concurring in part and dissenting in part.
I concur with the majority’s cogent analysis and conclusion respecting Stephens’s challenge to the sufficiency of the evidence contained in Part I.
I must, however, part ways with the majority in its treatment of Stephens’s Batson claim in Part II. It is unwise to consider Stephens’s Batson claim in the first instance when he failed to preserve the issue by objection during jury selection, and the district court did not raise the issue until long after it could have fashioned any relief. Even on the merits, however, I do not believe that Stephens has established a prima facie Batson claim warranting a remand. Moreover, the majority’s analysis muddles the three-part framework and the allocation of the burden of persuasion as set forth in the Bat-son line of cases.
At the outset, it bears repeating that this case comes to us in an unusual pos*519ture, in which Stephens’s Batson challenge was considered in the first instance at the appellate stage rather than at the district court level. At trial, the parties conducted jury selection — including the government’s peremptory strikes at issue here — without a whiff of complaint or objection on the part of Stephens’s counsel or, for that matter, the district court itself.
The only reason Stephens’s Batson claim is before us and not forfeited is the government’s agreement to waive any forfeiture argument for purposes of this appeal, so long as Stephens brought the Bat-son challenge on appeal rather than as a collateral attack under 28 U.S.C. § 2255. The rationale, we learn, is the government’s hope that we clarify something that should by now be perfectly obvious — under the Batson framework, a defendant has a meaningful burden to establish a prima facie case of discrimination before a court may require explanation from the government. This burden is not to be taken for granted or ignored. See United States v. Stewart, 65 F.3d 918, 925 (11th Cir.1995); United States v. Bergodere, 40 F.3d 512, 516 (1st Cir.1994).
Even though the government acquiesced to Stephens’s belated Batson challenge,1 I believe that a first-time consideration of his claim at this late stage is particularly unwise. The majority and I agree that one of the reasons we deferentially review district judges’ factual determinations with regard to Batson is the fact that they “sit[ ] in the unique position to make credibility assessments of the actions of trial attorneys ... [and] ha[ve] the opportunity to observe patterns and practices of particular attorneys during prior jury selections.” United States v. Cooper, 19 F.3d 1154, 1161-62 (7th Cir.1994); accord Bergodere, 40 F.3d at 517. This deference is in part due to judicial recognition that jury selection is a nuanced process “that is not an exact science. Its watchwords are judgment, flexibility, and discretion.” Bergodere, 40 F.3d at 517; see also Dunham v. Frank’s Nursery & Crafts, Inc., 967 F.2d 1121, 1126 (7th Cir.1992). Therefore, it is axiomatic that trial judges are normally in the best position to observe jury selection and resolve Batson challenges in the first instance and in the appropriate time frame. It is at this point that the majority and I must part ways. The reasons for deference have little force when the district court fails to act contemporaneously in response to an objection or even its own perception of possible discrimination during jury selection. See United States v. Chandler, 12 F.3d 1427, 1431 (7th Cir.1994) (“Contemporaneous objection is imperative with respect to Batson claims because the trial court frequently is in a position to rule on the objection, and in all probability[,] to resolve such claims.... ”).
The jury’s verdict in the case was returned on February 21, 2003. Then, over two months after completion of the trial, the district judge sua sponte raised the *520Batson issue. It is baffling that the district judge brought up a jury selection matter for which he no longer could provide a remedy. Moreover, it seems a distinctly bad idea for us to compound the problem by taking up where the district judge left off. Stephens’s counsel never made a timely objection during jury selection (or at any time, for that matter), nor did he speak up when the district court asked whether either side had any concerns following the peremptory strikes and before empaneling the jury. Cf. Aki-Khuam v. Davis, 339 F.3d 521, 527 (7th Cir.2003); United States v. Brisk, 171 F.3d 514, 523 (7th Cir.1999); Holder v. Welborn, 60 F.3d 383, 388 (7th Cir.1995); Doe v. Burnham, 6 F.3d 476, 481 (7th Cir.1993).
