At issue is whether Delma Banks, Jr., is entitled to habeas relief for his capital-murder conviction, because of the State’s non-disclosure to Banks’ counsel of the transcript of the prosecution’s pre-trial interview of a key witness. See Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963) (addressing the suppression of favorable, material evidence).
The murder occurred in April 1980. That fall, a Texas state-court jury convicted Banks; he was sentenced to death. Legal proceedings in the nearly 30 years since have included a direct appeal, three state habeas petitions with two evidentiary hearings, two federal habeas proceedings with an evidentiary hearing, two opinions from our court, and one from the Supreme Court.
In 2004, the Court held: the State’s suppressing a key punishment-phase witness’ (Robert Farr’s) police-informant status affected “the reliability of the jury’s verdict regarding punishment”, Banks v. Dretke, 540 U.S. 668, 703, 124 S.Ct. 1256, 157 L.Ed.2d 1166 (2004); and, therefore, Banks was entitled to habeas relief for his sentence. Id. at 705, 124 S.Ct. 1256. It also held: a certificate of appealability (COA) should have been granted regarding the State’s suppressing a transcript of a key guilt-phase witness’ (Charles Cook’s) pre-trial interview with prosecutors and a law-enforcement officer (the Cook transcript); and, therefore, this matter was remanded to our court to consider whether, because of this suppression, Banks is also entitled to habeas relief for his conviction. Id. at 705-06, 124 S.Ct. 1256.
Accordingly, that same year, we rendered an opinion on an issue related to the claim for which the Court had granted the COA, Banks v. Dretke, 383 F.3d 272 (5th Cir.2004), and remanded this matter to district court “in order for it (1) to determine whether Banks’ Cook-transcript Brady claim was tried by implied consent of the parties; and (2) if it was, to decide that claim”. Id. at 281.
In this third proceeding in our court, the State contests the district court’s on-remand April 2008 Brady-based grant of habeas relief for Banks’ conviction. The district court concluded, inter alia: habeas relief should be granted because the State failed to disclose the Cook transcript to Banks for use in his trial.
*300Because Brady’s materiality prong is not satisfied, the habeas relief for Banks’ conviction is VACATED. The Court’s grant of habeas relief for Banks’ sentence is, of course, not affected by this opinion. This matter is remanded to district court for further proceedings consistent with this opinion.
VACATED IN PART and REMANDED.
I.
The facts and procedural history have been extensively discussed in numerous previous opinions. See, e.g., Banks v. Cockrell, 48 Fed.Appx. 104 (5th Cir.2002) (unpublished), rev’d, 540 U.S. 668, 124 S.Ct. 1256, 157 L.Ed.2d 1166 (2004). They are provided in great detail again, however, because proper analysis of the instant Cook-transcript Brady claim necessarily involves a careful review of them.
A.
On Friday evening, 11 April 1980, 16-year-old Richard Whitehead (the victim) and a 14-year-old female friend encountered 21-year-old Banks at a bowling alley, and agreed to give him a ride home. Departing in the victim’s distinctive, multicolored Mustang, the trio ended up drinking Coors beer together in a secluded park near Nash, Texas. Nash is approximately four miles west of Texarkana, Texas, and is located in Bowie County.
Around 11:00 p.m., the trio left the park and drove the victim’s friend home. The victim and Banks left the friend’s house in the Mustang; they briefly visited another of the victim’s friends; and, shortly before midnight, the victim and Banks left that house together in the Mustang.
A few hours later, at about 4:00 a.m. on Saturday, 12 April, two gunshots were heard coming from the part of the park where the victim and Banks had been drinking beer. On 15 April, the victim’s body was found in that portion of the park. He had been shot three times with what was later determined to be a .25 caliber Galesi pistol (one of the shots was between his eyes from very close range); his Mustang was missing; and empty cans of Coors beer were strewn about.
On Saturday morning, 12 April, Banks traveled approximately 175 miles to Dallas, Texas. He arrived by 8:30 a.m., about four and one-half hours after the gunshots had been heard. When Banks arrived in Dallas, he was driving a distinctive, multicolored Mustang.
In Dallas, on the morning Banks arrived, Charles Cook and his wife were standing outside, waiting for a bus. Banks, who did not know them, pulled up in the Mustang and asked for directions. Cook talked Banks into giving his wife a ride to work, and the three departed in the Mustang. After dropping Cook’s wife off at work, Cook and Banks continued to ride around together for much of the day; they visited some of Cook’s acquaintances; and Banks stayed the next two nights with Cook and his family at Cook’s grandparents’ house in Dallas (Cook’s house).
That Saturday, 12 April, while riding around Dallas in the Mustang, and after Cook noticed blood on one leg of Banks’ trousers, Banks said he had shot a “white boy”. That evening, Banks told Cook that he had “decided to kill the white boy for the hell of it and take his car and come to Dallas”. Cook noticed that Banks had a pistol; the next evening (Sunday), Cook took the pistol away from Banks and hid it. It was later identified as the .25 caliber Galesi murder weapon.
On Sunday, 13 April, Banks made a collect call to his mother in Texarkana, from Cook’s house; Banks’ mother urged *301him to turn himself in. Later that weekend, Banks shared this information with Cook’s neighbor, Bennie Lee Jones.
After spending Saturday and Sunday night with Cook, Banks was given money by Cook’s wife; and, on Monday, 14 April, he boarded a Greyhound bus bound for Texarkana. (Nash, where Banks lived, is near Texarkana.) Early on Tuesday morning, 15 April (the day the victim’s body was found near Nash), Cook abandoned the multi-colored Mustang in West Dallas; it was never recovered. Shortly thereafter, Cook sold Banks’ .25 caliber Galesi pistol, along with some jumper cables and tools from the Mustang, to his neighbor, Jones.
Later that month, Banks telephoned Cook twice, in an attempt to recover his (Banks’) pistol. On 23 April, Banks returned to Dallas. He traveled with two acquaintances, Farr and Marcus Jefferson; and, unknown to Banks, he was followed by law-enforcement personnel. Farr was Banks’ girlfriend’s brother-in-law. (As noted, he was also a paid police informant; this suppressed status was the basis for the Court’s granting habeas relief for Banks’ sentence.) Marcus Jefferson was Banks’ girlfriend’s brother.
Upon arriving in Dallas, Banks drove around, looking for Cook’s house. Upon locating it, Banks went to the door of the house, while Farr and Marcus Jefferson waited in the vehicle; Banks asked Cook for his (Banks’) gun; he returned to the vehicle; and they departed. In the vehicle, Banks told Farr and Marcus Jefferson: Cook didn’t have his (Banks’) gun because he had given it to someone else; and Cook, instead, gave Banks a .22 caliber pistol.
Departing from Cook’s house, Banks, still traveling with Farr and Marcus Jefferson, apparently got lost trying to find the way back to the highway. The still-trailing law-enforcement personnel, after observing several traffic violations, initiated a traffic stop of Banks’ vehicle. A .22 caliber pistol was recovered from the vehicle; Banks was arrested; and, after his arrest, Farr and Marcus Jefferson were released.
On 24 April, the day after Banks had attempted to retrieve his pistol from Cook, law-enforcement personnel: visited Cook; asked for Banks’ pistol; and were taken, by Cook, to Jones’ house. There, Jones returned the .25 caliber Galesi pistol to Cook, who provided it to the police. (This was the pistol later determined to be the murder weapon.)
B.
Banks was tried for capital murder in the fall of 1980. Between the April 1980 murder and Banks’ trial, Cook provided an affidavit/statement on 24 April 1980 to law-enforcement personnel; and, prior to trial, he was interviewed by a law-enforcement officer and prosecutors, after which a transcript was made (the earlier-referenced Cook transcript).
The 24 April affidavit, provided the day after Cook was visited by Banks, Farr, and Marcus Jefferson, was given on the day Cook provided the murder weapon to the police. This statement consisted of a typed, two-page, single-spaced sworn affidavit (based on an initial, hand-written statement by Cook at the police station that same day), describing his interactions with Banks during the weekend following the murder; and an addendum, consisting of a one-paragraph sworn affidavit, confirming that the pistol reclaimed from Jones was Banks’ pistol. At Banks’ trial, his counsel received and reviewed the April 1980 affidavit prior to cross-examining Cook.
The pre-trial Cook-transcript at issue in this appeal is an undated, 38-page, typed *302transcript, not signed by Cook. It is the transcript of the pre-trial interview of Cook, conducted by members of the prosecution team and a law-enforcement investigator. It was not provided to Banks’ counsel for the 1980 trial. It was, instead, produced pursuant to a discovery order in conjunction with Banks’ 1999 federal habeas evidentiary hearing. See Banks, 48 Fed.Appx. 104. The total document, as produced to Banks, was a 75-page packet consisting of one complete copy of the 38-page Cook transcript, and a nearly-complete second copy of that transcript (pages 24 through 38, followed by pages 2 through 23, and missing page 1). The complete copy has some underlining and other handwritten notations, made by the lead prosecutor at Banks’ trial (who died before the 1999 federal court evidentiary hearing).
1.
As discussed, on 24 April 1980, the day he obtained the murder weapon for the police, Cook also provided them with the two-page, single-spaced, typed statement (and a one-paragraph addendum), both in the form of sworn affidavits. Among other things, Cook, in that April 1980 affidavit, stated: he noticed blood on Banks’ leg while they were riding around in the multicolored Mustang; Banks said “he got in to [sic] it with a white boy”; Banks said he shot the “white boy” and “the bitch is dead”; Banks, later in the evening, talked about “killing [the white boy] and taking his car to Dallas just for the hell of it”; Cook took Banks’ pistol away from him; Cook later sold it to his neighbor, Jones, along -with booster cables and a small tool box taken from the Mustang; and Cook further assisted Banks by abandoning the Mustang in Dallas, after Banks had returned to Texarkana by bus.
This April 1980 affidavit is consistent with, and very closely matches, Cook’s trial testimony. (That testimony is described in detail, infra.) At trial, this affidavit was provided to, and reviewed by, Banks’ counsel at the conclusion of Cook’s direct examination and prior to his being cross-examined. Banks’ counsel did not object to the manner in which it was produced by the State, such as not receiving it before trial. He requested, and was given, time to review it, with the jury in recess, after noting to the court: “[I]t’s a lengthy statement. It’s going to take me a few minutes to read it”.
2.
As noted, before Banks’ trial (apparently during the prior week), Cook was interviewed by members of the Banks trial’s prosecution team, along with Investigator Huff of the Bowie County (Texas) Sheriffs Department. (Nash is located in that county.) A transcript was made of the interview- — the earlier-described “Cook transcript”. That undated, unsigned, 38-page transcript forms the basis for the instant Brady claim. The transcript begins with an Assistant District Attorney (interviewer) stating to Cook:
What we want to do first is go through it and let you just tell us best [sic] you can remember everything you know about Delma Banks and about this case and we’ll try not to stop you any more than we have to — let you just kinda [sic] run through it as you remember it, then go back and pick up some specifics.
The transcript’s first 18 pages consist largely of Cook’s description of the events at issue, beginning with the moment he met Banks and continuing through 24 April 1980, when he provided the police with the affidavit; the interviewer’s interruptions in this portion of the transcript are minimal and clarifying in nature. The balance of the transcript consists primarily of the interviewer’s probing and testing *303Cook’s story; pointing out inconsistencies in his statements; pressing him for further details; and questioning him about his criminal history and legal problems.
Cook’s description of the events at issue, as stated in the transcript, is quite consistent with his April 1980 affidavit. In the transcript, Cook stated, inter alia: he noticed blood on Banks’ leg while they were riding around in the multi-colored Mustang; Banks said “I got into it with a white boy” and “I think I knocked him off’; and Banks, later in the evening, talked about killing the “white boy” in a park while drinking beer after “deciding] he wanted his car”. Cook further stated: he took Banks’ pistol away from him; he later sold it to his neighbor, Jones, along with “battery cables” and a small tool box taken from the Mustang; and he further assisted Banks by abandoning the Mustang in Dallas, once Banks had returned by bus to Texarkana.
As noted, the prosecution did not provide the Cook transcript to Banks’ attorney prior to, or during, trial. (During the June 1999 federal habeas evidentiary hearing, another Assistant District Attorney (the ADA), one of the prosecutors at the 1980 trial, testified that he did not believe the State had been required to produce this pre-trial interview to Banks’ attorney. See Transcript of Evidentiary Hearing at 45-46, Banks v. Johnson, No. 5:96-CV-353 (E.D. Tex. 7-8 June 1999).) For the trial, Banks and his attorney apparently were not aware of the Cook transcript; and, accordingly, unlike Cook’s April 1980 affidavit, the transcript was not available for trial use by Banks’ attorney, including for cross-examining Cook.
C.
