dissenting:
The government’s “primary” piece of evidence, Appellants’ Br. 10, is a single report: [redactions] (the Report) [redactions] After carefully laying out the parties’ arguments about the Report’s internal and external indicia of reliability, the district court found it “not sufficiently reliable to support a finding by a preponderance of the evidence that Latif was recruited by an A1 Qaeda member or trained and fought with the Taliban.” Abdah' (Latif) v. Obama, No. 04-cv-01254, slip op. at 25 (D.D.C. July 21, 2010). According to the district court, “there is a serious question as to whether the [Report] accurately reflects Latifs words, the incriminating facts in the [Report] are not corroborated, and Latif has presented a plausible alternative story to explain his travel.” Id. at 26. The government concedes that its case for lawfully detaining Latif “turn[s]” on the Report. Appellants’ Br. 5. This, then, represents a first among the Guantanamo habeas appeals in this circuit: never before have we reviewed a habeas grant to a Guantanamo detainee where all concede that if the district court’s fact findings are sustained, then detention is unlawful. Cf. Almerfedi v. Obama, 654 F.3d 1, 5-7 (D.C.Cir.2011) (reversing habeas grant and finding detention lawful based on conceded facts and facts found by the district court); Uthman v. Obama, 637 F.3d 400, 402 (D.C.Cir.2011) (same); Al-Adahi v. Obama, 613 F.3d 1102, 1103, 1111 (D.C.Cir.2010) (same).
But rather than apply ordinary and highly deferential clear error review to the district court’s findings of fact, as this circuit has done when district courts have found the government’s primary evidence reliable, the court, now facing a finding that such evidence is unreliable, moves the goal posts. According to the court, because the Report is a government-produced document, the district court was required to presume it accurate unless Latif could rebut that presumption. Maj. Op. at 1181. In imposing this new presump*263tion and then proceeding to find that it has not been rebutted, the court denies Latif the “meaningful opportunity” to contest the lawfulness of his detention guaranteed by Boumediene v. Bush, 553 U.S. 723, 779, 128 S.Ct. 2229, 171 L.Ed.2d 41 (2008).
Compounding this error, the court undertakes a wholesale revision of the district court’s careful fact findings. Flaws in the Report the district court found serious, this court now finds minor. Latif s account, which the district court found plausible and corroborated by documentary evidence this court now finds “hard to swallow,” Maj. Op. at 1194. [redactions] the district court found did not implicate Latif, this court now finds do in fact implicate him. And on and on, all without ever concluding that the district court’s particular take on the evidence was clearly erroneous. But see Fed.R.Civ.P. 52(a)(6) (“Finding of facts, whether based on oral or other evidence, must not be set aside unless clearly erroneous.... ”).
In Part I, I explain why the district court committed no error in declining to apply a presumption of regularity to the Report. In Part II, I apply the deferential clear error standard this circuit has used throughout these Guantanamo habeas cases. Finding no clear error, I would affirm the district court’s grant of the writ of habeas corpus.
I.
All agree that this case turns on whether the district court correctly found that the government’s key piece of evidence, the Report, was unreliable. And all agree that the “question whether evidence is sufficiently reliable to credit is one we review for clear error.” Al Alwi v. Obama, 653 F.3d 11, 19 (D.C.Cir.2011). Our disagreement centers on whether the district court was required to afford the Report a presumption of regularity.
The presumption of regularity stems from a humble proposition—that “[public officers] have properly discharged their official duties.” Sussman v. U.S. Marshals Serv., 494 F.3d 1106, 1117 (D.C.Cir.2007) (quoting United States v. Chem. Found., Inc., 272 U.S. 1, 14-15, 47 S.Ct. 1, 71 L.Ed. 131 (1926)). The contours of the presumption are best understood by how courts typically apply it. For example, courts assume that “official tax receipts]” are properly produced, Riggs Nat’l Corp. v. Comm’r, 295 F.3d 16, 21 (D.C.Cir.2002), that state court documents accurately reflect the proceedings they describe, Hobbs v. Blackburn, 752 F.2d 1079, 1081 (5th Cir.1985), that mail was duly handled and delivered, Legible v. Dann, 544 F.2d 1, 7 n. 39 (D.C.Cir.1976), and that agency actions in the ordinary course of business are undertaken on the basis of fact, Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 415, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971) (citing Pacific States Box & Basket Co. v. White, 296 U.S. 176, 185, 56 S.Ct. 159, 80 L.Ed. 138 (1935)).
These cases—in fact every case applying the presumption of regularity—have something in common: actions taken or documents produced within a process that is generally reliable because it is, for example, transparent, accessible, and often familiar. As a result, courts have no reason to question the output of such processes in any given case absent specific evidence of error. Such a presumption rests on common sense. For instance, courts have no grounds to credit a defendant’s allegation that “the state court trial docket” or “the waiver of trial by jury form” contain inaccurate information when that defendant has no support other than a self-serving allegation. See Thompson v. Estelle, 642 F.2d 996, 998 (5th Cir.1981) (noting that the “district court could properly rely upon the regularity of the state court’s docu*264ments in preference to [the appellant’s] self-serving testimony”). Courts presume accuracy because they can trust the reliability of documents produced by such processes. Courts and agencies are hardly infallible, but for the most part we have sufficient familiarity and experience with such institutions to allow us to comfortably rely on documents they produce in the ordinary course of business.
In saying that “[c]ourts regularly apply the presumption ... [to] processes that are anything but ‘transparent,’ ‘accessible,’ and ‘familiar,’ ” Maj. Op. at 1182, this court cites a single case where we presumed the accuracy of a tax receipt from the Central Bank of Brazil for purposes of claiming foreign tax credits under the Internal Revenue Code. See id. at 1182 (citing Riggs Nat’l Corp., 295 F.3d at 20-22). As the Supreme Court has held, the presumption of regularity applies to “the actions of tax officials,” and the “records of foreign public officials.” See Riggs Nat’l Corp., 295 F.3d at 20 (citing Supreme Court cases). But again, this is because we have no reason to question or be concerned with the reliability of such records.
By contrast, the Report at issue here was produced in the fog of war by a clandestine method that we know almost nothing about. It is not familiar, transparent, generally understood as reliable, or accessible; nor is it mundane, quotidian data entry akin to state court dockets or tax receipts. Its output, a [redactions] intelligence report, was, in this court’s own words, “prepared in stressful and chaotic conditions, filtered through interpreters, subject to transcription errors, and heavily redacted for national security purposes.” Maj. Op. at 1179. Needless to say, this is quite different from assuming the mail is delivered or that a court employee has accurately jotted down minutes from a meeting.
To support its approach here, this court invokes presumptions of regularity for state court fact-finding and for final judgments in criminal habeas proceedings. See id. at 1181-82. Aside from the abstract and uncontroversial proposition that courts should be sensitive to the separation of powers as well as to federalism, id. at 1181-82, the analogy makes little sense. State court judgments and fact findings arise out of a formal and public adversarial process where parties generally have attorneys to zealously guard their interests, and where neutral state court judges, no less than federal judges, pledge to apply the law faithfully. That federal courts give a presumption of regularity to judgments and fact findings that emerge from such a process, where criminal defendants have ample opportunity to challenge adverse evidence, see Lackawanna Cnty. Dist. Att’y v. Coss, 532 U.S. 394, 402-03, 121 S.Ct. 1567, 149 L.Ed.2d 608 (2001), provides no reason for habeas courts also to presume the accuracy of [redactions] intelligence reports prepared in the fog of war. Indeed, unlike statutory habeas, where federal review follows state court proceedings, constitutional habeas is the only process afforded Guantanamo detainees. Cf. Boumediene, 553 U.S. at 780, 128 S.Ct. 2229 (“It appears that the common-law habeas court’s role was most extensive in cases of pretrial and noncriminal detention, where there had been little or no previous judicial review of the cause for detention. Notably, the black-letter rule that prisoners could not controvert facts in the jailer’s return was not followed (or at least not with consistency) in such cases.”).