The majority aptly describes Stephens’s Batson claim as coming before us “via a circuitous path not typically seen.” And with good reason — abundant caselaw instructs that a party failing to make a timely Batson objection forfeits the right to do so on appeal. See, e.g., Chandler, 12 F.3d at 1431-32; accord Brown v. Kinney Shoe Corp., 237 F.3d 556, 561-62 (5th Cir.2001); McCrory, 82 F.3d at 1249. This is a sensible rule for any number of reasons, including judicial economy and fairness. See Ford v. Georgia, 498 U.S. 411, 422, 111 S.Ct. 850, 112 L.Ed.2d 935 (1991); McCrory, 82 F.3d at 1247 (“If ... a Batson objection may be raised after the jury has been sworn and trial has begun, there can be no remedy short of aborting the trial. This would permit the defendant to manipulate the system to the extreme prejudice of the prosecution and give the defendant a strong inducement to delay raising the objection until trial is underway.”). In addition, the opportunity to vindicate the rights of an excluded venireperson will be lost if an error is not corrected prior to empaneling the jury. See Galarza v. Keane, 252 F.3d 630, 642-43 (2d Cir.2001) (Walker, C.J., dissenting). These considerations do not disappear simply because the government affirmatively relinquished its forfeiture arguments, nor should the district court’s post-trial ruminations about possible discrimination — at a point when its ability to do something about the strikes had long since expired — weigh in favor of our hearing Stephens’s Batson challenge in the first instance.
This is not to imply that the district court’s belated order was on a par with a defendant’s tardy Batson objection brought after a guilty verdict. The district judge indeed was in a position to evaluate the government’s peremptory strikes (as was Stephens’s counsel) and to take action in a timely fashion. It is an open question whether a trial court must sua sponte raise a Batson issue at the first sign that something may be amiss. See Burnham, 6 F.3d at 481 (“We are aware of no case which authorizes a judge to invoke Batson when a party has never objected on that basis.”); accord Clark v. Newport News Shipbuilding & Dry Dock Co., 937 F.2d 934, 939 (4th Cir.1991) (“Neither Bat-son nor its progeny suggests that it is the duty of the court to act sua sponte to prevent discriminatory exclusion of jurors. Rather, even in criminal eases, the objection is deemed waived if not timely raised.”). But even if there is such an obligation, a court’s intervention should be timely, just as courts require of Batson objections from parties. Putting off action on potential Batson problems ill serves the parties and excluded venirepersons, and also increases costs to the judicial system. See Brisk, 171 F.3d at 523 (“When a new trial must be granted because of an untimely Batson decision, the error imposes an additional and unnecessary expenditure of judicial and litigant resources since a new trial could have been avoided by a timely decision.”) (citation and internal quotation marks omitted).
*521Finally, the nuanced nature of voir dire itself demands prompt resolution of potential Batson problems. See McCrory, 82 F.3d at 1248. Delay increases the risk that the lawyers involved in jury selection will not be available or will not adequately recall details of voir dire and the actual reasons for striking particular venirepersons. See Holder, 60 F.3d at 388. Indeed, I note that in the government’s motion to vacate the district court’s post-trial Batson ruling, filed less than four months after voir dire, the government represented that its prosecutors could not reliably recall details of the racial composition of the venire. (R. 39 at 16.) Assuming these same prosecutors are available to appear at a Batson hearing on remand, it is unlikely that their memories have improved in the years since jury selection in this ease. Cf. Carter v. Hopkins, 151 F.3d 872, 875 (8th Cir.1998).
These prudential concerns aside, a review of Stephens’s belated challenge on the merits must comport with the well-established Batson framework. In the first stage of this three-part framework, a defendant seeking to establish a prima facie case “must point to facts and circumstances raising an inference that the potential jurors were excluded because of race.” Cooper, 19 F.3d at 1159. Only after the defendant makes out a prima facie case do matters proceed to the second stage, in which the government must offer race-neutral actual reasons for its challenges, see McCain v. Gramley, 96 F.3d 288, 290 (7th Cir.1996), and the final stage, in which the trial court decides whether the government’s proffered reasons are pretextual, indicating that race discrimination is afoot. See United States v. Alanis, 265 F.3d 576, 584 (7th Cir.2001). At all times during this analysis, the burden of persuasion rests with, and never shifts from, the opponent of the peremptory strike. See Purkett v. Elem, 514 U.S. 765, 768, 115 S.Ct. 1769, 131 L.Ed.2d 834 (1995) (per curiam); United States v. George, 363 F.3d 666, 673 (7th Cir.2004).
