Floyd v. Vannoy

Full title: JOHN DAVID FLOYD, Petitioner - Appellee v. DARREL VANNOY, WARDEN…

Court: UNITED STATES COURT OF APPEALS FOR THE FIFTH CIRCUIT

Date published: Apr 6, 2018

887 F.3d 214 (5th Cir. 2018)

Opinion

No. 17-30421

04-06-2018

John David FLOYD, Petitioner-Appellee v. Darrel VANNOY, Warden, Louisiana State Penitentiary, Respondent-Appellant

Richard Martin Andrew Davis, Emily Lucy Ann Maw, Innocence Project New Orleans, New Orleans, LA, for Petitioner-Appellee. Donna Rau Andrieu, Assistant District Attorney, District Attorney's Office for the Parish of Orleans, New Orleans, LA, for Respondent-Appellant.

RHESA HAWKINS BARKSDALE, Circuit Judge

Richard Martin Andrew Davis, Emily Lucy Ann Maw, Innocence Project New Orleans, New Orleans, LA, for Petitioner-Appellee.

Donna Rau Andrieu, Assistant District Attorney, District Attorney's Office for the Parish of Orleans, New Orleans, LA, for Respondent-Appellant.

Before SMITH, BARKSDALE, and HIGGINSON, Circuit Judges.

RHESA HAWKINS BARKSDALE, Circuit Judge:

For two murders in New Orleans, Louisiana, in 1980, within days of, and in close proximity to, each other and involving extremely similar facts, John David Floyd was convicted in a state-court joint bench trial of the first, but acquitted of the second, murder, with state post-conviction relief’s being denied for the first time in 2011, but federal habeas relief’s being granted in 2017 because, after concluding the habeas application was not time-barred, the district court concluded: material evidence, favorable to Floyd, had been withheld prior to trial; and the state courts' contrary decisions had unreasonably applied clearly-established federal law, as proscribed by the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA). For the State’s challenge to that relief, at issue is whether: Floyd established "actual innocence" to overcome the statute of limitations for his application; the State withheld favorable, material evidence, in violation of Brady v. Maryland , 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963) ; and the district court correctly applied AEDPA in concluding that, in denying Floyd post-conviction relief in state court, those courts unreasonably applied clearly-established federal law. AFFIRMED.

I.

On 26 November 1980, William Hines, a white male, was found nude, stabbed to death inside the bedroom of his apartment on Governor Nicholls Street, in the French Quarter. The apartment had no signs of forced entry or evidence of burglary. One glass of alcohol was in Hines' bedroom; another, in his kitchen; and his wounds indicated he was stabbed while lying down.

Detective John Dillmann, the lead detective for the murder investigation, found the scene demonstrated a strong likelihood Hines was murdered by a welcome visitor with whom he shared a drink and had sexual relations. Accordingly, police dusted for fingerprints whiskey bottles, the glass of alcohol in Hines' kitchen, and the glass of alcohol on his nightstand.

Along that line, a crime-scene photograph of Hines' kitchen shows, among other items, a wine bottle and two whiskey bottles on the kitchen table. In addition, the crime-scene technician’s report states:

TECH. T. SEUZENEAN DUSTED

SEVERAL WHISKEY BOTTLES —Neg. RESULTS

DUSTED /—WHISKEY BOTTLE AND LIFTED—2 PARTIAL LATENT PRINTS

DUSTED /—WHISKEY GLASS FROM NIGHT TABLE IN BEDROOM—Neg. RESULTS

DUSTED /—WHISKEY GLASS FROM KITCHEN TABLE—Neg. RESULTS

Accordingly, it appears the "DUSTED ... WHISKEY GLASS FROM KITCHEN TABLE", but not shown in the photograph, was on the table where the dusted whiskey bottles were located. (To repeat, and as emphasized by the dissent at 248–49, no whiskey glass is visible on the table in the photograph. Myriad items shown on the table prevent conclusively determining whether a whiskey glass was there. But, as shown above, the technician’s report states: "DUSTED /—WHISKEY GLASS FROM KITCHEN TABLE".)

In any event, the relatively close proximity of the whiskey glass and the dusted whiskey bottle from which two prints were lifted (the whiskey bottle) is critical in our analysis. This is especially true for Detective Dillmann’s erroneous related testimony at trial, in which he: stated "there were two highball glasses filled with a liquid on each side of the bed"; and made no reference to the whiskey glass in the kitchen.

Along that line, the dissent at 249 states "the majority has decided, because it fits its narrative, to credit the tech over Dillmann". The dissent’s conclusion that the detective’s testimony and the technician’s report have comparable credibility is contrary to the State’s narrative, not ours. The State, in its opening brief at 16, acknowledges that the detective’s testimony about the glasses, "rendered for the first time a full year and a half after the crime, [and] directly contradicted by Crime Scene Tech Tim Suzeneau’s report", is less credible than the technician’s report. Likewise, at oral argument in our court, the State maintained the technician’s report, "generated on the day of the offense", was more accurate than the detective’s testimony, "recollected at trial ... a little over a year after the incident".

In the alternative, the dissent at 248–49 asserts a possibility the detective’s testimony and contradictory technician’s report were both accurate because there may have been one glass in the kitchen and two in the bedroom. But, nothing in the record supports this theory of three whiskey glasses being discovered at the Hines scene.

In sum, in its opening brief and at oral argument, the State maintained the crime scene technician’s report included a detailed list of all collected evidence. Again, the report included only two whiskey glasses: one from the kitchen and one from the bedroom.

Police also collected hair, appearing to be a black person’s, from Hines' bedsheets. But, because Hines had been dead for at least 24 hours prior to his body’s being discovered, any evidence of seminal fluid or spermatozoa on, or in, his body was undetectable.

Following multiple interviews, Detective Dillmann learned Hines was gay and frequented gay establishments in the French Quarter. And, the detective’s report, and subsequent testimony, provided that John Clegg, a close friend of Hines and the last known person to see him alive, had advised the detective that Hines "frequently had sexual relations with both black and white males".

At 4:45 a.m. on 28 November, only two days after the discovery of Hines' body, Rodney Robinson, a black male, was found dead at the Fairmont Roosevelt Hotel in downtown New Orleans, just one mile from Hines' apartment. In the hours preceding his death, Robinson had visited several bars with his friend David Hennessy. After Robinson, according to Hennessy, drove him to his home at around 3:15 a.m., Robinson said he was returning to his hotel for the night. Just 90 minutes later, he was found nude, stabbed to death, in a hallway in his hotel.

In their investigation, officers found the locks on Robinson’s hotel-room door functional; glasses containing alcohol remained on end tables next to his bed; and articles of clothing were scattered about the room. Consequently, they believed Robinson was murdered after sharing a drink and having sexual relations with his killer. Detectives' interview of Hennessey revealed Robinson was gay.

Police discovered physical evidence of: blood stains along the hallway wall; a blood-stained blue-knit cap in the hallway relatively near Robinson’s body; seminal fluid on a tissue discovered near his bed; and spermatozoa and seminal fluid in his body. Additionally, police discovered a black person’s hair—determined later not to be Robinson’s—on the blue-knit cap. Further, hotel guests staying nearest Robinson’s room reported hearing screams and rapid footsteps in the hallway; and a hotel security guard reported seeing a black male running from the back door of the hotel shortly before the police arrived. Detective Michael Rice, lead detective for the murder investigation, believed the guard "witnessed the perpetrator ... making good his escape".