In its analysis, this court ignores a key step in the logic of applying a presumption of regularity, namely, that the challenged document emerged from a process that we can safely rely upon to produce accurate information. Reliability, not whether an *265official duty was performed, cf. Maj. Op. at 1178-79, is the touchstone inquiry in every case this court cites. For example, in a probation revocation decision by the Seventh Circuit—which, incidentally, never uses the term “regularity,” see United States v. Thomas, 934 F.2d 840 (7th Cir.1991)—the court found that the probation report “was of the type that generally carries a presumption of reliability,” id. at 846 (emphasis added). A probation officer not only “testified [and was cross-examined] about the preparation, maintenance, and interpretation of special reports prepared by the probation office” but also “applied that ... knowledge to [the report at issue].” Id. at 842. Given that testimony, and given also that “the district court ... had reviewed the report ‘many times,’ ” the Seventh Circuit saw no reason to think the report was “inaccurate.” Id. at 846. Reinforcing its emphasis on the importance of assessing reliability, the Seventh Circuit cited an earlier decision, United States v. Verbeke, where it had found admissible a report produced by a drug treatment center because the report was found to be “reliable,” because the defendant had an opportunity to cross-examine its author, and because no evidence discredited it. 853 F.2d 537, 539 (7th Cir.1988). These decisions do not, as this court now does, ask only whether an official duty was regularly performed; rather, they examine the reliability of the proffered evidence and the process that produced it. As yet another decision the court cites puts it, courts will permit “the introduction of ‘demonstrably reliable ’ hearsay evidence in probation revocation proceedings.” United States v. McCallum, 677 F.2d 1024, 1026 (4th Cir.1982) (emphasis added).
To be sure, the government in this case has produced a declaration stating [redactions] see Maj. Op. at 1186 (quoting this affidavit). But we have no idea what the [redactions] is, nor anywhere near the level of familiarity or experience with that course of business that would allow us to comfortably make presumptions about whether the output of that process is reliable. Cf. Bismullah v. Gates, 501 F.3d 178, 185-86 (D.C.Cir.2007) (finding that it was “not at all clear” that even the Combatant Status Review Tribunal was “entitled to a presumption of regularity ... because a CSRT does not have the transparent features of the ordinary administrative process and the [military officer charged with obtaining and reviewing evidence] is not the final agency decisionmaker”). Of course, we may take some assurance from the fact that the Executive Branch acts in good faith when carrying out its duties. But the very point of Boumediene is to ensure that detainees have a “meaningful opportunity” to subject the Executive’s detention decisions to scrutiny by an independent Article III court.
This is not to say that reports similar to the one at issue here are necessarily unreliable. Perhaps after careful scrutiny district courts will conclude that many are reliable. See, e.g., Khan v. Obama, 655 F.3d 20, 24-26 (D.C.Cir.2011). My point is far more modest: because we are unfamiliar with this highly secretive process, and because we have no basis on which to draw conclusions about the general reliability of its output, we should refrain from categorically affording it presumptions one way or the other. This approach does not reflect “skepticism]” or “cynic[ism]” about the Executive Branch, Maj. Op. at 1179-80—it is nothing more than what Boumediene directs us to do. See Boumediene, 553 U.S. at 786, 128 S.Ct. 2229 (requiring habeas court “to assess ” not presume, “the sufficiency of the Government’s evidence” (emphasis added)). And indeed, from time immemorial courts have been skeptical of hearsay evidence without implying bad *266faith or cynicism about the Executive (or whoever is attempting to present that evidence).
Nor am I suggesting that district courts should give no weight to sworn declarations by government officials that such reports are [redactions] because [redactions] Up to a point, the declaration does provide support for the Report’s reliability. For one thing, it suggests that the Report is in fact authentic, i.e., that it really is an interrogation summary. Relying on similar declarations, many district courts that have heard Guantanamo habeas cases— including the district court here—have adopted a presumption of authenticity for government records like the Report even while consistently rejecting a presumption that such records are accurate. See, e.g., Alsabri v. Obama, 764 F.Supp.2d 60, 66-67 & n. 8 (D.D.C.2011); Hatim v. Obama, 677 F.Supp.2d 1, 10 (D.D.C.2009), vacated on other grounds, 682 F.3d 720 (D.C.Cir.2011); Ahmed v. Obama, 613 F.Supp.2d 51, 54-55 (D.D.C.2009). But see, e.g., Al Kandari v. United States, 744 F.Supp.2d 11, 19-20 (D.D.C.2010) (declining to adopt a presumption of either authenticity or accuracy). Going one step further, habeas courts might also properly rely on the analogy between intelligence reports and business records to conclude that “[t]he fact that these reports were prepared by government agents in the course of their normal intelligence gathering duties provides a degree of support for their reliability.” Alsabri, 764 F.Supp.2d at 68. I thus have no problem with the observation, made in a decision cited by the concurrence, Con. Op. at 1204, that “the basic fact that public officials usually do their duty ... has ... that quality and quantity of probative value to which it is entitled.” Stone v. Stone, 136 F.2d 761, 763 (D.C.Cir.1943). As that decision goes on to say, however, “the probative strength of the evidence is for the [factfinder] to consider.” Id. Nor do I quarrel with the observation that, as a general matter, government records consisting of interrogation summaries with inculpatory admissions are more likely to be reliable evidence than documents reporting third-party (and sometimes anonymous) hearsay.
But this court goes well beyond these modest conclusions—and well beyond what the government actually argues in its briefs—when it relies on the bare fact that government officials have incentives to maintain careful intelligence reports as a reason to require district courts to presume that such reports are not only authentic, but also accurate, despite circumstances casting their reliability into serious doubt. See Appellants’ Br. 30-31. (arguing in passing that the district court in this case erred by failing to give any weight to the general presumption that government officials carry out their duties properly but never urging adoption of a categorical, burden-shifting presumption of regularity); Appellants’ Reply Br. 22-24 (same). One need imply neither bad faith nor lack of incentive nor ineptitude on the part of government officers to conclude that [redactions] compiled in the field by [redactions] in a [redactions] near an [redactions] that contain multiple layers of hearsay, depend on translators of unknown quality, and include cautionary disclaimers that [redactions] are prone to significant errors; or, at a minimum, that such reports are insufficiently regular, reliable, transparent, or accessible to warrant an automatic presumption of regularity.
It is thus not at all surprising that our court has never before applied the presumption of regularity in Guantanamo Bay habeas cases despite numerous opportunities to do so. For instance, in Barhoumi, the government, seeking to establish that the petitioner was “part of’ an al Qaida associated militia, relied on an intelligence *267report that included an English translation of a diary allegedly authored by a member of that militia. Barhoumi v. Obama, 609 F.3d 416, 420 (D.C.Cir.2010). Among other challenges to this evidence, we considered petitioner’s argument that the government’s failure to make a copy of the diary available in its original Arabic or to provide information regarding the qualifications or motives of the translator raised doubts about reliability. Although we characterized this objection as “troubling” and “aecept[ed] that the additional layer of hearsay added by the diary’s translation render[ed] it somewhat less reliable than it otherwise would [have] be[en] (particularly if the government had provided information regarding its translation),” we nonetheless reviewed the diary’s internal and external indicia of reliability and concluded that the district court had not clearly erred by relying on it. Id. at 430-32. Had we believed that a presumption of regularity applied to the translation recorded in the intelligence report, none of that extended analysis would have been necessary. Instead, we would have simply presumed the document’s accuracy—and expected the district court to do the same. As my colleagues begrudgingly admit, Maj. Op. at 1183-84, that is exactly what the government asked us to do in Barhoumi, but to no avail. See Appellees’ Br. 52, Barhoumi, 609 F.3d 416 (No. 09-5383) (arguing that “translations are presumed to be accurate in the absence of evidence to the contrary” (emphasis added)).
We followed exactly the same playbook in Bensayah and Al Alwi, two cases in which we reviewed district court reliability determinations about the precise type of government intelligence document at issue here [redactions] Bensayah v. Obama, 610 F.3d 718, 725 (D.C.Cir.2010). In Bensayah, rather than granting the government’s evidence a presumption of regularity on the grounds that it consisted of government records regularly kept, we carefully evaluated other evidence purporting to corroborate the document’s contents, ultimately concluding that the district court committed clear error by finding that document reliable. See id. at 726-27. Nor did we apply a presumption of regularity in Al Alwi even though the government’s evidence, as here, consisted of interrogation summaries allegedly reporting the petitioners’ own statements and even though those documents had greater indicia of reliability than the Report at issue in this case. Indeed, in Al Alwi we adopted a rule—that “the [district] court must take the absence of corroboration into account in assessing the reliability of petitioner’s out-of-court statements,” Al Aim, 653 F.3d at 19 (emphasis added)—that directly conflicts with this court’s observation that “[b]y definition, a presumptively reliable record needs no additional corroboration unless the presumption is rebutted.” Maj. Op. at 1192.