Banks was tried in Texas state court for capital murder during the fall of 1980. During the trial’s guilt and punishment phases, the Rule was invoked; accordingly, witnesses were instructed not to communicate with each other and were sequestered until called to testify.
1.
The two-day guilt phase began on 29 September 1980, the jury’s having been selected the prior week. Cook’s importance as a witness was noted in the State’s opening statement. (Banks did not make one.) The State called 16 witnesses, including Cook, Farr, and Marcus Jefferson. (The guilt-phase testimony is described in detail in part II.B., concerning the instant Brady claim.)
On 30 September 1980, the State rested; Banks rested without presenting a defense. In closing arguments, each side focused extensively on Cook’s testimony and credibility. The jury retired at approximately 8:00 p.m. and returned its verdict about three hours later, finding Banks guilty of capital murder.
2.
On 1 October 1980, the punishment phase began. The State called two witnesses: Farr and Vetrano Jefferson (the older brother of the above-referenced Marcus Jefferson).
As noted, Farr’s suppressed-informant status provided the basis for punishment-phase habeas relief. Farr testified: he, Banks, and Marcus Jefferson drove to Dallas so that Banks could reclaim his pistol to commit armed robberies.
Vetrano Jefferson was Banks’ wife’s brother, and, as noted, the older brother of Marcus Jefferson. Vetrano Jefferson testified: one week before the victim’s death, Banks struck him (Vetrano Jefferson) with a pistol and threatened to kill him.
*304The defense then called ten witnesses, including Banks. (Cook was not called as a witness during this phase.)
The jury found the requisite special issues, and the death penalty was imposed.
D.
In 1982, on direct appeal, the Texas Court of Criminal Appeals affirmed the conviction and sentence. Banks v. State, 643 S.W.2d 129 (Tex.Crim.App.1982). The Supreme Court denied review. Banks v. Texas, 464 U.S. 904, 104 S.Ct. 259, 78 L.Ed.2d 244 (1983).
E.
Banks next filed three state habeas petitions, all of which were denied. Evidentiary hearings were conducted for the first and second petitions.
The first raised, inter alia, a Swain-based jury-discrimination claim and a sufficiency-of-the-evidence claim. Ex parte Banks, No. 13,568-01 (Tex.Crim.App.1984) (unpublished); see also Swain v. Alabama, 380 U.S. 202, 85 S.Ct. 824, 13 L.Ed.2d 759 (1965), overruled by Batson v. Kentucky, 476 U.S. 79, 106 S.Ct. 1712, 90 L.Ed.2d 69 (1986). The second raised, inter alia, another sufficiency-of-the-evidence claim. Ex parte Banks, 769 S.W.2d 539, 540 (Tex. Crim.App.1989). The third raised, inter alia, claims for ineffective assistance of counsel, jury discrimination, withholding material impeachment evidence on Cook and Farr, and insufficient evidence supporting future dangerousness. Ex parte Banks, No. 13,568-03 (Tex.Crim.App. 11 Jan. 1996) (unpublished).
F.
In March 1996, Banks filed a federal habeas petition, after having contacted Farr and Cook. As described in our 2002 opinion, Banks’ communications with Farr and Cook took place roughly two decades after Banks’ conviction and were claimed to have the following results:
Farr revealed he had been a paid informant. And, Cook stated: significant portions of his testimony were false and given under pressure from authorities; Deputy Huff [the Bowie County investigator for the murder] and others assured him that, in exchange for favorable testimony, a pending charge in Dallas County would be dismissed; and his testimony had been rehearsed on several occasions.
Banks, 48 Fed.Appx. 104.
1.
In his federal petition, in claiming numerous violations of his constitutional rights, Banks raised: failure to grant change of venue; at least 11 instances of ineffective assistance of trial counsel; improper district-attorney contact with prospective jurors and their being improperly excluded, based upon their responses and race; improper reference to Banks’ decision not to testify; improper jury instructions related to burden of proof and Banks’ intent; prosecutorial misconduct; improper admission of prior-offense evidence during the punishment phase; failure to prove Banks posed a threat of future violence; arbitrary and discriminatory imposition of the death penalty; unconstitutionality of Texas’ death-penalty statute; and ineffective assistance of appellate counsel. See Banks v. Johnson, No. 5:96-CV-353 (E.D. Tex. 11 May 2000) (unpublished report and recommendation of the magistrate judge).
The magistrate judge to whom the matter was referred for a report and recommendation granted Banks limited discovery and an evidentiary hearing on, inter alia, his prosecutorial-misconduct Brady claims (including: Farr’s informant status; and whether, with prosecutors, Cook had a *305deal for his testimony). At this point, pursuant to a discovery order, the Bowie County District Attorney’s office produced documents related to Banks’ trial; the Cook transcript was among them. Subsequently, at the June 1999 evidentiary hearing, the ADA, one of the prosecutors at the 1980 trial, confirmed: at trial, his earlier-referenced co-counsel (who had died before the evidentiary hearing) was in possession of the transcript and several pages of handwritten notes; and they were not disclosed to Banks prior to, or during, the trial. See Banks, 48 Fed.Appx. 104.
Cook testified at the June 1999 evidentiary hearing. His testimony sharply contradicted the State’s. Among other things, he claimed: his April 1980 affidavit was, in many respects, incomplete and untruthful; and portions of his trial testimony were untruthful. See id.
Essentially pursuant to the recommendations of the magistrate judge following the June 1999 evidentiary hearing, the district court concluded, inter alia, that Banks was entitled to habeas relief with respect to his death-penalty sentence, because Farr’s informant status had not been disclosed by the State; and was not entitled to habeas relief with respect to his conviction. Banks, at this point, urged two Cook-related Brady claims.
One was the instant Cook-transcript Brady claim. As our 2002 opinion noted: “The district court refused to consider the Brady claim concerning the [Cook] transcript”. Id. (emphasis omitted). Although Banks did not expressly raise this claim in his habeas petition, he included allegations pertaining to it in his proposed findings and conclusions, objection to the magistrate judge’s report and recommendation, and Rule 59 motion. Banks’ Rule 59 motion, which sought to amend the judgment, inter alia, to include the Cook-transcript Brady claim, was denied by the district court. See id.
The other Brady claim was that Cook had a deal with prosecutors to give false testimony at trial in exchange for the dismissal of a pending, unrelated arson charge. The district court, adopting the magistrate judge’s recommendations, was unpersuaded by this deal-for-Cook’s-testimony Brady claim because, inter alia, the arson had not yet been committed when Cook provided his sworn April 1980 affidavit to the police. See Banks v. Johnson, No. 5:96-CV-353 (E.D. Tex. May 11, 2000) (unpublished report and recommendation of the magistrate judge); Banks v. Johnson, No. 5:96-CY-353, 2000 WL 35482430 (E.D.Tex. Aug. 18, 2000) (unpublished district court opinion and order), rev’d, 48 Fed.Appx. 104 (5th Cir.2002).
Because Banks’ federal petition was filed prior to the 1996 effective date of the Antiterrorism and Effective Death Penalty Act (AEDPA), that Act was not applicable to the claims raised in his federal habeas petition. E.g., Banks, 540 U.S. at 687 n. 9, 124 S.Ct. 1256 (citing Lindh v. Murphy, 521 U.S. 320, 336-37, 117 S.Ct. 2059, 138 L.Ed.2d 481 (1997)). Banks was required, however, to obtain a certificate of appealability (COA), pursuant to AEDPA, to appeal any denied claim, including the instant Cook-transcript Brady claim. Banks, 540 U.S. at 705, 124 S.Ct. 1256; see Green v. Johnson, 116 F.3d 1115, 1120 (5th Cir.1997); 28 U.S.C. § 2253(c)(1)(A).
The district court denied a COA for both Cook Brady claims. Accordingly, among other issues presented to our court, Banks requested a COA for each of those Brady claims.
2.
On appeal in 2002, our court reversed the district court in part, denying, inter alia, habeas relief for the sentence, con*306cerning the Farr Brady claim. As had been done by the district court, a COA for both Cook Brady claims was denied. Banks, 48 Fed.Appx. 104. As a result, they were not among the numerous claims considered on the merits.
In denying a COA for the Cook-transcript Brady claim, we held, inter alia: the district court had correctly refused to consider this claim, because it was not raised in Banks’ habeas petition; and Banks never sought leave to amend his petition to add this claim. Id. Regarding this issue, Banks had urged in our court that, during the federal habeas evidentiary hearing, the Cook-transcript Brady claim had been tried by “express or implied” consent, pursuant to Federal Rule of Civil Procedure 15(b) (providing, inter alia: issues not raised in the pleadings, but litigated by implied consent of the parties, are treated as if they had been pleaded). We determined, however, that Rule 15 did not apply to that evidentiary hearing. Id.
3.
Following our 2002 denial of all habeas relief, the Supreme Court granted review, Banks v. Cockrell, 538 U.S. 977, 123 S.Ct. 1784, 155 L.Ed.2d 665 (2003); and it subsequently reversed part of our 2002 holdings. Banks, 540 U.S. at 705, 124 S.Ct. 1256. The Court reversed our denial of Banks’ Farr Brady claim, holding, instead, that Banks was entitled to habeas relief for his sentence, and reversed our denial of a COA for Banks’ Cook-transcript Brady claim. Id. at 689, 124 S.Ct. 1256.
With respect to the Cook-transcript Brady claim, and addressing our holding that this claim was barred from consideration because it was not raised in the federal habeas petition, including our holding Rule 15 inapplicable under the circumstances, the Court'noted: “while AEDPA forbids a finding that exhaustion had been waived unless the State expressly waives the requirement, 28 U.S.C. § 2254(b)(3), under pre-AEDPA law [applicable to Banks’ petition], exhaustion and procedural default defenses could be waived based on the State’s litigation conduct”. Id. at 705, 124 S.Ct. 1256. “At least as to the application of Rule 15(b), this case surely fits [the criteria for obtaining a COA]. A [COA], therefore, should have issued.” Id. Accordingly, the Court remanded this matter to our court for further proceedings.
4.
On remand, we held, inter alia, that, “on this record, Rule 15(b) applies to the Cook-transcript Brady claim as addressed in Banks’ evidentiary hearing”. Banks, 383 F.3d at 280 (emphasis omitted). We further stated, however, that “[w]e decline ... to decide in the first instance whether that Brady claim was tried by implied consent of the parties .... [W]e conclude that remand to the district court is required in order for it (1) to determine whether [that] claim was tried by implied consent of the parties; and (2) if it was, to decide that claim”. Id. at 281 (emphasis in original).
5.
For the August 2004 remand from our court, the district judge was the judge who had ruled on the habeas petition in 2000. Likewise, the magistrate judge to whom the Brady claim was referred on remand, in order to make a report and recommendation, was the magistrate judge who had: conducted the federal habeas evidentiary hearing in 1999; and provided a report and recommendation in 2000.
In March 2006, the magistrate judge provided a report and recommendation for the two remanded issues. Banks v. Dretke, No. 5:96-CV-353, 2006 WL 4914890 (E.D.Tex.Mar. 23, 2006) (unpublished). *307The magistrate judge recommended: the Cook-transcript Brady claim was litigated by implied consent of the parties; but the claim did not warrant habeas relief.
In April 2008, the district court adopted the magistrate judge’s implied-consent recommendation; but it disagreed with the substantive recommendation, concluding, instead: the Cook-transcript was, inter alia, material under Brady, and, therefore, Banks is entitled to habeas relief for his conviction. Accordingly, regarding Banks’ conviction and sentence, the district court ordered: “A writ of habeas corpus shall issue ..., ordering [the State] to release Banks from custody unless the State of Texas, within 240 days from [1 April 2008], commences new trial proceedings against Banks”. Banks v. Quarterman, No. 5:96-CV-353 (E.D. Tex. 1 Apr. 2008) (unpublished order and judgment); see also Banks v. Quarterman, No. 5:96-CV-353, 2008 WL 906716 (E.D. Tex. Apr. 1, 2008) (unpublished memorandum opinion). That judgment was stayed pending appeal.
II.
At issue are: whether, pursuant to Rule 15(b), the Cook-transcript Brady claim was litigated by implied consent; and, if it was, whether that claim should be granted. In the light of our denial of the Brady claim, infra, we could simply assume arguendo that the claim was litigated by implied consent. Instead, we decide this implied-consent issue in order to provide a ruling on both issues, should further review be requested.
A.
Presented at the June 1999 federal habeas evidentiary hearing was, inter alia, the Brady-based claim, raised in the petition, essentially asserting Cook had a deal with prosecutors for his testimony. During this hearing, however, Banks’ counsel interjected suppression of the transcript of a pre-trial interview with Cook. As noted, that claim had not been raised in the petition.