Detective Dillmann considered the similarities in the Hines and Robinson crimes—comparable defensive wounds, lack of forced entry, each victim’s being gay, glasses of alcohol near each victim’s bed (again, for Hines' murder, only one glass was near his bed; the other was in the kitchen, as was the whiskey bottle), and evidence of sexual relations between the perpetrator and victim—to conclude the same perpetrator was responsible for both murders. Initially, investigators unsuccessfully pursued black, male suspects. John Floyd, a white male, then 32, lived as a "drifter" in New Orleans at the time of the murders. He was a heavy drinker and drug-user, and frequented numerous bars in the French Quarter. On 29 November, one day after the discovery of Robinson’s body, Detective Dillmann received a tip from Harold Griffin that Floyd had recently made incriminating statements linking him to Robinson’s murder.

Griffin reported that, after drinking with Floyd at the Louisiana Purchase Bar from 10:00 p.m. on 28 November (approximately 17 hours after Robinson’s body was found) until 5:00 a.m. the next day, 29 November, Floyd asked Griffin to accompany him to the detoxification center at Charity Hospital. Griffin testified that, during their walk to the hospital, Floyd told him "he heard that perhaps going to the Detox Center would be the next best thing to keep from being held accountable for doing something wrong"; Floyd then asked Griffin if he had "heard of the stabbing at the Fairmont"; and he replied "No".

Later that day, Griffin learned of Robinson’s murder as covered in the 29 November morning edition of the Times Picayune , and reported his conversation with Floyd to the New Orleans Police Department (NOPD), finding it peculiar Floyd knew of the murder prior to the paper’s publication. But, the paper had published a story on Robinson’s murder in its 28 November evening edition, prior to Floyd’s statements to Griffin on the 29th.

Following up on Griffin’s tip, Detective Dillmann questioned French Quarter bar owner Steven Edwards, who advised that Floyd made incriminating statements linking him to Hines' murder. According to Edwards, in late November he encountered Floyd "drinking heavily" and refused him service at the Mississippi River Bottom bar. Edwards testified: he told Floyd, "you know you are barred from the f...ing bar"; Floyd then threatened, "[d]on't come f...ing with me. I already wasted one person"; Edwards asked, "Who? Bill Hines?"; and Floyd replied, "Yeah, on Governor Nichol[l]s".

Based on these statements to Griffin and Edwards, Floyd was made a suspect in the two murders. After receiving a positive identification from both Griffin and Edwards, Detective Dillmann and a NOPD officer found Floyd drinking at the Louisiana Purchase Bar. They purchased Floyd at least one drink before arresting and transporting him to NOPD’s homicide office.

There, Detective Dillmann began interrogating Floyd. He testified Floyd initially denied any involvement in the two murders, but, within 30 minutes, became very emotional about his drinking and drug-use, and confessed verbally to killing Hines and Robinson.

Following Floyd’s admissions, the detective called Detective Rice, and they procured Floyd’s signed confessions to both murders. Detective Rice witnessed Detective Dillmann take the Hines confession, and Detective Dillmann did the same for Detective Rice’s taking the Robinson confession. The confessions were taken on the evening of 19 January 1981, and had markedly similar descriptions such as: drinking and having sexual relations with the victims before fatally stabbing them in response to each man’s wanting to "f... [him]".

Indicted on two counts of second-degree murder, Floyd waived his right to a jury trial, and proceeded to a joint bench trial in Orleans Parish Criminal District Court, maintaining a defense of third-party guilt. For the Hines murder the State presented: Floyd’s confession to murdering Hines; Detective Dillmann’s testimony that the confession was credible; and Edwards' testimony regarding Floyd’s threats to him. For the Robinson murder, the State presented: Floyd’s confession to murdering Robinson; Detective Rice’s testimony related to Floyd’s Robinson confession; Griffin’s testimony regarding Floyd’s statements to him; and testimony by Byron Reed, Floyd’s acquaintance and former sexual partner, that Floyd made an incriminating statement about the Robinson murder to him.

For the Hines charge, the defense presented NOPD criminalist Daniel Waguespack’s testimony that Floyd was excluded from the blood and hair discovered at Hines' residence. (The hair from the Hines scene has since been lost, preventing DNA testing. It appears this was part of the evidence destroyed during Hurricane Katrina in 2005, after Detective Dillmann took the police files to use in writing a book about, inter alia , the investigation, as discussed infra .) For the Robinson charge, the defense presented: NOPD criminalist Alan Sison’s testimony, discussed infra , that the blood and seminal fluid from the Robinson scene were not attributable to Floyd; testimony from Patricia Daniels, the Parish of Orleans coroner’s office’s medical technologist, that Floyd was excluded from all seminal fluid discovered in Robinson’s body; and the Fairmont’s security guard’s testimony that she repeatedly attempted to report seeing a black male running from the hotel on the night of the murder. For both charges, the defense presented: Floyd’s testimony his confessions were untrue and a result of Detective Dillmann’s "beating" him during the interrogation; and testimony by Dr. Marvin Miller about Floyd’s susceptibility to coercion.

In short, the State did not present any physical evidence linking Floyd to Hines' murder. Rather, Detective Dillmann testified the evidence of the glasses of whiskey discovered in Hines' apartment (as discussed supra , the detective erroneously testified the glasses were discovered "on each side of the bed"; instead, the crime-scene technician’s report demonstrates one glass was found in the kitchen, where the whiskey bottle was located, and one glass was found in the bedroom), the placement of clothing in his residence, and the position of Hines' body corroborated "perfectly" the descriptions in Floyd’s confession, and supported its credibility. For example, the detective testified: Floyd’s statement in his confession that "[w]e were both drinking" was consistent with the fact that investigators "found two drinking glasses in the bedroom of the apartment"; and Floyd’s descriptions in his confession of Hines' falling "on the floor next to the bed" after he stabbed him, corroborated the "position of the body where it fell off the bed".

And, as noted, Edwards testified about Floyd’s incriminatory threats to him. The trial judge found Floyd’s incriminating statements, including in his confession, sufficient to support his guilt for Hines' murder, and convicted him of second-degree murder.

Analogous to the Hines charge, the State did not present any physical evidence linking Floyd to Robinson’s murder. To support his guilt, the State presented evidence of Floyd’s confession, and of the incriminating statements linking him to that murder.

The defense presented physical evidence to contradict Floyd’s confession to murdering Robinson after sexual relations. NOPD Criminalist Alan Sison testified the seminal fluid discovered in Robinson’s hotel room was attributable to an individual with type-A blood; medical technologist Daniels, the seminal fluid found in Robinson’s body was also attributable to an individual with type-A blood. Floyd, however, has type-B blood; Robinson had type-O. Further, Sison testified the black person’s hair discovered in the blue-knit cap, found in the hallway relatively near Robinson’s body, was "dissimilar" to Floyd’s long blonde hair.

Obviously, there was more exculpatory evidence to present for Robinson’s murder than for Hines', in part because Hines' body was not discovered until at least 24 hours after his death. Although Floyd contemporaneously confessed to murdering Hines and Robinson, and investigators presumed the same perpetrator committed both crimes, the trial judge found Floyd’s confession and alleged incriminating statements insufficient to support his guilt for the Robinson murder.

After Floyd was found guilty of Hines' murder, but simultaneously acquitted of Robinson’s, he was sentenced to life imprisonment without parole. The Supreme Court of Louisiana affirmed his conviction and sentence. State v. Floyd , 435 So.2d 992 (La. 1983).