And most recently, in Khan v. Obama, we reviewed the district court’s finding that the government’s informant reports were reliable. Again, rather than applying a presumption of regularity, we spent page after page carefully evaluating the reliability of the reports. In affirming the district court’s determination that the documents were reliable, we emphasized external indicia of reliability, such as photographs and items seized from petitioner’s home, as well as detailed government declarations explaining why the reports were reliable. Khan, 655 F.3d at 27-32.
Our approach in Barhoumi, Al Alwi, Bensayah, and Khan reflects a careful and conscious balancing of the important interests at stake. While federal courts typically exclude hearsay unless it falls within a specific exception, see Fed.R.Evid. 803, we understand that in the context of ene*268my combatant proceedings such evidence may be the best available. Barhoumi, 609 F.3d at 427. Thus, rather than acting on our deep, historically rooted skepticism of hearsay by excluding such evidence altogether, we admit it but are careful to assign it no more weight than it is worth as measured by any available indicia of reliability. See id. (holding that hearsay evidence is “always admissible” in such proceedings, but that it “must be accorded weight only in proportion to its reliability”); see also Al-Bihani v. Obama, 590 F.3d 866, 879 (D.C.Cir.2010). The presumption of regularity, which this court expressly premises on “defer[ence] to Executive branch expertise,” Maj. Op. at 1182, disturbs this careful balance, substituting a presumption in place of careful district court “review and assess[ment of] all evidence from both sides.” Al-Bihani, 590 F.3d at 880. Given the degree to which our evidentiary procedures already accommodate the government’s compelling national security interests by admitting all of its evidence, including hearsay; given the heightened risk of error and unlawful detention introduced by requiring petitioners to prove the inaccuracy of heavily redacted government documents; and given the importance of preserving “the independent power” of the habeas court “to assess the actions of the Executive” and carefully weigh its evidence, id., I find this court’s departure from our practice deeply misguided.
To be clear, I make no claim that anything in Barhoumi Bensayah, Al Alwi, Khan, or any of our other Guantanamo habeas cases affirmatively rules out the possibility of applying a rebuttable presumption of accuracy to certain kinds of government evidence in some circumstances. My point is only that our cases, proceeding in the very common-law-like fashion that my colleagues describe, see Maj. Op. at 1185, have endorsed and applied a careful and fine-grained approach to the assessment of reliability. We have applied that approach to claims that a document was mistranslated (Barhoumi) and to claims that a document is insufficiently corroborated (Al Alwi, Khan)— two of the issues in this case. We have applied that approach to a (Bensayah, Al Alwi), and to government interrogation summaries (Al Alwi)—the same type and category of documents as the Report. Following that approach, we have both upheld (Barhoumi, Al Alwi, Khan) and overturned (Bensayah) district court findings that a government document is reliable. The only feature of this case not previously encountered is that here the government lost: the district court found the dispositive government Report unreliable and granted a writ of habeas corpus.
Moreover, the presumption discards the unanimous, hard-earned wisdom of our district judges, who have applied their fact-finding expertise to a wide array of government hearsay evidence. In doing so, they have developed a uniquely valuable perspective that we ought not so quickly discard. These judges, including the district judge in this case, have unanimously rejected motions to give government evidence a presumption of accuracy. See, e.g., Alsabri, 764 F.Supp.2d at 66 (noting “ample reason” to decline to presume the accuracy of the government’s exhibits and explaining that circuit precedent supported its approach); Al Kandari, 744 F.Supp.2d at 19 (“Simply assuming the Government’s evidence is accurate and authentic does not aid [the reliability] inquiry.”); Ahmed, 613 F.Supp.2d at 55 (“[T]here is absolutely no reason for this Court to presume that the facts contained in the Government’s exhibits are accurate.”); see also Benjamin Wittes, Robert M. Chesney & Larkin Reynolds, The Emerging Law of Detention 2.0, at 52 (May 12, 2011) (indicating *269that “none of the publicly available rulings on the issue have favored the government”), available at http://www.brookings. edu/papers/2011/05_guantanamo_wittes. aspx. Rather than ignoring serious doubts about government evidence by presuming its accuracy, our district courts have instead done exactly what we expect of careful factfinders and precisely what our case law demands: scrupulously assess the reliability of each piece of evidence by applying “a long, non exclusive list of factors ... such as: consistency or inconsistency with other evidence, conditions under which the exhibit and statements contained in it were obtained, accuracy of translation and transcription, personal knowledge of [the] declarant ..., levels of hearsay, recantations, etc.” Ahmed, 613 F.Supp.2d at 55; see also Sulayman v. Obama, 729 F.Supp.2d 26, 42 (D.D.C.2010) (“As to many of the intelligence reports [the government] relies upon ... there is nothing in the record regarding the qualifications of the interpreters used in those interrogations to render a reliable interpretation. There are other intelligence reports ... in which the government has failed to provide foundational evidence that those statements ‘were made under circumstances that render them intrinsically reliable or were made by reliable sources.’ ” (citation omitted)).
Brushing aside these district court rulings, my colleagues think that those courts “may” have been denying a presumption of accuracy because they “[c]onfus[ed]” it for a presumption of truth, Maj. Op. at 1180, the difference being that the latter presumes the content of a report is true, whereas the former presumes that the government official filling out the report did so accurately—i.e., that in doing the interview, he correctly heard, translated, recorded, and summarized the content embodied in the report. The district courts have suffered from no such confusion, nor do I, for the core question presented in this case is whether the Report accurately reflects Latif s words. Unsurprisingly, my colleagues cite not a single case where a district court refers to a presumption of truth or, for that matter, a single instance in which the government argued for a presumption of truth rather than a presumption of accuracy. They cite Ahmed, but nowhere did the district court there say that “the requested presumption would go to the truth of ‘the facts contained in the Government’s exhibits.’ ” Maj. Op. at 1180 (citing Ahmed, 613 F.Supp.2d at 55). Rather, the district court denied a presumption of accuracy, doing so for several reasons, including the need to assess the “accuracy of translation and transcription,” and not just because of alleged torture, as this court now implies. 613 F.Supp.2d at 55; see also Al Mutairi v. United States, 644 F.Supp.2d 78, 84 (D.D.C.2009) (expressing concern that the government’s evidence “is based on reports of interrogations (often conducted through a translator) where translation or transcription mistakes may occur”). In Al Mutairi, the district court even pointed to evidence in that very case exemplifying such problems: “for over three years” the government had, “based on a typographical error in an interrogation report,” erroneously insisted “that Al Mutairi manned an anti-aircraft weapon in Afghanistan.” Id.; see also Al Rabiah v. United States, 658 F.Supp.2d 11, 18 (D.D.C.2009) (noting “discrepancies]” between two reports summarizing the same interrogation that the government had made no attempt to reconcile); Al Odah v. United States, 648 F.Supp.2d 1, 6 (D.D.C.2009) (noting that “interrogators and/or interpreters included incorrect dates in three separate reports that were submitted into evidence based on misunderstandings between the Gregorian and the Hyri calendars”). Indeed, *270the same district court whose decision we now review explained in another Guantanamo case that it “has learned from its experience with these cases that the interrogation summaries and intelligence reports on which [the Government] rel[ies] are not necessarily accurate and, perhaps more importantly, that any inaccuracies are usually impossible to detect.” Odah v. Obama, No. 06-cv-1668, slip op. at 3 (D.D.C. May 6, 2010); see also id. (“[T]here are many steps in the process of creating these documents in which error might be introduced[:] ... the interpreter must understand the question posed and correctly translate it; the interviewee must understand the interpreter’s recitation of the question; the interpreter must understand the interviewee’s response and correctly interpret it; the interrogator must understand the interpreter’s translation of the response; the interrogator must take accurate notes of what is said; and the interrogator must accurately summarize those notes when writing the interrogation summary at a later time.”). Of course, concerns about the accuracy of the reports necessarily raise concerns about their truth. But there are no grounds for assuming the district courts are confused about this distinction.