As discussed, following our remand in 2004, the district court held the issue was litigated by consent. Banks, 2008 WL 906716, at *1. Its Rule 15(b) ruling is reviewed for abuse of discretion. E.g., Smith v. EMC Corp., 393 F.3d 590, 595 (5th Cir.2004). That Rule provides, inter alia:
When an issue not raised by the pleadings is tried by the parties’ express or implied consent, it must be treated in all respects as if raised in the pleadings. A party may move — at any time, even after judgment — to amend the pleadings to conform them to the evidence and to raise an unpleaded issue. But failure to amend does not affect the result of the trial of that issue.
Fed.R.Civ.P. 15(b)(2). In deciding whether an issue was litigated by implied consent, the court considers “[1] whether the parties recognized that the issue entered the case at trial, [2] whether the evidence supporting the issue was introduced at trial without objection, and [3] whether a finding of trial by consent would prejudice the opposing party”. Smith, 393 F.3d at 596.
As discussed, the district court adopted the magistrate judge’s implied-consent recommendation and held the Cook-transcript Brady claim, although not raised in Banks’ petition, was litigated by such consent. Banks, 2008 WL 906716, at *1. It held: “[1] the questioning during the [June 1999 federal habeas] evidentiary hearing put the [State] on notice of the Cook[-t]ranscript Brady claim, ... [2] there was no objection to the introduction of the transcript, and ... [3] the [State] *308had a fair opportunity to litigate the issue”. Banks, 2008 WL 906716, at *1.
Regarding whether the State had notice of the claim — an issue that tangentially affects all three prongs of the implied-consent analysis — the district court, id., adopted the magistrate judge’s following recommendation:
[T]he crux of the questioning surrounding the transcript related to whether Cook was instructed how to testify at trial .... The transcript and testimony about the transcript at the federal evidentiary hearing squarely brought to light evidence that Cook was coached by law enforcement. If the [State] did not have notice of [Banks’] Cook-transcript Brady claim prior to the hearing, [it] certainly should have reasonably believed that the claim was presented once defense counsel introduced the transcript and began examining [the ADA],
Banks, 2006 WL 4914890, at *3-4.
The State maintains the implied-consent ruling was an abuse of discretion. It essentially contends that the requisite three factors for such consent are not satisfied.
Banks, of course, maintains otherwise. Accordingly, he contends there is no reason to disturb the district court’s ruling that the Brady claim based on the impeachment value of the Cook transcript was both recognized and litigated by the State at the federal habeas evidentiary hearing, and, therefore, tried by implied consent.
1.
For the implied-consent’s recognition prong, the ADA’s evidentiary-hearing testimony demonstrates there was sufficient questioning pertaining to the Cook transcript, and to the Brady obligation to disclose it, to have made the State aware that an unpleaded claim, separate and distinct from the deal-for-Cook’s-testimony Brady claim, had been inserted in the proceedings. Among other things, the evidentiary-hearing transcript contains the following exchanges with the ADA, exploring the State’s obligation to disclose the Cook transcript:
Q: Now if that document contains— well, was that document provided to [Banks’ counsel] prior [to] trial?
A: According to Texas procedure he’s not entitled to that document.
Q: Well, how about if that witness statement contains exculpatory material or impeaching material?
A: Well, when you say impeaching material, I’m not sure that’s as easily qualified as exculpatory material. I mean, you know, you can impeach somebody with anything.
Q: If that document contains statements that were inconsistent with the testimony that you expected Mr. Cook to give at trial, were you obligated to turn that over to [Banks’ counsel] prior to trial pursuant to Brady v. Maryland?
A: How would I know that? I mean, how would you know what a person was going to testify to at trial ....
Transcript of Evidentiary Hearing at 45-46, Banks v. Johnson, No. 5:96-CV-353 (E.D. Tex. 7-8 June 1999) (emphasis added).
2.
With respect to the implied-consent’s second prong, it is undisputed that the State did not object to the introduction of the transcript, or to its use by Banks’ counsel to show that, according to Banks’ counsel, Cook had been coached. See Banks, 2008 WL 906716, at *4. The State *309urges: its failure to object is excusable because the transcript is also relevant to the raised deal-for-Cook’s-testimony Brady claim; and, therefore, it did not object because it expected the transcript to be used for that claim.
3.
Finally, with respect to the implied-consent’s third prong, whether the State had a fair opportunity to litigate the Cook-transcript Brady claim so that there was no prejudice to the State, the evidentiary hearing spanned two days and involved 15 witnesses. Notably, Cook testified after the ADA; and Cook’s testimony elaborated upon the “coaching” issues prompted by the Cook transcript. Along that line, Cook was asked numerous leading questions by Banks’ counsel related to whether his testimony had been coached, including the following:
Q: They made it very clear that this was a very — that this case was very important, that I would testify as they wanted me to, and that I would spend the rest of my life in prison, if I did not.
A: That’s true.
Q: They repeatedly told me how to say things and insisted repeatedly that I would testify about events that were not true.
A: That’s true.
Q: ... That [Banks] wanted to know what it was like to kill a white person. Is that true or false?
A: I was told to say that.
Q: ... [TJhat he wanted to kill someone for the hell of it.
A: I was told to say that.
Transcript of Evidentiary Hearing at 144, 150, Banks v. Johnson, No. 5:96-CV-353 (E.D. Tex. 7-8 June 1999).
Following the hearing, the magistrate judge made this recommendation: “The record shows that the [State] had a fair opportunity to present evidence that Cook was not coached and to explain what transpired during Cook’s pre-trial interview”. Banks, 2006 WL 4914890, at *4. The district judge adopted that recommendation. Banks, 2008 WL 906716, at *1.
Having performed the three-pronged Rule 15(b) analysis, we hold the district court did not abuse its discretion in ruling that the Cook-transcript Brady claim was litigated by implied consent. (The Dissent, of course, does not dispute this holding.) Accordingly, we turn to the merits of the Cook-transcript Brady claim.
B.
As discussed, the district judge rejected the magistrate judge’s recommendation that the Cook transcript was not material under Brady. Therefore, the district judge ruled, based upon this claim, that Banks is entitled to habeas relief for his conviction. Id. at *2, *6.
“[F]or ‘a habeas corpus appeal, we review the district court’s findings of fact for clear error and review its conclusions of law de novo’ ”. Henderson v. Quarterman, 460 F.3d 654, 659 (5th Cir. 2006) (quoting Martinez v. Johnson, 255 F.3d 229, 237 (5th Cir.2001)). (For such review, under the pre-AEDPA law applicable here, we “generally accord[ed] a presumption of correctness to any state court factual findings”. Mann v. Scott, 41 F.3d 968, 973 (5th Cir.1994). Of course, because the instant Brady claim was not at issue in state court, no presumption of correctness is involved.) Claimed Brady violations are mixed questions of law and fact, reviewed de novo. Bower v. Quarterman, 497 F.3d 459, 466 (5th Cir.2007).
*310 “Brady claims” serve to protect the due-process rights of criminal defendants. As the Court has explained:
Under the Due Process Clause ..., criminal prosecutions must comport with prevailing notions of fundamental fairness. We have long interpreted this standard of fairness to require that criminal defendants be afforded a meaningful opportunity to present a complete defense. To safeguard that right .... [a] defendant has a constitutionally protected privilege to request and obtain from the prosecution evidence that is either material to the guilt of the defendant or relevant to the punishment to be imposed. Even in the absence of a specific request, the prosecution has a constitutional duty to turn over exculpatory evidence that would raise a reasonable doubt about the defendant’s guilt.
California v. Trombetta, 467 U.S. 479, 485, 104 S.Ct. 2528, 81 L.Ed.2d 413 (1984) (internal citations omitted).
Accordingly, a Brady violation can occur even if the evidence was withheld in good faith. See Brady, 373 U.S. at 87, 83 S.Ct. 1194. In short, the “violation occurs when the [State] fails to disclose evidence materially favorable to the accused”. Youngblood v. West Virginia, 547 U.S. 867, 869, 126 S.Ct. 2188, 165 L.Ed.2d 269 (2006). As the Court stated in 2004, when addressing the merits of the punishment-phase Farr Brady claim and whether Banks should have been granted a COA for his guilt-phase Cook-transcript Brady claim:
Strickler v. Greene, 527 U.S. 263, 281-282[, 119 S.Ct. 1936, 144 L.Ed.2d 286] (1999) ... [sets out] the three components or essential elements of a Brady prosecutorial misconduct claim: “The evidence at issue must be favorable to the accused ...; that evidence must have been suppressed by the State ...; and prejudice must have ensued” .... [C]oincident with the third Brady component (prejudice), prejudice ... exists when the suppressed evidence is “material” for Brady purposes.
Banks, 540 U.S. at 691, 124 S.Ct. 1256 (internal citations omitted).
Suppressed evidence is material for Brady purposes “ ‘if there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different’ ”. Strickler, 527 U.S. at 280, 119 S.Ct. 1936 (quoting United States v. Bagley, 473 U.S. 667, 682, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985)) (emphasis added). Along this line, “a ‘showing of materiality does not require demonstration by a preponderance [of the evidence] that disclosure of the suppressed evidence would have resulted ultimately in the defendant’s acquittal’”. Youngblood, 547 U.S. at 870, 126 S.Ct. 2188 (quoting Kyles v. Whitley, 514 U.S. 419, 434, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995)). Instead, the reversal of a conviction “require[s] ... a ‘showing that the favorable evidence could reasonably be taken to put the whole case in such a different light as to undermine confidence in the verdict’ Youngblood, 547 U.S. at 870, 126 S.Ct. 2188 (quoting Kyles, 514 U.S. at 435, 115 S.Ct. 1555) (emphasis added).
On remand in district court, the State did not contest the suppression prong; instead, it urged Brady’s remaining prongs were not satisfied. Banks, 2008 WL 906716, at *2. In response, the district court ruled: “the Cook pre-trial transcript is both favorable and material under Brady”, id. at *6; and, accordingly, Banks “satisfied his burden of showing a reasonable probability that there would have been a different result in this case had the transcript been disclosed to the defense”, id. (emphasis added). The State’s likewise *311not contesting here Brady’s second prong (suppression), we turn to the other two. Pursuant to our de novo review, although Brady’s first prong (whether document favorable) has been satisfied, its third (materiality) has not.
1.
Regarding whether the withheld Cook transcript was favorable, Banks essentially urges it could have been used at trial to impeach Cook. Impeachment evidence “falls within the Brady [disclosure] rule” because “[s]uch evidence ... may make the difference between conviction and acquittal”. Bagley, 473 U.S. at 676, 105 S.Ct. 3375. Moreover, any evidence that may have affected the jury’s verdict, including impeachment evidence, satisfies Brady's first prong. Id.
As discussed in more detail infra, the Cook transcript could have been used to attempt to impeach Cook with respect to his trial testimony that, inter alia, he had not spoken pre-trial to anyone about his testimony. Accordingly, the Cook transcript clears the Brady favorable vel non threshold.
2.
Turning to Brady’s third prong, we must consider whether “the suppressed evidence is ‘material’ ”. Banks, 540 U.S. at 691, 124 S.Ct. 1256 (quoting Strickler, 527 U.S. at 282, 119 S.Ct. 1936). Again, it is material “if there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different”. Bagley, 473 U.S. at 682, 105 S.Ct. 3375 (emphasis added). As also discussed, a reasonable probability of a different result is shown when the suppression of evidence “undermines confidence in the outcome of the trial”. Kyles, 514 U.S. at 434, 115 S.Ct. 1555 (quoting Bagley, 473 U.S. at 678, 105 S.Ct. 3375).
Toward that end, the Supreme Court has emphasized four aspects of the Brady materiality inquiry. First, a materiality showing does not require demonstration by a preponderance of the evidence that disclosure of the suppressed evidence would have resulted in acquittal; second, the materiality inquiry is not a sufficiency-of-the-evidence test; third, once materiality is established, harmless error analysis has no application; and fourth, materiality must be assessed collectively, not item by item. Kyles, 514 U.S. at 434-36, 115 S.Ct. 1555.
For the instant Brady claim, we can consider only the Cook transcript, and the effect, if any, that its suppression had on the jury trial. Concomitantly, it is important to emphasize what we cannot ■consider.
We cannot reconsider already-disposed-of habeas issues, including, inter alia, sufficiency-of-the-evidence challenges, ineffective-assistance-of-eounsel claims, and Brady-based allegations that Cook had a plea-bargain-type deal with prosecutors. Along this line, the Dissent at 334 and 342-44 erroneously spends pages analyzing what it describes as: “The Farr Brady evidence”.