Despite the government’s hopes, the majority’s analysis does nothing to clarify the prima facie burden of a defendant making a Batson challenge. The majority instead finds a pattern of racially disproportionate strikes by the government despite the small numbers involved and then effectively places the burden on the government to rebut the inference of discrimination, even though it is Stephens who must carry the burden at this stage. Under the majority’s analysis, the statistically disproportionate strikes alone apparently are enough to satisfy that burden, so the challenging party really does not need to come forward with any evidence at all. Thus, a party must explain its actions any time it strikes a minority venireperson in a manner disproportionate to the racial makeup of the venire, no matter that the numbers of minorities on the venire are so small that even a single strike may be facially disproportionate.
In this case, the government used 33% of its strikes (2 out of 7) against African-Americans, who represented 9.7% of the venire (3 out of 31, following strikes for cause). According to the majority, the inference of discrimination arises from this disproportionate use of peremptory challenges against African-Americans, and statistical evidence of the government’s strikes against the other minorities constitute additional “relevant circumstances” supporting an inference of discrimination. But we have expressed doubt that statistical evidence alone rises to the level of establishing a prima facie case. See McCain, 96 F.3d at 291 (“[I]t is illogical to infer in every case a discriminatory intent from noting that the percentage of challenges used by one party against members of a racial group is either more or less *522than the percentage of that group’s total percentage of the venire panel.”); accord Allen v. Lee, 366 F.3d 319, 330 (4th Cir.2004) (en banc) (“Though statistics are not utterly bereft of analytical value, they are, at best, manipulable and untrustworthy absent a holistic view of the circumstances to which they apply.”).
Rather than looking to circumstances that actually might be relevant, the majority simply looks to more raw numbers- — the government’s strikes against Hispanics and the sole Asian venireperson — and concludes that the pattern of discrimination is established. It is, however, problematic to infer that strikes may be discriminatory simply because peremptory strikes fall disproportionately among members of a certain group. See United States v. Roberts, 163 F.3d 998, 999 (7th Cir.1998) (“Batson establishes a rule of disparate treatment, not of disparate impact....”); United States v. Cooke, 110 F.3d 1288, 1301 (7th Cir.1997) (“[The defendant] must do more than merely point to the fact that the government excluded an African-American venireperson .... ”); Alverio v. Sam’s Warehouse Club, Inc., 253 F.3d 933, 941 (7th Cir.2001) (“[T]he exclusion of all members of a specific minority group does not, on its own, establish that the peremptory strikes were discriminatory.”).2 An apparently disproportionate pattern of strikes may have relevance, but even so, the exis-fence of such a pattern alone does not end the inquiry. See McCain, 96 F.3d at 292.
A “pattern” is more appropriately understood from a party’s actions and the way in which it conducts peremptory strikes, as opposed to raw numbers alone. See id. at 291-92 (“[A] ‘pattern’ does not necessarily correlate to the racial proportions of the venire panel.... A ‘pattern’ is more likely demonstrated by the manner in which a party uses its strikes as compared to its total strikes or to the total number of members of the racial group.”) (emphasis in original). Thus, “even if a ‘pattern’ could be said to exist, that fact is not dispositive. Courts must look to the totality of the circumstances, including the final make-up of the jury and the questions asked by the party.” Id. (citing Batson and Splunge v. Clark, 960 F.2d 705, 707 (7th Cir.1992) (finding relevant that the prosecutor demanded certain responses only from African-American venire-per-sons)); accord Bergodere, 40 F.3d at 516 (“A defendant who advances a Batson argument ordinarily should come forward with facts, not just numbers alone.”) (citation and internal quotes omitted).
The totality of the circumstances present in this case reveals that there is no prima facie evidence of race discrimination. The district judge alone conducted voir dire, so there could not have been discriminatory questioning by the prosecu*523tors, and there is nothing in the record to indicate any discriminatory behavior or statements on the prosecutors’ part. The only hint of discrimination to be found comes from racial statistics, and these are inadequate to carry Stephens’s prima facie burden.
A closer look at the parties’ use of peremptory strikes reveals why it would be inappropriate to conclude that “numbers alone” can constitute a pattern satisfying the prima facie burden. Because the number of minority venirepersons was so small, the magnitude of the statistically disproportionate strikes against minorities was exaggerated to such an extent that it would have been impossible for the government to strike any minorities without creating an inference of discrimination under the majority’s expansive formulation. For example, even if the government had used all 7 of its strikes, and eliminated even one African-American, 14.2% of its strikes would have eliminated a minority representing only 9.7% of the venire. The effect is even more pronounced if (as happened here) the government strikes the lone member of a particular minority group — in this case, the government used one of its six strikes to remove the sole Asian-Ameriean venireperson (thus using 17% of its challenges to eliminate 3% of the venire).