In support of a presumption of regularity, this court relies on the plurality opinion in Hamdi, which, applying Due Process analysis, states that “the Constitution would not be offended by a presumption in favor of the Government’s evidence” in enemy combatant proceedings for citizen detainees “so long as that presumption remained a rebuttable one and fair opportunity for rebuttal were provided.” Hamdi v. Rumsfeld, 542 U.S. 507, 534, 124 S.Ct. 2633, 159 L.Ed.2d 578 (2004) (plurality opinion). According to this court, because the Hamdi plurality provisionally blessed such a general presumption, its own presumption requiring deference to official government documents must pass constitutional muster. Maj. Op. at 1179. But the Hamdi plurality made clear that the presumption it sanctioned would apply only if the government “puts forth credible evidence that the habeas petitioner meets the enemy-combatant criteria.” 542 U.S. at 534, 124 S.Ct. 2633 (emphasis added); see also Almerfedi, 654 F.3d at 6 & n. 7 (explaining the Hamdi framework requires the government to “put forth credible facts” tending to show that the petitioner meets the detention standard, such as that he received military training at an al Qaida camp, which the petitioner can then rebut with his own facts and explanation). In other words, a presumption is acceptable if the government can first show that its evidence is credible, but the Hamdi plurality never suggested that the government could make that showing by relying on a presumption that government-produced evidence is credible and accurate. It is the latter presumption that is at issue here and about which the Hamdi plurality had nothing to say. Given that the district court in this case concluded that the Report was “not sufficiently reliable,” Latif slip op. at 25—i.e., that it was not credible—the court’s reliance on the Hamdi plurality to defend its presumption of regularity is misplaced.
This court believes that our decisions in Alr-Bihani, 590 F.3d 866, and Parhat v. Gates, 532 F.3d 834 (D.C.Cir.2008) support the “continuing viability” of applying a presumption of regularity to Guantanamo habeas cases. Maj. Op. at 1183. In AlBihani, however, although the district court “reserved [the] authority” granted by its case management order to presume the government’s evidence accurate, it went on to “assess[] the hearsay evidence’s reliability as required by the Supreme Court.” Al-Bihani, 590 F.3d at 880. Even the government agrees with *271this view of Al-Bihani See Appellees’ Br. 52, Barhoumi, 609 F.3d 416 (No. 09-5383) (“In this case, as in Bihani [Al-Bihani], the district court did not presume the accuracy or authenticity of the government’s evidence.” (emphasis added)). The most one can say about Al-Bihani on this issue is that we suggested—in dicta—that a district court could apply a presumption to a particular piece of evidence if appropriate—a power the district court in that case declined to exercise. This is a far cry from the holding today—that all such reports and their underlying hearsay must be granted a presumption of regularity. As to Parhat, a pre-Boumediene case arising under the Detainee Treatment Act of 2005, it is true that the Act incorporated a “rebuttable presumption that the Government Evidence is genuine and accurate.” Maj. Op. at 1183 (emphasis removed). But in that case, we took the opportunity to clarify that, at a minimum, hearsay evidence “must be presented in a form, or with sufficient additional information, that permits [an assessment of] its reliability.” Parhat, 532 F.3d at 849. As we recently reiterated, “[t]he government’s evidence in Parhat was insufficient to enable the court to assess its reliability.” Khan, 655 F.3d at 26. This hardly supports the proposition that courts must assume government reports like the one at issue here are accurate, especially given that the Supreme Court in Boumediene specifically found that the process provided by the Detainee Treatment Act was an inadequate substitute for the writ of habeas corpus. See 553 U.S. at 792, 128 S.Ct. 2229.
In sum, given how and where we typically apply the presumption of regularity, and given the balance this circuit has already struck on how to deal with hearsay evidence in Guantanamo Bay cases, and given the seasoned observations of our district courts about the reliability of such evidence, the question still unanswered to my satisfaction is “Why?” Wby does this court now require district courts to categorically presume that a government report—again, one created in a [redactions] near an [redactions] with multiple layers of hearsay, and drafted by unidentified translators and scriveners of unknown quality—is accurate? Whether the presumption can be overcome by a preponderance of the evidence or by clear and specific evidence— this court never says which—I fear that in practice it “comes perilously close to suggesting that whatever the government says must be treated as true,” see Parhat, 532 F.3d at 849. In that world, it is hard to see what is left of the Supreme Court’s command in Boumediene that habeas review be “meaningful.” 553 U.S. at 783, 128 S.Ct. 2229.
But the court’s assault on Boumediene does not end with its presumption of regularity. Not content with moving the goal posts, the court calls the game in the government’s favor. Instead of remanding to give Latif an opportunity to rebut the presumption of regularity, this appellate court engages in an essentially de novo review of the factual record, providing its own interpretations, its own narratives, even its own arguments, see Maj. Op. at 1185-99, and finds that “neither internal flaws nor external record evidence rebuts that presumption in this case,” id. at 1179. But see Pullman-Standard v. Swint, 456 U.S. 273, 292, 102 S.Ct. 1781, 72 L.Ed.2d 66 (1982) (where district court fact “findings are infirm because of an erroneous view of the law, a remand is the proper course”). To be sure, such a finding would be appropriate if the record supported “only one resolution of the factual issue.” 456 U.S. at 292, 102 S.Ct. 1781. But that cannot be the case where, as here, the question of reliability turns entirely on witness credibility, inferences drawn from errors and inconsistencies in the Report, and *272the resolution of conflicts in other record evidence, see infra Part II. Given the court’s conclusion that the presumption has not been rebutted, remand may well be a “pointless exercise.” Con. Op. at 1200.
II.
Rather than adopting a presumption of regularity, I would apply clear error review to the district court’s findings of fact just as we have consistently done throughout our Guantanamo cases. See, e.g., Almerfedi, 654 F.3d at 4-5 (reviewing district court fact findings for clear error); Al-Madhwani v. Obama, 642 F.3d 1071, 1074-75 (D.C.Cir.2011) (same); Salahi v. Obama, 625 F.3d 745, 750 (D.C.Cir.2010) (same); Al Odah v. United States, 611 F.3d 8, 14-15 (D.C.Cir.2010) (same); Bensayah, 610 F.3d at 723 (same); Barhoumi, 609 F.3d at 423-24 (same); Awad v. Obama, 608 F.3d 1, 7 (D.C.Cir.2010) (So long as “the district court’s account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it” and, critical to this case, “[wjhere there are two permissible views of the evidence, the factfinder’s choice between them cannot be clearly erroneous.” (citations omitted)). Under that standard, I would conclude that the district court committed no clear error by finding that the Report was insufficiently reliable; that it committed no clear error by crediting Latifs account of what happened only insofar as it needed to; and that it adequately addressed the other record evidence.
A
The starting point, of course, is the Report itself. See Awad, 608 F.3d at 6-7 (holding that the same clear error standard applies to fact findings based on documentary evidence and inferences drawn from that evidence). The district court’s primary concern about the Report related to the circumstances under which it was produced, circumstances that, according to the district court, increased the likelihood that mistakes had been made. In particular, the court emphasized the [redactions] summarized in the Report and another report produced around the same time. Latif, slip op. at [redactions] In addition, [redactions] The Report’s heavy redactions—portions of only [redactions] out of [redactions] pages are unredacted—make evaluating its reliability more difficult. The unredacted portions nowhere reveal whether the same person [redactions] or whether someone else performed each of these tasks. And because all the other [redactions] in the Report are redacted, the district court was unable to evaluate the accuracy of [redactions] by inquiring into the accuracy of the Report’s [redactions] In view of all these concerns, the district court found it especially troubling that neither the Report nor any of the Government’s other evidence “provider[s] any information with which to confirm [redactions] used the high level of care necessary to ensure [redactions] was accurate.” Id. at 26.
“[F]actual errors” in the Report reinforced the district court’s concerns. Id. Specifically, although the Report states “that Latif said he had been to Jordan to accompany a friend who needed medical care for his hand[,] ... Latif has repeatedly told interrogators, and has provided evidence to show, that he went to Jordan for treatment of an injury to his own, not a friend’s, head, not hand.” Id. at 15. In addition, the Report erroneously states that “Latif is unmarried and has no children,” even though “a declaration Latif submitted for use in this litigation states that he is married and has a son.” Id. Lastly, in what even my colleagues con*273cede is an “obvious mistake.” Maj. Op. at [redactions] the Report’s first page states that [redactions]
Also troubling the district court was the lack of “corroborating evidence for any of the incriminating statements in the [Report].” Latif, slip op. at 26. As the district court explained: “No other detainee saw Latif at a training camp or in battle. No other detainee told interrogators that he fled from Afghanistan to Pakistan, from Tora Bora or any other location, with Latif. No other type of evidence links Latif to A1 Qaeda, the Taliban, a guest house, or a training camp.” Id.