As the Supreme Court noted, the district court’s Farr-based grant of habeas relief was only for the punishment phase of the trial; in the Court’s own words, it was “solely with respect to the capital sentence”. Banks, 540 U.S. at 689, 124 S.Ct. 1256. The Court’s writ of certiorari pertaining to the Farr Brady claim was limited solely to considering the “Farr Brady claim as it trains on [Banks’] death sentence”, id.; in other words, a guilt-phase Farr Brady claim was never before the Court. See Petition for Writ of Certiorari, Banks v. Cockrell, 2002 WL 32135512 *312(granted only with respect to “Suppression of Impeachment Evidence of Key Penalty Witness Farr”, “Ineffective Assistance of Counsel at Penalty Phase”, and “Suppression of Impeachment Evidence of Key Guilt Phase Witness Cook” (emphasis added)). (The one requested issue for which the writ was not granted concerned claimed improper exclusion of minority jurors. See Banks, 540 U.S. at 689, 124 S.Ct. 1256.) In short, the Dissent’s reintroducing a guilt-phase Farr Brady claim is completely improper because that claim is an already-disposed-of habeas claim.
We note, as well, that the district court, in its 2008 opinion, clearly understood that Banks’ guilt-phase habeas claim rests only on the suppression of the Cook transcript. In the district court’s 2000 opinion, issued by the same district judge as the 2008 opinion, the court held Banks entitled to a Farr-based grant of habeas relief for the punishment phase, but further held Farr’s suppressed informant status immaterial to the guilt phase. Banks, 2000 WL 35482430, at *2. A COA was not granted for a guilt-phase Farr Brady claim. Accordingly, for the district court’s 2008 opinion: the magistrate judge, in recommending that the Cook transcript’s suppression was not material, made no mention of Farr’s guilt-phase testimony, Banks, 2006 WL 4914890; and, likewise, the district court, in holding that the suppression was material, also made no mention of Farr’s guilt-phase testimony, Banks, 2008 WL 906716.
We similarly note that Banks, in his brief for this appeal, does not attempt to use Farr’s guilt-phase testimony as a justification for granting guilt-phase habeas relief. Banks’ brief never even remotely suggests that the already-disposed-of Farr Brady claim should be reconsidered; in fact, that brief barely mentions Farr, and does so only in the context of discussing this case’s background and the Court’s punishment-phase grant of habeas relief. Likewise, Banks’ counsel made no mention of Farr during oral argument. Although “the part[y] alleging a Brady violation[] ha[s] the burden of establishing all three prongs of the Brady test”, United States v. Edwards, 442 F.3d 258, 267 n. 9 (5th Cir. 2006), it is the Dissent, and not Banks, which has sua sponte reintroduced the guilt-phase Farr Brady claim into this appeal. (The Dissent does not address this well-established rule.) As discussed in further detail infra, however, we need not decide whether consideration of Farr is waived by Banks’ not addressing it, and we need not rely on our view that Farr’s suppressed informant status is not properly before our court for the instant guilt-phase materiality analysis, because, even assuming that Farr should be included in our materiality analysis, our conclusion remains the same.
In sum, our court, for this appeal, must assess the collective impact of the Cook transcript’s being suppressed. Kyles, 514 U.S. at 436, 115 S.Ct. 1555. Already-disposed-of habeas claims are not properly before us, and may not be considered in our assessment of the Cook transcript’s materiality. Similarly, the already-disposed-of guilt-phase Farr Brady claim is not properly before us, but, because it has been resurrected by the Dissent, it is discussed in greater detail below to demonstrate that its inclusion does not change our holding Brady’s materiality prong is not satisfied.
Returning to what cannot be considered in this appeal, we also cannot consider newly-raised claims which, in effect, seek to re-try this case. For example, the Dissent at 334 complains: “The State failed to introduce the Mustang car or any of its contents into evidence”; and, at 345, comments on the “unexplained absence of *313the victim’s Mustang”. Describing the State’s not introducing the Mustang into evidence as both “unexplained” and a “fail[ure]” is, of course, simply not accurate — the record shows that the State was unable to introduce the Mustang into evidence because, following its abandonment, it was never located or recovered. The more important point, however, is that, although evidence must be assessed for the purpose of placing the Cook transcript into context for our materiality assessment, we cannot allow our assessment to morph into a complete re-trial of this case, where we look back to 1980 to speculate about, inter alia, how extensive the search was for the Mustang or why the prosecutor did not introduce its contents into evidence. As we review the record for the purpose of assessing the Brady materiality of the Cook transcript, we simply cannot allow that review to expand into an unlimited re-trial of this entire case, in which we consider newly-raised claims or, as discussed supra, in which we consider already-disposed-of claims.
As another example of a newly-raised claim that Banks and the Dissent use toward, in effect, re-trying this case, Banks, in his brief and during oral argument, questions the validity of the firearms expert’s trial testimony. That expert opined: “[T]he two bullets [from the victim’s body] which were submitted by Doctor DiMaio and the single bullet [found on the ground in a mass of congealed blood which was] submitted by Deputy Huff and also the two spent cartridge cases ... had [all] been fired in the .25 caliber Galesi pistol which [Deputy] Huff submitted”. This now-challenged expert testimony was not cross-examined at trial, let alone raised as a basis for habeas relief. (The Dissent at 334 similarly notes, among other things, that the firearms expert “was not asked ... to explain the reliability or margin of error of his methodology or results” and the coroner was not asked to estimate the time of death; and, at 345, describes that expert’s opinion, which was not cross-examined, as “conclusory”.) Needless to say, expert-witness challenges, first raised almost 30 years after trial, and other challenges of that kind, are completely beyond the scope of the Cook-transcript Brady issue at hand. (And again, as discussed supra, a sufficiency-of-the-evidence habeas claim was also considered and denied long ago, Ex parte Banks, 769 S.W.2d at 540; a COA for that claim was not granted; and this type of already-disposed-of claim cannot be considered as part of the instant appeal.)
Concerning the Cook transcript itself, we cannot consider the handwritten comments and other notations appearing in the transcript. During the June 1999 evidentiary hearing, the ADA (the “second chair” prosecutor at Banks’ trial) identified the handwritten items as having been made by the District Attorney (the DA), the lead prosecutor at Banks’ trial. (As discussed, the DA died before the evidentiary hearing.) The ADA testified that he was familiar with the DA’s handwriting, and that the DA had been in possession of the Cook transcript during the trial. Transcript of Evidentiary Hearing at 45, Banks v. Johnson, No. 5:96-CV-353 (E.D. Tex. 7-8 June 1999). That the DA had made notations for himself on the transcript makes sense, of course, as he was the prosecutor who conducted the direct and re-direct examination of Cook (and all other guilt-phase witnesses) during Banks’ trial. Obviously, had the transcript been disclosed to Banks’ counsel for trial, those handwritten items would not have been included. See, e.g., Rose v. State, 427 S.W.2d 609, 611 (Tex. Crim.App.1968) (no error where defense counsel was not permitted to examine work-product statement of prosecuting attorney).
*314Moreover, there is no evidence that, pri- or to trial, Cook ever saw these handwritten items, nor, for that matter, any evidence that he even saw the transcript. Banks notes, among other things, that various passages had stars drawn next to them or were underlined or bracketed, and that some passages had brief notes written by them. For example, for a passage in which Cook stated he had been at home (in Dallas) on Friday night, 11 April 1980 (when Banks first joined the victim in Texarkana), the DA wrote: “Stayed home Fri Night”.
Banks primarily takes issue with “Do not say” being written in the margin next to a criminal-history passage in which Cook admits to having assaulted a teacher, discussed infra. While Banks urges this shows improper witness coaching, we disagree. The Cook transcript was produced from Cook’s pre-trial interview; and, again, there is no evidence suggesting that Cook ever saw the transcript, nor the notes placed on it.
Finally, we cannot consider Cook’s post-trial statements, in which — almost 20 years later — he claimed his trial testimony was untruthful. This includes his 1998 written statement, in which he — in a statement obviously drafted by someone else, as shown by comparing it with his April 1980 affidavit — disavowed, inter alia, portions of his written and signed April 1980 affidavit and his trial testimony.
We emphasize, in the preceding paragraphs, what we may not consider largely because the Dissent overreaches regarding what it means to consider materiality collectively, as required, inter alia, by Kyles, 514 U.S. at 436, 115 S.Ct. 1555. In that regard, United States v. Sipe, 388 F.3d 471 (5th Cir.2004), is instructive. While the instant appeal addresses a Brady claim for one item of suppressed evidence (the Cook transcript), Sipe addressed Brady claims for five different suppressed items. Our court considered each of the five suppressions properly before it, and concluded:
Even if none of the nondisclosures standing alone could have affected the outcome, when viewed cumulatively in the context of the full array of facts, we cannot disagree with the conclusion of the district judge that the government’s nondisclosures undermined confidence in the jury’s verdict.
Id. at 477.
Unlike Sipe, there is only one Brady claim properly before this court. As discussed, neither COAs nor a writ of certiorari were granted for a guilt-phase Farr Brady claim, a sufficiency-of-the-evidence claim, or similar claims. Our sole task is to evaluate the effect that disclosure of the Cook transcript would have had on Banks’ trial. Doing so collectively, in this context, entails: examining the usefulness of the various statements favorable to Banks that are contained in that transcript; and then considering whether all of those statements, considered together, and not item-by-item, collectively undermine confidence in the outcome of the trial’s guilt-phase. In other words, was the Cook transcript’s suppression material for Brady purposes?
a.
For our materiality analysis, we begin by reviewing all trial evidence from the two-day guilt phase. As discussed, this review is performed to determine whether there is a “reasonable probability that, had the [Cook transcript] been disclosed to the defense, the result of the proceeding would have been different”. Bagley, 473 U.S. at 682, 105 S.Ct. 3375 (emphasis added). Performing this initial analysis, it becomes apparent that, had Cook not testified at all, the jury had *315ample evidence with which to find Banks guilty beyond a reasonable doubt.
During the two-day guilt phase, the following testimony relevant to our Brady analysis was presented to the jury by witnesses called by the State. (Again, Banks did not present a defense.) For most of these witnesses, cross-examination, if any, was extremely brief.
Patricia Hicks, the victim’s lk-year-old friend, testified: she was with the victim on Friday, 11 April 1980; they were in his Mustang, which was green with an off-color hood; they first encountered Banks that evening, outside a bowling alley (in Texarkana); she did not know Banks, but he approached them and asked for a ride home; the victim agreed, and the three left the bowling alley together; they next purchased two six-packs of Coors beer; they went to a park outside of Nash, Texas, where Hicks had two beers and the victim and Banks each had about four; the three of them left the park in the victim’s Mustang at approximately 11:00 p.m.; the victim and Banks soon dropped Hicks off at her house; and the victim and Banks departed in the victim’s Mustang.
On cross-examination of Patricia Hicks: Banks’ counsel asked three questions. The first was whether there was “animosity or hard feeling” between the victim and Banks (no); the second and third were whether Hicks saw “other blacks” at the bowling alley (no).
Patty Bungardt, the victim’s 20^year-old friend, testified: she saw the victim and Banks together on Friday evening, 11 April 1980; they were in the victim’s Mustang; and they stopped at her house around 11:30 p.m. for ten or 15 minutes before departing together in the Mustang.
Patty Bungardt was not cross-examined.
Mike Fisher, who had been asleep in a house near the park in Nash, Texas, testified: he was sleeping in a house located approximately 100 yards from the park (where Hicks, the victim, and Banks had been drinking beer); and he was awakened by the sound of two gunshots at approximately 4:00 a.m. on Saturday, 12 April 1980.
On cross-examination of Mike Fisher: Banks’ counsel asked two questions, which clarified that Fisher had heard two gunshot-like noises.
Larry Whitehead, the victim’s father, testified: his son’s automobile was an extremely distinctive 1969 two-door Mustang; it was primarily light metallic green in color, with a black vinyl top; but, due to ongoing repairs, it had a primer-red hood, a black left-front fender, a light blue area near the right-front fender, and no grill. (As discussed infra, witnesses testified that Banks was later seen driving this distinctly-painted, multi-colored vehicle in Dallas. The Dissent at 334 characterizes Banks as having been seen driving merely a “similar” Mustang.)
Larry Whitehead was not cross-examined.
Investigator Willie Huff, of the Bonne County Sheriffs Department, testified: the victim’s body was found on the morning of 15 April 1980 in a park outside Nash, Texas; Huff reached the scene at approximately 10:00 a.m., and he began to conduct an investigation; at the crime scene, he recovered seven empty cans of Coors beer, two spent .25 caliber cartridge casings, and a fired bullet, which he “found in a mass of congealed blood under the [victimj’s ear, on the ground”; and he provided the recovered cartridge casings and bullet to one of the State’s firearms examiners, Allen Jones.