The majority simply aggregates the statistically disproportionate strikes against members of several minority groups to find a “pattern,” and the majority’s view makes no allowance for the exaggerated effect due to small numbers. True, our caselaw does not suggest a numerical cutoff for when a disproportionate pattern of strikes by itself may become statistically significant and properly support an inference of discrimination. But surely this court’s demonstrated reluctance to rely on numbers alone (and our requirement that a Batson challenger point to additional “facts and circumstances”) is at least partly a response to making too much of statistics when the numbers involved are so small. Cf. Cooke, 110 F.3d at 1301; McCain, 96 F.3d at 291-92.
Instead of looking to numbers alone, the better approach is to look at the manner of the strikes, and other facts that indicate whether a party acted with discriminatory purpose, in order to put statistics into a useful context. See McCain, 96 F.3d at 291-92. In this case, for example, the majority focuses solely on the raw numbers of the government’s peremptory strikes, but does not address the inferences to be drawn by the strikes the government did not make. It must be relevant, for example, that the government did not exercise all of its peremptory challenges, although minorities remained in the venire and eligible for jury service. Cf United States v. Griffin, 194 F.3d 808, 826 (7th Cir.1999) (“[T]he fact that the [government did not challenge the other black juror further weakens the argument that the government’s strikes were based on a motive to discriminate.”) (citation and internal quotes omitted). Specifically, the government did not use its remaining peremptory challenge to remove Juror # 4, a Hispanic woman who was seated as a juror, or Juror # 13, an African-American woman. Indeed, with regard to these jurors, the prosecutors’ acquiescence to their service is consistent with the “apparent” reasons for its peremptory strikes, as discussed infra.
But even if numbers alone can trigger an inference of discrimination, Stephens is lucky that the government did not initiate a Batson challenge of its own. Stephens himself struck from the venire one African-American (Juror # 13), one Hispanic, and nine Caucasians. Looking only at statistics, as the majority does, one could *524arguably infer discrimination on Stephens’s part — after all, he complains on appeal that no African-Americans and only one Hispanic remained on the petit jury, yet he struck one third of the African-Americans and one quarter of the Hispanic venireper-sons himself.
The majority questions this observation and infers that I suggest that the government’s actions are immune from scrutiny because Stephens himself struck minority venirepersons. I certainly wish to dispel any illusion that I subscribe to the latter proposition. I have called attention to Stephens’s strikes not to suggest that they cancel out possible wrongdoing on the government’s part, but instead, to further illustrate the problem of inferring discriminatory intent from statistics alone. We have said that one relevant factor to consider in the totality of the circumstances is the final makeup of the jury, e.g., McCain, 96 F.3d at 292, and it is indisputable that Stephens himself has helped to bring about a jury with no African-Americans and one fewer Hispanic. Cf. Mahaffey v. Page, 162 F.3d 481, 484-85 (7th Cir.1998) (noting that the most important factor in the case was that the jury did not include any African-Americans, and thus “not a single member of Mahaffey’s own race was seated on the jury that decided his fate”).
Nevertheless, the majority instead praises Stephens’s strikes because they bear a symmetry to the racial composition of the venire and concludes that nothing about Stephens’s strikes suggests an effort to disproportionately eliminate a particular racial group. Racial proportionality is not, however, the standard by which we are to assess whether a party’s peremptory strikes run afoul of Batson. If that is what the majority proposes (that racially proportionate strikes are necessarily immune from Batson scrutiny), it is the majority that has endorsed an astounding proposition. There is no principle requiring racially proportional strikes. E.g., Batson, 476 U.S. at 86 n. 6, 106 S.Ct. 1712; McCain, 96 F.3d at 291. And individuals, not racial groups, have the right to serve on juries. Cf. Powers v. Ohio, 499 U.S. 400, 409-10, 111 S.Ct. 1364, 113 L.Ed.2d 411 (1991). Indeed, if a party purposely set out to empanel a jury directly proportional to the racial makeup of the venire, it very likely would be forced to discriminate against some venirepersons on the basis of race in order to achieve that balance. No matter how noble its intentions, such a strategy would offend Batson just as surely as one that struck jurors on the basis of discriminatory stereo-types. Cf. United States v. Nelson, 277 F.3d 164, 209-12 (2d Cir.2002) (concluding that racial or religious “jurymandering,” whether by the parties or by the trial court, is impermissible).