The district court properly weighed the cumulative effect of these subsidiary findings. See Al-Adahi, 613 F.3d at 1105-06. According to the district court, those findings “support[] an inference that poor translation, sloppy notetaking, [redactions] or some combination of those factors resulted in an incorrect summary of Latif s words.” Latif, slip op. at 26.
All of the concerns just described are obviously relevant to evaluating the Report’s accuracy. It goes without saying that the circumstances under which the Report was produced and the evidence, or lack of evidence, of care taken to avoid mistakes when the Report was produced shed light on that question. Likewise, it is undoubtedly probative of the Report’s reliability that it contains factual errors, for the presence of a known error increases the likelihood that other information in the Report is inaccurate as well. And of course, it is also relevant that the government has offered no independent corroboration for any of the Report’s incriminating facts. After all, skepticism about the trustworthiness of uncorroborated confessions has deep, historical roots, so much so that a criminal defendant “may not be convicted on his own uncorroborated confession.” Smith v. United States, 348 U.S. 147, 152-53, 75 S.Ct. 194, 99 L.Ed. 192 (1954) (noting that the rule’s “foundation lies in a long history of judicial experience with confessions”). And we recently made clear that in these Guantanamo habeas cases “the [district] court must take [such an] absence of corroboration into account in assessing the reliability of the petitioner’s out-of-court statements.” Al Alwi, 653 F.3d at 19 (emphasis added).
Moreover, none of the subsidiary fact findings the district court made about the Report itself were clearly erroneous. As this court acknowledges, “the [district] court cited problems with the Report itself, including its substantial redactions, [redactions] its reference Latif s ‘hand’ instead of his head injury, its internally inconsistent statements [redactions] and the perceived lack of corroboration.” Maj. Op. at 1191. And this court agrees that “[t]he inconsistencies in the Report may suggest a document produced on the field by imperfect translators or transcribers.” Id. at 1188.
Nonetheless, this court insists, “[i]t is almost inconceivable,” id. at 1187, that the inculpatory information in the Report could have resulted from “poor translation, sloppy notetaking [redactions] or some combination of those factors,” Latif slip op. at 26. According to the court, there is too high a “level of inculpatory detail” in the Report for it to have resulted from such mistakes. Maj. Op. at 1187. And because the incriminating statements “are intertwined with other details in the Report that persist in Latif s current account of his travels,” the court believes, [redactions]
My colleagues’ interpretation of the evidence is undoubtedly plausible. Yet when one accounts for all of the Report’s various problems, the fact that admittedly true facts “are intertwined” with contested inculpatory ones also supports another plau*274sible explanation, akin to what happens in the children’s game of telephone. In that game, one child whispers a phrase to another, who in turn whispers it to a third, and so on, until the last child announces what he or she has heard. As anyone who has played well knows, the whole point of the game is that what the final child hears is both recognizably similar to the original statement and yet amusingly transformed. Cf. Carol D. Leonnig & Josh White, An Ex-Member Calls Detainee Panels Unfair, Wash. Post, June 23, 2007 (reporting former-Combatant Status Review Tribunal member, Lieutenant Colonel Stephen Abraham, as “equating] the government hearsay presented [to the CSRTs] about detainees with a game of telephone” (internal quotation marks omitted)).
Now imagine the game when, as may have happened here, [redactions] and the Report produced “[i]n the field,” with the assistance of “imperfect translators or transcribers,” Maj. Op. at 1188. And imagine further, as may also have happened here, that the “imperfect” translator may have misheard Latif s exact words, or the interrogator may have misheard the “imperfect” translation, or the “imperfect” notetaker may have failed to transcribe precisely what was “imperfect[ly]” translated, or that whoever wrote up a summary based on those notes, [redactions] may have failed to understand what exactly that “imperfect” notetaker had written, or that some combination of all those things may have occurred. This problem is all the more exacerbated when—again as may have happened here—the same [redactions] and fails to [redactions] until [redactions] As Latif s [redactions]—as so often occurs in the game of telephone And so, Latif s statement about a charity-worker named Ibrahim Al-Alawi from Ibb encouraging him to travel from Yemen may have become a reference to a jihadi recruiter; his statement about traveling to an Islamic Center in Kabul run by an imam named Abdul Fadel may have transformed into one about serving north of Kabul under an Afghan commander with the homophonous name Abu Fazl; and his statement about three teachers named Abu Bakr of the Arab Emirates, Awba of Kuwait, and Hafs of Saudi Arabia, may have turned into one about fellow Taliban soldiers with the similar sounding (or identical) names Abu Bakr, Abu Hudayfa, and Abu Hafs—just as his statement about a trip to Jordan for treatment of an injury to Latifs head became a statement about treatment for his friend’s hand. Indeed, my colleagues nowhere disagree that all of the names and statements appearing in the Report sound similar to names and statements Latif later made. And, although only [redactions] that appear to have beer [redactions] are in the record and unredacted, both contain details that also [redactions] appear in [redactions], details which an [redactions] might have [redactions] if they were having trouble understanding [redactions] say [redactions] Obviously, moreover, we have no way of knowing whether the redacted [redactions] likewise contain [redactions] Given that the circumstances under which the Report was produced increased the probability of mistakes, given that the Report contains other “factual errors,” and given that the government has failed to corroborate any of the Report’s incriminating information, Latif, slip op. at 26, this explanation is at least plausible—the only question for us when reviewing fact findings, such as these, for clear error. See Awad, 608 F.3d at 7 (reiterating that “[i]f the district court’s account of the evidence is plausible in light of the record in its entirety, the court of appeals may not reverse it” (quotation omitted)), But see Maj. Op. at 1192 (conceding this explanation is “possible,” yet incorrectly asserting that “the relevant *275question is whether th[e] hypothesis is likely”).
B
The district court did not stop with the Report. It also “eonsider[ed] the explanation of events Latif has offered”—again in service of the critical question of whether the Report was “sufficiently reliable.” Latif, slip op. at 27. According to Latif, with the help of a charitable worker, he left Yemen in 2001 seeking free medical treatment for the lingering effects of a serious head injury suffered in a 1994 car accident. Although the government challenges Latifs claim that he left Yemen in 2001 seeking medical treatment, it never disputes that “in 1994, [Latif] sustained head injuries as the result of a car accident and [that] the Yemeni government paid for him to receive treatment” in Jordan at that time. Id. at 5.
Besides his own narrative, Latif also offered documentary evidence to corroborate his account. Three documents are particularly noteworthy. The first, “a letter, dated August 21, 1994, from a doctor at the Islamic Hospital in Amman, Jordan,” confirms “that Latif ‘was admitted’ on July 9, 1994 ‘following a head injury.’ ” Id. at 23 (quoting letter). The second, “a letter dated August 18, 1999 from Yemen’s Ministry of Public Health,” states “that ‘[w]e recommend that [Latif] return to the previous center outside for more tests and therapeutic and surgical procedures at his own expense.’ ” Id. (alterations in original) (quoting letter, which also states that Latif “is hard of hearing” and that “a wide circular hol[e] was detected in [Latifs] left eardrum”). And the third—the most important—is Latifs intake form dated December 31, 2001 (i.e., shortly after he was seized [redactions] Filled out when Latif was taken into United States custody, the intake form states that Latif was in possession of “medical papers” when seized traveling from Afghanistan to Pakistan. Id. at 23 & n. 12.
This documentary evidence, the district court found, “corroborated]” Latifs “plausible” story. Id. at 26-27. The district court also rejected the government’s contention that Latifs exculpatory account was a “cover story” and found the government’s “attack[s]” on the “credibility of [the] story ... unconvincing.” Id. at 26. This too was an obviously relevant evidentiary consideration. A petitioner’s version of events, should he choose to provide one, can be relevant when assessing the government’s evidence. After all, the more believable the petitioner’s exculpatory account, the greater the reason to doubt the government’s inculpatory one. Cf., e.g., Al-Adahi, 613 F.3d at 1107 (weighing petitioner’s “false exculpatory statements” in the government’s favor). Having thus assessed Latifs story positively, and given that the story contradicts incriminating information contained in the Report, the district court relied on the story to support its finding that the Report is “not suffi-ciently reliable.” Latif, slip op. at 25.
