ALONZO R. VAUGHN and CARL S. MORTON v. UNITED STATES

93 A.3d 1237, 2014 WL 2969004, 2014 D.C. App. LEXIS 191

• Court: District of Columbia Court of Appeals
• Decided: July 3, 2014
• Docket: 11-CF-228 & 11-CF-363
• Status: Published

Opinion

EASTERLY, Associate Judge:

Carl Morton and Alonzo Vaughn appeal their convictions for aggravated assault (D.C.Code § 22-404.01 (2012 Repl.)) and assault on a law enforcement officer (D.C.Code § 22-405(e) (2012 Repl.)). Both were charged in connection with an incident at the D.C. Jail in which a group of men attacked a fellow inmate, Deon Spencer, and a corrections officer who came to that inmate’s aid, Sergeant Charles White. The victims could not identify Mr. Morton and Mr. Vaughn; but the incident was recorded by multiple cameras, and, although these recordings were not exactly movie-quality, two corrections officers said they could identify Mr. Morton and Mr. Vaughn in the footage. The government presented the testimony of these corrections officers in conjunction with the recordings to the jury, and the jury convicted. But unbeknownst to the defendants, one of these identifying witnesses, introduced by the government as “Officer” Angelo Childs, had a significant credibility issue.

Six months earlier, Officer Childs had filed reports accusing a different inmate (“Inmate A”) of assault, thereby providing a potential justification for his use of a chemical agent on the inmate. His accusations were investigated by the Department of Corrections (DOC) Office of Internal Affairs (OIA). The DOC OIA determined in a “Final Report,” that, among other things, video footage of this incident did not show the alleged inmate assault.

The OIA Officer who wrote the OIA Final Report about the Inmate A incident stated in a sworn affidavit (submitted by the government during post-trial proceedings in this case) that he sent the OIA Final Report to the DOC Office of the Director, the entity in charge of disciplinary action, and he was later informed that the DOC Office of the Director demoted Officer Childs from Lieutenant to Sergeant after it received the report. In addition, this OIA Officer, who also assisted the government in the investigation of the Spencer-White attacks, stated in his affidavit that he “notified the U.S. Attorney’s Office for the District of Columbia of the Investigative Report concerning Lieutenant Childs and his subsequent demotion” on September 15, 2009, approximately two months before Mr. Vaughn and Mr. Morton’s trial.

The government did not disclose this favorable impeaching information to the defense. Instead, a week before trial, the government filed a motion in limine to preclude the defense from questioning Officer Childs about the misconduct detailed in the OIA Final Report. In that motion, the government provided a “summary” of the OIA Final Report that gave no indication that the OIA had investigated a potentially false allegation of an inmate assault by Officer Childs and others and determined that this allegation was false; the government also did not reveal Officer Childs’s resulting demotion. Rather, the government’s summary focused exclusively on only a portion of the OIA Final Report that considered whether Officer Childs (1) had properly used a chemical agent on the accused inmate, who the government (quoting the very portion of Officer Childs’s Incident Report the OIA had discredited) indicated had been acting aggressively, and (2) had falsely indicated in a report that the inmate was unrestrained. Portraying this investigation with skepticism, the government argued that it had little to do with Officer Childs’s credibility because the OIA had determined Officer Childs had only “suggested]” that the inmate was unrestrained.

The difference between the government’s summary of the OIA investigation and the actual OIA Final Report almost certainly would have come to light had the government provided the trial court with the full copy. It did not. Along with its summary, the government submitted to the trial court ex parte what it said was the OIA Final Report, but in fact was only the first five pages of the ten-page report (and included none of the documents in the appendix, 76 pages in all). The first five pages of the OIA Final Report contain “background” information, investigative notes, and a full reproduction of Officer Childs’s account of an inmate assault in his Incident Report without any indication that that account was being questioned; the findings adverse to Officer Childs begin on the sixth page.

Under Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), the government has a constitutionally mandated obligation to disclose to the defense, prior to trial, information in the government’s actual or constructive possession that is favorable and material. The government did not fulfill its due process disclosure obligations in this ease. Moreover, its failure to provide the court and the defense with complete and accurate information as to the contents of the OIA Final Report thwarted the trial court’s ability to “require strict compliance with the demands of Brady ... in the first instance.” Boyd v. United States, 908 A.2d 39, 62 (D.C.2006). As to Mr. Morton, we reverse his two convictions on this ground and remand for a new trial after the government has certified in writing that it has fulfilled its duty to learn of and disclose all the favorable information in the possession of the United States Attorney’s Office and the entire the prosecution team. As for Mr. Vaughn, we determine that a jury instruction conceded to be erroneous by the government requires reversal of his conviction for aggravated assault. We reject the remaining claims raised by Mr. Morton and Mr. Vaughn.

I. The Spencer-White Attacks, the DOC Investigation, and Mr. Morton and Mr. Vaughn’s Trial

Mr. Morton and Mr. Vaughn were prosecuted for their alleged involvement in an attack at the D.C. Jail on a corrections officer, Charles White, but it all began with an attack on a fellow inmate, Deon Spencer. 1 On December 27, 2007, Mr. Spencer was in the communal dining area of the Southwest 2 (“SW2”) housing unit when an inmate began hitting and kicking him. Corporal James Weathers, a corrections officer on duty in SW2, tried to assist Mr. Spencer, but he was outnumbered by the many inmates who rushed into the area and joined in the attack. Corporal Weathers radioed for help, and “that’s when everything just broke apart.” He described the scene as a “melee.”

Sergeant White, assigned to the Southeast 2 housing unit, responded to Corporal Weathers’s call for help. As he tried to quell the disturbance, he was attacked. Sergeant White was initially pushed onto a dining table, but he immediately got back up and moved away from the table. At some point, Sergeant White ended up on the floor. Several inmates then kicked him about the head and body and knocked him unconscious. As he lay in the walkway, inmates ran back and forth through the area, over and around his body.

In the meantime, a number of other corrections officers responded to the scene. 2 The incident ended when a chemical agent was dispersed, forcing inmates back into their cells. SW2 went on lock-down.

Mr. Morton and Mr. Vaughn were not initially identified as participants in the attack. The victims, Mr. Spencer and Sergeant White, were unable to identify their assailants from the 156 inmates housed on SW2 on the day of the attack. The officers who had seen and responded to the attack wrote Incident Reports immediately after the incident identifying a number of inmates they had seen participating in the violence — many of whom were charged and eventually pled guilty in connection with this incident. 3 Mr. Morton and Mr. Vaughn were not among these individuals.

OIA Investigator Benjamin Collins, who led the DOC investigation into this incident, sought to identify additional individuals involved in the attacks using recordings of the incident from fixed cameras capturing images of SW2 from different angles. The recordings were of limited utility on their own, however. In particular, the footage of the attack on Sergeant White is of extremely poor quality. Instead of a seamless “moving picture,” it is a choppy series of still images, capturing a chaotic sequence of events only at regular intervals. In addition, the images are highly pixelated and the faces are, as DOC staff conceded, “blurry.” 4 Accordingly, the DOC set out to find corrections officers who could identify individuals in the recorded footage.

The morning after the attacks, OIA Investigator Collins met with Officer Childs. 5 Officer Childs had not been on duty the day prior but he had already watched some footage of the incident with Officer Harper. Officer Childs told OIA Investigator Collins that “he was very familiar with the vast majority of the inmates in the unit and that he ... had named at least 11 inmates who he saw in the video that were involved in the altercation.” Apparently he then identified Mr. Morton as a participant in the attack on Sergeant White. Several months later OIA Investigator Collins brought Officer Childs to the U.S. Attorney’s Office, at which point Officer Childs identified Mr. Vaughn as another participant in this attack. OIA Investigator Collins also spoke to Sergeant Harper in the course of his investigation. Sergeant Harper ultimately identified Mr. Morton and Mr. Vaughn as participants in the attack on Sergeant White as well.

Beyond Officer Childs’s and Sergeant Harper’s identifications of participants from the video footage, the record does not reflect that the government developed any other evidence against Mr. Morton and Mr. Vaughn. Mr. Morton and Mr. Vaughn were each indicted in November 2008 for their alleged involvement in the attack on Sergeant White.

Mr. Morton and Mr. Vaughn went to trial a year later, in November 2009. The government argued that Mr. Vaughn was the only inmate dressed all in white and that he had given Sergeant White “one good shove.” The government argued that Mr. Morton was one of the inmates who kicked Sergeant White as he lay on the floor. In support of these theories, the government presented Officer Childs’s testimony. Explaining his ability to identify Mr. Morton and Mr. Vaughn as participants in the attack on Sergeant White, Officer Childs testified that he had been working on the unit for over a year, that Mr. Morton and Mr. Vaughn had each been on the unit for several months, that he knew them and their facial and physical features well, and that he could clearly see them in the video footage and in multiple stills taken from the video footage. Although defense counsel was able to highlight minor inconsistencies in Officer Childs’s testimony on cross-examination, Officer Childs’s testimony was largely un-impeached.

Sergeant Harper testified after Officer Childs. He could only identify Mr. Morton at one point in the recorded footage of the attack on Sergeant White and only recognized Mr. Morton in one of several video stills of that attack. 6 Sergeant Harper was impeached with the fact that, despite his claim at trial that he “kn[ew] on December 27 that all of [the individuals he identified at trial] were involved,” he had not named Mr. Morton or Mr. Vaughn in his initial Incident Report written December 27.

In closing, the government made clear that its case turned on the jury’s assessment of the credibility of the DOC officers who had testified and had identified the defendants. The government began by asking the jury “to consider all of the evidence”: “[t]he video, the documents!,] and importantly the testimony that you heard from the Corrections Officers.” The government informed the jury that its ultimate task was “to assess the credibility of those officers when they were on the stand. That’s what you’re asked to do here today.” Specifically addressing the identifications made by Officer Childs and Sergeant Harper, the government returned to the theme of credibility: “[P]art of your job in this case is to assess the credibility of these officers on the stand_ Angelo Childs told you repeatedly and strongly ... [t]hat is Alonzo Vaughn and that is Carl Morton. Jimmy Harper, same thing.” And the government concluded by “emphasizing] again, this case is ... about your credibility judgments.” It was a winning argument. Even if the jury was not convinced by Sergeant Harper’s somewhat inconsistent and weak testimony, it had little or no reason to doubt the testimony of Officer Childs.

II. The Brady Claim

Mr. Morton and Mr. Vaughn both argue that there were reasons to doubt Officer Childs’s credibility, but that the government did not timely or accurately apprise the defense of this information. We return to the facts on record.

A. The Pertinent Facts

1. The OIA Final Report

The following facts are taken from the OIA Final Report. In the spring of 2009 (as the parties prepared for trial in Mr. Morton and Mr. Vaughn’s case), Officer Childs participated in a mission to eliminate contraband at the D.C. Jail. Officer Childs, at that time a lieutenant, was a supervising officer on the Search and Recovery Team (SRT), along with Major Nora Talley and Lieutenant Gregory McKnight. Sergeant David Thomas and his drug-sniffing dog, “Reggie,” also assisted with the search for contraband.

In the course of this contraband recovery mission, Inmate A 7 and his cell were searched by members of the SRT. Specifically, Inmate A was strip searched in his cell and then restrained and taken by Lieutenant McKnight to the Body Orifice Security Scanner (BOSS Chair); 8 meanwhile the drug-sniffing dog inspected Inmate A’s cell. Both before and after his trip to the BOSS Chair, Inmate A told Lieutenant McKnight that he was scared of dogs and of the drug-sniffing dog in particular.

