrelated business established and solely operated by Reggie Berthiaume. Tr. 205; Feb. Tr. 745.
. Although this is not a major point, it is worth noting that, aside from the fact that the government has asserted that King had no legitimate "income” during the conspiracy period, the issue that is relevant for purposes of evaluating the quality of the government’s evidence is whether he had any legitimate cash flow during the conspiracy period— which he clearly did.
. In a case as massive as this one, involving scores of primary actors, millions of transactions and millions of documents, it is quite reasonable from the standpoint of fairness and manageability (and perhaps essential as a legal matter) to limit the forfeiture analysis to those factual and legal contentions that are actually made by the government. For example, although forfeiture in a case involving illegal gambling and related money laundering offenses can be based on outbound money rather than deposits of illegal gambling proceeds, e.g., United States v. Iacaboni, 363 F.3d 1, 4 (1st Cir. 2004), the government has, in this case, opted to hang its hat on deposits of funds it asserts were betting proceeds, as the government’s summaries (cited in the right-hand column of GX C89) suggest and as was discussed at the forfeiture trial. Tr. 225 (Government "tried to always use deposits in order to be consistent.”); Tr. 234 (Government "used deposits exclusively in compiling” its forfeiture overview.). See also, doc. no. 2242 (Government proposed findings of fact), ¶¶ 25, 29, 51-53, 56, 83, 88, 90-92, 96-98, 103, 105, 118-121, 135, 159, 172, 187, 189 *928and 190. (As is discussed below with respect to Line 16 of the government's money laundering forfeiture summary, it is not true that the government focused exclusively on deposits, but that is beside the point to be made here.) Once the government chooses to focus on deposits, as it has mostly done and was perfectly entitled to do in this case, the focus, in terms of forfeitability, must stay on deposits, lest the case descend into a morass of double counting problems. Thus, payments out of an account may be of general evidentia-ry value for the purpose of establishing the illicit nature of transactions involving the account, but once the government casts its lot with deposits, consistency requires that the focus remain on deposits for the purpose of identifying transactions (under the heading of Money Laundering Accounts in GX C89) that add to the sum total of the forfeiture money judgment.
As another example, with one exception— Bramley’s residence in Plano, Texas (see, Doc. no. 2242, at 32, ¶ 116) — the government does not assert a facilitation theory as to any of the properties it seeks to forfeit in this case. (The government originally asserted a facilitation theory as to Bramley’s Mercedes Benz car (doc. no. 1999-2, at 8) and as to Diebner's residence (doc. no. 1999-4, at 2). It no longer does so. See, Doc. no 2242, at 31-37 and 39-43.)
In any number of nooks and crannies of this case, it has occurred to the undersigned that perhaps some available but unasserted factual or legal theory might save a point for the government. But a case as massive and complex as this one simply does not permit that form of generosity. That is especially true where, as here, the government has had an exceptionally long time to formulate and substantiate its forfeiture theories.
. Aside from having an understanding that Grupo Legendz was "affiliated with the Le-gendz enterprise” (Tr. 179), Lowrance had no knowledge about that company. Id. at 180-82. He had no recollection of reviewing corporate documents as to Grupo Legendz. Id. at 180. Due to his lack of information as to the nature of the activities of Grupo Legendz, Lowrance disclaimed any opinion as to whether the deposits into the Banco Nacional account represented “legal lawful income.” Tr. 182. *929Of course, the facts as to Grupo Legendz were well known to the FBI and IRS CID agents who sat through the four jury trials, all of which occurred long before Lowrance took the stand at the forfeiture trial. The evident compartmentalization of the flow of information within the prosecution team is discussed elsewhere in this order.
. It is worth noting that, even though the government bases its characterization of these funds on the asserted nature of the payments into the HSBC account (it calls these deposits, summarized on GX C49A, “a known amount of illegal proceeds,” doc. no. 2242 at 14, ¶ 55), this characterization is not saved by reference to payments out of that account. Payments out of the HSBC account are summarized on GX C49B. That summary shows 193 payment transactions, all of which, except one, occurred between February, 2007 and January, 2009. Of those, the identity of the payee is unknown for 43 transactions and the payee is "Virtual Automated Technology” for 92 transactions. Payments to known payees (which the government asserts were Le-gendz bettors) other than Virtual Automated Technology account for less than five percent of the dollar amount of the payments shown by GX C49B. Within that five percent, Low-rance could identify only one person as a Legendz bettor. Tr. 202.