*316Huff further testified: his investigation determined Banks was the last person to be seen with the victim; he put Banks under surveillance; later that month, on or about 23 April 1980, he followed Banks, Farr, and Marcus Jefferson to Dallas, Texas; Banks was driving; the trio went to Cook’s house at approximately 3:00 a.m.; Banks exited the vehicle, went to the door of the house, and soon returned to the vehicle carrying a small object; after the trio left Cook’s house, the police initiated a traffic stop; and they found a .22 caliber pistol in the vehicle.
Huff further testified: he later returned to Cook’s house and asked Cook for Banks’ weapon; Cook then retrieved Banks’ .25 caliber Galesi pistol from his neighbor’s (Bennie Lee Jones’) house and handed it over to Huff; and Huff provided that pistol to one of the State’s firearms examiners, Allen Jones.
On cross-examination of Investigator Huff:: Banks’ counsel asked approximately 30 questions, which explored Huffs discovery of the spent shell casings and fired bullet, and primarily focused on their subsequent chain of custody. Later in the guilt-phase, after Huff had been re-called by the State, Banks’ counsel asked Huff three more questions. They were, essentially, whether Farr and Cook were “doper[s]”; Huff responded: “To my personal knowledge, I don’t know.”
Marcus Jefferson, Banks’ 19-year-old acquaintance, testified: he was with Banks and Farr on the April 1980 trip to Dallas, Texas; and, after they had visited Cook’s house, Banks returned to the vehicle and “said some broad had his gun, so they gave him another gun”. (Emphasis added.) (The Dissent’s attempt to discredit this testimony, including its mischaracterization of it, is discussed in detail, infra.)
Marcus Jefferson was not cross-examined.
Robert Farr, Banks’ acquaintance (not disclosed as a police informant until the federal habeas proceeding), testified: he was with Banks and Marcus Jefferson on the April 1980 trip to Dallas, Texas; and, after they had visited Cook’s house, Banks returned to the vehicle and “said that his gun was in West Dallas, and [Cook] gave him a .22”. (Emphasis added.) (The Dissent at 333 contends the State’s proving capital murder was “crucially dependent” upon Fair’s credibility. It was not. Farr was an important punishment-phase witness, because his testimony that Banks wanted to retrieve his gun for the purpose of committing armed robberies was a primary reason why the State was able to urge future dangerousness during the punishment phase; but, as noted here, Farr’s guilt-phase testimony was far less extensive, and was essentially duplicative of Marcus Jefferson’s above unchallenged guilt-phase testimony. See also Banks, 540 U.S. at 689, 124 S.Ct. 1256 (considering Farr “solely with respect to the capital sentence” (punishment phase)); Banks, 2000 WL 35482430, at *2 (adopting the magistrate judge’s conclusion that, “even if Farr’s informant status had been revealed at trial, the outcome of the guilt phase of the trial would have been the same”). In sum, and contrary to the Dissent’s assertions at 333 and 345, Farr’s guilt-phase testimony was not “crucial”, because it essentially just repeated the earlier, unchallenged testimony of Marcus Jefferson; and, for that same reason, Farr’s testimony was not “uncorroborated”.)
On cross-examination of Robert Farr: Banks’ counsel asked approximately 50 questions. More than half explored Farr’s criminal history and drug use (ie., “How long has it been since you shot up, Robert”; and “Are you holding now, Robert”). *317Farr admitted readily to extensive drug use. The other questions explored: whether Farr had a deal with prosecutors for his testimony; and whether he knew Cook.
Charles Cook, the witness central to the instant Brady claim, testified-, he first met Banks at about 8:00 a.m. on Saturday, 12 April 1980, while he and his wife were outside, waiting for a bus; Banks drove up in a green, two-tone Mustang with a brown hood and a black side; Banks said he did not know anyone in Dallas and asked for directions to a cheap place to stay; Cook talked Banks into driving his wife to work; and the trio departed in the Mustang.
Cook further testified: after Banks had dropped Cook’s wife off at work, Banks and Cook continued to drive around in the Mustang; Cook noticed blood on the right leg of Banks’ trousers; Banks said he “got into it on the highway with a white boy” and had shot him; they then went to Cook’s grandparents’ house (Cook’s house), where Cook gave Banks a change of clothes; Banks was introduced to Cook’s grandmother at the house; they next drove to a Mend’s house to see Cook’s sister’s new baby; and then, around mid-day, they drove to a rent-by-the-hour motel, so that Banks could take a bath and sleep.
Cook further testified: while Banks was at the motel, Cook used the Mustang to run errands; he then retrieved Banks from the motel; they picked Cook’s wife up from work in the Mustang; and the trio returned to Cook’s house, where Cook got permission from his grandparents for Banks to spend the night. (As noted, the house belonged to Cook’s grandparents; Cook lived there with his wife and child.)
Cook further testified: that Saturday night, he allowed Banks to sleep in his child’s room; around midnight, he observed Banks sitting on the bed looking depressed and asked him what was wrong; Banks responded: he had lied about the altercation taking place on the highway earlier in the day; “me and this white boy and his girl Mend was riding around”; “the white boy went to the car to get another beer”; Banks “decided to kill the white boy for the hell of it and take his car and come to Dallas”; and he shot the “white boy”.
Cook further testified: that Saturday night, he then noticed Banks had a pistol with him; the next morning (Sunday) he took Banks out early, intending to find a different place for him to stay; Banks telephoned his mother in Texarkana from Cook’s house during the day; Banks’ mother told him to turn himself in; Cook took Banks’ pistol away from him on Sunday evening and hid it; and Cook then allowed Banks to sleep at his house that night.
Cook further testified: the next day (Monday), he and Banks drove the Mustang to the bus station; he watched Banks board a Greyhound bus bound for Texarkana; Cook drove the Mustang back to his house; early on Tuesday morning, Cook abandoned the Mustang in West Dallas (it was never recovered following this abandonment); and, shortly thereafter, he sold Banks’ pistol, along with some tools taken from the Mustang, to his neighbor, Bennie Lee Jones.
Cook further testified: Banks telephoned him later in the week, asking Cook to mail the pistol to Banks, but Cook took no action; several days later, Banks telephoned again, saying he needed to see Cook and that he had decided not to turn himself in; Banks then showed up at Cook’s house, again asking for his pistol; and Cook told Banks he had gotten rid of it. (As discussed supra, Farr and Marcus Jefferson both traveled to Dallas with *318Banks for this encounter, and were waiting in a vehicle outside Cook’s house while Banks spoke to Cook near the door of the house. At trial, as discussed above, Farr testified: upon returning to the vehicle, Banks “said that his [Banks’] gun was in West Dallas, and [Cook] gave him a .22”. The Dissent at 334 incorrectly states that “Farr’s testimony provided the only cogent corroboration ... that Banks had brought the murder weapon to Dallas and had left it with Cook”. The Dissent inexplicably dismisses the above-discussed, nearly-identical corroborative statement provided by Marcus Jefferson, who testified (prior to Farr’s testifying): upon returning to the vehicle, Banks “said some broad had his [Banks’] gun, so [Cook] gave him another gun”. Marcus Jefferson is not the subject of a Brady claim, and his testimony is not challenged on appeal.)
Cook further testified: soon after Banks had attempted to retrieve his pistol from Cook, police officers showed up at his house asking for Banks’ pistol; Cook retrieved it from Jones’ house; and he gave it to the police.
On cross-examination of Charles Cook: Banks’ counsel reviewed Cook’s April 1980 affidavit and then asked approximately 55 questions. Approximately half of them focused upon impeachment of Cook’s credibility by raising past criminal offenses and drug use (i.e., “Number one, you fell [sic] in Dallas County in ’72, didn’t you”; and “What were you strung out on, Mr. Cook”). Cook admitted, inter alia, to drug use and having two convictions. Cook denied, three times, talking to anyone about his testimony prior to trial. (Cook’s denials during cross-examination relate to the instant Brady claim; as discussed supra, and as discussed in more detail, infra, the ' Cook transcript shows that he had participated in a pre-trial interview with prosecutors and Investigator Huff.) Banks’ counsel also pursued, with little success, a line of questioning that apparently was intended to advance a defense theory that Cook and Farr were partners in a narcotics operation, and that they had framed Banks for the victim’s murder. (During oral argument before our court for the instant appeal, however, Banks’ counsel conceded that he does not contend Cook is a suspect in the victim’s murder.)
Ida Marie “Rita” Cook, Charles Cook’s wife, testified: she first met Banks at about 8:30 a.m. on Saturday, 12 April 1980, while she and Cook were outside waiting for a bus; Banks drove up in a Mustang that was green, black, and “another color”; Banks drove her to work; he stayed with the Cooks over the weekend; and, on Monday, she gave him money so that he could return to Texarkana on a Greyhound bus.
As noted in the discussion of Cook’s cross-examination, supra, Banks’ counsel had attempted to advance a defense theory that Cook and Farr were partners in a narcotics operation, and that they had framed Banks for the victim’s murder. In that regard, during direct examination, Rita Cook further testified: she was originally from Texarkana (near Nash), but did not know Banks prior to Saturday, 12 April 1980; her husband, Cook (who was not from Texarkana), had last visited Texarkana in late summer of 1979; and she and Cook had been together at home (in Dallas) on the night of Friday, 11 April 1980 (when Banks met the victim in Texarkana).
On cross-examination of Ida Marie “Rita” Cook: Banks’ counsel asked approximately 35 questions. Approximately 28 of them explored her and her husband’s drug use and criminal background. She admitted to drug use and having a drug-related conviction. The other questions *319primarily explored the Cooks’ income. No questions were asked about their connections to Texarkana.
Bennie Lee Jones, Charles Cook’s neighbor, testified: he saw and spoke to Banks on the street outside of his (Jones’) house on the weekend of 12 April 1980; Banks then mentioned that he was in some sort of trouble and that his mother wanted him to turn himself in; Jones later purchased a pistol and some tools from Cook; and he gave the pistol back to Cook when Cook, accompanied by police officers, asked for it. Jones identified the murder weapon, a .25 caliber Galesi pistol, as the pistol he had purchased from, and returned to, Cook.
On cross-examination of Bennie Lee Jones: Banks’ counsel asked nine questions, which essentially clarified that Jones knew Cook and had purchased the murder weapon, jumper cables, and a tool set from him.
Bennie Whiteurs, Charles Cook’s grandfather, testified: Banks stayed in his house with Cook during the weekend in question.
On cross-examination of Bennie Whiteurs: Banks’ counsel asked four questions. After clarifying that Cook is Whiteurs’ grandson, Banks’ counsel asked whether Cook had been at home with Whiteurs (in Dallas) on the night of the murder; Whiteurs answered he had.
Carol Cook, Charles Cook’s sister, testified: she gave birth on Easter Sunday, 6 April 1980; the next Saturday, 12 April, Banks and Cook visited her during the day to see her baby; and Banks and Cook arrived, and departed, in a green Mustang.
On cross-examination of Carol Cook: Banks’ counsel asked seven questions. The first six essentially established that she was an unwed mother on welfare; the seventh was: “Okay. You know that your brother, Charles [Cook], is in a heap of trouble, don’t you?” She responded: “No, I don’t know what you mean. You — no, I don’t.”
Lou Ann Hamby, of the General Telephone Company in Texarkana, Texas, testified: she was in charge of customer records for the telephone company; she had in her possession a customer bill for Delma Banks (Sr.), detailing long-distance calls; and that bill showed that a collect call had been placed from Cook’s house to Banks’ house during the weekend in question.
Lou Ann Hamby was not cross-examined.
Dr. Vincent DiMaio, a Dallas County medical examiner, testified: he examined the victim’s body on 15 April 1980; the victim died as a result of three gunshot wounds; “powder tattooing” on the victim’s face indicated that one shot, entering between his eyes, had been fired at a distance of only 18 to 24 inches; the victim’s blood contained a small amount (.036) of alcohol, which was the equivalent of “something like one and a half beers”.
DiMaio further testified: during his examination of the victim, he recovered one bullet from the victim’s back, and a second bullet from the victim’s chest cavity; and he gave these bullets to the State’s firearms examiners, Larry Fletcher and Allen Jones. (As noted, the third bullet, which had exited the victim’s head, was recovered by Investigator Huff at the crime scene.)
On cross-examination of Dr. Vincent DiMaio: Banks’ counsel asked two questions, which clarified that the victim’s blood alcohol, as assessed after his death, was the “equivalent [of] something like one and a half beers”.
Larry Fletcher, a firearms examiner at the Dallas County Institute of Forensic Sciences, testified: Investigator Huff provided his laboratory with, inter alia, a *320recovered bullet and two spent .25 caliber cartridge casings for testing.
Larry Fletcher was not cross-examined.