From 1983 until 2006, Floyd wrote numerous letters to individuals and organizations, asserting his innocence. In 2006, 23 years after his conviction was affirmed by the Supreme Court of Louisiana, the Innocence Project of New Orleans (IPNO) assisted Floyd in filing his first state-court application for post-conviction relief. It was supported by newly-discovered evidence, including: pre-trial fingerprint-comparison results from the Hines scene marked "NOT JOHN FLOYD" and "NOT VICTIM"; pre-trial fingerprint-comparison results from the Robinson scene listed "NOT DAVID HENNESSEY", "NOT VICTIM", and "NOT JOHN FLOYD"; post-trial DNA-test results from hair discovered at that scene; Clegg’s post-trial affidavit, stating Detective Dillmann misrepresented Clegg’s pre-trial statement that Hines had a distinct sexual preference for black males (the Clegg statement); Detective Dillmann’s post-conviction statements, including the statement in his 1989 book, Blood Warning: The True Story of the New Orleans Slasher , that he showed Floyd "two of the grisliest shots" in an attempt to "crack him"; evidence of the detective’s subsequent mistreatment of suspects; and Floyd’s I.Q. score of 59, discovered through tests not existing at the time of trial.

In 2010, the Criminal District Court for the Parish of Orleans denied relief from the bench, without providing reasons. Likewise, the Supreme Court of Louisiana denied relief in a 4-3 decision, without providing reasons. Floyd v. Cain , 62 So.3d 57 (La. 2011). But, reasons were assigned in a detailed dissent, which opined, inter alia , "the exculpatory value of the fingerprint evidence is sufficient to undermine confidence in the outcome of Floyd’s trial, thus satisfying the requirements for a new trial set forth in Brady ". Id . at 59. (Johnson, J., dissenting).

Following the state-court decisions, Floyd filed in 2011 for federal habeas relief under 28 U.S.C. § 2254, maintaining, inter alia , the State withheld favorable, material evidence in violation of Brady . But, in December 2012, the district court adopted the magistrate judge’s report and recommendation (R & R) to deny Floyd’s petition as untimely under AEDPA.

Floyd’s January 2013 motion to alter and amend the decision was considered in the light of the Supreme Court’s superseding McQuiggin v. Perkins decision. 569 U.S. 383, 386, 133 S.Ct. 1924, 185 L.Ed.2d 1019 (2013) (holding AEDPA’s time-bar overcome by a valid actual-innocence claim). To overcome the time-bar, Floyd presented such a claim: in the light of newly-discovered exculpatory evidence related to the Hines and Robinson murders, he was actually innocent of murdering Hines. The district court vacated the denial and remanded the petition to the magistrate judge for a R & R in the light of McQuiggin .

The magistrate judge’s resulting R & R recommended: Floyd failed to meet his burden to demonstrate actual innocence; and, accordingly, his petition should be dismissed with prejudice, without considering the merits of his constitutional claims. Floyd v. Cain , 2016 WL 4799093, at *26 (E.D. La. 14 Sept. 2016). But, in a 67-page opinion providing an exhaustive analysis of Floyd’s actual-innocence claim, the district court concluded that, in the light of the newly-discovered evidence, "any reasonable, properly instructed juror, evaluating this case with the requisite caution and care, would reasonably doubt Floyd’s guilt of the murder of William Hines". Id . Having concluded that Floyd had overcome the time-bar, the court remanded the petition to the magistrate judge for a R & R on the merits. Id .

Regarding Floyd’s constitutional claims, the subsequent R & R recommended granting Floyd’s Brady claim. Floyd v. Vannoy , 2017 WL 1837676, at *4 (E.D. La. 8 May 2017). In a 33-page opinion, the district court approved and adopted the R & R, but added additional reasons for the decision. Id . at *1. For example, although the R & R did not find it necessary to consider Clegg’s affidavit and his pre-trial statement in the light of the fingerprint-comparison results being sufficient to support Floyd’s Brady claim, the district court opinion considered them to conclude Clegg’s statement to Detective Dillmann was additional Brady material. Id . at *12–16.

The two district-court opinions, totaling 100 pages, provide far greater, and much more graphic, factual detail than does this opinion. As with its decision regarding the time-bar, the district court’s merits opinion provides an exhaustive analysis of Floyd’s Brady claims and the unreasonableness of the state courts' contrary decisions. Id . at *5–16. In granting relief, the court concluded: the State withheld favorable, material evidence in violation of Brady (the fingerprint-comparison results from the Hines scene and the Clegg statement); and the state-court decisions denying relief were an unreasonable application of clearly-established federal law. Id. at *16. Accordingly, Floyd was awarded habeas relief, with the State’s being ordered to retry, or release, him within 120 days of the decision. Id . The district court stayed its order, pending resolution of this appeal. Floyd v. Vannoy , 2017 WL 2688082, at *2–4 (E.D. La. 22 June 2017).

II.

"In a habeas corpus appeal, we review the district court’s findings of fact for clear error and its conclusions of law de novo ." Lewis v. Thaler , 701 F.3d 783, 787 (5th Cir. 2012) (quoting Busby v. Dretke , 359 F.3d 708, 713 (5th Cir. 2004) ). The State claims: Floyd failed to meet the necessary actual-innocence burden to overcome the time-bar for his habeas application; and, in the alternative, the state-court denials of post-conviction relief were, pursuant to AEDPA, neither "contrary to", nor "involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States". 28 U.S.C. § 2254(d)(1).

Accordingly, our review encompasses three legal standards. First, actual innocence is established through demonstrating that, in the light of newly-discovered evidence, "it is more likely than not that no reasonable juror would have found petitioner guilty beyond a reasonable doubt". Schlup v. Delo , 513 U.S. 298, 327, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995) ; see also McQuiggin , 569 U.S. at 399, 133 S.Ct. 1924. Second, Brady is violated when: the State suppresses evidence; that is favorable to his defense; and material to guilt or punishment. E.g. , Brady , 373 U.S. at 87, 83 S.Ct. 1194. And third, a state-court decision is an unreasonable application of clearly-established federal law only if fairminded jurists could not disagree that the decision was inconsistent with Supreme Court precedent. E.g. , Harrington v. Richter , 562 U.S. 86, 101, 131 S.Ct. 770, 178 L.Ed.2d 624 (2011).

A.

Floyd filed for state post-conviction relief in March 2006, over 23 years after his conviction became final, and contrary to AEDPA’s requiring seeking such relief within one-year of the conviction. 28 U.S.C. § 2244(d)(1). Moreover, where, as here, the conviction preceded AEDPA’s 26 April 1996 enactment, the limitations period expired one-year from that date. Flanagan v. Johnson , 154 F.3d 196, 200 (5th Cir. 1998) (citing United States v. Flores , 135 F.3d 1000, 1006 (5th Cir. 1998) ).

Nonetheless, in the "extraordinary case", McQuiggin , 569 U.S. at 393, 133 S.Ct. 1924 (quoting Schlup , 513 U.S. at 324, 115 S.Ct. 851 ), in which a prisoner asserts a "credible showing of actual innocence", he may overcome the time-bar, and have his claims considered on the merits, id. at 392, 133 S.Ct. 1924 ; House v. Bell , 547 U.S. 518, 537, 126 S.Ct. 2064, 165 L.Ed.2d 1 (2006) ; Schlup , 513 U.S. at 316, 115 S.Ct. 851. In that regard, the district court concluded: Floyd’s actual-innocence claim was valid; and, accordingly, his petition was not time-barred. Floyd , 2016 WL 4799093, at *26.