Although agreeing that Latifs story is relevant, my colleagues nonetheless conclude that by describing it as “plausible” and “not incredible,” the district court never actually credited that account. But “reading the district court’s explanation in [such] a parsed manner that overlooks its meaning in context” is inconsistent with clear error review. United States v. Brockenborrugh, 575 F.3d 726, 741 (D.C.Cir.2009). Here is what the district court actually said about Latifs story:
The Court makes this ruling [i.e., about the accuracy of the Report] having taken into consideration the explanation of events Latif has offered. Latifs story is not without inconsistencies and unanswered questions, but it is supported by *276corroborating evidence provided by medical professionals and it is not incredible. [The district court then rejected the government’s theory that Latif had told inconsistent stories over the course of his detention and was therefore telling a “cover story.” The district court reasoned that the government’s theory was based on just “two isolated statements,” one of which “does not contradict Latifs version of events.” Finally, the district court found the government’s] other arguments attacking the credibility of Latifs story ... similarly unconvincing. The smaller inconsistencies to which [the government] ha[s] pointed may be no more than misstatements or mistranslations; even if some details of Latifs story have changed over time, for whatever reason, its fundamentals have remained the same.
Latif, slip op. at 27-28. What else could the district court have meant other than that it found Latifs account convincing enough, plausible enough, consistent enough, and corroborated enough to give it at least some weight against the government’s evidence? And as we have held, “[m]erely because a particular piece of evidence is insufficient, standing alone, to prove a particular point does not mean that the evidence ‘may be tossed aside and the next [piece of evidence] may be evaluated as if the first did not exist.’ ” Salahi, 625 F.3d at 753 (alteration in original) (quoting Al-Adahi, 613 F.3d at 1105). After all, it is the government that bears the burden to demonstrate the lawfulness of detention, and here the district court concluded that the government had failed to meet that burden because (1) “there is a serious question as to whether the [Report] accurately reflects Latifs words” given (la) the circumstances under which it was produced and (lb) the “factual errors” it contains; (2) “the incriminating facts in the [Report] are not corroborated^] and [ (3) ] Latif has presented a plausible alternative story to explain his travel.” See Latif, slip op. at 26. It is in just this circumstance—where doubts about the government’s evidence and confidence in the detainee’s story combine with other evidence to fatally undermine the government’s case—that a detainee may prevail even without the district court needing to credit the detainee’s story by a full preponderance of the evidence. To require otherwise would, in effect, inappropriately shift the burden of proof to Latif.
Given that the district court found Latifs story entitled to at least some weight and given that such a finding could properly guide its evaluation of the government’s evidence, the only remaining question for us is whether that finding was clearly erroneous. It was not. As this court itself acknowledges, Latifs story, on its own terms, is not “intrinsically] implausible].” Maj. Op. at 1193-94. And that observation is reinforced by corroborating evidence showing that Latif needed to leave Yemen for medical care in 1994, that Yemen’s Ministry of Public Health recommended he do so again in 1999, and that Latif had medical papers with him when seized crossing into Pakistan. That a trip abroad for medical care had been necessary, not once but twice, makes it more likely that Latif would have needed to travel abroad for medical care in 2001 as well. And the fact that Latifs condition was still serious enough to require such a trip in 1999, five years after he was first injured, increases the odds that the injury continued to be that serious two years later in 2001. Equally important, the most plausible reason for why Latif would have had medical papers in his possession when first seized is that his trip in fact had a medical purpose.
Attempting to cast doubt on the district court’s favorable assessment of Latifs ac*277count, this court insists that the district court “toss[ed] ... aside” inconsistencies in Latifs account. Id. at 1220; see also id. at 1195-96. But the district court did no such thing. It expressly recognized those inconsistencies, Latif, slip op. at 24-25 (summarizing the alleged inconsistencies); id. at 27 (“Latifs story is not without inconsistencies and unanswered questions.”), ultimately finding the government’s “attaek[ on] the credibility of Latifs story” based on those inconsistencies “unconvincing.” Latif, slip op. at 27. Particularly significant to the district court was the fact that the “fundamentals [of Latifs story] have remained the same.” Id. As Latif points out, those fundamentals—appearing in more than a dozen interrogation summaries and statements from December 31, 2001 [redactions] until May 10, 2009—“include [Latifs] adamant denials of any involvement with al Qaida or the Taliban; his serious head injury from a car accident in Yemen; his inability to pay for the necessary medical treatment; and his expectation and hope that Ibrahim Alawi would get him free medical care.” Appellee’s Br. 57. Indeed, at least some in the government apparently agree. The commanding officer of the Defense Department’s Criminal Investigative Task Force noted in a June 16, 2004 memo that Latifs statements to interrogators had “been relatively consistent.” Ex. 80, Memorandum from Criminal Investigative Task Force to General Counsel, Department of Defense (June 16, 2004). Moreover, before making too much of smaller inconsistencies it is important to remember that they appear not in verbatim transcripts prepared by a court reporter with the aid of an audio or video recording, but rather in brief summaries of translated interrogations. As mentioned above, it would be unsurprising to discover that minor errors crept in as Latifs account passed from his mouth to a translator (of unknown ability) to an interrogator to the interrogator’s notes and finally to the interrogator’s summary of those notes—the last of which represents the only evidence in the record of what Latif actually said in each of his interrogations. As we remarked in another Guantanamo Bay habeas case, “[t]he task of resolving discrepancies among the various accounts offered into evidence is quintessentially a matter ... for the district judge sitting as the fact-finder.” Al-Madhwani, 642 F.3d at 1076 (internal quotation marks omitted).
Rather than applying clear error review to the district court’s resolution of such discrepancies, this court suggests its own story—a story not found by the district court, never argued by the government, and based on its own review of the raw evidence—about how Ibrahim may have “promised Latif the medical treatment he needed to induce him to join the Taliban.” Maj. Op. at 1188. Exhibiting heretofore unknown expertise in al Qaida recruitment strategies, the court posits that “[s]uch a recruiting tactic (or cover story)” would make sense. Id. “Latifs medical records and his professed desire for medical treatment,” the court thus finds, “are therefore consistent with the Report, not inconsistent.” Id. at 1188. But see United States v. Microsoft, 253 F.3d 34, 117 (D.C.Cir.2001) (en banc) (“[District court factfindings receive either full deference under the clearly erroneous standard or they must be vacated. There is no de novo appellate review of factfindings and no intermediate between de novo and clear error, not even for findings the court of appeals may consider sub-par.”).
C
The government points to several additional pieces of evidence that, it believes, buttress its argument that the Report is *278reliable. The district court considered all of this evidence. Some items it found insufficient to outweigh its concerns about the Report and its positive assessment of Latifs story. Others it found failed to implicate Latif or prove the point the government hoped to make. As a reviewing court, our job is to determine only whether those assessments were clearly erroneous. They were not.
First, consider the circumstances leading up to Latifs seizure by Pakistani authorities—circumstances to which the district court gave less weight than the government would have liked. Latif left Kabul in November 2001 and then traveled through Jalalabad before eventually arriving at the Pakistani border where Pakistani authorities detained him. According to the government, this path mirrors that of Taliban soldiers retreating from Kabul. Although not contending that this evidence is dispositive, the government argues that because Latifs admitted route is consistent with that of Taliban soldiers and with information in the Report, it is a helpful piece in the puzzle, bolstering its claim that the Report’s inculpatory statements are accurate.
Fair enough, but how helpful? If this route is commonly used by innocent civilians, then the evidence is not that helpful at all. To understand why, consider a simple hypothetical. Suppose the government were to argue in a drug case that the defendant drove north from Miami along 1-95, “a known drug route.” Familiar with 1-95, we would surely respond that many thousands of non-drug traffickers take that route as well. Given what we know about our own society, the 1-95 inference would be too weak even to mention. Cf. Almerfedi, 654 F.3d at 6 n. 7 (noting that some conduct such as possessing an AK-47 is so “commonplace in Afghanistan [that it] does not meaningfully distinguish an al Qaeda associate from an innocent civilian”). On the other hand, if the alleged drug trafficker had driven along an infrequently traveled country road, then a contention that that road was “a known drug route” would carry more weight. The burden of proof is on the government to demonstrate whether travel on a particular route to the Pakistani border, when considered in context, is more like the lonely country road and thus worthy of consideration when it comes to distinguishing between enemy combatants and innocent civilians.