Allen Jones, a firearms examiner at the Dallas County Institute of Forensic Sciences, testified: he examined and tested the items received from Dr. DiMaio and Investigator Huff, consisting of the .25 caliber Galesi pistol, the pistol’s magazine, and the recovered bullets and cartridge casings; and his opinion was that all recovered bullets and casings were fired by that .25 caliber Galesi pistol.
Allen Jones was not cross-examined.
Accordingly, among other evidence: Hicks and Bungardt testified Banks was with the victim on the night of the murder; Huff testified his investigation concluded Banks was the last person to be seen with the victim; Fisher testified he heard gunshots coming from the park at approximately 4:00 a.m.; and the firearms expert opined the recovered bullets that killed the victim were fired by the .25 caliber Galesi pistol. Moreover, and perhaps even more importantly, other witnesses provided evidence similar to Cook’s, including: Ida Marie Cook, who testified that Banks approached her and her husband (Cook) at around 8:00 a.m. on 12 April 1980 in a Mustang matching the distinctive description of the victim’s vehicle, that he stayed the weekend with them, and that he left on a Greyhound bus; Whiteurs, who testified that Banks spent the weekend with Cook at his house; Carol Cook (Cook’s sister), who testified that Banks and Cook visited her on 12 April 1980 in the distinctive, multi-colored Mustang; Bennie Lee Jones, who testified that Cook sold to him, and subsequently reclaimed, the .25 caliber Galesi murder weapon, and who further testified that Banks stated to him that he was in some sort of trouble and that his mother wanted him to turn himself in; Hamby, who testified that a collect call had been placed from Cook’s house to Banks’ parents’ house on the weekend in question; and Farr and Marcus Jefferson, who testified that, during the late-April visit to Cook’s house, Banks was trying to get “his gun” back from Cook. Compare to Strickler, 527 U.S. at 293, 119 S.Ct. 1936 (detailing evidence provided by other witnesses in its materiality finding, after noting that, while the key witness (the subject of that ease’s Brady claim) “provided the only disinterested, narrative account of what transpired ... [, that witness’] vivid description of the events ... was not the only evidence that the jury had before it”).
The Dissent at 334 urges that the evidence set forth above, without Cook’s testimony, provided the jury with nothing more than “suspicion” that Banks committed capital murder. This is simply an inaccurate description of the overwhelming evidence of Banks’ guilt that the jury was able to consider. Had Cook not testified, the jury still would have had evidence showing, inter alia: Banks was the last person seen with the victim; the victim was found murdered in the portion of the park where he and Banks had earlier been drinking beer together; Banks then, after gunshots were heard in the park, traveled approximately 175 miles to Dallas, Texas, in a distinctive green, red, and black Mustang matching the description of the victim’s vehicle; Banks told Cook’s neighbor he was in some sort of trouble; the murder weapon showed up in Cook’s house— 175 miles from the scene of the crime — on exactly the same weekend Banks did, despite Banks’ and Cook’s not having known each other until that weekend; and Banks later returned to Cook’s house to reclaim “his gun”. Compare to id. at 292-93, 119 S.Ct. 1936 (deciding, in its materiality finding, that “[e]ven if [the witness at issue] and her testimony had been entirely dis*321credited, the jury might still have [reached the same conclusion]”).
Of course, our analysis of the trial evidence, besides Cook’s testimony, is not meant to suggest that Cook was not a key witness for the prosecution. Among other evidence, he offered valuable testimony at trial that was not corroborated by other witnesses: Banks had blood on his trousers; he confessed to killing a “white boy”; he confessed to taking the victim’s vehicle; and he possessed the .25 caliber Galesi pistol up until Cook took it away from him on Sunday night. As demonstrated above, however, even if Cook’s credibility were to have been further impeached at trial, the majority of his extensive testimony was corroborated by other witnesses; and, even if Cook’s testimony was completely discredited by the jury, the jurors still had an abundance of other evidence with which to find Banks guilty beyond a reasonable doubt.
b.
The Dissent at 338 strongly criticizes the above analysis, in which we have fully disregarded Cook’s trial testimony, as merely applying a “sufficiency-of-evidence test”. The above analysis, however, is only the starting point for our materiality evaluation. Just as with the Brady analysis performed in Strickler, 527 U.S. at 292-93, 119 S.Ct. 1936, a thorough examination of the trial evidence, besides Cook’s testimony, provides a starting point for assessing the Cook transcript’s materiality. As our court explained in Sipe,
a Brady determination is inevitably a contextual inquiry, involving questions of both law and fact. Moreover, it is intimately intertwined with the trial proceedings: because the court must judge the effect of the evidence on the jury’s verdict, the Brady decision can never be divorced from the narrative of the trial. In addition, the court must consider not simply the withheld evidence in isolation, but also the quantity and quality of other evidence in the record.
388 F.3d at 479 (emphasis omitted); see also Smith v. Black, 904 F.2d 950, 967 (5th Cir.1990) (“The materiality of Brady material depends almost entirely on the value of the evidence relative to the other evidence mustered by the state.”), vacated on other grounds, 503 U.S. 930, 112 S.Ct. 1463, 117 L.Ed.2d 609 (1992).
The Dissent itself, in its materiality analysis at 340, engages in a similar evidentiary review to support its assertions that “Cook’s testimony ... was the only direct evidence warranting a guilty-as-charged verdict”, and that “the prosecution would have had only inconclusive circumstantial evidence”, had Cook’s testimony been “impeached or discredited”. (On the one hand, the Dissent, in urging the Cook transcript’s suppression was material, is completely correct, at 339, to review and analyze the evidence in attempting to support a materiality finding. See also Dissent at 345 (“To determine whether confidence in the verdict is undermined by the suppression, we must necessarily evaluate the strength or weakness of the State’s other evidence of guilt.”). On the other hand, the Dissent at 338 then inexplicably faults this opinion for engaging in the same analysis to explain our ruling that the suppression was not material.)
As discussed in Sipe, supra, our evidentiary review assists in evaluating the Cook transcript’s materiality by exploring the extensive corroboration of almost all of Cook’s testimony and by placing that testimony into context with all of the other evidence supporting Banks’ conviction. This contributes to our determination that “the favorable evidence could [not] reasonably be taken to put the whole case in such a different light as to undermine confi*322denee in the verdict”. Kyles, 514 U.S. at 435, 115 S.Ct. 1555; see also Dissent at 338.
Having examined the trial evidence, our evaluation continues in order to further confirm that a materiality finding does not rest upon disallowed sufficieney-of-the-evidence or harmless-error analyses. See Kyles, 514 U.S. at 434-35, 115 S.Ct. 1555. Because the jury heard Cook’s testimony, we next consider what Banks’ counsel might have been able to accomplish on cross-examination of Cook, had the Cook transcript been available.
Our cross-examination consideration of the Cook transcript focuses on its value as an impeachment tool. The Dissent at 339 cites Tassin v. Cain, 517 F.3d 770 (5th Cir.2008), Sipe (5th Cir.2004), and United States v. Fisher, 106 F.3d 622 (5th Cir.1997), as examples of instances where Brady materiality was found because suppressed evidence would have permitted impeachment of key witnesses. Tassin concerned the prosecution’s witness’ testimony contradicting the defendant’s; and, the suppressed evidence would have shown that the prosecution’s witness had a deal for her testimony. See Tassin, 517 F.3d at 772. (Whether Cook had a deal for his testimony is, of course, an already-disposed-of Brady claim.) Sipe, as previously discussed, involved the collective assessment of five different suppressions. See Sipe, 388 F.3d at 477. And Fisher involved the suppression of an FBI report in which the described events “directly contradicted]” a key witness’ trial and grand-jury testimony. See Fisher, 106 F.3d at 634. (A far cry from the instant case, where the Cook transcript, while containing some inconsistencies, certainly does not directly contradict, and generally mirrors, Cook’s trial testimony and his April 1980 affidavit.) The primary import from these cases is that an evaluation of materiality is a fact-intensive examination done on a careful, case-by-case basis.
The transcript could have been used primarily to show Cook was not truthful when, during his trial testimony, he denied, three times, talking before trial to anyone about his testimony. Even without the Cook transcript, but possessing Cook’s April 1980 affidavit, Banks’ counsel — evidently, and quite logically, knowing no prosecutor would place a key witness like Cook on the stand without first interviewing him — expressed great doubt when Cook denied talking to anyone before trial about his testimony.
While the Cook transcript would have allowed Banks’ counsel, as discussed more fully below, to address Cook’s trial-testimony denials, we first note that, consistent with the above doubt expressed by Banks’ counsel, there is absolutely nothing improper about the prosecutors’ having interviewed Cook prior to trial. Again, as shown above, Banks’ counsel recognized this. This is in keeping with one of the most basic and well-known rules of trial advocacy: “Never ask a question for which you do not know the answer”. Ward v. Whitley, 21 F.3d 1355, 1362 (5th Cir.1994) (addressing that concept in the context of cross-examination).
The Dissent at 340 criticizes this portion of our analysis as confusing improper attorney witness coaching with the higher duties owed by prosecutors. Acknowledging, of course, that prosecutors are held to a higher standard, we nevertheless observe, again, that there is absolutely nothing improper about the prosecutors’ having interviewed Cook prior to trial. (Indeed, not having done so would have been extremely questionable.) This pretrial-interview practice, of course, was in existence long before the instant 1980 capital-murder trial. As the Court noted in *323United States v. Ash — a case which specifically dealt with prosecutorial pre-trial preparation — “the interviewing of witnesses before trial is a procedure that predates the Sixth Amendment. In England in the 16th and 17th centuries counsel regularly interviewed witnesses before trial”. 413 U.S. 300, 318, 93 S.Ct. 2568, 37 L.Ed.2d 619 (1973) (addressing prosecutorial pre-trial preparation in the context of “photographic displays”) (citing 9 W. Holdsworth, History of English Law 226-28 (1926)).
Along that line, there is little, if any, impeachment value to a mere showing that Cook had met with prosecutors prior to trial. As shown by Banks’ counsel’s reaction at trial, such a meeting was expected; moreover, it was completely proper under Texas law as it existed at the time of the trial. See, e.g., Tex.Code Crim. Proc. Ann. art. 39.14 (Vernon 1983) (effective 1 Jan. 1966) (addressing the State’s pre-trial possession of, inter alia, “written statements of witnesses”); cf. Mathis v. State, 469 S.W.2d 796, 799 (Tex.1970) (prosecutor has a duty to interview a co-indictee prior to trial); Jackson v. State, 501 S.W.2d 660, 663 n. 2 (Tex.Crim.App.1973) (recognizing the likelihood of pre-trial witness interviews by prosecutor); Whitfield v. State, 492 S.W.2d 502, 504 (Tex.Crim.App.1973) (noting the prosecutor interviewed the witness prior to trial); see generally Goldberg v. United States, 425 U.S. 94, 96 S.Ct. 1338, 47 L.Ed.2d 603 (1976) (discussing prosecutor’s pre-trial witness interviews in the context of the Jencks Act, 18 U.S.C. § 3500).
Returning to our analysis of the Cook transcript’s impeachment value, the transcript, as the Dissent notes at 335, is, obviously, evidence that Cook did discuss his testimony with prosecutors before trial. Banks’ counsel could have used the Cook transcript to address Cook’s testimony that he did not discuss his testimony with anyone prior to trial.
On the other hand, Cook’s April 1980 affidavit, which Banks’ counsel possessed at trial, showed he had met with police long before trial that September. Indeed, Banks’ counsel used the April 1980 affidavit to make this very point during closing argument. After Cook had denied, three times, not talking to anyone prior to trial about his testimony, Banks’ counsel specifically referenced the April 1980 written affidavit that had been disclosed to him at trial by the prosecution in arguing to the jury that this affidavit was proof that Cook had perjured himself on the witness stand when he denied not talking to anyone prior to trial. Banks’ counsel told the jury:
Even though [Cook] says I’ve never talked to anybody in my life about this case, you know that’s not the truth. You know that is not true, because I stood up and I said, “Your Honor, under the rules of law I’m entitled to see the statement that he gave them.” He had it. The District Attorney’s office produced [the April 1980 affidavit].
Along this line, the jury was already aware, for its credibility determination, that Cook had not been truthful about talking to others prior to trial. See Felder v. Johnson, 180 F.3d 206, 213 (5th Cir. 1999) (“Suppressed evidence is not material when it merely furnishes an additional basis on which to impeach a witness whose credibility has already been shown to be questionable”.) (quoting United States v. Amiel, 95 F.3d 135, 145 (2d Cir.1996)). Had the Cook transcript been produced, it would merely have allowed Banks’ counsel to add one more example to the Cook-lied-about-not-talking-to-others point that counsel was already making to the jury.