Of considerable note, in its reply brief on appeal, the State does not expressly challenge Floyd’s innocence. Instead, it has offered him two pleas during the pendency of his federal habeas application, and concedes "it does not take issue with Floyd being permanently released from custody". The State also concedes it challenges the actual-innocence ruling only because of the precedent it sets. (A strong argument can be made that, for the actual-innocence ruling, the State’s concessions constitute judicial estoppel, precluding its being challenged.)

In any event, the "fundamental miscarriage of justice exception" permits prisoners with an otherwise untimely application to pursue their constitutional claims. McQuiggin , 569 U.S. at 392–93, 133 S.Ct. 1924. This exception’s demanding standard requires "evidence of innocence so strong that a court cannot have confidence in the outcome of the trial unless the court is also satisfied that the trial was free of nonharmless constitutional error". Id . at 401, 133 S.Ct. 1924 (quoting Schlup , 513 U.S. at 316, 115 S.Ct. 851 ). The standard is seldom met. House , 547 U.S. at 538, 126 S.Ct. 2064 (citing Schlup , 513 U.S. at 327, 115 S.Ct. 851 ).

An actual-innocence claim is only established when it is shown that, in the light of newly-discovered evidence, "it is more likely than not that no reasonable juror would have found petitioner guilty beyond a reasonable doubt". Schlup , 513 U.S. at 327, 115 S.Ct. 851 ; see also McQuiggin , 569 U.S. at 399, 133 S.Ct. 1924. Therefore, a credible claim must be supported by "new reliable evidence—whether it be exculpatory scientific evidence, trustworthy eyewitness accounts, or critical physical evidence—that was not presented at trial". Schlup , 513 U.S. at 324, 115 S.Ct. 851. Actual innocence is then demonstrated only when the court scrutinizes the likely impact on reasonable jurors of "the overall, newly supplemented record", House , 547 U.S. at 538, 126 S.Ct. 2064, to conclude that, in the light of all evidence—both the evidence presented at trial and that newly discovered—"no juror, acting reasonably, would have voted to find [petitioner] guilty beyond a reasonable doubt", McQuiggin , 569 U.S. at 386, 133 S.Ct. 1924 (quoting Schlup , 513 U.S. at 329, 115 S.Ct. 851 ). As re-stated in McQuiggin , the court must conclude "it is more likely than not that no reasonable juror would have convicted [the petitioner]". Id . at 395, 133 S.Ct. 1924 (quoting Schlup , 513 U.S. at 329, 115 S.Ct. 851 ) (alteration in original).

Our court does not consider habeas relief based on "freestanding claims of actual innocence". In re Swearingen , 556 F.3d 344, 348 (5th Cir. 2009). Instead, a successful actual-innocence claim provides a "gateway" for the petitioner to proceed on the merits. House , 547 U.S. at 536, 126 S.Ct. 2064. (Therefore, as also discussed at length in the dissent at 246–47, Floyd’s successful actual-innocence claim permits our considering the merits of his constitutional claim: the State withheld favorable, material evidence, in violation of Brady . McQuiggin , 569 U.S. at 386, 133 S.Ct. 1924. And, for review of the Brady claim and the concomitant AEDPA unreasonableness standard for that claim, a great deal of the newly-discovered evidence and the withheld evidence overlaps.)

To establish actual innocence, Floyd presents substantial exculpatory evidence related to both murders. As discussed supra , although he confessed to murdering Hines and Robinson, he was convicted solely of Hines' murder. Therefore, his habeas petition centers on that conviction. But, the district court concluded, and we agree, that, because Floyd’s confessions are intertwined, evidence demonstrating Floyd falsely confessed to murdering Robinson supports his assertions he likewise did so for Hines. Floyd , 2016 WL 4799093, at *2. In other words, newly-discovered evidence further and conclusively exculpating Floyd of Robinson’s murder—undermining both confessions—is relevant to his actual-innocence claim because it supports Floyd’s assertions his confessions were false.

At trial, the State did not present any physical evidence linking Floyd to either murder. His conviction for Hines' murder was based solely on his confession and threat to Edwards. Accordingly, Floyd’s actual-innocence claim hinges on whether, in the light of the items he advances as newly-discovered evidence, any reasonable juror could rely solely on the evidence presented at trial—Floyd’s confession and threat to Edwards—to find Floyd guilty beyond a reasonable doubt. McQuiggin , 569 U.S. at 386, 133 S.Ct. 1924 (citing Schlup , 513 U.S. at 329, 115 S.Ct. 851 ).

The claimed newly-discovered evidence is: fingerprint-comparison results of fingerprints lifted from the Hines scene; fingerprint-comparison results and DNA-test results from fingerprints and hair discovered at the Robinson scene and on his vehicle; Detective Dillmann’s misconduct in later interrogations and tests demonstrating Floyd’s susceptibility to coercion; and an affidavit from Clegg.

1.

The fingerprint-comparison results exclude Floyd and Hines as contributors of the fingerprints lifted from the whiskey bottle discovered at the Hines crime scene. In 2008, IPNO obtained an envelope containing the fingerprints, and copies of the NOPD logbook chronicling them. The envelope and logbook conveyed that police initially lifted the fingerprints from the Hines scene, performed a fingerprint-comparison test, and logged the fingerprints "NOT VICTIM" and "NOT JOHN FLOYD". Although police possessed this information at the time of trial, it was neither presented as evidence nor disclosed to the defense.

For the requirement that actual-innocence claims be supported by "new reliable evidence", Schlup , 513 U.S. at 324, 115 S.Ct. 851, the State’s assertion that this fingerprint evidence is not "new", and, therefore, cannot support Floyd’s claim, distorts the clear meaning of the Schlup standard. Id . at 332–33, 339–40, 115 S.Ct. 851. Although the fingerprint-comparison results existed at the time of the joint bench trial, the results were not presented, were withheld from both the prosecution and the defense, and could not, therefore, have affected the trial judge’s analysis of Floyd’s guilt. Accordingly, because this information was not presented at trial, and remained unknown to the prosecution, defense, and trial judge throughout the trial, it is "new" evidence. Id. at 339, 115 S.Ct. 851.

Along that line, the Court, in McQuiggin , held no threshold diligence requirement applies to actual-innocence claims; the delay is simply a factor in the court’s reliability evaluation. 569 U.S. at 399, 133 S.Ct. 1924. Scientific-based evidence, like the fingerprint-comparison results, is less susceptible to manipulation and, therefore, is appropriately considered reliable evidence despite the time lapse. See id . at 399–400, 133 S.Ct. 1924.

2.

The Robinson DNA-test results and fingerprint-comparison results exclude Floyd and Robinson as the contributors of the hair and fingerprints discovered at the Robinson scene. Parallel to the Hines charge, the State did not present physical evidence linking Floyd to Robinson’s murder, and his defense centered on third-party guilt. The newly-discovered evidence of the fingerprint-comparison results exclude Robinson, Hennessey, and Floyd as contributors of the fingerprints lifted from the drinking glasses next to Robinson’s bed and the passenger-side door of his vehicle.