Once Inmate A returned from the BOSS Chair, Lieutenant McKnight left him, still in restraints, standing near the bubble in the presence of Officer Childs and Major Talley. At this point, Inmate A “stated emphatically that he would not go anywhere near the canine.” Although the protocol for the contraband recovery mission did not include using the drug-sniffing dog to inspect inmates, Officer Childs grabbed Inmate A’s arm and signaled to Officer Thomas to bring the dog over to Inmate A. As the dog approached, Inmate A tried to back away but Officer Childs “halted his movement.” Inmate A “stomped his foot in the direction of the canine”; Officer Childs “responded by placing his container of chemical agent near [Inmate A’s] face.” Sergeant Thomas then “positioned” the dog behind [Inmate A] and “placed his right hand against the lower calves and then on the lower back of [Inmate A].” The dog “followed with his muzzle where Sergeant Thomas placed his hand.” The dog “stood up on his hind legs and attempted to rest his paws on [Inmate A’s] back.” Inmate A recoiled from this contact, but “his movement was impeded when Lieutenant Childs, who maintained physical control of [Inmate A], sprayed a single burst of chemical agent directly into [Inmate A’s] face.”

That same day, Officer Childs prepared an Incident Report explaining the reason for his use of force and a Disciplinary Report accusing Inmate A of “Lack of Cooperation [and] assault without serious injury.” In both reports, Officer Childs asserted that Inmate A had “refuse[d] to be searched] by the [dog]” and had “started kicking at” the drug-sniffing dog and that Officer Childs had sprayed the chemical agent in Inmate A’s face to “seize [sic] his disruptive behavior.” Officer Childs then stated in the Incident Report that Inmate A was “placed in restraints” and escorted out of the unit, thereby effectively stating that Inmate A had previously been unrestrained. Officer Childs concluded his Incident Report with the assertion that “[t]his incident stemmed from the violent/disruptive behavior of [Inmate A].” Major Talley and Sergeant Thomas also filed reports in which they told a similar narrative of Inmate A’s unprovoked, aggressive behavior.

According to the OIA Final Report, video footage contradicted all of these reports. Four days after this incident, at the request of the Director of the DOC, the OIA opened a formal investigation. The lead investigator was Benjamin Collins, the same OIA investigator who had led the inquiry into the Spencer-White attacks and had brought Officer Childs to the U.S. Attorney’s Office so that he could serve as a witness. After OIA Investigator Collins completed his investigation, the OIA issued the Final Report on June 27, 2009.

The OIA Final Report is ten pages long and includes appendices totaling 76 pages. 9 The first five pages of the OIA Final Report contain “background” information, investigative notes, and a full reproduction of Officer Childs’s account, in his Incident Report, that Inmate A had tried to thwart a legitimate search for contraband and acted with unprovoked aggression. The findings adverse to Officer Childs and his colleagues do not begin until page six. At that point, the OIA memorialized its determination that Inmate A had not — as Officer Childs had claimed — engaged in “disruptive behavior” necessitating use of force. Specifically, the report states that “[v]ideo footage of the incident does not support the allegation that [Inmate A] assaulted any Correctional Officer or canine.” On subsequent pages, the report reviewed and then similarly discredited the stories told by Sergeant Thomas and Major Talley. With respect to Major Talley’s report, the OIA determined that “[vjideo footage ... does not depict any malicious aggressive behavior towards the canine or any staff member by [the] inmate ..., and therefore contradicts Major Talley’s written account.” With respect to Sergeant Thomas, the OIA likewise determined that notwithstanding his assertions that Inmate A “kick[ed] at” the dog, “video footage capturing this occurrence on the tier shows no contact or direct interaction between [Inmate A] and canine Reggie.”

The OIA Final Report further notes that, although Officer Childs’s “narrative suggests that at the time of the incident, [Inmate A] was not restrained,” “[u]pon review of the facts and circumstances of this incident, it is evident that [Inmate A] was in restraints and not a threat to ‘normal operations’ when he was sprayed with chemical agent by Lieutenant Childs.” In other words, the OIA determined that, to the extent the false story of an inmate assault was meant to justify the use of a chemical agent to subdue Inmate A, it failed not only because it was false, but also because it did not account for the fact that Inmate A was already restrained. 10

The OIA Final Report concludes with four summary “Findings” reiterating that Officer Childs “us[ed] ... a chemical agent on a restrained inmate who posed no immediate danger to himself or others” thereby violating the DOC “use of force continuum”; that he “submitted a false and or misleading Incident Report of the facts in stating that the inmate was placed in restraints after being sprayed with chemical agent” and that Major Talley and Sergeant Thomas had also filed “false and[/]or misleading” Incident Reports asserting that Inmate A had attempted to kick the drug sniffing dog and tried to “fight” Officer Childs. 11

The OIA merely conducts investigations; it does not recommend disciplinary action. Thus, according to the affidavit by OIA Investigator Collins later filed in this case by the government, see infra p. 1252, after .the OIA Final Report was completed, he sent it to the DOC’s Office of the Director. Again according to Investigator Collins’s affidavit, the Office of the Director of the DOC issued to Officer Childs “formal and written notification” on August 26, 2009, that he had been demoted from Lieutenant to Sergeant. It was Investigator Collins’s understanding that Officer Childs had signed this notice, 12 and that Officer Childs’s demotion became effective September 13, 2009. OIA Investigator Collins stated in his affidavit that “[o]n September 15, 2009, [he] notified the U.S. Attorney’s Office for the District of Columbia of the [OIA Final Report] concerning [then] Lieutenant Childs and his subsequent demotion.”

2. The Government’s Nondisclosure

The government did not turn over the OIA Final Report to the defense prior to the November 2009 trial, nor did it disclose to the defense that it had information Officer Childs had been demoted by DOC in connection with the incident with Inmate A. Instead, less than a week before trial began, the government, represented by the two Assistant United States Attorneys who prosecuted the case, filed (1) a motion in limine to prevent the defense from cross-examining Officer Childs on any issue related to the OIA Final Report, and (2) an accompanying ex parte motion requesting that (a) the court review in camera the first five pages of the ten-page report (“the five-page ex parte submission”), and (b) the court permit the government to file this attachment under seal. The government did not inform the court that the five-page ex parte submission was not the complete OIA Final Report or that it was missing its appendices. 13

The government’s stated goal in filing its motion in limine was “to preclude the defense from referring to the fact DOC Office of Internal Affairs may have made potentially adverse credibility findings” against Officer Childs. The government told the court and the defense that the OIA “investigation resulted in two findings related to Officer Childs: (1) Officer Childs’ use of force violated DOC policy and (2) Officer Childs submitted a false or misleading statement in reciting the facts,” specifically, that “DOC Internal Affairs found that Officer Childs’ statement that Inmate A was placed in handcuffs after being sprayed with a chemical agent was false or misleading.” The government pushed back against the second of these “two findings” and refused to “conced[e],” that Officer Childs “in fact made a false and/or misleading statement.”

To support this position, the government provided the court and the defense with a “summary of an incident ... in which DOC Internal Affairs issued a Final Report.” This summary, however, did not mention that the OIA had determined that Officer Childs and his fellow corrections officers had falsely accused Inmate A of assault. Instead the government’s summary indicated that Inmate A had acted aggressively. The government noted that there had been a “heated discussion between Inmate A and Officer Childs,” and that Officer Childs had thought Inmate A was engaging in a “deliberate attempt ... to circumvent the search process.” The government then quoted the OIA Final Report’s reproduction of Officer Childs’s (discredited) account of the incident in which he stated that when he and his colleagues “attempted to search [Inmate A], [Inmate A] started kicking at the dog” and then stated that “Because [Inmate A’s] actions interfered with the normal operations of the facility, I sprayed one burst of chemical agent ... [and] then instructed [Inmate A] to seize [sic] his disruptive behavior.”

The government not only credited a narrative of inmate aggression that the OIA had expressly discredited without disclosing that fact, thereby providing an incomplete recapitulation of the content of the OIA Final Report, it also cast doubt on the report even as to its purported focus— namely, whether Officer Childs had falsely indicated in reports that the inmate was unrestrained when Officer Childs sprayed him with a chemical agent. The government asserted that the text of Officer Childs’s report was “ambiguous at best” and that “it is not apparent that [Officer Childs] lied in his report.” The government further stated that “[t]he conclusion that Officer Childs made a false or misleading statement is at odds with the body of the [Final OIA Report] and does not appear evident from the text of Officer Childs’s [Incident Report].”

The government buttressed its intimations that the OIA investigation was unreliable when it argued its motion on the first day of trial. Athough the government acknowledged that it “expect[ed] [Officer Childs] to say that he [had been] demoted related to this incident,” it immediately undercut the force of this new disclosure by indicating that Officer Childs had had little opportunity to address or contest the report; indeed, the government asserted that Officer Childs “ha[d] never seen” the OIA Final Report, and thus he would not be able to speak to “the particulars.” Based on its description of the OIA Final Report as limited in scope and undependable in outcome, the government took the position that the April 2009 incident did not “bear[ ] directly upon” Officer Childs’s veracity and thus was not the proper subject of cross-examination.

The trial court relied on the government’s representations regarding the OIA Final Report, and it denied repeated requests by the defense 14 during the trial for disclosure of the actual report. Without the actual report, the defense had no ability to call the government’s characterization of the report into question or to persuade the court to order its disclosure. 15 The court itself was in no position to assess the adequacy of the government’s summary of the OIA investigation. The court had the government’s five-page ex parte submission, but that incomplete document concluded with Officer Childs’s (subsequently discredited) account of the incident with Inmate A; thus it said even less about Officer Childs’s misconduct than the government’s summary did.

Mr. Morton and Mr. Vaughn did not receive a copy of the OIA Final Report until three months after trial, in February 2010, after they had already moved for a new trial on other grounds. 16 In a “Sup plemental Motion for a Judgment of Acquittal and, in the Alternative, Defendant’s Motion for a New Trial,” the defense argued that the government had violated its pretrial disclosure obligations under Brady. Sentencing was delayed and a series of hearings over a period of months followed.

During this time, the government took no steps to inform the court that the five-page ex parte submission had been incomplete. The court was left to figure this out by itself, in the midst of a status hearing two months after the defense filed its Brady claim. When the court was shown the defense copy of the OIA Final Report, which included all ten pages, the court noted that the final pages looked unfamiliar. The government initially could not “represent what may ... have happened.” It later informed the court that it had failed to provide the court with the full OIA Final Report “by inadvertence.” Even so, the government asserted that the “erroneously omitted portion” of the OIA Final Report changed nothing. Instead the government asserted that it had “fully disclosed and discussed” in its motion in limine “all of the information detailed in pages six through nine” of the report.