. See, United States v. King, 2014 WL 12623415, at **17-19 (W.D. Okla. Dec. 1, 2014), and United States v. Bramley, 2015 *931WL 12852048, at **18-19, *21 (W.D. Okla. Sept. 16, 2015).
. GX C49A also includes two deposits of funds from Bet Jamaica, an organization affiliated with Spiros Athanas, formerly a defendant in this case. The most obvious characterization of those deposits is that they were what is known in the trade as “book to book” transfers, see, United States v. King, 2014 WL 12623415, at **14-15 (W.D. Okla. Dec. 1, 2014), and the government does not assert otherwise, either in testimony or in briefs. The court has previously found (and adheres to that finding) that book-to-book transfers between these offshore entities are not within the scope of the Indictment. Id. at *15.
. And this is aside from the fact that, in several instances, the source document, GX1405A, shows the "originator” to be Reginald Berthiaume, not Platinum Advantage (contrary to what is shown on the government's summary). In those instances, there is no mention of Platinum Advantage other than a notation on the wire transfer instructions to the effect that Platinum Advantage is to be notified of the completion of the transfer. For discussion purposes, the court is content to ignore these discrepancies even though the government's erroneous transfer of data from the source documents to the summary makes the summary questionable as a summary admissible under Rule 1006.
. Why the government made this representation in the first place is unfathomable. The evidence at the forfeiture trial established beyond any doubt that the government has never known the source of 300 of the 324 deposits at issue here, and the government does not assert otherwise.
. In the cited paragraph of its proposed findings, the government proposes a total of $46,981,070.31 for the money laundering accounts. To support that, the government cites GX C89 and Lowrance's testimony at Tr. 159- — both of which propose a total of $43,417,030.14 for this category (about $3.6 million less). The difference is not explained, and the $46,981,070.31 figure was not mentioned at the forfeiture trial or in the government's Forfeiture Brief. That is of no moment for present purposes, because the point to be made here is that the government still proposes to include in that category all of the money attributable to the deposits as to which it has no clue of the source.
. E.g., United States v. Bramley, 2015 WL 12852048, at *14 (W.D. Okla. Sept. 16, 2015).
. Amegy Bank cashier’s check no. 131655, dated February 13, 2102, was received in evidence twice — as GX 8871 and as GX 693. In its proposed findings, the government cites those two exhibits separately, incorrectly suggesting two separate $45,000 payments to FLO. Doc. no. 2242 at 25, ¶ 93. In a case involving as many documents as this one, that is an understandable oversight.
. Other checks cited by the government in its proposed findings are GX 693, GX 887I and GX 887M. Doc. no. 2242, at 25, ¶ 93.
. This is an entirely hypothetical assumption. The court gets no sense from the massive body of evidence in this case that King found devious maneuvers to be necessary with respect to movement of money within Panama.
. Although the undersigned will make no claim to be a reader of Spanish, it appears from the source documents that Banco Panama is not the same institution as Banco Na-cional de Panama, an institution which is discussed elsewhere in this order. Compare, GX 1259 (Banco Nacional) with GX 1262 (Banco Panama).
. On motion of the government, the indictment was dismissed as to Rigoberto Nolan a/k/a Rigoberto Nolan Forbes on October 5, 2016. Doc. no. 2336.
. See, note 67, above, and materials there cited.
. The government does not, in its proposed findings, so assert. See, doc. no. 2242, at 26-27. However, it is necessary to address KHML here, lest it be suggested to a reviewing court that this court disregarded testimony that would buttress the government's contentions in Line 10.
. Lowrance was initially a little confused about this, testifying at first that the Line 11 account (2657) was closed. Tr. 115. Upon questioning by the court, Lowrance confirmed the obvious, to wit: that $229,951.78 withdrawal closed the Line 10 account (0395), not the Line 11 account. Tr. 116. The court was (and is) persuaded that Lowrance had appropriately corrected himself, but the AUSA was still not persuaded until Lowrance set him straight a few questions later. Tr. 116.
. For the reasons discussed elsewhere in this order, the testimony of Karlo Stewart should be viewed with great caution. The court credits Stewart’s testimony on this point because it is corroborated by documentary evidence.
. Compare Lowrance’s approach: "But I think from a money laundering standpoint, we would look at it as it's — the deposit here is one set of money. The deposit, if it moves, this is another set of money.” Tr. 235.