Although not presented at trial, police recorded the fingerprint-comparison results of fingerprints lifted from the glasses as belonging to neither Robinson, Hennessey, nor Floyd. Additionally, police labeled the fingerprints lifted from Robinson’s vehicle, "NOT ... DAVID HENNESSEY", "NOT VICTIM", and "NOT JOHN FLOYD". Further, NOPD’s initial analysis of hair lifted from Robinson’s bed concluded it belonged to a black male other than Robinson; and Floyd presents the post-trial DNA evidence, further excluding him as the source of that hair.

Similar to the earlier-discussed newly-discovered evidence of fingerprint-comparison results from the Hines scene, this evidence meets the "new reliable" Schlup standard because: it is scientific-based evidence that is not easily manipulated; was unknown to the defense at the time of the trial; and was not presented at trial. McQuiggin , 569 U.S. at 400, 133 S.Ct. 1924 ; Schlup , 513 U.S. at 324, 115 S.Ct. 851.

Regarding the requirement that evidence presented at trial must be considered in the light of all newly-discovered evidence, House , 547 U.S. at 538, 126 S.Ct. 2064, any evidence exculpating Floyd of Robinson’s murder—undermining his confession—supports his assertion he falsely confessed to, and is actually innocent of, Hines'. Floyd confessed to killing Robinson after having sexual relations with him. The physical evidence presented at trial by the defense, however, refuted Floyd’s confession, and demonstrated a likelihood Robinson was killed by a black male with type-A blood. Floyd’s newly-discovered evidence regarding Robinson further excludes him from the Robinson scene, invalidates his confession, and links a third party to that scene.

At trial, no physical evidence was presented to contradict Floyd’s confession about Hines. Detective Dillmann testified the evidence discovered at the Hines scene corroborated Floyd’s statements, and proved his confession credible. Specifically, the detective testified the evidence of the "glasses filled with a liquid on each side of the bed" corroborated Floyd’s confession to drinking with Hines before killing him.

But, as discussed supra , the testimony about the location of the glasses is incorrect; one was found in Hines' bedroom and one in his kitchen, where the whiskey bottle was found. According to the detective’s testimony, these glasses were one of the three details proving Floyd’s confession credible. Again, however, his testimony was incorrect regarding the location of the glasses: one of the glasses, which Detective Dillmann testified corroborated Floyd’s statement that he and Hines had been drinking together, was found not by the bed, but in the kitchen with the whiskey bottle, which had partial prints from neither Floyd nor Hines but a third party.

The newly-discovered evidence of the fingerprint-comparison results from the whiskey bottle in Hines' residence could be found by a reasonable juror to refute Floyd’s confession, link a third-party to the crime scene, and impeach the detective’s testimony. (Although the dissent at 248–49 states the murder scene excluded the kitchen, investigators considered Hines' entire apartment in their crime-scene investigation. Moreover, police selected multiple items from the kitchen to dust for prints, and Detective Dillmann testified about the importance of the evidence of "two highball glasses filled with a liquid". Again, one of the glasses, according to the State and the crime scene technician’s report, was discovered in Hines' kitchen.)

Confessions are generally considered strong evidence of guilt, and a sound confession alone may significantly influence a juror’s decision. Murray v. Earle , 405 F.3d 278, 295 (5th Cir. 2005). "Confession evidence (regardless of how it was obtained) is so biasing that juries will convict on the basis of confession alone." Id . Nonetheless, the credibility of Floyd’s confession must be evaluated in the light of the newly-discovered evidence excluding the possibility Floyd committed the crimes to which he confessed. McQuiggin , 569 U.S. at 386, 133 S.Ct. 1924 (citing Schlup , 513 U.S. at 329, 115 S.Ct. 851 ). It follows that, in the light of this newly-discovered contradictory physical evidence, it is more than likely a reasonable, informed juror would reasonably doubt the credibility of Floyd’s confessions.

3.

Floyd testified at trial that Detective Dillmann "slapp[ed] [him] on the side of the head"; "hit [him] a bunch of times"; "kick[ed] [him] on the side of the head with his boots" and "threatened to put [his] head through the brick wall and throw [him] out through the window". He further testified he immediately began agreeing to anything the detective asked of him after the detective told him that he "could kill [Floyd] and get by with it".

In that regard, Floyd asserts newly-discovered evidence of, inter alia , the detective’s abuse during an interrogation for a crime after the Hines and Robinson murders, his later admissions to showing Floyd crime-scene photographs, and Dr. Gregory DeClue’s related examination, discussed infra , undermine the validity of Floyd’s confession, in support of his actual-innocence claim.

Floyd presents newly-discovered evidence of the detective’s subsequent mistreatment of suspects. In State v. Seward , the Supreme Court of Louisiana ruled a confession coerced, finding the State failed to prove the defendant was not beaten during an interrogation led by Detective Dillmann. 509 So.2d 413, 415–18 (La. 1987). The suspect testified to similar descriptions of being hit in the head, kicked, and forced to the floor during the interrogation. Id . at 415.

Further, at trial, the State asserted Floyd’s detailed descriptions of both crimes proved his confessions credible. Now, Floyd asserts newly-discovered evidence of Detective Dillmann’s subsequently published 1989 book, Blood Warning: The True Story of the New Orleans Slasher , in which the detective describes showing Floyd "two of the grisliest shots" of the Hines crime scene in an effort to "crack him".

Along that line, the State asserted at trial that the credibility of Floyd’s confessions was demonstrated through his volunteering specific crime-scene details. These assertions are severely weakened by evidence that, during the interrogation, detectives provided Floyd with significant details about the crime scenes. Notably, Floyd’s descriptions regarding the position of Hines' body do not accurately describe the scene as found by police, but, rather, correspond to crime-scene photographs taken after Hines' body was moved.

Additionally, evidence of forensic psychologist Dr. DeClue’s 2009 examination of Floyd, employing methods not available at the time of trial, found Floyd had an I.Q. of 59 and communication skills of a "second or third grade[r]", rendering him "extremely vulnerable" to police coercion.

The credibility of Floyd’s confessions, and his trial testimony he was coerced by Detective Dillmann, are appropriately considered in the light of the newly-discovered evidence of: the detective’s conduct during a subsequent interrogation; Floyd’s observing photographs of the crime scene; and Dr. DeClue’s findings regarding Floyd’s susceptibility to coercion. House , 547 U.S. at 538, 126 S.Ct. 2064 (quoting Schlup , 513 U.S. at 327, 115 S.Ct. 851 ). Although jurors are likely to find confessions compelling, our court must make a "probabilistic determination" of the hypothetical jurors' opinions of the newly-discovered evidence, and voluntariness of Floyd’s confession. Id. (quoting Schlup , 513 U.S. at 329, 115 S.Ct. 851 ). Considering the evidence as a whole, it is likely a reasonable juror would doubt Floyd’s confession was "freely and voluntarily made", State v. Trudell , 350 So.2d 658, 661 (La. 1977), and, therefore, lacked credibility to alone establish his guilt beyond a reasonable doubt, House , 547 U.S. at 538, 126 S.Ct. 2064.

4.

The final newly-discovered evidence is presented through Clegg’s 2008 affidavit. According to Floyd, it undermines his guilt and casts doubt on Detective Dillmann’s investigative practices. At trial, the State supported Floyd’s guilt with the detective’s testimony that Clegg, a friend of Hines', stated Hines "frequently had sexual relations with both black and white males". But, in his 2008 affidavit, Clegg maintained: Hines' preference was for black males; pre-trial, he informed the detective of that preference; and Clegg was "very surprised" when Floyd (a white male) was arrested.