Even after the court and the defense received complete copies of the OIA Final Report, the government stood by its characterization of the OIA investigation as one solely concerned with Officer Childs’s use of force and his alleged false reporting related to the use of force. It further denied that it was “aware” that Officer Childs had ever been formally disciplined in connection with this alleged misconduct. The government took the position that Officer Childs had “accepted a voluntary demotion” as an informal resolution to an inconclusive investigation regarding “allegations of false and/or misleading statements,” and that the only formal discipline Officer Childs had received in connection with the Inmate A incident was the Letter of Direction for improper use of force issued by Major Talley. 17

Almost a year after the defense filed its supplemental motion for a new trial raising a Brady claim, the government filed a document that could have clarified matters: the affidavit from OIA Investigator Collins. 18 In this affidavit, OIA Investigator Collins explained that (1) Major Talley’s Letter of Direction did not “supersede or impede” the OIA investigation; (2) He “officially completed” the OIA Final Report “[o]n June 27, 2009 ... and forwarded it to Office of the Director for the DC Department of Corrections,” because the DOC OIA does not, “as a matter of routine or in this specific case ... recommend disciplinary action”; (3) “Several months after the completion of [the OIA Final Report] [he] was verbally informed that Lieutenant Childs was demoted to the rank of Sergeant as a result of his actions on April 7, 2009”; and (4) “On September 15, 2009, [he] notified the U.S. Attorney’s Office for the District of Columbia of the [OIA Final Report] concerning Lieutenant Childs and his subsequent demotion.” OIA Investigator Collins also stated that he acquired (in January 2011) more details about Officer Childs’s demotion. Specifically, he learned that Officer Childs had “received formal and written notification of his demotion on August 26, 2009 ... from the Office of the Director of the Department of Corrections and [that this notification] was signed by Lieutenant Childs.” OIA Investigator Collins also learned ¡from “a DC Department of Corrections, Human Resources Specialist” that Officer Childs had been “formally demoted ... on September 13, 2009.” The government, however, filed this affidavit, without any accompanying explanation, after more than a year of post-trial litigation in this case, and it never acknowledged that OIA Investigator Collins’s sworn statements contradicted its earlier representations.

When the parties returned to court a week after this filing, the court ruled on the defendants’ supplemental motion for a new trial. The court determined that there had never been a “finding of un-tru[th] telling against Officer Childs.” The court further determined that the affidavit “essentially vindicates” that, “although the Internal Affairs Division did do an investigation of Officer Childs, no report was issued by that office and no action was taken by that office.” Accordingly, the trial court denied the motion from the bench “for the reasons stated in the government’s opposition” to the defendants’ motion. 19

B. Brady Analysis

Our adversarial system is premised on the belief that “[s]oeiety wins not only when the guilty are convicted but when criminal trials are fair.” Brady, 373 U.S. at 87, 83 S.Ct. 1194. Prosecutors have a critical role in ensuring the fairness of criminal trials. They are the representative of the sovereign, whose “interest ... in a criminal prosecution is not that it shall win a case, but that justice shall be done.” Berger v. United States, 295 U.S. 78, 88, 55 S.Ct. 629, 79 L.Ed. 1314 (1935); see also Miller v. United States, 14 A.3d 1094, 1107 (D.C.2011) (explaining that prosecutors must “seek justice before victory”). Prosecutors are thus obligated to play a dual role at trial; they must advocate for the government “with earnestness and vigor,” Berger, 295 U.S. at 88, 55 S.Ct. 629, but they also have an obligation under Brady “to assist the defense in making its case.” United States v. Bagley, 473 U.S. 667, 675 n. 6, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985). In this “limited departure from a pure adversary model,” id., prosecutors have a constitutionally imposed duty to disclose to the defense pretrial information that is “favorable to an accused ... [and] material either to guilt or to punishment.” Brady, 373 U.S. at 87, 83 S.Ct. 1194.

To determine on appeal whether the government, through its representatives in the trial court, has violated its obligations under Brady, we consider: (1) whether the information in question is “favorable to the accused”; (2) whether this information was possessed and suppressed by the government, “either willfully or inadvertently”; and (3) whether that information was material, i.e., whether there is “a reasonable probability that, had the evidence been disclosed, the result of the proceeding would have been different.” Miller, 14 A.3d at 1109. If the information was favorable, suppressed, and material, then reversal is required, “irrespective of the good faith or bad faith of the prosecution.” Brady, 373 U.S. at 87, 83 S.Ct. 1194.

Reviewing the trial court’s legal conclusions de novo and its underlying findings of fact for clear error, Miller, 14 A.3d at 1120, we consider each criterion for a Brady violation in turn.

1. Favorable Information Subject to Disclosure

At least in the abstract, it is easy to articulate what constitutes “favorable” information subject to disclosure under Brady. It is information “of a kind that would suggest to any prosecutor that the defense would want to know about it” because it helps the defense. See Miller, 14 A.3d at 1110 (quoting Leka v. Porbuondo, 257 F.3d 89, 99 (2d Cir.2001)). The defense perspective controls. See id. (“[T]he critical task of evaluating the usefulness and exculpatory value of the information is a matter primarily for defense counsel, who has a different perspective and interest from that of the police or prosecutor.” (quoting Zanders v. United States, 999 A.2d 149, 164 (D.C.2010))).

Favorable information includes impeaching information. See Giglio v. United States, 405 U.S. 150, 154-55, 92 S.Ct. 763, 31 L.Ed.2d 104 (1972). Indeed, although the Supreme Court held in Bagley, 473 U.S. at 676, 105 S.Ct. 3375, that the failure to disclose impeaching information is not “more egregious” than a failure to disclose affirmatively exculpatory information, both the Supreme Court and this court have repeatedly made clear that impeaching information does not have a lesser standing in the context of the government’s Brady disclosure obligations. Rather, “[t]he jury’s estimate of the truthfulness and reliability of a given witness may well be determinative of guilt or innocence.” Bagley, 473 U.S. at 676, 105 S.Ct. 3375 (quoting Napue v. Illinois, 360 U.S. 264, 269, 79 S.Ct. 1173, 3 L.Ed.2d 1217 (1959)); see also Bennett v. United States, 797 A.2d 1251, 1256 (D.C.2002) (“Impeaching evidence is exculpatory.” (internal quotation marks omitted)).

In this case, whether favorable information existed that was subject to disclosure under Brady turns on how one characterizes the OIA investigation and its outcome. When the trial court denied Mr. Morton’s and Mr. Vaughn’s motion for a new trial, it accepted the government’s representations that (1) the OIA investigation concerned Officer Childs’s use of force and alleged false reporting related to his use of force, and (2) Officer Childs’s demotion was something more in the nature of an administrative plea deal to resolve an inconclusive inquiry that ultimately was unrevealing regarding Officer Childs’s credibility. As an appellate court, we ordinarily defer to the trial court’s findings of fact, unless those findings are clearly erroneous. See Miller, 14 A.3d at 1120. Here, record documents, the OIA Final Report and OIA Investigator Collins’s affidavit, contradict the court’s findings. We recognize that the trial court was constrained in its ability to assess these documents by the government’s late production and continued misrepresentation or nondisclosure of the information in its possession. Unlike the trial court, however, we have had, from the outset of our review, the entire OIA Final Report with its appendices. With these advantages that the trial court did not share, we conclude that the trial court was misled and that its adoptive fact-finding was clearly wrong.

Although the purpose of the OIA investigation as articulated in the OIA Final Report was “to identify the facts and circumstances regarding the use of [a chemical agent] by Lieutenant Angelo Childs” in April 2009, the report itself reveals that the investigation went beyond an inquiry into whether Officer Childs complied with use-of-force guidelines. Ultimately, the OIA determined that Officer Childs and his fellow DOC Officers had falsely accused Inmate A of assaultive behavior, and they had all filed false reports to this effect. The OIA also determined that Officer Childs falsely suggested that the inmate was unrestrained and Officer Childs sprayed him with a chemical agent to stop the (fabricated) assault. OIA Investigator Collins’s affidavit completes the story. The inescapable inference is that the Office of the Director credited all the conclusions of the OIA Final Report and as a consequence meted out significant discipline by demoting Officer Childs from Lieutenant to Sergeant. 20

Once we clarify the actual subject and the apparent outcome of the OIA investigation, the determination that this information was favorable information subject to disclosure under Brady is not difficult. The OIA’s determination of Officer Childs’s false reporting was clearly impeaching, and was the sort of information in which any competent defense lawyer would have been intensely interested. See Milke v. Ryan, 711 F.3d 998, 1007 (9th Cir.2013) (“That [a law enforcement officer] was disciplined for lying on the job obviously bears on his credibility and qualifies as Giglio evidence.”).

Of course the favorability of the OIA Final Report does not turn on its ultimate truth or the government’s assessments thereof. The government could not withhold this information because it did not trust the conclusions of the OIA Final Report, or because it did believe its witness, Officer Childs, who professed innocence of false reporting 21 and asserted ignorance of the reason for his discipline. 22 As we said in Zanders and Miller, “[i]t is not for the prosecutor to decide not to disclose information that is on its face exculpatory based on an assessment of how that evidence might be explained away or discredited at trial, or ultimately rejected by the fact finder.” Miller, 14 A.3d at 1110 (quoting Zanders, 999 A.2d at 164); see also Smith v. Cain, — U.S. -, 132 S.Ct. 627, 630, 181 L.Ed.2d 571 (2012) (holding that a witness’s inconsistent statements were subject to disclosure notwithstanding government’s “argument ... that the jury could have disbelieved [the] undisclosed statements”).

2. The Government Suppressed Favorable Information in its Possession.

When the government possesses favorable information subject to disclosure under Brady, it has an obligation to disclose this information to the defense in a timely and complete manner. In this case, there is nothing in the record to indicate that the government ever responded to the Brady request made by the defense months before trial 23 or that it volunteered Brady information to the defense in the typical manner by sending the defense a disclosure letter. Ultimately, the defense did not obtain a copy of the OIA Final Report until three months after trial, albeit without the appended material, and it was not given OIA Investigator Collins’s affidavit until over a year after trial, a week before the trial court ruled on its post-trial Brady claim. Nevertheless, the government argues that it fulfilled its disclosure obligations when it filed, a week before trial, its motion in limine to preclude cross-examination of Officer Childs about the OIA Final Report and provided a “summary of the complete Final Report.” We cannot agree.

To begin with, there was nothing about the motion in limine that put the defense on notice that the government was disclosing Brady information. The motion never cited Brady, much less indicated that the government was attempting “to assist the defense in making its case.” Bagley, 473 U.S. at 675 n. 6, 105 S.Ct. 3375. Instead, the object of the motion in limine was to preclude defense cross-examination of a government witness. This is not an elevation of form over substance. As we said in Miller, Brady does not authorize the government to engage in a game of hide-and-seek, or require the defense to “scavenge for hints of undisclosed Brady material.” 14 A.3d at 1113 (quoting Banks v. Dretke, 540 U.S. 668, 695, 124 S.Ct. 1256, 157 L.Ed.2d 1166 (2004)). But that is precisely what the defense was forced to do in this case. Thus, at his first opportunity to address the court about the government’s motion in limine, defense counsel told the court that he was “just basically hitting in the dark” and he was “not sure what the finding [of the DOC OIA] says”:

[I]f I hadn’t read the motion closely, I wouldn’t even have known [that the government might possess Brady information], It wasn’t as if [the prosecutors] called me and told me, I have a report you need to see. I just got this in the motion, and the motion, ironically, was to actually suppress or limit the use of this information while it’s being given to me, even though it wasn’t really being given to me; it was just being alluded to or stated, but it’s — I don’t have the actual report to this day.