. This refers to the total ($43,417,030.14), shown on the second page of GX C89, for the government’s proposed money laundering forfeiture.
. As is discussed later in this order, the government also seeks to forfeit Bramley's residence. It seeks to do so on a facilitation theory. The government'has also asserted that the use of bettor payments to make some of the mortgage payments (which the government has proven quite clearly) "taints up the property.” Tr. 515 (Jansen). Those instances of the use of bettor payments to service the mortgage show up, as discrete items, on the government’s forfeiture summary as to Bram-ley (totaling $117,968.00 — GX C81, the sum of the fourth and fifth lines). It is unclear whether the government invokes this "taint up” theory as a separate basis for forfeiture of the Bramley residence.
. The "more” in Bornfield was facilitation, id., which the government argues as to Bramley’s residence (doc. no. 1999-2 at 5; doc. no. 2242, at 32, ¶ 116).
. Doc. no. 1999-1, pp. 14-18 and (as to Robles) 1999-8, at 1.
. The government's original proposed extrapolation period, for individuals who were alleged to have been involved in the Legendz enterprise from its inception to its demise, was ten years. See, e.g., doc. no. 1999 (Forfeiture Brief), at 14-17; doc. no. 1999-1, at 4, and doc. no. 2130 (pretrial proposed findings of fact and conclusions of law), at 7-10, 23. Post-trial, the government's proposed extrapolation period for those individuals was nine years. Doc. no. 2242, at 13 (¶53), 15 (¶ 59), 16 (¶¶ 61, 62), 17 (¶¶ 64, 66), 18 (¶ 68), 19 (¶¶ 72, 75, 76), 39 (¶ 149) and 45 (¶ 174).
. The government's approach to extrapolation in this case is not supported by any accounting principle or scholarly work that Lowrance could cite. Tr. 410.
. As to some of the government’s proposed extrapolations, the government also proposes alternative attributions that are not based on extrapolation. Those will also be addressed here.
. No "Agent” attribution is urged as to King, for two reasons. First, he was the owner and undisputed chief executive officer of the Le-gendz enterprise. He was not a mere agent. Secondly, the government understandably attributes to King all of the proceeds attributable individually to the agents.
. As explained in Part IX(C)(1), above, these attributions of betting proceeds to Legendz agents (under the heading of "Agents” on GX C89) are shown to come to a total of $142,281,24737 on the first page of GX C89. The government revised GX C89 to insert the updated numbers shown in red on the first page of that exhibit, but did not update either the total at the bottom of the "Agents” box, or the grand total on p. 4 of GX C89.
. This is of more than passing interest when viewed in light of the fact that, as noted, Lowrance also did not consider it necessary to talk to Campbell to determine whether the government’s extrapolation had any resemblance to reality.
. This is true regardless of which source documents the government might cite. Low-rance finally settled on GX 1257 as the source document for the Campbell extrapolation, Tr. 418. That 115-page document, though perhaps easily explainable by Campbell, defies interpretation, with any reasonable degree of confidence, by a novice reader.
. In its proposed findings of fact, the government straddles the gap between Lowrance’s testimony and the clear import of the source document by simply referring to Hernandez’s “shipment activity” without taking a position as to who generated the funds shipped. Doc. no. 2242, at 19, ¶ 72.
. On the basis of the evidence at the four jury trials: “Cliff” on GX 198 is Paul Wilson. Feb. Tr. 265. "Bigmike” is Mike Lawhorn. Feb. Tr. 169. "Gato” is Terry Campbell. Feb. Tr. 199. “Paul” is asserted by the government to be Paul Tucker. "Ski” is asserted by the government to be Kqry Koralewski. “Doc” is Rodger Bramley. Feb. Tr. 246.
. The court makes no finding of intentional deception by Lowrance. Careful consideration suggests that he looked at the summary and made an erroneous assumption as to the nature of the activity it represented. Under the circumstances (given the fact the Lowrance was the government's sponsor of the summary), that is serious enough, but decidedly less serious than intentional deception.
. Lowrance asserted that this was a "calculation,” not an extrapolation. Tr. 142. Regardless, this $16,770,000 figure is the product of multiplying $215,000 by seventy-eight.
