dissenting from the denial of rehearing en banc:
With respect, I dissent from my colleagues’ refusal (7-7) to grant rehearing en banc in this case. Judge Wynn’s excellent dissent from the panel opinion well explains why we should have affirmed the judgment of the district court. See United States v. Foster, 662 F.3d 291, 298 (4th Cir.2011). I write now only to point out that the panel opinion is directly and inexplicably at odds with what all recognize is controlling law — Shepard v. United States, 544 U.S. 13, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005).
In Shepard, the Court grappled with the very question at issue here — whether a defendant’s state burglary convictions necessarily involved entry into buildings, and *396thus qualified as ACCA predicate offenses. Id. at 17, 125 S.Ct. 1254. Shepard held that only certain “conclusive records made or used in adjudicating guilt” — now widely known as Shepard-approved documents— can establish with the requisite certainty that the state predicate convictions “ ‘necessarily’ rested on the fact” that the defendant burglarized a building, and thus committed the ACCA predicate offense of generic burglary. Id. at 21, 125 S.Ct. 1254.
In Shepard itself, the Shepard-approved documents set forth the actual street addresses, e.g., “30 Harlem St.,” of the burglarized establishments. Id. at 31-32, 125 S.Ct. 1254 (O’Connor, J., dissenting). Yet both the Government and the Shepard Court (majority and dissent) agreed that the actual street addresses were insufficient to demonstrate that the defendant had burglarized buildings.1
The panel in the case at hand, nonetheless, concludes that the mere proper names of burglarized establishments are sufficient to prove that the defendant burglarized buildings. The panel puts forward no reason, and I can discern none, why the proper names of burglarized places in this case establishes that they are buildings while the actual street addresses of burglarized places in Shepard did not.
The panel’s error stems from its understandable urge to avoid an unlikely conclusion by applying “common sense” or “logic.” See Foster, 662 F.3d at 291-98 (panel majority and concurrence invoking “common sense” no less than fifteen times, and “logic” six times). The First Circuit had precisely the same understandable inclination in Shepard. See United States v. Shepard, 348 F.3d 308, 314 (1st Cir.2003), rev’d 544 U.S. 13, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005). Just as the First Circuit concluded that the offenses qualified as ACCA predicates because it was “unlikely — to the point of nearly impossible” that the burglaries were not of buildings, id., the panel here concludes that the Government has met its burden of establishing the burglaries were of buildings because there is only a “more-than-remote possibility” that they were not, Foster, 662 F.3d at 295. But in reversing the First Circuit in Shepard, the Supreme Court prohibited ACCA courts from fact-finding by “common sense,” “logic,” or the law of probabilities. Rather, Shepard requires ACCA courts to limit their role to determining whether a small group of reliable documents prove beyond question that the prior offense constitutes a proper ACCA predicate.
Specifically, Shepard holds that only “conclusive records made or used” by the original sentencing court can provide the reliability needed to determine that the prior state convictions “necessarily ” (not very, very likely) qualified as proper ACCA predicates. 544 U.S. at 21, 125 S.Ct. 1254 (emphasis added). In reaching this holding, no member of the Shepard Court suggested that common sense or logic or probabilities (or judicial notice) could establish that buildings stood at the burglarized addresses. Indeed, even the Government disavowed the view that a court could consult extrinsic evidence. See Transcript of Oral Argument, Shepard, 544 U.S. 13, 27-28,125 S.Ct. 1254.
*397Moreover, the Shepard Court clearly understood that its holding would sometimes require courts to turn a blind eye to common sense and logic. Both the majority and the dissent recognized the overwhelming likelihood that the predicate burglaries in that case involved buildings, not boats. See, e.g., 544 U.S. at 22, 24 n. 4, 125 S.Ct. 1254; id. at 29, 125 S.Ct. 1254 (O’Connor, J., dissenting); Transcript of Oral Argument at 5 (“[T]here isn’t any question ... that the police reports in fact gave addresses of particular buildings.”) (Ginsburg, J.); id. at 8-9 (“There are no boats, you know, in Watertown. It’s not a dock----”) (Breyer, J.); id. at 24 (“This is a case where we all know what the truth is.... ”) (Kennedy, J.). Nevertheless, the Court required that proof that the burglarized entities were buildings come from a small group of reliable documents that “necessarily” established this fact.
Shepard set forth a bright-line rule that prohibits courts from “mak[ing] a disputed finding of fact about what the defendant and state judge must have understood as the factual basis of the prior plea.” 544 U.S. at 25, 125 S.Ct. 1254. Instead, Shepard requires that Shepard-approved documents alone conclusively establish that the state offense “ ‘necessarily’ rested on the fact[s]” qualifying it as an ACCA predicate. Id. at 21, 125 S.Ct. 1254. Even though the Court has recently recognized that the “absence of records will often frustrate application of the [ACCA],” Johnson v. United States, — U.S. -, 130 S.Ct. 1265, 1273, 176 L.Ed.2d 1 (2010), it has resisted efforts to back away from Shepard’s bright-line rule. See, e.g., United States v. Gutierrez-Ramirez, 405 F.3d 352 (5th Cir.), cert. denied 546 U.S. 888, 126 S.Ct. 217, 163 L.Ed.2d 198 (2005) (forbidding use of abstracts of judgment).