Regarding our court’s considering only "new reliable evidence" to support a claim of actual innocence, Schlup , 513 U.S. at 324, 115 S.Ct. 851, Clegg asserts in his affidavit that the detective manipulated Clegg’s initial statements. He was a close friend of Hines' and has no apparent connection to Floyd. The reliability of this new evidence is strengthened by the unlikelihood Clegg, a friend of the murder victim, would falsely assert a particular defendant did not fit the profile of the likely killer, in order to support the defendant’s innocence. House , 547 U.S. at 551, 126 S.Ct. 2064 (ruling witness' disinterest in aiding defendant supports credibility of post-conviction testimony). Further, reliability is not affected by the passage of time as Clegg has neither died, nor otherwise become unavailable for further questioning. E.g. , McQuiggin , 569 U.S. at 399–400 & n.4, 133 S.Ct. 1924.

The likely impact on reasonable jurors of Clegg’s pre-trial statements, as presented at trial by the detective, is considered with the newly-discovered evidence of Clegg’s contradictory affidavit. Id. at 386, 133 S.Ct. 1924. It is more than likely the evidence of the detective’s testimony, asserting a possibility Floyd’s profile aligned with that of men with whom Hines frequently had sexual relations, would have little persuasive value in the light of Clegg’s pre-trial statement that he understood his friend to have a distinct preference for black males. In other words, in the light of the newly-discovered evidence through Clegg’s affidavit, no reasonable juror would have relied upon Clegg’s pre-trial statement—that Floyd did not fit the likely profile of the perpetrator—to adequately support Floyd’s guilt.

Additionally, a statement from the victim’s friend, asserting the defendant did not fit the profile of the likely killer, would more than likely affect a reasonable juror’s analysis of Floyd’s guilt. In the light of the newly-discovered evidence of the detective’s alleged misrepresentations, Clegg’s stating Hines' preference for black males casts doubt on Floyd’s guilt, and supports his third-party-guilt defense.

In sum, for the actual-innocence claim, Floyd’s guilt was contingent solely on his confession and alleged threat to Edwards. And, the persuasive impact of Floyd’s confessions must be scrutinized in the light of all the evidence, presented at trial and new. Id. at 386, 133 S.Ct. 1924 (citing Schlup , 513 U.S. at 329, 115 S.Ct. 851 ). Floyd overcomes the time-bar if, in the light of the newly-discovered evidence, no reasonable juror would determine the confession and alleged threat to Edwards were sufficient to establish Floyd’s guilt beyond a reasonable doubt. Id . at 395, 133 S.Ct. 1924 (citing Schlup , 513 U.S. at 329, 115 S.Ct. 851 ).

In the light of the newly-discovered evidence of: the fingerprint-comparison analysis excluding Floyd from the Hines scene; the Robinson-related fingerprint-comparison results and DNA tests further discounting Floyd’s confession; Detective Dillmann’s improper interrogation techniques; Floyd’s vulnerability to coercion; and Clegg’s affidavit maintaining Floyd did not fit the likely profile of the perpetrator, no reasonable juror would find Floyd’s confession and Edwards' testimony about a threat sufficient to support Floyd’s guilt beyond a reasonable doubt. Re-stated, because, in the light of the newly-discovered evidence, no reasonable juror, considering the record as a whole, would vote to convict Floyd of Hines' murder, Floyd’s actual-innocence claim is sufficient to overcome the untimeliness of his habeas application. Id . at 386, 133 S.Ct. 1924.

B.

Again, "a claim of ‘actual innocence’ is not itself a constitutional claim, but instead a gateway through which a habeas petitioner must pass to have his otherwise barred constitutional claim considered on the merits". Herrera v. Collins , 506 U.S. 390, 404, 113 S.Ct. 853, 122 L.Ed.2d 203 (1993). In considering Floyd’s constitutional/Brady claim on the merits, we leave behind the foregoing actual-innocence analysis and start afresh. The claim by the dissent at 246 that our ensuing AEDPA/Brady analysis is "colored by the gateway" actual-innocence issue is totally incorrect.

In order to obtain habeas relief, Floyd must show the state-court denial of post-conviction relief was, as required by AEDPA and as discussed in part II. C., infra , an unreasonable application of clearly-established federal law. To determine whether Floyd meets that demanding burden, the better procedure in this instance is to first determine whether he successfully demonstrates the merits of his constitutional claim: the State, in violation of Brady , withheld favorable, material evidence of fingerprint-comparison results and the statement from Clegg.

For determining whether, pursuant to AEDPA, the state-court denial of habeas relief was an unreasonable application of clearly-established federal law, there is no state-court opinion available for our review. As a result, we must consider the arguments or theories that "could have supported, the state court decision; and ... whether it is possible fairminded jurists could disagree that those arguments or theories are inconsistent with the holding in a prior decision of [the Supreme Court]". Richter , 562 U.S. at 102, 131 S.Ct. 770.

The dissent at 247 asserts our bifurcated analysis demonstrates a fundamental misunderstanding of AEDPA. This assertion, however, ignores the Supreme Court’s holding that, when our court determines a previously denied claim is meritorious, AEDPA demands we provide a "further explanation that the state court’s contrary decision was unreasonable". See id . at 101, 131 S.Ct. 770. For doing so, we may employ any methodology, including an initial "review [of] the state court decision de novo before applying the AEDPA standard of review", to determine whether AEDPA forecloses federal habeas relief. Lockyer v. Andrade , 538 U.S. 63, 71, 123 S.Ct. 1166, 155 L.Ed.2d 144 (2003). In that regard, the Supreme Court has not foreclosed an initial de novo review of a state-court decision before addressing whether that decision was unreasonable, as required under AEDPA.

Contrary to the dissent’s assertion at 247–48, we do not contend the Court has "blessed" this bifurcated analysis or any other methodology—only that it has not foreclosed our approach. In that regard, Lockyer holds there is no mandatory standard-operating procedure for applying the AEDPA standard:

The Ninth Circuit requires federal habeas courts to review the state court decision de novo before applying the AEDPA standard of review. We disagree with this approach. AEDPA does not require a federal habeas court to adopt any one methodology in deciding the only question that matters under § 2254(d)(1) -whether a state court decision is contrary to, or involved an unreasonable application of, clearly established federal law. See Weeks v. Angelone , 528 U.S. 225 [120 S.Ct. 727, 145 L.Ed.2d 727] (2000).

Id . (citations omitted).