. Aside from these deficits, which alone compel rejection of the $16.7 million extrapolation, it is worth noting that there is no satisfactory basis for assuming that the volume of Diebner's sports betting business was anywhere near constant, even on a year-to-year basis, during whatever period it was that he was associated with Legendz. As Jansen acknowledged, Myler’s recollection of his dealings with Diebner was that Diebner started small and it was not until "some later unnamed date” that "the numbers got larger.” Tr. 582. This is most vividly illustrated by Diebner’s dealings with the Cabrera family of betting clients. As will be seen, they bet very large amounts of money through Diebner. Testifying in May, 2015, Sergio Cabrera testified that he met Diebner "[a]bout three, four years” ago. May Tr. 737. That would be in 2011 or 2012. The government wants to start the Diebner extrapolation about 6 ½ years before Myler last dealt with Diebner (doc. no. *9622242, at 40, ¶ 152), which would begin the extrapolation period in about 2006, long before Diebner acquired the Cabreras as betting customers. The government has not attempted to explain or account for what would obviously be long term- trends (to say nothing of variations within that long term) in Diebner's betting volume.
. Nor does the government offer a weekly average, for the twenty-nine weeks, in its proposed findings of fact. In its proposed findings, the government does posit that the twenty-nine notes evidence $743,632.80 in illegal proceeds. Doc. no. 2242, at 39, ¶ 151. The government does not go on to explain how that number, for twenty-nine weeks, comes *963out to $17,432,674.40. No witness has ever vouched for that number as being supported by GX 325. And perhaps for good reason: If GX 325 represents $743,632.80 for twenty-nine weeks, the total for 78 months is about $8.6 million, not $17,432,674.40.
. As has been noted, that seventy-eight month period is subject to serious question, but that is beside the present point.
. In its Forfeiture Brief, the government offers a total for the 29 weeks. Doc. no. 1999-4, at 4. But this figure, divided by twenty-nine and multiplied over seventy-eight months, does not square with either of the extrapolations the government witnesses advocated at the forfeiture trial. And this is aside from the fact that no witnesses have testified to that figure or to the basis upon which it was calculated.
. Consistent with this, it was unmistakably clear from Lowrance’s testimony that the FBI fed him a $215,000 per month number for Diebner (Tr. 373), which comes to exactly *964$50,000 per week when divided by 4.3. This leads, of course, to the $16,770,000.00 extrapolation that the court has rejected for other reasons.
. The government would argue that total betting volume, not just losses, should be reckoned here. Correct though that may be, the court, having rejected the government's two extrapolations relating to Diebner, is simply proceeding on the basis of obvious, unassailable numbers. The court is not obligated to come up with alternative extrapolations that have not been tested in the courtroom.
. The government also seeks to include betting proceeds from Tipton Rowland in the forfeiture money judgment. Doc. no. 2242, at 42, ¶ 155. Rowland was certainly a gambling client of Diebner. But, during the May trial (the only trial in which he testified), he was at pains, even in the face of leading questions suggesting otherwise, to make it clear that his payments of gambling losses included payments for video poker losses. May Tr., 715, 716, 718, 719. This case does not involve video poker, and it is clear from Rowland’s testimony that, to the extent that his payments to an entity controlled by Diebner represented video poker losses, that activity was separate from Diebner's Legendz-related bookmaking. The government ignores this distinction in its proposed findings (doc. no. 2242 at 42), and the record provides no basis for a differentiation between Rowland’s Legendz-related losses and his video poker losses.
. Lowrance testified, believably, that he had heard of 5 Dimes but knew no more than that about 5 Dimes. Tr. 337, 399-400. Given the government's reliance on a government-prepared spreadsheet referring to 5 Dimes (based on Tanner's handwritten records), it is worthy of note that the case agents for this case were in the courtroom during the February and May, 2015 trials at which the facts as to 5 Dimes were plainly established. Feb. Tr. 139-40, 483, 485-86, 1115. May Tr. 122.
. The government also cites Lawhorn's testimony at the April, 2015 trial of King, doc. no. 2242, at 19 (citing Apr. Tr. at 1843), but that testimony intimates no dollar volume for Lawhorn’s activity. It only describes the "50 Red” arrangement (and does so consistently with Lawhorn's testimony at the October trial).
. In the parlance of this case, Lawhorn is referred to as a master agent because he had other agents working under him. Oct. Tr. 1133. In this case, retail-level agents have commonly been referred to as bookies regardless of whether, in any individual instance, the person in question was making book on a risk-bearing basis, as would be the case with a traditional bookie. Adding to the confusion, the court, and others, have also at times referred even to master agents, very generically, as bookies.