Based on nothing more' than a few anecdotes, this court suggests that Latifs route was akin to the country road. It asserts that the details of Latifs post-Kabul travels are “analogous” to those we found “strong[ly] suggestive]” of al Qaida membership in Uthman. Maj. Op. at 1197. But how analogous are they really? Uthman was captured “in the vicinity of Tora Bora” at a time when “most, if not all, of those in the vicinity of Tora Bora ... were combatants.” Uthman, 637 F.3d at 404. By contrast, the record in this case contains no evidence that Latif ever traveled through the Tora Bora mountains, and the city we know he did travel through—Jalalabad—has over 160,000 residents, most of whom were presumably not combatants, see Jalalabad, Britannica Academic Edition, http://www.Britannica.com/ EBcheeked/topic/299643/Jalalabad (last visited Sept. 20, 2011) (estimating population as of 2006 at 168,600). In Uthman, the detainee had not only taken a particularly suspicious route, but also was captured with a “small group” that included two “confessed ... bodyguards for Osama bin Laden” and another admitted Taliban fighter, all three of whom Uthman had studied with at the Furqan Institute, “a religious school at which other men were recruited to fight for Al Qaeda.” Uthman, *279637 F.3d at 404-05 (internal quotation marks omitted). One of the bodyguards “described the group as ‘brothers’ retreating from battle.’ ” Id. at 405. Here, Latif told interrogators that his Afghan guide was the only person who accompanied him to the Pakistani border, Ex. 25, Summary Interrogation Rep. (Mar. 6, 2002), and the only evidence to the contrary are [redactions] My colleagues accuse the district court of “failfing] even to consider” Latif s route. Id. at 1195. But the district court did consider it, expressly acknowledging that “Abu Khalud arranged travel for other detainees along the same route Latif reportedly took to Afghanistan.” Latif, slip op. at 10. Given that the government failed to demonstrate that route was towards the country road end of the 1-95-country road continuum—i.e., that the evidence was sufficiently probative—the district court committed no clear error by failing to “factorf ] [it] into [its] decision,” Maj. Op. at 1195.
Second, consider the government’s argument that “Latif was recruited by an al Qaeda member” in Yemen, a theory the district court found the government had failed to prove. Latif, slip op. at 25. To support its theory, the government pointed to evidence allegedly showing that Latif s charitable benefactor, Ibrahim Alawi, is actually an al Qaida facilitator known as Abu Khalud, whose real name is Ibrahim Ba’alawi. Some of this evidence could certainly have led a reasonable factfinder to accept the government’s interpretation, including that “Ba’alawi” and “Alawi” have similar spellings and that the route Latif took to Afghanistan at Ibrahim’s urging was the same path reportedly taken by other detainees who, unlike Latif, admit to having taken that trip to fight alongside the Taliban and some of whom have also admitted, again unlike Latif, to being Abu Khalud-recruits. That evidence, however, hardly forecloses the district court’s contrary finding that the government had failed to prove by a preponderance of the evidence that Ibrahim Alawi was Abu Khalud. To repeat, although we have treated evidence that a petitioner reached Afghanistan along a “route similar to the paths of admitted al Qaeda members now in U.S. custody” as a plus factor in determining whether that petitioner was “part of’ al Qaida, Uthman, 637 F.3d at 405, we have never suggested nor has the government shown that this particular path is so uniquely associated with al Qaida recruits that a district court clearly errs when it treats such evidence as more akin to traveling along 1-95 than a lonely country road.
The record contains ample additional evidence that supports the district court’s finding. Latif introduced expert declarations explaining that “Ba’alawi” and “Alawi” are distinct Arabic names and that both are common in Yemen. Latif, slip op. at 18-19. Notably, therefore, Latifs interrogation summaries all refer to some variation of the name Ibrahim Alawi but none include the “Ba,” and none mention Abu Khalud. By contrast, interrogation summaries for seven of the eight detainees mentioning the al Qaida facilitator named Abu Khalud refer either to “Abu Khalud” or “Ibrahim Ba’alawi” but never “Ibrahim Alawi,” id., and the eighth, who apparently used the name “Alawi,” is a detainee this very district court, in a different case, found not credible because his statements conflicted with those of several other detainees, id. at 19 n. 10 (citing Abdah v. Obama, 717 F.Supp.2d 21, 35 (D.D.C.2010)). But see Maj. Op. at 1193-95 (ignoring the district court’s adverse credibility finding about that detainee). Moreover, Latif described Ibrahim to interrogators as “skinny,” with a “big beard” and as “30-40 yrs. old,” as having two children—[redactions] a boy, and [re*280dactions] a girl, and as being from Ibb. Latif, slip op. at 19-20. By contrast, other detainees described Abu Khalud as short, fat, with a short beard and moustache, and around 27 years old, with a visible injury on his face caused by a bullet injury sustained in Bosnia, with one daughter named [redactions] and as being from Ta’iz, not Ibb. Id. But see Maj. Op. at 1198 (dismissing these differences because Latifs descriptions of Ibrahim Alawi appear in interrogation summaries produced after Latifs initial interview). In light of this mixed record it is self-evident that “there are two permissible views of the evidence,” meaning that “the factfinder’s choice between [those two views] cannot be clearly erroneous.” Awad, 608 F.3d at 7.
Next, consider record evidence that, according to this court, shows “that Latif stayed at al-Qaida guesthouses.” Maj. Op. at 1196. That evidence consists of [redactions] which the government claims include [redactions] The [redactions] contain [redactions] along with [redactions] Seeking to connect these [redactions] to [redactions] the government argues that the [redactions] the [redactions] and the [redactions] The government also points to an [redactions] Finally, the government contends that the fact that [redactions] buttresses its [redactions] given its theory that the [redactions] contain [redactions]
But the district court found that the [redactions] the government points to “is not [redactions] The district court based that finding on the differences between [redactions] and the [redactions] as well as the fact that [redactions] is not [redactions] The district court also noted Latifs innocent explanation for not having his passport—that he “gave it to Ibrahim to use in arranging his stay at a hospital.” Id.
Ample record evidence supports the district court’s factual analysis. At the most basic level as the district court noted, [redactions] Moreover, [redactions] explaining that even the apparent [redactions] between the [redactions] is misleading given that [redactions] Indeed, [redactions] implying [redactions] In addition, as [redactions]
This court nonetheless accuses the district court of overlooking the government’s expert evidence that [redactions] By that theory, Latif could be any [redactions] and maybe he goes [redactions] But the district court committed no clear error when, after considering the [redactions] it concluded that [redactions]
Finally, the district court’s reliance on Latifs explanation for not having his passport is plausible in light of other record evidence about the practice of at least one hospital, the Islamic Hospital in Jordan, of taking foreign patient’s passports “to guarantee that [those] patients will not leave the country before settling their bills.” Pet’r Trial Ex. No. 7. Moreover, although leaving behind one’s passport with an al Qaida operative at an al Qaida run guesthouse might suggest al Qaida affiliation, see Al Alwi, 653 F.3d at 17, such a scenario is several inferential steps removed from the only relevant fact we know about Latif—that he did not have his passport with him when seized. To be sure, a reasonable factfinder might have interpreted this evidence differently. Yet again, the record contains enough evidence to support “two permissible views of the evidence,” Awad, 608 F.3d at 7 (quotation omitted), meaning that “the factfinder’s choice between [those two views] cannot, [therefore,] be clearly erroneous.” Id.
D
The court groups many of its criticisms about the district court’s fact finding under the catch-all header of Al-Adahi. Accord*281ing to my colleagues, the district court took an “unduly atomized” approach to the evidence. Maj. Op. at 1176. The district court did no such thing.
Absent some affirmative indication to the contrary, we “presum[e] that the district court knew and applied the law correctly.” United, States v. Mouling, 557 F.3d 658, 668 (D.C.Cir.2009). Such affirmative evidence of legal error was quite obviously present in Alr-Adahi, as the “fundamental mistake” we identified in that district court’s opinion makes clear:
Al-Adahi’s ties to bin Laden “cannot prove” he was part of Al-Qaida and this evidence therefore “must not distract the Court.” The fact that Al-Adahi stayed at an al-Qaida guesthouse “is not in itself sufficient to justify detention.” Al-Adahi’s attendance at an al-Qaida training camp “is not sufficient to carry the Government’s burden of showing he was a part of’ al-Qaida.