The dissent at 247–48 misemploys Lockyer to support its contention that the Court prohibits our bifurcated analysis. The dissent correctly notes, as we provide above, that Lockyer states: "The Ninth Circuit requires federal habeas courts to review the state court decision de novo before applying the AEDPA standard of review. We disagree with this approach". Id . But, the dissent fails to include the above-provided crucial sentence from Lockyer that follows the dissent’s quotation: "AEDPA does not require a federal habeas court to adopt any one methodology in deciding the only question that matters under § 2254(d)(1) —whether a state court decision is contrary to, or involved an unreasonable application of, clearly established federal law", 538 U.S. at 71, 123 S.Ct. 1166 (internal citation omitted). Again, in this instance, due to the complexity of Floyd’s Brady claim, it is the better practice to first consider the merits of the claim before considering whether the state-court denial stemmed from an unreasonable application of Brady . The Brady Court held "suppression by the prosecution of evidence favorable to an accused upon request violates due process where the evidence is material either to guilt or to punishment, irrespective of the good faith or bad faith of the prosecution". 373 U.S. at 87, 83 S.Ct. 1194. But as the Brady rule was developed, its scope was extended to all favorable evidence possessed by the prosecution, even without a request from the defense, United States v. Agurs , 427 U.S. 97, 107, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976), and established a prosecutorial duty "to learn of any favorable evidence known to the others acting on the government’s behalf in the case, including the police", Kyles v. Whitley , 514 U.S. 419, 437, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995). On the other hand, the prosecution’s failure to disclose such evidence constitutes a Brady violation only where the "nondisclosure was so serious that there is a reasonable probability that the suppressed evidence would have produced a different verdict". Strickler v. Greene , 527 U.S. 263, 281, 119 S.Ct. 1936, 144 L.Ed.2d 286 (1999). Accordingly, to prevail on his Brady claim, Floyd must prove: the State suppressed evidence; it was favorable to his defense; and it was material to his guilt or punishment. E.g. , Brady , 373 U.S. at 87, 83 S.Ct. 1194 ; see also United States v. Cessa , 861 F.3d 121, 128 (5th Cir. 2017). The district court concluded the fingerprint-comparison results from the Hines scene, fingerprint-comparison results from the Robinson scene, and Clegg’s pre-trial statement were Brady material. Floyd , 2017 WL 1837676, at *5.

1.

Prior to trial, the State disclosed police and crime-scene reports related to the two murders. Additionally, the State proffered a partial list of the evidence seized from each scene. As discussed supra , the crime-scene technician report for Hines' murder established an NOPD evidence technician dusted for fingerprints the whiskey bottles, the whiskey glass from the kitchen table, and the whiskey glass from the night table in the bedroom, but simply listed the fingerprints as "Laboratory-Exam—No".

Likewise, the crime-scene technician report for the Robinson murder established an NOPD evidence technician dusted for prints: a drinking glass containing alcohol on each of the nightstands in Robinson’s hotel room; the passenger side of his vehicle; and a glass, a cup, and a whiskey bottle inside the vehicle. Like the fingerprints lifted from the Hines scene, these fingerprints were marked "Laboratory-Exam—No".

Floyd’s first Brady claim stems from the State’s failure to disclose the fingerprint-comparison results: the logbook noting Floyd was excluded from the fingerprints collected from both crime scenes, the envelope registering the lifted fingerprints from the Hines scene as "NOT VICTIM" and "NOT JOHN FLOYD", and the envelope registering the lifted fingerprints from the Robinson scene as "NOT VICTIM", "NOT JOHN FLOYD", and "NOT ... DAVID HENNESSEY".

a.

For Brady ’s first prong, prosecutors had a duty to ensure all favorable evidence in the State’s possession was disclosed to Floyd, Kyles , 514 U.S. at 437, 115 S.Ct. 1555 ; and, it is undisputed that, prior to trial, the State failed to disclose the envelopes or the results of the fingerprint-comparison testing. Moreover, the district court concluded this issue was waived by the State’s failing to advance it before the magistrate judge. Floyd , 2017 WL 1837676 at *6 (citing Warren v. Bank of Am., N.A. , 566 Fed.Appx. 379, 381 n.1 (5th Cir. 2014) ). The State claims it did not waive this issue when failing to object to the magistrate judge’s recommendation that the fingerprints were analyzed, and the analysis excluded Floyd, or by conceding "[t]he record supports Floyd’s contention that neither the envelopes nor the results of any testing that may have been done on the lifted fingerprints were disclosed to the defense pretrial". The State, however, does not dispute Floyd received neither the envelopes nor the fingerprint-comparison results, but contends these results were effectively disclosed through the crime-scene report.

Along that line, the State maintains the comparison results were effectively disclosed because, although neither the envelopes nor fingerprint-comparison results were provided through the crime-scene report and partial list of seized evidence, Floyd was given proper notice of, and access to, the fingerprint-comparison results prior to trial. The State claims Brady required no more than disclosure of the crime-scene reports and list of collected evidence, asserting this alone was sufficient to render defense counsel "plainly aware" the fingerprints were either not attributable to Floyd or not suitable for comparison. Further, the State maintains Floyd’s access to the report and the list of evidence provided him the opportunity to further inspect the evidence and perform his own analysis. In short, the State maintains no Brady violations stemmed from the failure to disclose this evidence because the comparison results were sufficiently disclosed by providing the defense with the crime-scene technician’s report listing the fingerprints as "Laboratory-Exam—No", and a partial list of evidence seized (the list included the lifted fingerprints).

Brady requires the prosecution disclose evidence when it is "of such substantial value to the defense that elementary fairness requires it to be disclosed even without a specific request". Agurs , 427 U.S. at 110, 96 S.Ct. 2392. The State’s assertion the fingerprint-comparison results were effectively disclosed through the crime-scene report and list of evidence distorts Brady ’s requiring prosecutors to offer exculpatory evidence absent a specific request by the defense. E.g. , id. Floyd’s Brady claim does not stem from the fingerprints themselves, but from the results of the State’s fingerprint-comparison test.

The State does not demonstrate compliance with Brady ’s disclosure requirement by asserting a possibility Floyd could deduce that, based on the general evidence provided to him, additional evidence likely existed. E.g ., Starns v. Andrews , 524 F.3d 612, 619 (5th Cir. 2008). To the contrary, the State’s nondisclosure may have reasonably led the defense to conclude no additional evidence existed. United States v. Bagley , 473 U.S. 667, 682–83, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985). Further, the State’s assertions the evidence was not withheld because Floyd could have conducted his own analysis are in direct contrast to clearly-established Brady law rejecting the defense’s ability to conduct their own analysis as justification for prosecutorial non-disclosure. Banks v. Dretke , 540 U.S. 668, 696, 124 S.Ct. 1256, 157 L.Ed.2d 1166 (2004) (holding "a rule thus declaring ‘prosecutor may hide, defendant must seek’, is not tenable in a system constitutionally bound to accord defendants due process"). Because the State possessed fingerprint-comparison results at the time of trial, and failed to disclose them to the defense, the evidence was withheld.

b.

Brady ’s second prong requires Floyd to show the withheld evidence was favorable to his defense. E.g. , Brady , 373 U.S. at 87, 83 S.Ct. 1194. Favorable evidence includes any exculpatory or impeachment evidence. United States v. Stanford , 823 F.3d 814, 841 (5th Cir. 2016) (quoting United States v. Barraza , 655 F.3d 375, 380 (5th Cir. 2011) ).

The State contends the fingerprint-comparison results, proving fingerprints lifted from the Hines scene were attributable to someone other than Floyd or Hines, are not favorable to Floyd because "fingerprint evidence devoid of context is no evidence at all". Nonetheless, at trial, Detective Dillmann testified the "highball glasses filled with a liquid on each side of the bed" (again, although one glass was by the bed, the other was instead in the kitchen) corroborated Floyd’s confession that he drank with Hines prior to killing him.

Moreover, the Kyles Court in 1995 held a withheld list of license-plate numbers, which excluded defendant’s vehicle from the crime scene—interestingly, the investigation was led by Detective Dillmann—was exculpatory and impeachment evidence. 514 U.S. at 450, 115 S.Ct. 1555.

On the police’s assumption, argued to the jury, that the killer drove to the lot and left his car [at the crime scene]

during the heat of the investigation, the list without [defendant’s] registration would obviously have helped [defendant] and would have had some value in countering an argument by the prosecution that a grainy enlargement of a photograph of the crime scene showed [defendant’s] car in the background.