. The government’s Forfeiture Brief, although citing Lawhorn’s testimony in the October trial, ignores his unequivocal testimony at that trial about having worked on a pay-per-head basis. Doc. no. 1999-1, at 17. The government's proposed findings also ignore this uncontradicted testimony. Doc. no. 2242, at 19, ¶¶ 73-75.
. Lawhorn also worked (apparently during the conspiracy period) with other offshore bookmaking operations (Oct. Tr. 1307), competitors of Legendz. It is reasonably clear that the estimate Lawhorn gave in his testimony encompassed only his activity in association with Legendz. The court assumes that to be the case.
. At the forfeiture trial, the court (after expressing skepticism about some of the government’s proposed extrapolations) invited the government to propose alternatives based on "discrete events,” as distinguished from extrapolations. Tr. 774 (and citing Lyons, 870 F.Supp.2d 281, 287 to illustrate resort to discrete events where the extrapolation of gambling proceeds fails). Perhaps for that reason, the government refers, in its proposed findings, to some $255,799.00 in "checks relating to Mr. Lawhorn." Doc. no. 2242, at 19, ¶ 75. *970Those checks were indeed admitted at the May trial, as part of an en masse offer of exhibits. May Tr. 13. But they were never mentioned in any other way at the May trial, at any of the other three jury trials (other than as part of an en masse offer at the October trial, Oct. Tr. 55-56), or at the forfeiture trial. The government quite understandably cites no testimony substantiating its characterization of these checks. They are, accordingly, of no consequence for present purposes.
. On direct examination, Lowrance gave an unqualified "yes” answer when asked whether he had assisted in preparing the government’s chart, GX C55. Tr. 72. But on cross examination, he would allow only that he participated in preparing it "[i]n a sense.” Tr. 343. Further cross examination of Lowrance demonstrated that he in fact had no significant role in preparing that summary (Tr. 343-45), nor did he even have enough familiarity with the basis for that summary to enable him to give any meaningful testimony shedding light on the reliability of the summary. This is another example of a pattern that is a serious concern to the court in this case. Admission of summary exhibits under Rule 1006 of the Federal Rules of Evidence is serious business and is, to some degree (in many instances, to a very great degree, as every trial judge knows), an act of faith. Government witnesses, as a practical matter, often get the benefit of the doubt when sponsoring summary exhibits. The court’s experience in this case is, to put it mildly, unsettling.
. The BCS National Championship Series system of college football playoffs ended with the January, 2014 bowl games, a few months after the Superseding Indictment in this case was returned. It was replaced by the NCAA Division I College Football Playoffs.
. The court considered an additional attribution on the basis of internal Legendz records for a different period (GX 68-GX 108—beginning in December, 2007 and ending in November, 2008), but those records have the same problem with respect to the reliability of the identification of "Paul.”
. GX 1134H was admitted into evidence at the instance of the government at the October trial. Oct. Tr. 362. It is the only source document relied upon by the government to attribute $4.5 million in betting proceeds to Middlebrook. Yet when it was offered by a defendant at the forfeiture trial, the government professed to be "at a loss to understand the purpose of this.” Tr. 728. This document, the singular basis for the attribution of proceeds to Middlebrook, was, over the government’s objection, admitted into evidence at the forfeiture trial as well.
. Although not stated by Lowrance (or Jansen), the $288,384 figure is the total of the dollar amounts under the heading of “Debit” on GX 1134H, exclusive of the carryover number ($152,068) at the top of that column. This is not a major point, but the government overlooks, for forfeiture purposes, the fact that the $152,068 is bad debt, owed but unpaid by losing bettors, as the government brought out in direct examination of Middle-brook at the October trial. Oct. Tr. 364. Although the $152,068 is not included in the government's extrapolation, it is noteworthy, given that level of bad debt, that the government's extrapolation from the seven months of activity shown on GX 1134H implicitly assumes a 100% collection rate.
. At the forfeiture trial, there was a suggestion, on behalf of Bramley, that Myler testified at the James hearing to the effect that he "received 20 pick-ups from Mr. Bramley.” Tr. 644. The court did not find this in Myler’s James hearing testimony, and this is not mentioned in Bramley’s proposed findings of fact. Doc. no. 2315.