In any event, the government’s disclosure was not timely, accurate, or complete. Accordingly, we conclude that the government suppressed favorable information in its possession.

a. As a Brady Disclosure, the Motion in Limine Was Not Timely

The government’s obligation to make timely disclosures is grounded in the very reason Brady disclosures are required: to provide protection against miscarriages of justice. Miller, 14 A.3d at 1107. The goal of ensuring that our “adversary system of prosecution [does not] descend to a gladiatorial level unmitigated by any prosecutorial obligation for the sake of truth,” Kyles v. Whitley, 514 U.S. 419, 439, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995), is not achieved by last-minute information dumps. Rather, where disclosure of Brady is concerned, there is no time for strategic delay and “as soon as practicable” should be the approach. See Miller, 14 A.3d at 1108 (explaining that “a strategy of delay and conquer ... is not acceptable” (internal citation and quotation marks omitted)); see also id. at 1111 (rejecting the rationale that “better late than never” is good enough and endorsing the ABA standards requiring Brady disclosure “at the earliest feasible opportunity ... as soon as practicable following the filing of charges” (internal quotation marks omitted)). Certainly, Brady disclosures are required “well before the scheduled trial date,” Zanders, 999 A.2d at 164; Perez v. United States, 968 A.2d 39, 66 (D.C.2009) (Brady requires “timely, pretrial disclosure”). Only in this way can we ensure “defense counsel [has] an opportunity to investigate the facts of the case and, with the help of the defendant, craft an appropriate defense.” Perez, 968 A.2d at 66; see also Miller, 14 A.3d at 1111 (“[A]s we have repeatedly recognized, exculpatory evidence must be disclosed in time for the defense to be able to use it effectively, not only in the presentation of its case, but also in its trial preparation.”); Boyd, 908 A.2d at 57 (“[T]imely disclosure ... can never be overemphasized.”).

By no means can the government’s motion in limine constitute a timely pretrial disclosure of the information it possessed about Officer Childs’s discipline as a result of the OIA investigation. The motion in limine provided no information on this subject although — according to the affidavit from OIA Investigator Collins that the government filed with the court— the “U.S. Attorney’s Office” was informed of the OIA Final Report “concerning Lieutenant Childs and his subsequent demotion” nearly two months before the government filed this motion. The government did not reveal that Officer Childs had been demoted until the first day of trial, when it briefly noted that Officer Childs was demoted “related to” the April 2009 incident that it had incompletely summarized in its motion in limine. The government did not provide any further details about when or how this discipline had been imposed until' it filed OIA Investigator Collins’s affidavit thirteen months after trial.

But the government’s disclosure obligations were triggered well before the DOC decided to demote Officer Childs. The government had an obligation to notify the defense that Officer Childs was under investigation by the DOC OIA. See United States v. Borne, 198 F.3d 905, 908 (D.C.Cir.1999) (determining that the prosecution had a duty to disclose the fact that one of its police officer witnesses had be come “the subject of an investigation into the truthfulness of his testimony” in another case); see also Bullock v. United States, 709 A.2d 87, 92-93 (D.C.1998) (remanding to trial court to develop record on whether government should have disclosed that testifying law enforcement officer was under investigation). As to the OIA investigation, which began in April 2009 and concluded in June 2009, the government’s motion in limine, filed a week before the November 2009 trial, was not an “as soon as practicable” Brady disclosure. 24

Accepting for the sake of argument the government’s assertion that its motion in limine filed a week before trial was a Brady disclosure, the belatedness of this filing is not excused by the government’s representation to the trial court that it did not learn of the DOC OIA investigation and its resulting report until “late summer.” Brady does not tolerate the “government's] failure to turn over an easily turned rock.” United States v. Brooks, 966 F.2d 1500, 1503 (D.C.Cir.1992). “[T]he individual prosecutor has a duty to learn of any favorable evidence known to the others acting on the government’s behalf in the case.” Kyles, 514 U.S. at 437, 115 S.Ct. 1555; see also Robinson v. United States, 825 A.2d 318, 324 (D.C.2003) (“[T]here is a duty to search branches of government ‘closely aligned with the prosecution.’ ”). In this case, OIA Investigator Collins — the same investigator who investigated Officer Childs — was working closely with the U.S. Attorney’s Office on the Spencer-White Case and had identified Officer Childs as central to the investigation of the Spencer-White attacks. For such an important witness so closely tied to the investigation, the government should have had the systems in place to ensure that it was alerted immediately about impeaching information. See Kyles, 514 U.S. at 419, 438, 115 S.Ct. 1555 (observing that “the prosecutor has the means to discharge the government’s Brady responsibility if he will,” and that “procedures and regulations can be established to carry [the prosecutor’s] burden and to insure communication of all relevant information on each case to every lawyer who deals with it” (quoting Giglio, 405 U.S. at 154, 92 S.Ct. 763)); see also Brooks, 966 F.2d at 1502-03 (observing that “the prosecutor’s own interest in avoiding surprise at trial gives him a very considerable incentive to search accessible files for possibly exculpatory evidence, quite independent of Brady ”).

In short, if it was a Brady disclosure at all, the government’s motion in limine was not a timely one.

b. As a Brady Disclosure, the Motion in Limine Was Neither Accurate Nor Complete

In addition to its timing, we also consider the content of the motion in li-mine, in particular the “summary” of the OIA Final Report. The defense attorneys were not satisfied with the government’s summary; they pushed for disclosure of the “actual report.” Even if it is theoretically possible for the government to fulfill its disclosure obligations under Brady by means of summaries of preexisting documents, such summaries must be “sufficiently specific and complete.” United States v. Rodriguez, 496 F.3d 221, 226 (2d Cir.2007). 25 Again we consider the requisite level of detail from the perspective of the defense; where source documents exist, the government must “summarize[] ... [them] with every detail that might have been relevant to defense counsel’s preparation as counsel viewed the case.” Wiggins, 386 A.2d at 1178 (Ferren, J., concurring). This may be challenging for the government, which presumably is not privy to defense counsel’s thoughts and theories pretrial. Accordingly, the government withholds source documents at its peril.

Here, the government’s motion in limine did not come close to satisfying our standards for the content of Brady disclosures. What the government called a “summary of an incident on April 6, 2009, in which DOC Internal Affairs issued a Final Report” was not a trae summary of the OIA Final Report at all, much less a summary of all the favorable information in the government’s possession regarding Officer Childs.

The OIA Final Report had exposed what the OIA determined was an untrue, self-serving story by Officer Childs and two colleagues of an assault by an inmate. The government’s summary, however, indicated that Inmate A had acted aggressively. The summary quoted Officer Childs’s explanation, reproduced in the OIA Final Report, that “[b]ecause [Inmate A’s] actions interfered with the normal operations of the facility,” he used a chemical agent on Inmate A and instructed him “to seize [sic] his disruptive behavior”— without qualifying that OIA had discredited this account because “video footage of the incident [did] not support” it. Thus, the government reproduced and represented as essentially undisputed substantial portions of the very Incident Report the OIA determined contained a false account of the incident.

The government’s motion in limine not only presented as true that which OIA had determined false, it used that false story as the backdrop for its account that the OIA investigation was simply an inquiry as to whether Officer Childs had used excessive force on a restrained Inmate A and whether Officer Childs had engaged in possibly sloppy report-writing to the extent he incorrectly “suggested]” that Inmate A was unrestrained. 26 The govern ment disputed in its motion in limine that this suggestion was “evident” from Officer Childs’s incident report, and it refused to “concede” that Officer Childs had “in fact” made false or misleading statements with respect to whether Inmate A was handcuffed, even though the OIA had determined that, as part of his fabricated story of inmate assault, Officer Childs had misleadingly indicated that Inmate A was unrestrained. The government’s omission of the disciplinary consequences of the OIA Final Report bolstered the inaccurate account of the OIA investigation in the government’s summary.

c. As a Brady Disclosure, the Motion in Limine Was Not Usable

The adequacy of a Brady disclosure ultimately turns on “the sufficiency, under the circumstances, of the defense’s opportunity to use the evidence when disclosure is made.” Miller, 14 A.3d at 1111 (quoting Leka, 257 F.3d at 100). Beyond the timing and the content of the motion in limine, the trial record reflects that the government never afforded defense counsel a meaningful “opportunity to use” the favorable information in the government’s possession. Id. The defense was never able to impeach Officer Childs with the OIA investigation or its disciplinary consequences. Moreover, the government and Officer Childs made representations at trial that blocked the court from acting to ensure that any favorable information in the government’s possession would come to light and the defense would be able to use it.

The government maintained throughout trial that the OIA investigation was limited to an inquiry into use of force and alleged false reporting related to the use of force, which had no bearing on Officer Childs’s credibility as an after-the-fact identification witness in the Spencer-White trial. The defense, ill-informed by the government’s summary, could not expose the government’s representations about the OIA investigation and its outcome as inaccurate, and could not persuade the trial court that a probing cross-examination was warranted. Instead, as counsel put it, they were forced to “tak[e] the [government’s] representations for what the report said or didn’t say” and did not “really have any facts about what happened or what was ... alleged to even ask [Officer Childs].”

Nevertheless, when Officer Childs took the stand, defense counsel attempted on cross-examination to “probe” the basis for the OIA investigation. The government objected, and, relying on the government’s representations, the trial court limited the defense to two questions: (1) if it was true Officer Childs had “submitted a false report to the Department of Corrections while [he] [was] working at the D.C. Jail” and (2) if it was true he had been “disciplined by the Department of Corrections for filing a false report.” To both questions, Officer Childs responded, “no.” The defense had no means of challenging these responses, much less any basis to ask the court for more leeway in impeaching Officer Childs with his prior false reporting.

Although the defense had been unsuccessful in its questioning, the trial court nonetheless tried to fulfill its “obligation to assure” itself that no Brady information had been withheld from the defense and that the defense had had a meaningful opportunity to confront Officer Childs. See Boyd, 908 A.2d at 59. These efforts were for naught.

The trial court questioned Officer Childs about false reporting out of the presence of the jury. In response, Officer Childs represented that DOC OIA “felt that I missed the inmate while he was in handcuffs” and that he had explained to DOC OIA that “that was an error. I do a lot of cutting and pasting. But I said if you look at the rest of my report, that I stated the inmate was standing in front of the bubble with the handcuffs on. So, how can I say that I applied handcuffs, right?” The court asked Officer Childs whether he had been disciplined in connection with the OIA investigation. Officer Childs represented that he had received “a voluntary demotion” but that he “didn’t understand it to be the result[] of any disciplinary action” and that “no one [had] told [him] the reason for the demotion.” The government did not seek to qualify or correct anything Officer Childs told the trial court, even though the government (1) had a complete copy of the OIA Final Report which documented the OIA’s determinations both that Officer Childs had made a false report about whether Inmate A was unrestrained and this false report was part of a larger false narrative of an inmate assault, and (2) had been informed by OIA Investigator Collins that the DOC Office of the Director had demoted Officer Childs after receiving the OIA Final Report. 27

The trial court then called only the government to the bench to ask about the five-page ex parte submission. The court stated that it “just wantfed] to make sure.... I have the entire filing, because mine stops at page 5, and there was no ... discipline or resolution mentioned in the document ... that I have a copy of.” The government did not alert the trial court that the ex parte submission was incomplete; instead, one of the trial prosecutors told the court, “[y]eah, mine is five pages long,” incorrectly indicating that the court had the full OIA Final Report. The same prosecutor further informed the court, “there is no discipline listed in the report itself,” but the prosecutor did not explain that the OIA only conducts investigations and does not make disciplinary decisions; nor did the prosecutor inform the court that the decision to demote Officer Childs for making false reports had been made by the proper authority, the DOC Office of the Director.

The trial court subsequently put its understanding of the OIA investigation on the record:

[W]e have a little more clarity concerning this incident. For your purposes, the report that was submitted to me, which is what I was just confirming with the government, is not in the form of a disciplinary action, which is why I followed up with the questions I had of Officer Childs, because — and I confirmed with the government that they weren’t aware of any, quote, disciplinary action the way we would — I had envisioned it could be; that is, there’d be an adjudication and then a finding of false report. I think it sounds like there was an administrative resolution between the officer and the Department of Corrections.