Banks correctly notes that the Cook transcript would have been useful with respect to an assault to which Cook admit*324ted on direct examination. When, utilizing another, well-recognized prudential trial tactic, the State preemptively questioned Cook about his extensive criminal history, Cook was relatively candid, acknowledging, for example: “I pleaded guilty [to theft in 1976] because I was guilty”. This criminal-history line of questioning was, of course, likely pursued by the State in order to lessen the effectiveness of anticipated cross-examination. Banks notes, however, that when Cook was asked about a 1972 assault conviction, he admitted to the assault, but stated he could not remember whom he had assaulted. In the Cook transcript, however, as noted supra, he had told the interviewer that the victim was a school teacher. (This is the above-described point at which “Do not say” was written in the margin by the DA.)
Had Banks’ counsel been in possession of the Cook transcript, he could have used it, as the Dissent agrees at 337: to show an inconsistency in Cook’s testimony regarding the identity of the person assaulted; to urge that Cook lied, at trial, when he said he could not remember the identity of that person; and to further insinuate that Cook was not a trustworthy witness. The jury, however, was already well aware that Cook was a drug user with an extensive criminal history.
In addition, Banks urges the Cook transcript could also have been used to impeach Cook during cross-examination by pointing to inconsistencies between it and Cook’s April 1980 affidavit. Among other things, he notes that Cook’s recollections differ regarding the colors of the clothing that he gave to Banks, and when, precisely, he first noticed Banks’ pistol. Despite the existence of these minor inconsistencies, however, the Cook transcript and the April 1980 affidavit are both quite consistent with Cook’s trial testimony on the major points at issue. All provide largely-identical accounts: of Banks’ weekend with the Cooks in Dallas, supported in part, inter alia, by the testimony of Ida Marie Cook, Bennie Lee Jones, Bennie Whiteurs, Carol Cook, and Lou Ann Hamby; and of Banks’ subsequent return to Dallas to get “his gun”, supported in part, inter alia, by the testimony of Marcus Jefferson, Robert Farr, and Investigator Willie Huff.
As discussed in more detail infra, for such reasons, the Cook transcript would also have been helpful to Banks because, during closing argument, the prosecution stressed to the jurors that Cook had been honest with them. Beyond these points, however, the transcript provides very little impeachment value. Contrary to Banks’ assertions, it consists largely of nothing more than pages of recollections by Cook, with inconsistencies in those recollections later pointed out by those conducting the interview. As our examination, below, of the question-and-answer exchanges between Cook and the interviewers demonstrates, the extensive “coaching” claimed by Banks is simply not present. (In disagreeing with this assessment, the Dissent at 338-39 points to the Court’s 2004 opinion in this case, which, for this issue, dealt only with the applicability of Federal Rule of Civil Procedure 15(b). In the introductory paragraphs of that opinion, the Court, inter alia, described Cook’s trial testimony as “intensively coached by prosecutors and law enforcement officers”. Banks, 540 U.S. at 675, 124 S.Ct. 1256. Obviously, the factual description by the Court is not the law of the case with respect to the Cook-transcript Brady claim; the Court’s 2004 opinion did not address the merits of that claim. (Of course, had it decided the Cook-transcript Brady claim on the merits, the Court could have rendered with respect to that claim, instead of remanding.) The Court’s 2004 opinion, as it relates to Cook, answered a purely legal question, determining: Rule 15(b) is appli*325cable to this habeas proceeding; and, accordingly, a COA should have issued for the Cook-transcript Brady claim. Id. at 703-05, 124 S.Ct. 1256. The merits of the Cook-transcript Brady claim were addressed, for the first time, by the district court on remand. See Banks, 383 F.3d at 281.)
Turning to our above-referenced examination of the Cook transcript’s question- and-answer exchanges, the extensive “coaching” claimed by Banks is not reflected. Along that line, we first observe that the transcript begins with the interviewer telling Cook:
What we want to do first is go through it and let you just tell us best [sic] you can remember everything you know about Delma Banks and about this ease and we’ll try not to stop you any more than we have to — let you just kinda [sic] run through it as you remember it, then go back and pick up some specifics.
(Emphasis added.) This is consistent with how to interview a witness before trial. The witness is asked to tell his version of the facts without coaching or other suggestion, and then the interviewer circles back to probe, test, and further explore any portions of that version that may be inconsistent with the witness’ earlier statements, or other expected evidence.
True to this method, the first half of the Cook transcript consists largely of Cook’s description of the events at issue, beginning with the moment he met Banks and continuing through 24 April 1980, when he provided the police with the affidavit; the interviewer’s interruptions during the first 18 pages of the transcript are minimal and clarifying in nature. For example:
Q: Now, what did ya’ll do that night?
A: Rode around and you know, showing him different places and things.
Q: Okay.
A: Had a few beers.
Q: Okay. Go ahead.
A: Now, that night I went in, I asked my grandmother ....
Next, and also true to this interview method, the interviewer takes a more active role during the latter half of the interview. It is here that Banks claims several instances of improper “coaching” by the interviewer. The interviewer’s questions are not, however, coercive or otherwise improper. Instead, and quite understandably, the interviewer probes and tests Cook’s story; points out inconsistencies in his statements; presses him for further details; and questions him about his own criminal history and legal issues. For example, when Cook was asked about the circumstances surrounding his April 1980 affidavit (which, again, was provided to Banks at trial), the following exchange took place:
Q: Did [the police] tell you he was a white boy?
A: Yeah, they told me he was a white boy, [sic]
Q: Now, see you told me that you that [sic] they didn’t tell you anything. They just put you in a room and told you to start writing and now you’ve come up with two or three things they’ve told you. Now, I wanna [sic] know what they told you.
This is not coercive or otherwise improper, nor is it valuable impeachment material for Banks’ attorney. It merely illustrates an instance where the interviewer identifies a fact that had been omitted by Cook earlier in the interview; and, quite naturally, presses Cook, here, to find out if there have been other omissions as well. The interviewer did not put words into Cook’s mouth in this exchange, nor did he suggest other items that might be added to the above-discussed list of things that the police had told Cook; this is a completely *326proper attempt, by the interviewer, to ensure that this list of items provided to him by Cook is complete.
Similarly, Banks points to the interviewer’s later comment to Cook that “[y]our statement [in the April 1980 affidavit about when you noticed blood on Banks’ pants] is obviously screwed up”. The phrase “obviously screwed up”, of course, causes close examination for coaching or other coercion. As with the exchange discussed supra, however, this comment — when read in context — is merely illustrative of another instance where the interviewer presses Cook to clarify inconsistencies.
This exchange occurred only after the interviewer noticed that, earlier in the interview, Cook had stated that he had seen blood on Banks’ leg and then provided Banks with a change of clothing, whereas, in his April 1980 affidavit, Cook had swapped the order of these two events. Accordingly, the interviewer read to Cook from Cook’s April 1980 affidavit; pointed out the inconsistency regarding when Cook first saw the blood on Banks’ trousers; and, further, pointed out that, logically, Cook must have seen the blood on Banks’ leg before Banks changed into new clothes. The interviewer’s full comment to Cook was:
Q: Your [April 1980] statement [about when you noticed blood on Banks’ pants] is obviously screwed up. And they are going to ask you about it and they are going to rake you over the coals about it because it is screwed up. It doesn’t make any sense that he [cha]nged clothes and you got back in the car and went riding and then you noticed blood on his pants because if he changed clothes he wouldn’t have any blood on his pants. So your [April 1980] statement doesn’t make any sense. What you told me before [in today’s interview] does make sense, that you noticed the blood on his pants, and then you took him to change clothes. They are going to ask you about it and you are just going to have to explain it. That you might a[sic] mistake and you got your facts out of sequence.
These exchanges, and others like it, are not instances where Cook is pressured or improperly coached into changing his story. They amount to logical clarifications raised in a completely permissible pre-trial witness-interview session. Moreover, the details provided by Cook in the Cook transcript very closely match the details that Cook provided in his April 1980 affidavit, which, again, was provided to Banks’ counsel.
Banks also points to Cook’s testimony as it relates to Banks’ possible motive — even though the jury was not required to determine a motive in order to convict. Cook, in his April 1980 affidavit, said Banks’ statement to him was: “[M]e and this white boy was [sic] in the woods in Texarkana drinking when I thought about killing him and taking his car to Dallas just for the hell of it”. Cook, in the transcript, said: “The white boy went [sic], goes to the car to get another beer. Delma Banks say [sic] this is when he decided he wanted his car. He say [sic] he took out the [J25 automati[e] that he had and shot him”; and, later, “I say [sic] why you shoot the boy man? ... [Banks] said man uh I don’t know man I wanted his car man”. Finally, at trial, Cook testified: “[Banks] said the white boy went to the car to get another beer. Delma [Banks] say [sic] this was when he decided to kill the white boy for the hell of it and take his car and come to Dallas”.
Banks, and the Dissent at 337-38, essentially urge Cook’s April 1980 affidavit and his trial testimony set forth a “for the hell of it” motive, while the Cook transcript sets forth a different, “[Banks] wanted his *327car” motive. This point is urged despite all three of these statements referencing both the killing of the victim and the taking of his vehicle. In that regard, we disagree that the variations in the wording of Cook’s statements pertaining to motive (which, again, was not an element of the crime) display inconsistencies of significant impeachment value. To the contrary, the statements quoted above — which, again, all reference both the killing of the victim and the taking of his vehicle (the two acts necessary for the capital-murder conviction) — serve, if anything, as an example of the overall consistency among Cook’s April 1980 affidavit, the Cook transcript, and his testimony.
Our analysis of Brady materiality in this case rests, in part, on our assessment of the Cook transcript’s usefulness to Banks as a cross-examination tool. This is, of course, entirely appropriate because, inter alia: had the transcript been produced, and assuming Banks’ counsel would have used it at trial, it appears that it would have been used to attempt to impeach Cook on cross-examination; indeed, Banks’ brief here roots many of its assertions in this impeaehment-on-cross-examination context. We now, however, also examine the State’s opening statement (Banks did not make one) and both sides’ closing arguments to ensure that the materiality vel non of the Cook-transcript is fully evaluated. See, e.g., Strickler, 527 U.S. at 290-91, 119 S.Ct. 1936 (considering the prosecutor’s closing argument, which emphasized the importance of the testimony of the witness at issue, in deciding that Brady’s materiality prong was not satisfied); see also Gibbs v. Johnson, 154 F.3d 253, 256 (5th Cir.1998) (discussing failure to disclose evidence contrary to a prosecutor’s closing-argument assertions, and citing West v. Johnson, 92 F.3d 1385, 1399 (5th Cir.1996)).
The State’s opening statement, describing what each witness would be called to show, noted that Cook “is an important witness”, and that his testimony “is critical”. The State also emphasized Cook’s criminal history. Among other things, the State told the jury: “Now, this man, Charles Cook, has a bad past. He has been to prison. He has two prior convictions”.
The State essentially repeated these themes in its closing argument. It reviewed the importance of each witness’ testimony; and, in the course of doing so, it again devoted significant time to Cook’s testimony. Moreover, as discussed supra, for assessing the Cook transcript’s impeachment value, the State told the jury that “Cook brought you absolute truth”. On the other hand, the State also emphasized that: Cook’s testimony “was not that of a Baptist preacher and ten Deacons”; “Cook is not President of the Chamber of Commerce .... [H]e has made mistakes”; and Cook readily admitted, in his testimony, to having two convictions.
As noted, Banks did not make an opening statement. At the conclusion of the State’s opening statement, Banks’ counsel elected not to make a statement until after the State had rested on its case in chief; when the State did so, Banks rested as well.
Banks’ counsel, in his closing argument, devoted a significant amount of time to discrediting Cook. Among other things, and as discussed supra, he referred to the April 1980 affidavit to assert that Cook had lied when he testified that he had not talked to anybody about the case prior to trial. Banks’ counsel also devoted a significant amount of time: challenging Cook’s credibility by citing his admitted drug use and criminal past; and re-urging the (now-abandoned) trial theory that Cook and Farr were part of a drug ring and involved *328with the victim’s murder. (Banks’ counsel, at oral argument, confirmed that the above-referenced trial theory has been abandoned, and that Banks does not claim Cook is suspected of the murder.)
In reviewing the opening statement and the closing arguments, we first note that, obviously, we do not condone the prosecutor’s failure, 30 years ago, to correct Cook’s statements about his not having talked to anyone prior to trial. (Perhaps the same can be said about Cook’s testifying he could not recall the identity of the person he admitted assaulting.) We are, nevertheless, bound by the task of reviewing the materiality of the Cook transcript’s suppression. The Dissent, at times, appears to improperly merge the suppression and materiality prongs of a Brady analysis. At 339, for example, its discussion of the Cook transcript’s suppression seems to take the position that, because there was a suppression, finding materiality is required as a means of sanctioning the prosecution. Of course, it is simply not the case that materiality must be found whenever a Brady suppression occurs. In the instant case, it is conceded that a suppression occurred; and, while this suppression was improper, materiality remains a separate prong of the Brady analysis.