Alr-Adahi, 613 F.3d at 1105 (emphasis added) (quoting district court opinion). By contrast, here the district court placed the Report, which the government concedes represents its “primary” piece of evidence, Appellants’ Br. 10, and on which the government admits its “case turned,” Appellants’ Br. 5, at the center of its analysis. The district court devoted two and a half pages to analyzing the Report and then another fifteen pages to summarizing other evidence introduced by the parties to prop it up or knock it down. Finally, the district court examined the cumulative effect of various evidentiary concerns on the Report’s reliability. When read in its full context, the district court’s opinion suffers from nothing like the flaws that we reviewed in Al-Adahi.
This court uses Al Adahi to turn the presumption of district court lawfulness on its head. Rather than giving the district court the benefit of the doubt, it seems to assume that the district court considered the evidence in isolation and ignored key facts. Take, for example, the contention that the district court tossed aside and considered in isolation alleged inconsistencies between statements attributed to Latif in different interrogation reports. Maj. Op. at 1220. This argument fails to recognize the leeway we have afforded district courts to resolve discrepancies among various accounts in other Guantanamo cases. In Al-Madhwani, we found no error in the district court’s decision to credit two different detainees’ interrogation summaries even though the detainees’ statements contradicted each other in certain respects, reasoning that the “task” of “resolving” such discrepancies “quintessentially” belonged to the district court. Al-Madhwani, 642 F.3d at 1076-77. Yet the only indication that the district court in that case had actually resolved the relevant contradictions between the two reports is its bald assertion that those reports are reliable; the discrepancies are never mentioned, let alone analyzed. By contrast, as this court concedes, the district court here expressly noted that it had “taken into consideration the explanation of events Latif has offered” in assessing the Report and expressly acknowledged that Latifs story is not -without “inconsistencies and unanswered questions.” Maj. Op. at 1195. The district court then specifically assessed the two primary inconsistencies the government relied on, as my colleagues implicitly acknowledge. Id. at 1195-96. Finally, the district court explained that any concern about “smaller inconsistencies,” most of which it had earlier summarized, was outweighed by the possibility that they had resulted from translation or transcription errors and by the fact that the “fundamentals [of Latifs story] have remained the same.” Latif, slip. op. at 27. For its part, this court reluctantly recognizes all this as “a welcome step toward *282the holistic approach to the evidence we called for in Al-Adahi.” Maj. Op. at 1195. But it is in fact more than that. If the district court’s implicit resolution of discrepancies in Al-Madhwani was adequate, then it follows a fortiori that so too was this district court’s far more explicit treatment. My colleagues acknowledge that their approach is in tension with “the usual practice” of “assum[ing] the [district] court considered all the evidence,” but nonetheless find this justified by the “unusual posture of this case”—i.e., a he-said, she-said case involving detainees at Guantanamo Bay. Id. at 1198. But if we take seriously the notion that district courts are better at finding facts and determining credibility, then we should be all the more eager to defer to their expertise when the stakes are high and when the case comes down to he-said, she-said—that is, when it rests entirely on credibility and how one interprets the facts.
The only affirmative indication this court identifies allegedly showing that the district court took an unduly atomized approach to the evidence relates to the circumstances of Latifs capture and to [redactions] The court makes much of the fact that in weighing the former, the district court employed language similar to the language used at one point by the district court in Alr-Adahi—specifically that “the timing of [Latifs] departure ... is not sufficient to create an inference that he was involved in fighting.” Latif, slip op. at 27 (emphasis added). The court, however, neglects to mention that this sentence appears in the middle of a paragraph evaluating the credibility of Latifs account, which itself appears in the middle of an extended assessment of the combined impact of multiple pieces of evidence on the Report’s reliability. This “pars[ing]” of the district court’s words “overlooks]” what those words “mean[ ] in context,” an approach that is, again, inconsistent with clear error review. See Brockenborrugh, 575 F.3d at 741.
As for the district court’s decision [redactions] my colleagues offer no convincing explanation for why the district court should have considered evidence that it found does not implicate Latif—unless, of course, that finding was clearly erroneous, something they never claim. Suppose, for example, that a witness in a burglary case testifies to having seen a man with a similar build as the defendant walk away from the site of the crime. If the factfinder concludes that the person the witness saw was not the defendant, then surely the factfinder can reasonably set aside the witnesses’ testimony in assessing whether the defendant was the burglar. So too here. Once the district court had determined that [redactions] did not implicate Latif, it was entirely proper for it to put them aside when evaluating the rest of the evidence.
The remainder of the court’s Al Adahi critique rests entirely on the claim that the district court “ignorefd] relevant evidence.” Maj. Op. at 1193. Not so. The district court expressly considered virtually all the evidence this court points to— including every single item of evidence the government claims is of primary or even secondary relevance. Compare id. at 1193-95 (“correspondence” between names appearing in the Report and names Latif later mentioned to interrogators), with Latif, slip op. at 15-16 (discussing same); compare Maj. Op. at 1194-95 (Latifs travel route from Yemen to Afghanistan), with Latif, slip op. at 10-11 (discussing same); compare Maj. Op. at 1195-96 (purported inconsistencies in Latifs statements), with Latif, slip op. at 27-28 (discussing same); compare Maj. Op. at [redactions] compare Maj. Op. at 1197-98 (circumstances of Latifs departure from Kabul and subsequent *283seizure by Pakistani authorities), tuith, Latif s slip op. at 12-13, 25, 27 (discussing same); compare Maj. Op. at 1198-99 (evidence that Latif s benefactor, Ibrahim AlAlawi, is in fact the A1 Qaida facilitator Abu Khalud), with Latif, slip op. at 17-21, 23-28 (discussing same). As for the claim that Latif may have (or may not have) traveled across the Pakistani border with Taliban-affiliated men, the district court’s silence is easily explained: [redactions] both of which [redactions] already chosen not to credit. But see Maj. Op. at 1196 (unreflectively treating this omission as an error distinct from the district court’s analysis of the Report).
To determine, as this court apparently does, that an experienced district court judge has totally ignored relevant evidence and so committed legal error because his twenty-seven page opinion omits mention of a handful of tertiary items plucked from thousands of pages of record evidence not only ignores the presumption of district court lawfulness, but also imposes on that court a virtually impossible burden. As the First Circuit put it, “[t]he district court could have written a 200-page decision on this case, but the far more compact assessment it made was entirely adequate under Rule 52(a).” Addamax Corp. v. Open Software Foundation, Inc., 152 F.3d 48, 55 (1st Cir.1998) (“[T]he district court was not required to make findings on every detail, was not required to discuss all of the evidence that supports each of the findings made, and was not required to respond individually to each evidentiary or factual contention made by the losing side.”). See also Schilling v. Schwitzer-Cummins Co., 142 F.2d 82, 84 (D.C.Cir.1944) (“While counsel may be disappointed that findings do not discuss propositions sincerely contended for, that, alone, does not make them inadequate or suggest that such propositions were not understood by the court.”); Medtronic, Inc. v. Daig Corp., 789 F.2d 903, 906 (Fed.Cir.1986) (“We presume that a factfinder reviews all the evidence presented unless he explicitly expresses otherwise.”); cf. Puerto Rico Maritime Shipping Authority v. Federal Maritime Comm’n, 678 F.2d 327, 351 (D.C.Cir.1982) (“It is frivolous to contend that the Commission did not consider the evidence because it did not catalogue every jot and tittle of testimony. Nothing is gained by a laundry-list recital of all evidence on the record supporting each view on every issue.”).
The district court’s opinion is by no means perfect. But clear error review demands a good deal less than perfection. See Microsoft, 253 F.3d at 118. That said, had the district court otherwise committed legal error or made some other mistake requiring remand, then I would have asked it to clarify whether it had indeed considered this evidence holistically. See, e.g., Salahi, 625 F.3d at 753 (noting that “the district court generally” considered all the evidence together but that “its consideration of certain pieces of evidence may have been unduly atomized” and that “since we [were] remanding” we would encourage the district court to clarify (emphasis added)). But nothing in our ease law requires, nor would I now hold, that the mere fact that a district court that obviously and carefully considered the entire record failed to mention a couple items of tertiary importance reflects undue atomization of the evidence.
III.
For the foregoing reasons, I would affirm the grant of the writ of habeas corpus.