The court explained that in allowing any questions about the investigation, it had been “operating under the assumption that ... we were talking about some sort of disciplinary action where there was a finding” but “we’re far afield from that.” The court then observed that there was no “factual basis that would support going further on this matter, in terms of questioning.” Concluding the discussion, the court reiterated that “based on what I heard” it would not permit “additional questioning ... on what I, at this point, assess ... to be a collateral matter, which was an informal resolution ... of a disciplinary matter, without a finding of ... a nontruth-telling event. 28 Again the government stood silent.

Based on the foregoing, we cannot say that the government, through its motion in limine, disclosed to the defense the favorable information in its possession in such a way as to allow the defense to use this information effectively. The government not only failed to give the defense (or the court) accurate or complete information, it then stood by at trial and allowed the defense’s ignorance and the court’s erroneous understanding of the pertinent facts to persist. The upshot was that the trial court did not permit thorough impeachment of Officer Childs on the subject of the OIA investigation and its disciplinary consequences because it was convinced that, on these topics, there was nothing to impeach Officer Childs about. In short, the government’s motion in limine to preclude impeachment of Officer Childs cannot be construed as a Brady disclosure because it worked — the government’s motion prevented an effective cross-examination of Officer Childs on the subject of his prior false reporting.

[[Image here]]

In the absence of a timely, accurate, complete, or usable disclosure, we conclude that the government did not disclose to the defense the favorable information in its possession in this case.

3. The Favorable Information Suppressed by the Government Was Material

We next consider whether the favorable information withheld by the government was “material.” See Miller, 14 A.3d at 1115. To assess materiality, we consider whether there is “a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different.” Id. (quoting Bagley, 473 U.S. at 682, 105 S.Ct. 3375). The reasonable probability standard does not require a showing that it is more likely than not the defendant would have been acquitted. Kyles, 514 U.S. at 434, 115 S.Ct. 1555. Rather, since Brady is a rule of fairness, the materiality threshold is met if, in the absence of proper disclosure, we question whether the defendant received a fair trial and our “confidence” in the outcome of the trial is thereby “undermine[d].” 29 Id.; see also Cain, 132 S.Ct. at 630 (quoting Kyles, 514 U.S. at 434, 115 S.Ct. 1555); United States v. Cuffie, 80 F.3d 514, 517 (D.C.Cir.1996) (explaining that the proper focus “is on the ‘potential impact that the undisclosed evidence might have had on the fairness of the proceedings’ rather than on the overall strength of the government’s case” (quoting United States v. Smith, 77 F.3d 511, 515 (D.C.Cir.1996))).

Our materiality analysis in this case must begin with an assessment of Officer Childs’s role in the government’s case. Officer Childs was a critical witness, as the government conceded in its brief to this court. Indeed, the government signaled Officer Childs’s importance by filing its motion in limine to preclude his cross-examination about the OIA Final Report. At trial the government’s need for his testimony became apparent: Officer Childs was only one of two government witnesses who inculpated Mr. Morton and Mr. Vaughn in the assault on Sergeant White, and he was objectively the stronger witness. 30 Similarly, it cannot reasonably be disputed that Officer Childs’s credibility was a central issue in the case. In closing, the government argued that its case turned on the jury’s assessment of the credibility of its witnesses and of Officer Childs in particular. See Lewis v. United States, 408 A.2d 303, 308 (D.C.1979) (noting that the “substantial possibility ... that impeachment with prior convictions will affect the outcome, whether it goes to general credibility ... or to a more specific target of reliability, such as bias” is “especially strong in the case of a key government witness” (citation omitted)).

Next, we consider the nature of the withheld information: It was powerfully impeaching. It did not simply establish that Officer Childs had a track record for untruthfulness. It established that he was willing to make false reports implicating inmates in assaults on law enforcement agents — the precise context of this case. See Milke, 711 F.3d at 1007 (concluding that suppressed information about a police officer’s untruthfulness would have undermined his credibility and demonstrated that he “had no compunction about abusing his authority with a member of the public” who was similarly situated to the appellant).

This leads us to the ultimate question: Does the suppression of this information undermine our confidence in the fairness of the trial? We conclude it does. Had the defense and the court known the full details of the OIA’s actual findings and of the discipline meted out by DOC as a result — and had the government known the defense knew — we think it likely that this case would have played out very differently.

Preliminarily, we question whether the government would have made the same representations and whether Officer Childs would have given the same testimony. But, at the very least, defense counsel would have had a firm foundation to press for much more leeway in cross-examining Officer Childs on the subject of his prior false reporting and resulting discipline, instead of “hitting in the dark.” See Cuffie, 80 F.3d at 517 (explaining that undisclosed evidence of untruthfulness is material if it “could have substantially affected the efforts of defense counsel to impeach the witness, thereby calling into question the fairness of the ultimate verdict” (quoting Smith, 77 F.3d at 515).) “[A]s a general rule, a defendant is entitled to wide latitude in presenting evidence tending to impeach the credibility of a witness, especially where[, as here,] that evidence relates to a key government witness.” Dockery v. United States, 746 A.2d 303, 306 (D.C.2000) (internal quotation marks omitted); Shorter v. United States, 792 A.2d 228, 233 (D.C.2001) (noting that trial courts may not curtail cross-examination so as to “keep[ ] from the jury relevant and important facts bearing on the trustworthiness of crucial testimony” (internal quotation mark omitted)). Here it appears that the trial court would have granted the defense such leeway; indeed, the court all but said that had Officer Childs been disciplined for making a false report, it would have allowed the defense to engage in a very different cross-examination of Officer Childs. This court can readily envision a series of permissible, pointed questions by defense counsel that would have seriously damaged Officer Childs’s credibility in the eyes of the jury — whether he answered them truthfully or tried to avoid giving a truthful answer. 31 See Milke, 711 F.3d at 1009 (explaining how police officer could have been confronted with prior instances of untruthfulness).

The government asserts, however, that it “would not have influenced the verdict” if Officer Childs had either (1) admitted that he had been “demoted because he had deployed a chemical agent on an inmate” or (2) explained that he had not “purposefully file[d] a false report” indicating the inmate was unrestrained. We reject the factual foundation for the government’s argument; as explained above, the OIA Final Report documents an investigation that was centrally concerned with a false report of an inmate assault by Officer Childs and two fellow corrections officers to justify an improper use of force, Officer Childs was demoted because of this investigation, and we infer that he was so disciplined as a sanction for all the conduct detailed in the OIA Report, including the determinations of false reporting.

The government also argues that disclosure of the OIA Final Report would have had little effect on the outcome of the trial because the defense could neither have inquired into it on cross-examination in any further depth, nor could it have introduced the report itself as extrinsic evidence.

In the first alternative, the government takes the position that the defense “could not have confronted Sergeant Childs with the report itself in order to impeach his credibility” because “appellants have not shown that Sergeant Childs’s statements were false.” Thus, the government argues the defense could not have laid the requisite foundation for cross-examination about a prior bad act. See Sherer v. United States, 470 A.2d 732, 738 (D.C.1983) (explaining that “a witness may be cross-examined on a prior bad act that has not resulted in a criminal conviction only where: (1) the examiner has a factual predicate for such question, and (2) the bad act bears directly upon the veracity of the witness in respect to the issues involved [in] the trial” (internal quotation marks omitted) (quoting United States v. Akers, 374 A.2d 874, 878 (D.C.1977))). Again, the government seeks to dispute the validity of the DOC’s determination of false reporting. But, the government’s disbelief in the accuracy of the OIA’s findings is irrelevant, both to an assessment of its obligation to disclose this information under Brady and to an assessment of the use the defense could have made of this information at trial. The point is that, unlike in Sherer, where the defense proffer that a government witness had previously lied “was conclusory and based almost entirely on inadmissible hearsay,” 470 A.2d at 739, here there was an official finding of false reporting by a government agency. This official determination provided the defense with an ample “factual predicate” for questioning Officer Childs about false reporting. 32 See id. at 738; see also Wagner v. Georgetown Univ. Med. Ctr., 768 A.2d 546, 563 (D.C.2001) (finding that a professional association’s findings recommending censure constituted a “more than sufficient factual predicate for the proposed cross examination” of a physician witness).

In the second alternative, the government argues that the defense would not have been able to present extrinsic evidence of the OIA Final Report because it concerned a collateral issue. See Rowland v. United States, 840 A.2d 664, 680 (D.C.2004). Thus, the government argues that even if the defense had had the full report and information about the subsequent demotion, it would have been stuck with the testimony Officer Childs gave at trial, namely, his denials that he had filed a false report and that he had been demoted for filing a false report. Whether the defense could have introduced the actual OIA Final Report into evidence is an interesting but ultimately academic issue. 33 Effectively, the government argues that Brady disclosure or no, nothing would have changed in how this case played out at trial. As explained above, we think that argument blinks reality. See Milke, 711 F.3d at 1009 (explaining how prior judicial determinations of police officer’s untruthfulness “would have been a game-changer” on cross-examination even without proof of extrinsic evidence).

Based on the record before us, whether the government had an obligation to accurately and completely disclose the contents of the OIA Final Report and the DOC’s consequent decision to demote Officer Childs should not have been a hard call for the government. 34 And had the defense been able to impeach Officer Childs with the DOC’s determination of his prior false reporting and consequent demotion, there is at least a reasonable probability that the jury would have weighed Officer Childs’s testimony and the government’s case differently. This concludes our materiality analysis for Mr. Morton. But for Mr. Vaughn, there is a coda. The suppressed impeachment evidence for Officer Childs was material only to the extent that Mr. Vaughn contested Officer Childs’s identification. But Mr. Vaughn submitted a post-trial affidavit in which he admitted that he was the inmate Officer Childs had identified him to be and argued only that his actions had been misinterpreted — that he had not pushed Sergeant White; he had been trying to help him. This admission negates our materiality determination. In light of Mr. Vaughn’s affidavit, which the government would be free to use as a party admission, we see little chance of a different result were Mr. Vaughn to be given a new trial. Thus, the government’s Brady violation is reversible only with respect to Mr. Morton.

C. Remand

By now government prosecutors should know; “Betray Brady, give short shrift to Giglio, and you will lose your ill-gotten conviction.” United States v. Olsen, 737 F.3d 625, 633 (9th Cir.2013) (Kozinski, C.J., dissenting from denial of petition for rehearing en banc). So it is for the government with respect to Mr. Morton’s conviction; we reverse and remand his case for a new trial.

Well before any such trial takes place, 35 the court must confirm that the government has in fact provided to the defense all the favorable information in its possession. As far as we can tell, the government has never represented in this case that it has fulfilled its constitutional duty to learn of and disclose to the defense all the Brady information in its actual or constructive possession. See Kyles, 514 U.S. at 487, 115 S.Ct. 1555. The government should be directed to make such a representation, in writing, filed with the trial court. See, e.g., Milke, 711 F.3d at 1019 (requiring, on remand, the state to make complete Brady disclosures regarding the subject police officer and to “provide a statement under oath from a relevant police official certifying that all of the records have been disclosed and none has been omitted, lost or destroyed”); United States v. Naegele, 468 F.Supp.2d 150, 155 (D.D.C.2007) (directing government to “conduct all necessary searches and produce to the defendant any and all materials that are responsive” to prior Brady requests and “certify to the [cjourt in writing that it has done so”). It also goes without saying that if any other instances of nondisclosure come to light, further “remedial sanctions” beyond a retrial may be needed. See Odom v. United States, 930 A.2d 157, 158-59 (D.C.2007) (recognizing a trial court’s broad authority to fashion “appropriate remedial sanctions” so long as they are just under the circumstances).