. These are shown as line items on GX C81: bettor checks deposited in an Edward Jones account, a bettor check deposited in a Linsco Private Ledger account, proceeds attributable to "Doc” from tracking records, funds attributable to Bramley from petty cash sheets, funds deposited in Panco Recursos, and proceeds supported by the testimony of Justin Walters and Michael O’Brien.
. In connection with the guilty pleas of McFadden and Zima, Zima forfeited this $208,000. Doc. no. 2009. None of the parties now before the court advance any arguments as to whether this creates a double counting problem, and the court sees no reason to undertake that analysis unaided by advocacy.
. McFadden speculated that this was $625,000, but the government appropriately leaves it at $600,000. Doc. no. 2242, at 21, ¶ 81.
. The exhibits cited by the government in support of this extrapolation (in GX C89) are far from self-explanatory. See, GX 393-398. Lowrance could not even say whether they represented bettor payments or bettor positions vis-a-vis the house. Tr. 353.
. As indicated by a leading question at the February trial, the government joins in this characterization. Feb. Tr. at 1104.
. The government asserts that Koralewski had no legitimate employment during the conspiracy period, and the government’s case against Koralewski for forfeiture of hard assets is based substantially on that premise, as discussed in Part IX(F), below. The merits of that assertion are irrelevant to the government’s forfeiture money judgment case against Koralewski, so there is no need to address that contention here.
. The government has not, in any of its papers, asked the court, in terms, to consider acquitted conduct for forfeiture purposes, nor has the government provided any argument, developed or otherwise, as to the permissibility of that. The permissibility of consideration of acquitted conduct for criminal forfeiture purposes is, to the knowledge of the undersigned, an open question in the 10th Circuit. The court declines to venture unbidden into that thicket. Accordingly, Koralewski's exposure to forfeiture is considered only within the framework of his Count 1 conviction.
. Lowrance had no substantive involvement in developing the government's attribution of proceeds to Koralewski. Tr. 155—56; 386-90.
. The court is at pains to add that it does not intimate here any suggestion that the gov-eminent would have been involved in or aware of any such activity on Stewart's part.
. The government has not pointed to any evidence indicating what happened with these accounts between 2011 and the spring of 2013, when the Indictment was filed and the takedown occurred.
. The evidence as to the source of funds gets no better with respect to the Jeep and the Maserati. Accepting, for the moment, the government's account of the flow of funds for the purchase of the Jeep (doc. no. 2242, at 9, ¶ 37), the trail begins only with two Banco General cashier’s checks. GX 1200G and H. The absence of evidence of a nexus to illegal gambling proceeds is at least equally plain with respect to the Maserati. GX C24.
.These are (all in Texas): 2 Palmer Crest in the Woodlands, 22 Nestlewood Place in The Woodlands, 30 Nestlewood Place in the Woodlands, 206 North Vesper Bend Circle in Spring, 86 Blue Creek Place in Spring and 82 Blue Creek Place in Spring.
. 22 Nestlewood Place, 30 Nestlewood Place, 206 North Vesper Bend Circle, 86 Blue Creek Place and 82 Blue Creek Place were all titled in the name of defendant Starting 5, LLC, which was acquitted on both of the counts in which it was charged in this case. Doc. no. 1768. (86 Blue Creek Place was titled first in defendant Serena King and then in Starting 5, LLC, but she, also, was acquit- . ted on all counts. Id.)
. The singular exception is 206 N. Vesper Bend, as to which some of the purchase money came from Olmos Overseas Ltd., which was a Legendz affiliate that was used as a repository for betting proceeds. GX 1205 I and J. But the other deficits in the government’s proof, noted in this section, are fatal to forfeiture of that property.
.GX 853 and the government exhibits'in the 1203 and 1204 series are cited by the government (doc. no. 2242, at 11-12, ¶¶41, 44, 45). GX 853 was admitted into evidence in the April and October trials. The 1203 and 1204 series were admitted into evidence in the October trial.
. The evident reasons for this are addressed in Part X, below.
. This order will result only in a preliminary order of forfeiture. Rule 32.2(b)(2)(A), Fed. R. Crim. P. Consequently, the possibility that King's assets may be subject to levy to satisfy the forfeiture money judgment is a matter beyond the scope of this order.
. This is of limited significance as to the defendants whose forfeitures have not yet been specifically addressed in this order, because, unlike Koralewski, they were all convicted on the money laundering conspiracy count, as well as on the RICO count. At least in the context of the facts of this case, forfeiture law, as applied on the basis of the money laundering convictions, is at least as expansive in its practical effect as it Would be if vicarious liability were available under RICO.