Having found a prima facie case on the numbers alone, the majority looks to other factors present in this case and concludes that none of them changes its conclusion. For example, the majority concludes that because Stephens is African-American, and the trial witnesses Caucasian, a circumstance arises that “does nothing to lessen the inference of discrimination.” To me, this puts the cart before the horse, because it suggests that the government must, at the prima facie stage, present evidence to contradict the majority’s conclusion that Stephens has met his burden.
In any event, the fact that the prosecution’s witnesses were all Caucasian does not present a racially sensitive situation whereby peremptory challenges against African American jurors (let alone jurors of other minority races) are to be viewed with even greater scrutiny. Moreover, the underlying facts and subject matter of the ease are not racially charged and do not suggest that strikes against minority veni-*525repersons might weigh in favor of finding an inference of discrimination. This case is therefore quite uniike those cases in which race had special significance — for example, the racially inflammatory situation presented in Mahaffey, in which an African-American defendant from Chicago’s south side was prosecuted for horrific crimes committed against a Caucasian family who lived in a north side neighborhood. See Mahaffey, 162 F.3d at 485; accord Stewart, 65 F.3d at 925.
In the present case, Stephens, who attended Yale, was holding down a white-collar management position at Accenture and was accused of using a computer function to defraud his employer of over $60,000 in unauthorized cash advances. Nothing about the crime alleged raises the specter of racial inflammation, and the fact that the prosecution’s witnesses all happened to be Caucasian does not change this conclusion. Cf. United States v. Grandison, 885 F.2d 143, 149 (4th Cir.1989) (“[T]o infer prosecutorial discrimination because of the race of government witnesses has serious implications.... [A]ny party is forced to take its witnesses as it finds them.”). This is simply not a case in which there is legitimate concern that racial issues could play a role in jury selection or the outcome of the trial.
The majority also finds additional support for Stephens’s prima facie case by engaging in a detailed parsing of what the government offers as “apparent” race-neutral reasons for its strikes (reasons that the government apparently felt compelled to offer in order to rebut the presumption of discrimination). In this regard, it is worth noting that Stephens, the government, and the majority all go to great lengths to stress that the government’s “apparent” rationale is not the same as its “actual” reasons for the strikes, which, it is alleged, the government has not yet presented. This seems to be nothing but a convenient fiction, in which everyone can pretend that we still are at the prima facie stage of the Batson inquiry simply by agreeing that the reasons “apparent” in the record are not the same as whatever actual reasons the government has yet to offer. Accord Stewart, 65 F.3d at 925 (“No party challenging the opposing party’s use of a peremptory strike ... is entitled to an explanation for that strike, much less to have it disallowed, unless and until a prima facie showing of racial discrimination is made.”); cf. Mahaffey, 162 F.3d at 483-84.
Though it claims to be doing otherwise, the majority weighs the very factors that it concedes to be appropriate in the next stage of review and finds the government’s “apparent” reasons insufficient to undermine the inference drawn from the disproportionate strikes alone. And in weighing the government’s “apparent” reasons, and giving them no credit whatsoever, the majority blurs the Batson framework by shifting the burden of persuasion to the government at the prima facie stage.
The majority does this by scouring the government’s “apparent” race-neutral justifications for purported contradictions— namely, that some Caucasian jurors shared certain traits with dismissed minority veni-repersons — and concludes that the record fails to provide evidence to negate the majority’s inference of discrimination. Cf. Alverio, 253 F.3d at 941 (“[Wjhere a party gives multiple reasons for striking a juror, it is not enough for the other side to assert that the empaneled juror shares one attribute with the struck juror.”) (citation omitted). But this conclusion seems to me nothing more than a thinly disguised finding that the government’s reasons, whether one calls them “apparent” or “actual,” are not to be credited because the supposed contradictions render them suspect.