III. Jury Instruction Error

Although Mr. Vaughn obtains no relief on his Brady claim, we conclude that his conviction for aggravated assault must be reversed on other grounds. Mr. Vaughn argues that the trial court erred in instructing the jury as to the elements of this offense on a theory of aider and abettor liability. Because counsel did not object, this claim is subject to plain error review. See Perez, 968 A.2d at 92. The government has conceded that the court’s aiding and abetting instruction was plainly incorrect, satisfying the first two criteria for reversal. See Perry v. United States, 36 A.3d 799, 818 (D.C.2011) (citing United States v. Olano, 507 U.S. 725, 732-36, 113 S.Ct. 1770, 123 L.Ed.2d 508 (1993)). But the government argues that Mr. Vaughn cannot show the third criterion for plain error review, namely that this error “af-feet[ed] substantial rights.” Id. We disagree, and because we also conclude that this error “seriously affect[ed] the fairness, integrity or public reputation of judicial proceedings,” id., the fourth criterion for plain error review, we exercise our discretion to reverse.

Mr. Vaughn was charged with both aggravated assault and felony assault. As the more serious offense, the D.C.Code appropriately sets a high bar for aggravated assault. Whereas felony assault requires proof of “significant bodily injury” 36 that a defendant caused “intentionally, knowingly, or recklessly,” D.C.Code § 22- 404(a)(2), aggravated assault requires proof of “serious bodily injury” 37 that a defendant either “knowingly or purposely cause[d]” or that resulted from conduct in which the defendant “[ujnder circumstances manifesting extreme indifference to human life ... intentionally or knowingly engage[d] ... [and] which create[d] a grave risk of serious bodily injury.” D.C.Code § 22-404.01; see also Perry, 36 A.3d at 817.

The problem in this case arose after the court correctly instructed the jury on the elements of principal liability for aggravated assault and felony assault. The trial court then explained that the jury could also convict if it determined a defendant had aided and abetted the principal offender. Whether the jury considers a defendant’s criminal liability as a principal or as an aider or abettor, the requisite mens rea for the charged assault is the same. Perry, 36 A.3d at 814-18; see also Wilson-Bey, 903 A.2d at 837-38 (holding that to be guilty as an aider or abettor, an accomplice must have the same mental state required for conviction as a principal). Although the court correctly charged the jury that it had to find the same mens rea to convict a defendant as a principal or an aider and abettor, it then gave an instruction for aider and abettor liability that inappropriately combined the standards for aggravated assault and felony assault. In effect, the court informed the jury that it could convict a defendant as an aider and abettor of an aggravated assault if it found that the defendant had the lesser mens rea for felony assault. 38 Under D.C.Code § 22-404.01 and Perry, this was error and plainly so.

To demonstrate that this plain error warrants reversal, Mr. Vaughn must show prejudice, i.e., a “ ‘reasonable probability’ of a different outcome if the jury had been properly instructed.” Perry, 36 A.3d at 818. This standard for assessing prejudice requires us to “make sense of the jury’s verdict in light of the evidence presented and the instructions given to the jury.” Id. at 821.

The first step of the inquiry is whether we can discern with any assurance that Mr. Vaughn was convicted of aggravated assault on a theory of principal liability— on which the jury had earlier been correctly instructed — or whether there is a reasonable probability that Mr. Vaughn was convicted on a theory of aiding and abetting liability — on which the jury never received a correct instruction. On the record before the court it is clear that the jury could not have convicted Mr. Vaughn of aggravated assault as a principal. No evidence was introduced at trial that he caused Sergeant White to suffer the serious bodily injury alleged, i.e., his subsequent unconsciousness and head injuries, either by personally inflicting these injuries or creating the grave risk that these particular injuries would occur. 39 The government’s theory at trial was that Mr. Vaughn was the inmate in white, whose only contact with Sergeant White was to give him “one good shove.” This push knocked Sergeant White against a table. But the recorded footage shows that Sergeant White immediately arose from the table and walked away; 40 meanwhile the inmate in white retreated and did not engage in any further aggressive action. The government acknowledged in closing that Mr. Vaughn was not present for the subsequent attack on Sergeant White; the government highlighted for the jury that Mr. Vaughn was not on the scene because “he [ran] back upstairs” and did not “stand there to help.” The complete lack of evidence that Mr. Vaughn principally caused serious bodily injury makes it clear that there is at least a reasonable probability that the jury relied on a theory of aiding and abetting to convict Mr. Vaughn of aggravated assault. See Perry, 36 A.3d at 821.

Focusing on aiding and abetting liability, we next consider whether it was reasonably probable that the jury’s determination of guilt turned on the mistakenly minimized intent element in the aiding and abetting instruction. See id. We think it was. To discern Mr. Vaughn’s mental state, the jury could only look to and draw inferences from his conduct. From Mr. Vaughn’s “one good shove” that caused Sergeant White to briefly lose his balance, it seems highly unlikely the jury would infer that Mr. Vaughn “himself intended to cause serious bodily injury.” Perry, 36 A.3d at 817. Thereafter, Mr. Vaughn did not seek to press his advantage and strike Sergeant White again. Instead, by immediately retreating, he allowed Sergeant White to recover and move away. Similarly, from Mr. Vaughn’s “one good shove,” it seems unlikely the jury would infer that Mr. Vaughn acted intentionally to “create[ ] a grave risk of serious bodily injury.” Id. at 817 (internal quotation marks omitted). At the time of Mr. Vaughn’s single push, no inmate had acted aggressively toward Sergeant White; nor is there any evidence in the record as to when the subsequent attack began. And, although one might argue that, at the time of his push, the ensuing escalation of events was reasonably foreseeable to Mr. Vaughn, diminishing the requisite aider and abettor mens rea for aggravated assault to reasonable foreseeability is precisely what this court held in Perry was disallowed under Wilson-Bey. Perry, 36 A.3d at 817-18 (rejecting under Wilson-Bey a jury instruction that would allow appellants to be “liable for aggravated assault for negligently having begun a[n] ... assault if it was ‘natural and probable’ that the melée would escalate to severe kicking by someone else, even if appellants did not themselves have the intent to cause serious bodily injury or ‘manifest extreme indifference to human life’ ”). Thus, we find that there is at least a “reasonable probability” that Mr. Vaughn would not have been convicted of aggravated assault absent the error in the aiding and abetting jury instruction.

Lastly, we conclude that the fourth element of plain error review is satisfied. 41 Mr. Vaughn was “wrongly convicted of aggravated assault on an aiding and abetting theory of liability, without a jury determination that [he] had the mens rea required for conviction of that offense,” and such “[a] wrongful conviction necessarily affects the integrity of this proceeding and impugns the public reputation of judicial proceedings in general.” Perry, 36 A.3d at 822.

Because Mr. Vaughn has satisfied all the elements of plain error review, we exercise our discretion to reverse his conviction for aggravated assault.

IV. Other Issues

A. Admission of Identification Testimony

Mr. Morton and Mr. Vaughn also argue that the court should not have permitted Sergeant Harper and Officer Childs to testify as identifying witnesses because their testimony did not satisfy the standards outlined in Sanders v. United States, 809 A.2d 584, 596 (D.C.2002). Discerning no abuse of discretion, id. at 590, we affirm the trial court’s Sanders ruling. 42

In Sanders, this court held that “lay witness opinion testimony regarding the identity of a person in a surveillance photograph or a surveillance videotape is admissible into evidence, provided that such testimony is: (a) rationally based on the perception of a witness who is familiar with the defendant’s appearance and has had substantial contact with the defendant; and (b) helpful to the factfinder in the determination of a fact in issue.” 43 809 A.2d at 596. Both Mr. Morton and Mr. Vaughn contend that the trial court erred when it allowed Officer Childs and Officer Harper to provide identification testimony where the evidence did not establish sufficient familiarity and substantial contact with Mr. Morton or Mr. Vaughn. 44

We consider it a close call whether, under Sanders, Officer Childs and Officer Harper had sufficient contact with and knowledge of the defendants. Certainly, this case does not involve the familial or close personal relationships seen in Sanders or other cases in which identification testimony has been admitted in federal courts under Rule 701. 45 But the government did present evidence that the officers, over a period of months, had daily interaction with Mr. Morton and Mr. Vaughn throughout the routine functions of their jobs — interaction which gave rise to familiarity and particular knowledge of their physical features. We conclude it was therefore within the discretion of the trial court to find that there was an adequate foundation for the proffered identification testimony. Both Mr. Morton and Mr. Vaughn point to weaknesses in the officers’ trial testimony, but such weaknesses are classic fodder for cross-examination; they do not bear on the court’s initial Sanders ruling. 46

B. Ineffective Assistance of Counsel

Mr. Vaughn also argues that the representation he received at trial was constitutionally ineffective. To establish ineffective assistance of counsel, the detendant must show that counsel’s performance was deficient and that this deficient performance prejudiced the defendant. Blakeney v. United States, 77 A.3d 328, 340 (D.C.2013) (citing Strickland v. Washington, 466 U.S. 668, 686-87, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984)). In reviewing the trial court’s determination that counsel was not ineffective, we accept the trial court’s factual findings unless they are without evidentiary support in the record, but review the legal conclusions regarding the constitutional significance of those findings de novo. Id. at 341. We affirm the trial court’s finding that Mr. Vaughn failed to establish deficient performance on the part of his trial counsel.

In order to show deficient performance, the defendant must “show[ ] that counsel made errors so serious that counsel was not functioning as the ‘counsel’ guaranteed the defendant by the Sixth Amendment.” Strickland, 466 U.S. at 687, 104 S.Ct. 2052. Mr. Vaughn first claims that his trial counsel’s representation was deficient because counsel failed to investigate and prepare witnesses who could have testified at trial. But the trial court found, consistent with trial counsel’s testimony in the post-trial proceeding, that counsel interviewed the witnesses that Mr. Vaughn provided, and ultimately determined that their testimony would not be helpful where it contradicted the evidence depicted in the videos. The ultimate decision not to call these witnesses “is a judgment left almost exclusively to counsel,” and we see nothing on this record to call that judgment into question. Oliver v. United States, 832 A.2d 153, 158 (D.C.2003) (internal quotation marks omitted).

Mr. Vaughn also claims that he would have exercised his right to testify but for counsel’s failure to prepare him to do so. The trial court, however, credited counsel’s testimony in post-trial proceedings that he had met with Mr. Vaughn several times prior to trial and asked a co-defendant’s counsel to conduct a mock cross-examination of Mr. Vaughn. This, too, is well within the range of reasonable attorney performance.

Finally, Mr. Vaughn argues that counsel presented defense theories that were inconsistent. Counsel argued that the person seen in the recorded footage wearing white was not Mr. Vaughn, but in any event, if the jury thought that person was Mr. Vaughn, the alleged push was insufficient to satisfy the element of serious bodily injury required for aggravated assault. Even if this argument in the alternative constituted a contradiction, a basic principle of criminal law is that a defendant may present inconsistent and contradictory defenses. See, e.g., McClam v. United States, 775 A.2d 1100, 1104 (D.C.2001).

For these reasons, we affirm the trial court’s denial of Mr. Vaughn’s ineffective assistance of counsel claim.