. This determination can make a substantial difference from one defendant to the next. In her order addressing forfeiture by the defendants who ran the geographically far-flung illegal sports betting operation in Lyons (a decision that was described by the First Circuit as "detailed, thoughtful, and well-researched," United States v. Lyons, 740 F.3d 702, 733 (1st Cir. 2014)), Judge Saris, assessing what was "reasonably foreseeable,” held one defendant accountable for about three times the amount of the gambling proceeds that she attributed to another defendant. Lyons, 870 F.Supp.2d at 292.
.This formulation is, as discussed in Part IV(A), above, a bit less stringent than the approach used by the court in Lyons, which also involved a far-flung sports betting organization.
. The court makes this estimate as to Tucker (and, where relevant, as to the other forfeiture defendants) on a “no less than” basis because it is mindful that where the exercise is already somewhat speculative, the court should estimate conservatively. Rudaj, 2006 WL 1876664 at *3.
. Under Rule 32(i)(3)(A), Fed.R.Crim.P., the court "may accept any undisputed portion of the presentence report as a finding of fact.” Criminal forfeiture is a component of sentencing. It is therefore appropriate for the court to rely, for present purposes, on the portions of Tucker’s presentence report as to which no objection has been lodged. See, e.g., United States v. Smith, 815 F.3d 671, 679 (10th Cir. 2016). At the forfeiture trial, the court indicated its intent to refer to the "unobjected-to portions of the presentence reports” on the issue of "future ability to earn a living.” Tr. 525. No party has registered any objection to proceeding on that basis.
. For the reasons indicated in Part 1(C), above, with respect to Tucker's company, Zapt Electrical Sales, Inc. (a defendant in this case), the court assumes that there are no remaining forfeiture issues as to Zapt.
. As has been noted, the court does not have the benefit of advocacy from the government as to any aspect of the application of the Excessive Fines Clause to the facts of this case, even though the government has expressly acknowledged the applicability of that constitutional limitation. It is, in any event, difficult to conceive that the government *1008would argue that the responsibilities of the defendant to his dependent minor children are not to be considered here.
. Bramley has asserted in multiple pleadings, and asserts on the basis of an exhibit introduced at the forfeiture trial, that his wife has a very substantial separate interest in the house, not a mere spousal interest (which she may well also have), but a separate interest owing to the fact that the acquisition of the house was very substantially financed out of Mrs. Bramley’s separate inheritance. Defendant’s Ex. 3 (Bramley); Tr. 637. Because the forfeiture proceedings are at the preliminary order stage, the extent of Mrs. Bramley's interest is not a matter that is ripe for adjudication. In any event, the issues raised by and on behalf of Mrs. Bramley are rendered moot by the court's determination, in this order, that the government has not made its case for forfeiture of the house.
. Per Bramley's presentence report, doc. no. 1624, he has a 1991 Mercedes 300E. The government does not seek forfeiture of that car. It appears likely that Bramley acquired that Mercedes after he was indicted in this case.
. In any event, forfeiture of Diebner’s house would not clear the incidental use bar. Any use by Diebner of his house for Legendz-related purposes received only fleeting mention at Diebner's jury trial. May Tr. 1633. See also, id. at 801 (Diebner picking up money at bettor’s house or at a restaurant); 1535 (Myler picking up money “in a designated parking lot or a restaurant”).
. As has been noted, King expressly did not adopt arguments based on the Eighth Amendment. See, doc. nos. 2283, 2288.
. The flow of information within a prosecution team is very rarely a matter that is (or, in any event, should be) of concern to the court. However, where,' as here, the government’s presentation to the court, by sworn witnesses and by government counsel as officers of the court, is demonstrably affected by the matters addressed here, scrutiny of that aspect of the case is appropriate.
146. The court’s aborted reference to "the government's own evidence” was erroneous in part. The testimony that overwhelmingly and without contradiction established the fact of Koralewski’s legitimate employment came from a witness called by Koralewski (Detec-live Meoni) as well as from a government witness (Middlebrook) — testimony that likely had more than a little to do with Koralewski’s acquittal on Counts 2 and 3. (The facts as to King's lawful entrepreneurial activities came entirely from government witnesses.)
. Campbell, Lawhorn, Middlebrook, Wilson and Ross.