*526The majority thus holds the government to a higher standard, at the prima facie stage, than the ordinarily low threshold the government must meet in the second stage of the Batson framework. See Purkett, 514 U.S. at 769, 115 S.Ct. 1769 (noting that a “ ‘legitimate reason’ is not a reason that makes sense, but a reason that does not deny equal proteetion[,]” and finding acceptable the prosecutor’s explanations that he struck jurors for “long, unkempt hair, a mustache, and a beard”); see also United States v. Evans, 192 F.3d 698, 701 (7th Cir.1999) (“Any neutral reason, no matter how implausible or fantastic, even if it is silly or superstitious, is sufficient to rebut a prima facie case of discrimination.”) (citation and internal quotes omitted). Certainly, the required showing is minimal compared to the burden the majority has effectively imposed on the government here at the prima facie stage, in which facially race-neutral “apparent” reasons are found wanting by the majority. This amounts to a finding that the government has offered pretextual reasons for its strikes — a determination appropriate only in the final stage of the Batson inquiry, not at the prima facie stage. Cf. Purkett, 514 U.S. at 768, 115 S.Ct. 1769 (“It is not until the third step that the persuasiveness of the justification becomes relevant[.]”)
Nevertheless, the government’s “apparent” reasons for its peremptory strikes— including, for example, level of education, employment, or errors on jury questionnaires — are not facially discriminatory, outlandish, or otherwise improper. See, e.g., Alanis, 265 F.3d at 584 (education); Alverio, 253 F.3d at 941 (employment); United States v. Smith, 324 F.3d 922, 927 (7th Cir.2003) (mistakes on juror questionnaire). In fact, given the nature of the case against Stephens, the “apparent” reasons for the government’s strikes are perfectly consistent with what we may assume was the government’s prosecution strategy. As the government pointed out, this case is a white-collar fraud case. Much of the evidence against Stephens involved accounting and computer functions, so it is not surprising that the government would seek jurors apparently better suited to understand the nature of the case and the evidence.
Despite the majority’s assumption from supposed inconsistencies in the record that the government was of a mind to discriminate, the government’s actions in exercising its strikes are perfectly consistent with a legitimate, race-neutral strategy given the nature of the wire-fraud ease against Stephens.
Indeed, as explained earlier, the government did not strike Juror # 13, an African-American woman, from the venire. The government’s reasons for wanting Juror # 13 on the jury are readily apparent from the record — on her juror questionnaire and during voir dire, Juror # 13 indicated that she had experience working in the business world and holds an MBA from the University of Chicago. In other words, Juror # 13 was an ideal juror from the government’s standpoint, because her education and experience would be helpful in understanding the evidence presented at trial. Likewise, the government did not strike Juror # 4, a Hispanic woman, who has an associate’s degree in accounting. As with Juror # 13, the government likely viewed Juror # 4 as being well suited to understand the evidence against Stephens. The prosecutors therefore did not strike either of these minority venirepersons, but did strike other minorities who did not have comparable traits.
Stephens, in contrast, apparently struck jurors according to an opposite strategy. As noted earlier, it was Stephens, not the government, who struck Juror # 13, again for reasons not difficult to divine. Juror *527# 13’s qualifications — particularly her education and white-collar business experience — made her more likely to understand the technical aspects of the prosecution’s case, possibly pro-business, and perhaps less sympathetic to Stephens.
The majority sidesteps all of these relevant factors in its zeal to find inconsistencies in the government’s “apparent” facts. Worse, the majority places no stock in any of these other “apparent” facts, so I wonder what “actual” reasons the government on remand could possibly offer in an attempt to undermine the apparently foregone conclusion that it has engaged in discrimination because the numbers are so far out of kilter. At any rate, a Batson hearing seems largely superfluous at this point, given the majority’s conclusions that the government’s “apparent” reasons for its strikes are inconsistent (and therefore, presumably, pretextual).