C. Sufficiency of the Evidence

Both Mr. Morton and Mr. Vaughn raise challenges to the sufficiency of the government’s evidence. Mr. Morton effectively reargues his Sanders challenge in his attack on the sufficiency of the evidence when he argues that he could not have been convicted “without the non-eyewitness testimony” of the corrections officers who viewed the videotape, and that even with this testimony “the jury must have been left to speculate about whether, in fact, the video identifications had any basis.” We have affirmed the trial court’s Sanders ruling, but even if we had not, we consider even wrongfully admitted evidence in assessing sufficiency, Mitchell v. United States, 985 A.2d 1125, 1134-35 (D.C.2009). Viewing the totality of the evidence presented in the light most favorable to the government, see id. at 1133, we reject Mr. Morton’s sufficiency challenge to his convictions for aggravated assault and assault on a law enforcement officer.

For his part, Mr. Vaughn attacks the sufficiency of the government’s evidence against him on the ground that the jury was “misled” by the poor quality of the video footage of his contact with Sergeant White and failed to discern that his “true act” was not to attack Sergeant White, as the government argued, but rather to come to his aid and defend him against attack by others. Again, we view the evidence in the light most favorable to the government. The jury could have viewed the video and concluded that it showed Mr. Vaughn engaging in an altercation with Sergeant White and taking an action that it could reasonably have interpreted as a push.

V. Conclusion

For the reasons set forth above we affirm Mr. Vaughn’s conviction for assault on a law enforcement officer; we reverse his conviction for aggravated assault; we reverse Mr. Morton’s convictions for assault on a law enforcement officer and aggravated assault; and we remand for further proceedings consistent with this opinion.

So ordered.

1 . Bobby Johnson and Lorenzo Woods, Mr. Morton and Mr. Vaughn's co-defendants at trial, were among those charged with the attack on Mr. Spencer. Mr. Woods was acquitted; Mr. Johnson is not a party to this appeal.

2 . Sergeant Jimmy Harper, the senior officer on duty, succeeded in restraining one inmate, but could not recall if he returned to the fray thereafter. Officer Tamira Robeson, who was working her first day on the unit and was assigned to the "bubble,” the control module for the unit, remained at her post.

3 . See United States v. Wilson, No. 2008-CF2-27313 (D.C.Super.Ct. Oct. 19, 2009) (guilty plea entered); United States v. Collins, No. 2008-CF2-27154 (D.C.Super.Ct. Oct. 19, 2009) (same); United States v. Tomlinson, No. 2008-CF2-27301 (D.C.Super.Ct. Sept. 15, 2009) (same); United States v. Lyles, No. 2008-CF2-27159 (D.C.Super.Ct. July 9, 2009) (same); United States v. Cheadle, No. 2008-CF2-27153 (D.C.Super.Ct. June 30, 2009) (same); United States v. Al-Delaema, No. 2008-CF2-27146 (D.C.Super.Ct. Mar. 3, 2009) (same). See also United States v. Owens, No. 2008-CF2-27313 (D.C.Super.Ct. Oct. 14, 2009) (charges dismissed).

4 .The vast majority of the inmates seen in the recordings are young African-American males within a range of average height and build. With the exception of one inmate dressed all in white, these inmates were wearing the DOC standard-issue one-piece orange jumpsuit. Presumably to facilitate identification, the inmates' jumpsuits bear armbands with an inmate's name and a number assigned to the individual inmate, but this identifying information is not visible in the videotapes.

5 . Although his precise rank proved to be a matter of some significance, the government generically introduced Mr. Childs as "Officer” at trial and for consistency's sake we follow suit.

6 . Even though Mr. Morton and Mr. Vaughn were not charged with any crime in connection with the attacks on Mr. Spencer, Sergeant Harper initially identified them as participants in the attack on Mr. Spencer. The government did not ask Sergeant Harper to identify Mr. Morton or Mr. Vaughn in the footage of that attack.

7 . We retain the trial court's and parties’ convention of calling this individual “Inmate A.”

8 . A BOSS Chair detects metal objects secreted inside a person’s body.

9 . The appendices include Officer Childs's Incident Report explaining his use of a chemical ■ agent and his Disciplinary Investigative Report formally accusing Inmate A of "assault without serious injury”; it also includes reports from Major Talley, Sergeant Thomas, and Lieutenant McKnight, a "Letter of Direction” from Major Talley to Officer Childs, see infra note 10, and three DOC Program Statements addressing (1) use of force and application of restraints, (2) the policies of the canine unit, and (3) the Employee Code of Ethics and Conduct.

10 . The OIA Final Report acknowledges that Major Talley issued Officer Childs a Letter of Direction reprimanding him for "Neglect of Duty and Incompetence” in connection with his use of a chemical agent on Inmate A when Inmate A was "restrained from the rear, and posed no immediate threat.” But the OIA Final Report indicates that this Letter of Direction was deficient because Major Talley, who herself had submitted a report recounting Inmate A’s alleged aggression, "did not [c]ite Lieutenant Childs for submitting a false and[/]or misleading Incident Report.” Instead she, like Officer Childs, adopted the narrative that Inmate A had been “combative.”

11 . Although these "Findings” highlight that Officer Childs "submitted a false and[/]or misleading Incident Report” in relation to whether Inmate A was restrained at the time Officer Childs sprayed him in the face with a chemical agent, the body of the report as a whole makes clear that all three officers told a similar story about Inmate A acting aggressively and assaulting Officer Childs and the dog that the OIA determined was untrue.

12 .Discipline of DOC employees is governed by the Comprehensive Merit Personnel Act, D.C.Code §§ 1-616.51 to -.54 (2012 Repl.). Implementing regulations require the agency to give such notice and the employee to acknowledge it. See 6-B DCMR § 1601.1 (2008) (covered employees may not be subject to discipline, including "reductions] in grade” "except as provided in this chapter”); 6-B DCMR § 1614.1 (2004) ("employee shall be given a notice of final decision in writing ... informing him ... of the reasons” for corrective or adverse action); 6-B DCMR § 1614.4 (2004) (notice shall be delivered to employee "on or before the time the action is effective”); 6-B DCMR § 1614.5 (2004) (employee "shall be asked to acknowledge its receipt”).

13 . To a reader unfamiliar with the full report, it is not obvious that the five-page excerpt is incomplete. No text is cut off mid-sentence; rather, the bottom of the fifth page reaches what might be considered a stopping point after reproducing Officer Childs's account of Inmate A's alleged assault from his Incident Report.

14 . Counsel for Mr. Woods played the lead role in advocating for disclosure of the OIA Final Report to the defendants.

15 . At the pretrial hearing, the trial court briefly considered ordering the government to disclose the report subject to a protective order, but it abandoned this option at the behest of the prosecution. The government asserted that, with respect to the other corrections officers named in the report, “there are employment issues,” and that it did not “believe that there is anything in the report that wasn’t disclosed in the motion [to limit cross-examination of Officer Childs].”

16 .Mr. Morton and Mr. Vaughn initially moved for a judgment of acquittal or in the alternative a new trial arguing, among other things, that their Sixth Amendment right to confrontation had been violated when they were denied a "meaningful” opportunity to cross-examine Officer Childs about the OIA Final Report. At a February 3, 2010 status hearing on that motion, the court ordered the government to disclose the full OIA Final Report pursuant to a protective order (the very procedure the government had successfully opposed during trial). The government subsequently disclosed the ten-page Final Report to the defense, but did not attach any of the appendices, apparently because the prosecutors did not have them; they had never asked DOC for a copy of the complete OIA Final Report with all of its attachments.

17 . When asked by the court at one post-trial hearing if this Letter of Direction “summarized aspects of the report,” the government answered, "that’s correct,” even though the Letter of Direction was issued before the commencement of the OIA investigation, and even though the OIA Final Report specifically determined that the Letter of Direction was inadequate because it failed to discipline Officer Childs for making a false report. See supra note 10.

18 . By this time, the trial court had figured out that, based on the dates, Major Talley’s Letter of Direction to Officer Childs could not summarize the OIA Final Report and directed the government to provide it with more information about what happened to Officer Childs after the completion of the OIA Final Report. Presumably, this is why the government filed OIA Investigator Collins’s affidavit.

19 . The trial court also referred to the unpublished memorandum opinion affirming co-defendant Bobby Johnson’s conviction. See Johnson v. United States, No. 10-CF-205, Mem. Op. & J. (D.C. Mar. 9, 2011). The government has argued on appeal that that opinion is "instructive” if not dispositive of Mr. Morton and Mr. Vaughn's Brady claim. We disagree. Setting aside that the panel chose not to publish its disposition, the panel in that case considered a different issue (Mr. Johnson’s assertion that his right to confront Officer Childs had been violated) on a very different record (the panel never had the benefit of the post-trial revelations that were made in this case).

20 .The government has never denied that Officer Childs was demoted in connection with the Inmate A incident detailed in the OIA Final Report; it simply presented the contents of that report as something other than they are. One might try to argue that Officer Childs's demotion may have been related to only some but not all of the misconduct actually found by the OIA (e.g., the improper use of force but not the false reporting). But such an argument would be unpersuasive as Officer Childs had already been disciplined (via the Letter of Direction) for his improper use of force, and the submission of false reports— in particular reports falsely accusing an inmate of criminal conduct- — are hardly insignificant.

21 . See infra pp. 1260-61.

22 . We note that the documentation of the notice of this action that Officer Childs should have received, see supra note 12, is not in the record. But the record does include OIA Investigator Collins’s sworn statements giving the specific dates on which notice was issued and became effective and affirming his understanding that Officer Childs signed this notice, thus indicating that the requisite notice requirements were followed. We further note that in conformance with these regulations Officer Childs was given notice (which he signed) of the far less consequential (and subsequently deemed inadequate) Letter of Direction issued by Major Talley.

23 . Coincidentally, counsel for Mr. Morton made a Brady request for, inter alia, any "information which tends to show a government witness’ bias or corruption ... or which otherwise impeaches the witness’ testimony” on April 10, 2009, one day before the OIA investigation into Officer Childs’s false reporting began. But whether or not defense counsel made a Brady request, the government had an independent obligation to disclose the favorable information in its possession. See United States v. Agurs, 427 U.S. 97, 107, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976).

24 . That this information was impeaching does not diminish the government's obligation to learn of and disclose this information in a timely manner. No less than exculpatory information, the defense is entitled to make thoughtful, effective use of impeaching information in the preparation of its case. See Miller, 14 A.3d at 1112 (explaining that the opportunity for use of favorable information guaranteed by Brady is the “opportunity for a responsible lawyer to use the information with some degree of forethought” (quoting Leka, 257 F.3d at 103)). The revelation that a witness may be untrustworthy or lying may open up new investigative avenues and require the contemplation of different strategies at trial. See, e.g., Sykes v. United States, 897 A.2d 769, 777-78 (D.C.2006). Moreover, where, as here, the government disputes the meaning or value of the information it possesses, time must be afforded, well before trial, to litigate the government’s disclosure obligations. As this case demonstrates, doing so on the eve of trial puts the defense at a strategic disadvantage and compromises the court’s ability to thoughtfully assess the information at issue. See Miller, 14 A.3d at 1111 (discussing the difficulty of integrating Brady information on the eve of trial and noting that ”[t]he defense may be unable to divert resources from other initiatives and obligations that are or may seem more pressing” (quoting Leka, 257 F.3d at 101)).