Having earlier examined the impeachment value of the transcript and the great deal of evidence, besides Cook’s testimony, supporting the jury’s guilty verdict, we now, for this portion of our materiality review, consider whether, and how, the suppression affected the opening statement and closing arguments. In this regard, it becomes clear that the availability of the transcript likely would not have changed the approach pursued by either party, nor would it have significantly altered the effectiveness of those approaches. We note, among other things: Banks’ counsel, by using Cook’s April 1980 affidavit, argued to the jury that Cook had lied about not talking to anyone prior to trial; and both parties made the jury more than well aware that Cook was a drug user with an extensive criminal past.
c.
Finally, as noted supra, materiality must be assessed collectively, not point by point. Kyles, 514 U.S. at 436, 115 S.Ct. 1555. As also discussed supra, however, we may not consider already-disposed-of habeas claims, including the guilt-phase Farr Brady claim and sufficiency-of-the-evidence claim. (In this regard, the Dissent at, e.g., 342 errs in considering such claims.) Our materiality assessment “take[s] into consideration all of the effects that flowed from the prosecution’s suppression [of the Cook transcript]”, Dissent at 341, and weighs those effects, not item-by-item, but as a whole.
Before summarizing our collective assessment, we first address the guilt-phase Farr Brady claim that the Dissent has sua sponte improperly reintroduced into this ease. As previously discussed, Farr’s suppressed informant status is the reason for the Supreme Court’s granting punishment-phase habeas relief; and, as also previously discussed, a guilt-phase Farr Brady claim was considered by the district court in 2000, but that court held Farr’s suppressed informant status immaterial to the guilt phase. Banks, 2000 WL 35482430, at *2. A COA was not granted for a guilt-phase Farr Brady claim; the Supreme Court, in 2004, addressed Farr as a “Penalty Witness”. See Petition for Writ of Certiorari, Banks v. Cockrell, 2002 WL 32135512. Likewise, the district court, in 2008, did not rely upon Farr’s guilt-phase testimony in finding that the Cook transcript’s suppression was material. See Banks, 2008 WL 906716.
*329As also previously discussed, Banks, in his brief and during oral argument, never even suggested that Farr’s suppressed informant status should be considered as a part of our guilt-phase materiality assessment. It is well established, of course, that an appellant abandons all issues not raised and properly presented in its initial brief on appeal. E.g., Cinel v. Connick, 15 F.3d 1338, 1345 (5th Cir.1994). Moreover, it is also well established that Brady violations not considered in district court are not properly before our court, and should not be considered. E.g., United States v. Skilling, 554 F.3d 529, 591 (5th Cir.2009). (The Dissent fails to address these reasons why Farr’s supposed informant status should not be considered.)
In sum, consideration of Farr’s suppressed informant status is an issue that has either been waived or is not properly before us. In the alternative, as shown below, even if that issue is considered as part of our materiality analysis, the evidence of Banks’ guilt is so overwhelming that his Brady claim would still fail.
While Farr was a key punishment-phase witness — a primary reason the State was able to urge future dangerousness during that phase — his guilt-phase testimony was, as discussed, minimal and essentially repetitive of Marcus Jefferson’s earlier testimony. Marcus Jefferson testified: upon returning to the vehicle, Banks “said some broad had his [Banks’] gun, so [Cook] gave him another gun”. Farr testified (following Marcus Jefferson’s testimony): upon returning to the vehicle Banks “said that his [Banks’] gun was in West Dallas, and [Cook] gave him a .22”. Clearly, Farr’s guilt-phase testimony, when viewed together with Marcus Jefferson’s, merely reiterates previously introduced evidence.
In that regard, the Dissent miseharacterizes Marcus Jefferson’s brief but important testimony. It maintains erroneously at 346 that Marcus “Jefferson was a 19 year old ineffectual witness who added nothing to Farr’s perjury-laden testimony”.
As stated, this is an incorrect description of the testimony and proceedings concerning it. Again, Marcus Jefferson testified before Farr. As shown by the quoted testimony in the Dissent at 346-47 n.9, on direct examination, Marcus Jefferson initially attempted to evade the State’s questions concerning Banks’ new-gun statement in the automobile, responding: Banks did not make any statement “while we were sitting still”. When the State narrowed its question, Banks’ counsel immediately asked to approach the bench, preventing Marcus Jefferson from responding. The jury was then excused and, inter alia, Banks’ counsel objected that Marcus Jefferson’s below-described testimony, that was given outside the presence of the jury, was inadmissible on several grounds, primarily “because it injectfed] an extraneous offense” and was “an attempt to try [Banks] as a criminal generally”. The objection was overruled.
Outside the presence of the jury, the State had again narrowly questioned Marcus Jefferson about Banks’ statement in the automobile, and Marcus Jefferson had responded: Banks “said some broad had his [Banks’] gun, so they gave him another gun”. Subsequently, Marcus Jefferson testified similarly in the presence of the jury. Nothing in the record suggests Marcus Jefferson had any problem recalling this version of what Banks said or that the State’s questioning Marcus Jefferson outside the presence of the jury in any way caused him to recall that version. Moreover, Banks’ counsel never even cross-examined Marcus Jefferson. Therefore, contrary to the position taken by the Dissent, there is no reason to question the strength *330of Marcus Jefferson’s uncontested testimony.
Before summarizing our collective assessment, we also review, one final time, facts that were established at trial. Among other things, evidence — other than Cook’s testimony — established: Banks and the victim were drinking Coors beer together in a secluded area of a park on the night of the murder; gunshots were heard coming from the park in the early hours of that morning; and the victim was subsequently discovered in that same area of the park, lying amongst empty cans of Coors beer. Additionally: Banks, in the hours immediately following the sound of the gunshots, traveled 175 miles, from Nash, Texas, to Dallas, Texas, for no known reason, and without a planned place to stay; despite not having a vehicle with him at the bowling alley the night before, he was now driving a distinctive, multicolored Mustang, matching the description of the victim’s; he spent that weekend in Dallas at the house of a complete stranger (Cook); and — in what Banks would apparently have the court believe is mere coincidence — the pistol undisputedly used to kill the victim showed up at that same stranger’s house (Cook’s house), 175 miles from the scene of the crime, on exactly the same weekend that Banks did. Moreover: Banks told Cook’s neighbor (Bennie Lee Jones) that he was in some sort of trouble; he made a collect telephone call to his mother from Cook’s house; he subsequently abandoned the Mustang that he had been driving (leaving it with Cook); and, weeks later, after having returned to Nash, Banks again traveled 175 miles back to Cook’s house in Dallas at 3:00 a.m. to reclaim “his gun” from Cook. (We again note, as well, that Banks’ referring to the gun at Cook’s house as “his gun” was established through both the testimony of Farr (who is not the subject of a guilt-phase Brady claim) and through the testimony of Marcus Jefferson (who is not the subject of any Brady claim, and whose testimony is not challenged in any way in this appeal)).
As discussed in greater detail supra, we provide this summary of the evidence, not to engage in a forbidden sufficiency-of-the-evidence analysis, but to place Cook and the Cook transcript into context. See Sipe, 388 F.3d at 479. Was the suppression of that transcript material? To answer this question, we must consider the trial’s guilt phase as a whole to properly assess the importance of the Cook transcript.
While Banks was in Dallas, Cook had more interaction with him than did any other witness, and, accordingly, was able to provide the most complete narrative of Banks’ weekend in Dallas. Additionally, he was the only witness who testified: to seeing blood on Banks’ trousers; to seeing Banks in actual possession of the murder weapon; and that Banks had confessed to the murder. In this regard, as earlier acknowledged, he was indeed a key witness. On the other hand, of course, not all key witnesses are of the same degree of importance. In some cases, the prosecution’s case is solely dependent upon a key witness, while in other cases — including this one — -the key witness is strongly buttressed by numerous other witnesses. See Wilson v. Whitley, 28 F.3d 433, 439 (5th Cir.1994) (“[W]hen the testimony of the witness who might have been impeached by the undisclosed evidence is strongly coiToborated by additional evidence supporting a guilty verdict, the undisclosed evidence generally is not found to be material”.), cert. denied, 513 U.S. 1091, 115 S.Ct. 754, 130 L.Ed.2d 653 (1995). Our placing Cook’s testimony into context by summarizing the other evidence presented is not engaging in a sufficiency-of-the-evidence showing; it, instead, demonstrates *331that the evidence provided by Cook was largely corroborated by other witnesses at trial. (The Dissent at 346 maintains erroneously that only “marginal portions of Cook’s testimony were corroborated”.)
For example: Ida Marie Cook and Carol Cook both confirmed that Banks was driving the multi-colored Mustang; Hamby, from the telephone company, confirmed that Banks had called his parents’ house, collect, from Cook’s house; Allen Jones, the firearms expert, confirmed that the pistol that ended up at Cook’s house on the same weekend that Banks did was the murder weapon; Bennie Lee Jones confirmed that he purchased that murder weapon from Cook shortly thereafter, and, additionally, that Banks said he was in some sort of trouble; and Farr and Marcus Jefferson both confirmed that Banks’ later visit to Cook’s house was for the purpose of reclaiming Banks’ gun from Cook. (Again, in describing Banks’ attempt to recover his gun from Cook, Marcus Jefferson testified: Banks “said some broad had his [Banks’] gun, so [Cook] gave him another gun” (emphasis added); and Farr testified: Banks “said that his [Banks’] gun was in West Dallas, and [Cook] gave him a .22” (emphasis added).) In sum, had the suppressed Cook transcript been available — and had it been used to impeach Cook more than he already was — the jurors still would have had to contend with the fact that Cook’s trial testimony, as detailed above, was supported at nearly every turn by some other witness’ testimony, or by some other item of evidence; and that it also fit with the other trial evidence provided by the witnesses from Nash, Texas.
The credibility of Cook’s testimony was not only buttressed by its consistency with the other evidence presented during the trial’s guilt phase; it was, as noted earlier, also buttressed by its consistency with his first written statement to the police — the April 1980 affidavit made available to Banks’ counsel at trial. In this regard (and considering, additionally, that the jury was already well aware — without the Cook transcript — that Cook had an extensive criminal history), failing to disclose the transcript of this single, completely permissible pre-trial interview with prosecutors cannot be said to “undermine[ ] confidence in the outcome of the trial”, Kyles, 514 U.S. at 434, 115 S.Ct. 1555 (quoting Bagley, 473 U.S. at 678, 105 S.Ct. 3375).
Assessing the Cook transcript’s materiality, we find that it, at best, could have been used to impeach Cook: with respect to his denial that he had discussed his testimony with others prior to trial (which, of course, Banks’ counsel used the available April 1980 affidavit to do); with respect to his not being able to identify the person he had assaulted in 1972; with respect to when, exactly, he first noticed the blood on Banks’ trousers; and possibly with respect to a few other factual inconsistencies. Along that line, even though the jury was well aware of Cook’s criminal history, the Cook transcript might have further diminished Cook’s credibility with the jury. Given, however, the jury’s existing knowledge of Cook’s criminal past, and further given the overall consistency of Cook’s testimony with his April 1980 affidavit and the extensive corroboration of his testimony by other witnesses at trial, it cannot be said that the assorted additional impeachment points that might have been raised, had the Cook transcript been disclosed, render its suppression material under Brady. The transcript falls far short of undermining confidence in the guilt phase’s outcome. Compare Strickler, 527 U.S. at 294, 119 S.Ct. 1936 (“The record provides strong support for the conclusion that petitioner would have been convicted of capital murder and sentenced to death, even if [the witness at issue] had been *332severely impeached.”); see also Youngblood, 547 U.S. at 870, 126 S.Ct. 2188; Kyles, 514 U.S. at 435, 115 S.Ct. 1555; Bagley, 473 U.S. at 682, 105 S.Ct. 3375.
In sum, for the Brady materiality test, and having assessed the contents of the Cook transcript, Cook’s April 1980 affidavit, his trial testimony, and the other trial evidence, “there is [not] a reasonable probability that, had the [Cook transcript] been disclosed to the defense, the result of the proceeding would have been different”. Bagley, 473 U.S. at 682, 105 S.Ct. 3375. Restated, the State’s suppression of the Cook transcript does not undermine confidence in the outcome of the guilt-phase of the trial. See Kyles, 514 U.S. at 434, 115 S.Ct. 1555 (citing Bagley, 473 U.S. at 678, 105 S.Ct. 3375).
III.
For the foregoing reasons, the habeas relief for Banks’ conviction is VACATED. The Court’s grant of habeas relief for Banks’ sentence is, of course, not affected by this opinion. This matter is remanded to district court for further proceedings consistent with this opinion.
VACATED IN PART and REMANDED.