Finally, I must reiterate the government’s concern that parties making Batson challenges should offer some meaningful quantum of proof, not merely statistics, in order to satisfy their prima facie burden. Batson and its progeny make clear that courts are not simply to skip over the prima facie stage. Aki-Khuam, 339 F.3d at 527 (finding that the trial court improperly “replaced the first step of the Batson analysis with [its] presumption of purposeful discrimination, thereby saddling [the party making the strikes] with the burden of overcoming that presumption”). As in the present case, local demographics and chance largely determine the racial makeup of any given venire, and peremptory strikes may as a consequence disproportionately affect certain racial groups simply as a matter of numbers (particularly where the numbers are small, as in this case). But courts must look beyond statistics and require litigants to carry their respective burdens under the Batson framework, or we risk paying lip service to clearly established caselaw and arguably set the stage for a rule that in practice (just as Stephens would have it) calls for a Batson hearing every time a party strikes a minority member from the venire. See Cooke, 110 F.3d at 1301 (“[The defendant] must point to facts and circumstances raising an inference that the potential juror was excluded because of race. Otherwise, every peremptory challenge used to exclude any cognizable minority from a petit jury would require a Batson-type hearing.”) (citation and internal revisions omitted).
Although my colleagues reprove me for suggesting a “parade of horribles,” I believe the majority’s Batson holding represents yet another step toward elimination of the peremptory challenge, which is undeniably an important and integral part of jury selection in our adversarial system. See Batson, 476 U.S. at 112, 106 S.Ct. 1712 (Burger, C.J., dissenting) (observing that the peremptory challenge is “a procedure which has been part of the common law for many centuries and part of our jury system for nearly 200 years.”); Burnham, 6 F.3d at 481 (“Tradition engraves the process of peremptory challenges into our system[.]”). At the least, a jury selection regime that places undue emphasis on racial proportions places form over substance and creates improper incentives. For example, parties may opt to use all of their peremptory strikes in order to mitigate any possible disproportionate impact on minorities (or anyone else protected by Batson and its progeny). Worse, parties could engineer their peremptory challenges to mirror the racial proportions of the venire, thus discriminating on the basis of race. And courts that fail sua sponte to respond to statistically disproportionate strikes risk remand, even if (as in this case) ample evidence supports the jury’s verdict and there is no indication *528that race was at issue in the trial. Before long, peremptory challenges will simply merge with challenges for cause if litigants must explain every strike of a protected venireperson, thus eliminating altogether a practice designed to assist litigants in the imprecise but necessary science of jury selection. Cf. Pruitt v. McAdory, 337 F.3d 921, 930-31 (7th Cir.2003) (“Picking jurors is a complex and multifaceted process. Individual factors or characteristics often do not provide the ‘silver bullet’ that will mean acceptance or rejection of any potential juror. Rather, it is a combination of factors that will determine whether a party believes a juror will be favorable to their side[.]”).
In conclusion, for the reasons set forth above, I respectfully dissent as to Part II of the majority opinion.
. Naturally, the government understood that Stephens could have brought a Batson challenge by collateral attack rather than on direct appeal. At argument, the government expressed its concern that trial courts, when resolving Batson challenges, often rubber-stamp defendants’ prima facie burden and move prematurely to the second and third stages, thus making it the government's burden to offer its actual reasons every time it strikes minority venirepersons. The government wanted Stephens's case to be heard on the merits in order to clarify a Batson challenger's burden at the prima facie stage, see discussion infra. The government certainly could have made a compelling argument that Stephens forfeited his Batson claim by not raising it in a timely fashion in the district court. See, e.g., McCrory v. Henderson, 82 F.3d 1243, 1249 (2d Cir.1996) (reversing grant of habeas writ, holding that “the failure to object to an adversary’s use of peremptory challenges until after the completion of jury selection waives the right to do so”).
. My colleagues dismiss these cases as inap-posite with regard to the prima facie stage. I cite these cases, however, for the broad point that racially disproportionate strikes alone will not suffice to show that peremptory strikes are discriminatory, whether at the pri-ma facie stage or beyond. But if these cases are inapposite for not addressing the quantum of proof required at the prima facia stage, so is Miller-El (cited by the majority), which itself concerned the third stage of the Batson framework. See Miller-El v. Cockrell, 537 U.S. 322, 339, 123 S.Ct. 1029, 154 L.Ed.2d 931 (2003). More relevant, the other case I cite, McCain, was decided by this court, directly addresses the prima facie question, and expresses this circuit's skepticism that numbers alone will satisfy the prima facie stage. In contrast, the majority cites cases from the Second and Ninth Circuits, which, unlike this court, have expressly adopted the position that statistical disparity alone may satisfy the burden at the prima facie stage. Cf. Brewer v. Marshall, 119 F.3d 993, 1004-05 (1st Cir.1997) (contrasting Second and Ninth Circuit views on statistical evidence with McCain’s reservations regarding same).