25 . See also Matthews v. United States, 629 A.2d 1185, 1199-1200 (D.C.1993) (finding alleged Brady information was not "improperly withheld" where the prosecutor "explained to [the defendant] exactly what the [exculpatory] statement said”), abrogated on other grounds by Wilson-Bey v. United States, 903 A.2d 818 (D.C.2006) (en banc); Wiggins v. United States, 386 A.2d 1171, 1173 (D.C. 1978) (holding that the government complied with Brady when it disclosed grand jury testimony in a "substantially verbatim” account).

26 . The government’s summary does not mention the other two corrections officers who the OIA determined also submitted similarly false accounts of an inmate assault. But its story that the OIA investigation was solely about Officer Childs's possible improper use of force and sloppy report-writing falls apart when the full scope of the DOC investigation into the actions of three officers to frame an inmate for assault is revealed.

We further question the government's attempt to summarize the report without obtaining the 76 pages of appendices that accompanied the report, where these appendices further demonstrated that the scope of the OIA investigation was not as the government represented.

27 . Consistent with the Supreme Court’s decision in Napue, 360 U.S. at 269, 79 S.Ct. 1173, we acknowledged in Longus v. United States, 52 A.3d 836, 844 (D.C.2012), that it is “[a] bedrock principle of due process in a criminal trial ... that the government may neither adduce or use false testimony nor allow testimony known to be false to stand uncorrected.” See also Thompson v. United States, 45 A.3d 688, 691 n. 4 (D.C.2012) (acknowledging the duty to correct false testimony extends beyond perjury to testimony that is false because it is "mistaken”). Neither Mr. Morton nor Mr. Vaughn raised a Napue claim in this case, however.

28 . The government asserts that because the trial court indicated that it would be open to further argument on the subject and the defense failed to raise the issue again, the defense "arguably waived any claim that they were entitled to confront Sergeant Childs at trial with a copy of the Final Report." We see no waiver where the defense- — because of the government’s suppression of favorable information — had nothing else to offer and no other means to persuade the court that its assessment of the OIA investigation was mistaken.

29 . The materiality assessment this court conducts on appellate review is necessarily different from the materiality assessment the government can make pretrial when assessing its Brady obligations, and we reiterate that prior to trial, the government must disclose information that is "arguably” material. Boyd, 908 A.2d at 61-62; see also Miller, 14 A.3d at 1109 (acknowledging there may be "a duty of disclosure even when the items disclosed subsequently prove not to be material.” (citing Boyd, 908 A.2d at 60)).

30 . The government argues on appeal that its case was "strong even without” Officer Childs. We cannot agree that a case based solely on the identification testimony of Sergeant Harper would have been "strong.” We further note that the government's assertion that "the jury was able to evaluate whether the tapes depicted appellants,” is inconsistent with its position that it needed the lay opinion testimony of Officer Childs and Sergeant Harper to assist the jury in identifying the individuals in the recordings of the attacks. See infra part IV.A.

31 . Even assuming Officer Childs would have given the same denials, his testimony that he had not been demoted as a result of false reporting would have provided the basis for request by the defense to demand that the government correct his false or mistaken testimony. See supra note 27.

32 . This factual predicate was not undermined by the absence of any record of Officer Childs’s false reporting and demotion in his personnel file, which the trial court apparently reviewed but is not in the record on appeal. It was never established that Officer Childs's personnel file was complete and in fact the personnel regulations allow for removal of documentation of discipline from an employee's file under certain circumstances. See 6-B DCMR § 1601.7 (2008). More fundamentally, there was never any question that Officer Childs was demoted in connection with the Inmate A incident. He admitted that he was. And even though he stated that he "didn't want to” take what he represented was a "voluntary demotion,” there is no indication in the record that he ever sought to administratively challenge this discipline or the underlying determination of misconduct on which it was based.

33 . Extrinsic evidence is admissible to substantiate a claim of witness bias because bias is never a collateral issue. Martinez v. United States, 982 A.2d 789, 795 (D.C.2009). Here we agree with the government that the defense could not have introduced the OIA Final Report to pursue a line of bias cross-examination on a theory that Officer Childs was testifying for the government to curry favor in his own misconduct proceedings. Because Officer Childs identified Mr. Morton and Mr. Vaughn before his false reporting incident took place and because he was demoted before he testified at their trial, the timing does not suggest an attempt to curry favor with the government while under investigation. Cf. id. at 795 (holding that officer’s status as subject of an ongoing MPD investigation was admissible as evidence of bias). But proffering a "well-reasoned suspicion” that a witness has a motive to curry favor with the government is only one means of establishing the requisite “proper foundation” to pursue a bias line of cross-examination. See Howard v. United States, 978 A.2d 1202, 1207 (D.C.2009). "Bias may be induced by a witnesses] like, dislike, or fear of a party, or by the witnesses] self-interest.” Dawkins v. United States, 41 A.3d 1265, 1271 (D.C.2012). And this court has recognized that corruption bias, i.e., a willingness to give false testimony, is its own separately cognizable form of bias. See Lon-gus, 52 A.3d at 851-53; In re C.B.N., 499 A.2d 1215, 1219-20 (D.C.1985). Although the government argues that the defense should be limited in its materiality argument to the theory of bias it advanced at trial, we are not inclined to find waiver in light of the fact that the defense did not have the full report at trial. Our inquiry is not about what the defense did without the Brady material, but about what the defense would have done had it had the Brady material.

34 . Indeed, we are left with many questions about the government's behavior in this case, including: (1) How could the government have so misconstrued the findings of the OIA investigation as memorialized in the full OIA Final Report as ultimately unrevealing regarding Officer Childs credibility? (2) How could the government have failed to realize at trial that it had not given the court the full OIA Final report, particularly when the trial court specifically asked if the five-page copy it had in hand was the complete report? (3) How could the government have made the representations it did about the consequences of the Inmate A incident or have allowed Officer Childs to testify without qualification about his lack of notice or understanding of those consequences, in light of the information contained in OIA Investigator Collins's sworn affidavit?

But these questions ultimately go to whether the government acted in bad faith, which, as we noted at the outset, is irrelevant to the issue raised before this court: whether the government violated its constitutionally imposed disclosure obligations. See Brady, 373 U.S. at 87, 83 S.Ct. 1194; Miller, 14 A.3d at 1107.

35 . Some courts have held that where the government engages in deliberate Brady misconduct, a new trial should not be held. See Virgin Islands v. Fahie, 419 F.3d 249, 254 — 55 (3d Cir.2005) (recognizing that “dismissal for a Brady violation may be appropriate in cases of deliberate misconduct”); United States v. Kojayan, 8 F.3d 1315, 1325 (9th Cir.1993) (remanding to the trial court to consider whether to allow a new trial after a Brady violation); see also Sanders v. United States, 550 A.2d 343, 344-46 (D.C.1988) (recognizing the trial court’s "supervisory power to dismiss the indictment as a sanction for government misconduct”). We neither endorse nor foreclose such an argument by the defense on remand.

36 . “Significant bodily injury” is defined as “an injury that requires hospitalization or immediate medical attention.” D.C.Code § 22-404(a)(2) (2012 Repl.).

37 . This court has defined "serious bodily injury” as an injury that causes "a substantial risk of death, unconsciousness, extreme physical pain, protracted and obvious disfigurement, or protracted loss of impairment of the function of a bodily member, organ, or mental faculty.” Perry, 36 A.3d at 815 n. 29.

38 . In its discussion of aiding and abetting, the court instructed the jury:

With respect to the charge of aggravated assault or assault with a significant injury, regardless of whether a Defendant is an aider and abettor or a principal offender, the Government must prove beyond a reasonable doubt that the Defendant personally acted with:

(A) inten[t] to cause significant bodily injury to the complainant; or (B) [knowledge] that significant bodily injury to the complainant would result from his conduct; or (C) [awareness] of and disregard[ ] [for] the risk of significant bodily injury that his conduct created.

39 . We question whether the government presented evidence that Sergeant White suffered serious bodily injury at all. The government presented evidence that Sergeant White briefly lost consciousness following the attack, that the head injuries he incurred did not cause substantial pain, and that, although he sought medical care, he fully recovered from these injuries without medical intervention. This appears to fall well below the "high threshold of injury,” Jenkins v. United States, 877 A.2d 1062, 1069 (D.C.2005), we have set to prove aggravated assault. See, e.g., Jackson v. United States, 940 A.2d 981, 983-84 (D.C.2008) (declining to find serious bodily injury where victim was forcibly raped and beaten in the head with a hammer); Bolanos v. United States, 938 A.2d 672, 679 (D.C.2007) (declining to find serious bodily injury where victim was stabbed in arm and stomach and required surgery to repair perforation to his intestine); see also Swinton v. United States, 902 A.2d 772, 775 (D.C.2006) (explaining that in cases finding serious bodily injury, the injuries were "usually ... life-threatening or disabling. The victims typically required urgent and continuing medical treatment (and, often, surgery), carried visible and long-lasting (if not permanent) scars, and suffered other consequential damage, such as significant impairment of their faculties. In short, these cases have been horrific.”). But Mr. Vaughn has not argued that the nature of Sergeant White’s injuries preclude his conviction for aggravated assault and we need not reach this issue in light of our disposition above.

40 . The recorded footage does not show, and the government presented no other evidence regarding, the point in time when Sergeant White was brought to the floor.

41 . The government asserts that "appellant is not entitled to reversal because he cannot satisfy the third and fourth plain-error requirements,” but save one citation, its brief contains no further reference to the fourth prong of plain error review.

42 . The government argues that appellants have waived their argument on this issue by failing to raise it before the trial court and asserts that this issue must be reviewed for plain error review if it is considered at all. We see no reason to delve into a preservation analysis in light of our determination that the trial court did not abuse its discretion.

43 . The evidentiary rule in Sanders is derived from Rule 701 of the Federal Rules of Evidence, which this court has adopted in substance. See King v. United States, 74 A.3d 678, 681 n. 13 (D.C.2013).

44 . Both Mr. Morton and Mr. Vaughn fault the trial court for not making specific findings on the Sanders issue, but trial counsel never requested such findings. See Tyson v. United States, 30 A.3d 804, 806 (D.C.2011) ("[W]hen a request for special findings is not timely made, the right to such findings is generally regarded as having been waived.”). Indeed, as noted above, there is some question wheth er this issue was waived or properly preserved. See supra note 42.

45 . See Sanders, 809 A.2d at 593 n. 10 (allowing identifying testimony from defendant’s neighbor of twenty-five years, former boss, and ex-girlfriend, among others); see also, e.g., United States v. Jackman, 48 F.3d 1, 2 (1st Cir.1995) (allowing identifying testimony from defendant's ex-wife and two personal acquaintances); United States v. Stormer, 938 F.2d 759, 762 (7th Cir.1991) (allowing identifying testimony where witnesses were acquainted with defendants for several years); United States v. Borrelli, 621 F.2d 1092, 1095 (10th Cir.1980) (allowing identifying testimony from defendant's stepfather).

46 . Mr. Morton also argues that Sergeant Harper and Officer Childs should not have been permitted to testify because the government failed to establish that their testimony would be helpful. But this argument is largely a repeat of his challenge to the quality and quantity of the officers’ “personal and direct contacts” with Mr. Morton. To the extent he focuses on other factors, he identifies one in particular, "the quality of the video or photo,” that defeats his argument. Indeed, elsewhere in his brief, Mr. Morton acknowledges the recorded footage used by the government in this case was “dark, pix[e]lated, 'jerky,' and 'blurry.' ” In Sanders, we stated that testimony may be helpful if the image of the defendant in the videotape or photograph is "obscured.” Sanders, 809 A.2d at 596.