Understandably judges will be tempted to rely on “common sense” or “logic,” for these are normal fact-finding tools. But Shepard outlaws their use when determining if a prior conviction “necessarily” rests on facts qualifying it as an ACCA predicate. 544 U.S. at 21, 125 S.Ct. 1254. It does so because of the “practical difficulties” of non-contemporaneous fact-finding, id. at 20, 125 S.Ct. 1254, and to avoid the potential erosion of the “Jones-Apprendi implementation of the jury right,” id. at 24-25, 125 S.Ct. 1254 (Souter, J.).2 By resorting to “common sense” and “logic” the panel engages in precisely the sort of fact-finding that Shepard forbids.
In response to these thoughts, the concurrence contends that I have “missed the whole point of what the debate in Shepard was about.” According to the concurrence, the Shepard Court had “much bigger fish to fry” than “opinfing] on the relative validity of a street address as proof of a burglary of a building.” What the concurrence steadfastly ignores, however, is that only because all involved (Shepard, the Government, and the Supreme Court majority and dissent) agreed that the actual street addresses of burglarized establishments did not prove that they were buildings did the Court need to “fry” the “bigger fish,” i.e., resolve wheth*398er police reports or complaint applications could be considered in determining if the burglary was of a building and so a proper ACCA predicate. I suppose the Shepard Court could have opined on this “big fish” question even if it believed that street addresses sufficed to establish that a building had been burglarized. But that would make the entire Shepard opinion dicta. No court has suggested this. Rather, until today, courts have uniformly regarded the bright-line rule established in Shepard as the Court’s holding and consistently followed that holding.3
In short, I believe the panel has failed to abide by Shepard. The Shepard-approved documents in this case, which set forth only the proper (not generic) names of the places burglarized, are even less adequate to prove those places are buildings than the documents held inadequate in Shepard, which contained the actual street addresses of the places burglarized. The panel’s resort to “common sense” and “logic” constitutes the sort of fact-finding held impermissible in Shepard.
The panel’s errors have serious real life consequences for the defendant, John Joel Foster. These errors result in treatment of Foster’s possession of a firearm (a hunting rifle in his truck) as the crime of an armed career criminal. Thus these errors subject Foster to imprisonment for fifteen, rather than two and a half, years. This seems a particularly tragic result given that after committing the disputed state burglaries at nineteen, Foster remained conviction-free for almost two decades. If common sense and logic had any role here, surely they would suggest that this is not the record of an armed career criminal.
I must dissent. Judge King, Judge Gregory, Judge Davis, Judge Keenan, and Judge Floyd have authorized me to indicate that they join in this dissent.
. The Government acknowledged this at oral argument, see 544 U.S. at 21, 125 S.Ct. 1254 (quoting Brief for United States at 25), and the Shepard dissenters agreed with the majority on this point. Id. at 31, 125 S.Ct. 1254 (O'Connor, J., dissenting) ("If these complaints [containing the addresses] were the only evidence of the factual basis of Shepard’s guilty pleas, then I would agree with the majority that there was no way to know whether those convictions were for burglarizing a building.").
. Though Justice Souter’s Sixth Amendment discussion only garnered a four Justice plurality, Justice Thomas, the fifth vote, took an even stronger view of the need for constitutional protection and Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000). Thus Justice Thomas went further than the majority, arguing that "the [judicial] factfinding procedure the [majority] rejects gives rise to constitutional error, not doubt.” 544 U.S. at 28, 125 S.Ct. 1254 (Thomas, J., concurring in part and concurring in the judgment). Moreover, Justice Thomas argued that even the "limited fact-finding” approved of by the majority and by Taylor v. United States, 495 U.S. 575, 110 S.Ct. 2143, 109 L.Ed.2d 607 (1990), also ran afoul of Apprendi.
. Alternatively, the concurrence presses an even less persuasive challenge to Shepard's unequivocal holding. The concurrence suggests that for "five straight years and counting” the Court has retreated from Shepard's bright-line rule and endorsed fact-finding by “practicality,” "plain[] reasonable logic,” “probabilistic logic,” and "common sense.” This mixes apples and oranges. The cases on which the concurrence relies dealt with an "inherently probabilistic” inquiry, James v. United States, 550 U.S. 192, 207, 127 S.Ct. 1586, 167 L.Ed.2d 532 (2007), i.e., whether prior offenses "involve[d] conduct that presents a serious potential risk of physical injury to another." See 18 U.S.C. § 924(e)(2)(B)(ii) (emphasis added). This "probabilistic” inquiry is far removed from the Shepard inquiry, which asks only whether a prior conviction " ‘necessarily’ rested on the fact[s]” qualifying it as an ACCA predicate. Shepard, 544 U.S. at 21, 125 S.Ct. 1254.