Id . Likewise, the fingerprint-comparison results excluding Floyd from the fingerprints lifted from the whiskey bottle "would obviously have helped [Floyd] and would have had some value in countering" the detective’s testimony and the State’s theory that Floyd shared a drink with Hines. Id .

Again, Detective Dillmann misstated the location of the subject glasses as being on each side of the bed; instead, one was found in Hines' bedroom, and the other was discovered in his kitchen, the same room in which the whiskey bottle was located. (Although the detective did not mention the whiskey bottle inherent to the suppressed fingerprint-comparison results, his testimony asserts the glasses found in Hines' apartment proved Floyd’s confession was credible.)

Therefore, fingerprint-comparison results proving fingerprints lifted from the whiskey bottle discovered at the Hines scene (in the kitchen, where one of the whiskey glasses was located) were not attributable to Floyd are particularly favorable to impeach the detective’s testimony that this evidence supported the credibility of Floyd’s confession. Because, in the context of the detective’s testimony, this evidence is favorable for impeaching the prosecution’s witness, it is favorable under Brady . Bagley , 473 U.S. at 676, 105 S.Ct. 3375 (citing Giglio v. United States , 405 U.S. 150, 154, 92 S.Ct. 763, 31 L.Ed.2d 104 (1972) ).

Further, the evidence at issue—exculpatory fingerprint-comparison results proving fingerprints uncovered at the Hines scene were attributable to someone other than Hines or Floyd—links a third party to the crime scene, and supports Floyd’s third-party-guilt defense. Because the fingerprint-comparison results could have advanced that defense, it is favorable evidence under Brady . Boyette v. Lefevre , 246 F.3d 76, 91 (2d Cir. 2001) (evidence supporting the defense’s suggestion of an alternative culprit is favorable); United States v. Safavian , 233 F.R.D. 12, 16 (D.D.C. 2005) (favorable Brady evidence includes all evidence "that tends to help the defense by [ ] bolstering the defense case").

As discussed supra , Floyd simultaneously confessed to both murders through confessions having markedly similar descriptions. In the joint bench trial, Floyd’s claimed guilt for Hines' murder was supported solely by his confession and threat to Edwards. Because it is more likely that, if Floyd falsely confessed to murdering Robinson, he simultaneously falsely confessed to murdering Hines, the fingerprint-comparison results from the Robinson scene are favorable to Floyd’s defense to discredit his Hines' confession.

In short, withheld exculpatory evidence of fingerprint-comparison results concluding the prints lifted from the scenes were attributable to neither the defendant nor the victims, especially where, as here, the defense is third-party guilt, falls squarely within the category of favorable evidence that Brady requires the prosecution to produce. Boyette , 246 F.3d at 91 ; Safavian , 233 F.R.D. at 16. (The third Brady prong, whether the evidence is material, is addressed infra, in subpart 3.)

2.

The other claimed Brady violation stems from Clegg’s 2008 affidavit: Detective Dillmann’s report and testimony misrepresented Clegg’s statement to the detective that Hines' preference was for sexual relations with black males. In that regard, the detective reported and testified that Clegg instead stated Hines "frequently had sexual relations with both black and white males"; but, in his 2008 affidavit, Clegg maintained: "Bill[ ] [Hines'] taste was for black men"; he knew "Bill’s taste was for black men"; he "saw Bill with black men on several occasions"; "Bill was often attracted to rough-looking black men"; the detective’s report misrepresented his statements; and he had advised the detective that Hines preferred black men. (Although the dissent at 250–51 considers this affidavit in its analysis of the reasonableness of the state courts' application of Brady , only Clegg’s pre-trial statement to Detective Dillmann, as presented in Clegg’s post-trial affidavit, not his entire affidavit, is properly considered favorable, material evidence withheld by the prosecution in violation of Brady . In short, and contrary to the dissent’s contention at 251 that we "cherry-pick[ed] certain sentences from Clegg’s affidavit", it is only those portions of Clegg’s statement, as contained in the affidavit, that are favorable to impeach the detective’s testimony that are relevant to our Brady analysis.)

a.

Similar to its assertions about the fingerprint-comparison results, the State contends the Clegg statement was effectively disclosed through the detective’s report’s naming Thomas Bloodworth as a reporting witness; Bloodworth identified Clegg and advised the detective to speak with him. The State claims the Clegg statement was effectively disclosed because "a reasonably diligent defense attorney would have similarly interviewed Bloodworth and, through him, learned of Clegg" and interviewed him. As discussed supra , the prosecutor’s Brady duty is not absolved through asserting various opportunities available for the defense to have uncovered the evidence. Banks , 540 U.S. at 696, 124 S.Ct. 1256.

b.

Additionally, the Clegg statement’s presenting a likelihood Hines' killer was a black male is exculpatory evidence that Floyd did not fit the profile of the likely killer, and, therefore, favorable to his defense. Brady , 373 U.S. at 87–88, 83 S.Ct. 1194. As with the fingerprint-comparison results, this evidence is particularly favorable in the light of the detective’s testimony that Hines was likely killed by a welcome visitor with whom he shared a drink and had sexual relations.

The Clegg statement is also favorable evidence because the fact that the statement was misrepresented in Detective Dillmann’s report could have been used to impeach his testimony and call into question the "thoroughness and even the good faith of the investigation". Kyles , 514 U.S. at 445, 115 S.Ct. 1555 ; accord id. at 446, 115 S.Ct. 1555 ("A common trial tactic of defense lawyers is to discredit the caliber of the investigation or the decision to charge the defendant, and we may consider such use in assessing a possible Brady violation" (quoting Bowen v. Maynard , 799 F.2d 593, 613 (10th Cir. 1986) ) ). Moreover, the Clegg statement could have been used to impeach Detective Dillmann’s testimony that, despite the fact that only hairs from a black person had been found at the Hines scene, he did not "under the circumstances" think that investigators "ought to be looking for a black" male because he "had learned that [ ] Hines' sexual preference was not to any one race".3.

Understandably, the materiality of the "suppressed evidence [is] considered collectively, not item by item". Id . at 436, 115 S.Ct. 1555. Therefore, to succeed on his Brady claim, Floyd must also demonstrate the fingerprint-comparison results and the Clegg statement are collectively material. E.g. , id. at 432–34, 115 S.Ct. 1555. Evidence is material if it demonstrates "a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different". Youngblood v. West Virginia , 547 U.S. 867, 870, 126 S.Ct. 2188, 165 L.Ed.2d 269 (2006) (quoting Strickler , 527 U.S. at 280, 119 S.Ct. 1936 ). A reasonable probability is a likelihood sufficient to "undermine confidence in the outcome". Bagley , 473 U.S. at 682, 105 S.Ct. 3375 (quoting Strickland v. Washington , 466 U.S. 668, 694, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984) ).

The materiality vel non of exculpatory evidence is "almost entirely" dependent on its value relative to the evidence presented in the prosecution’s case for guilt. United States v. Sipe , 388 F.3d 471, 478 (5th Cir. 2004) (quoting Smith v. Black , 904 F.2d 950, 967 (5th Cir. 1990) ). Therefore, where the undisclosed evidence renders the verdict suspect, evidence of "relatively minor importance might be sufficient to create a reasonable doubt". Wearry v. Cain , ––– U.S. ––––, 136 S.Ct. 1002, 1006, 194 L.Ed.2d 78 (2016) (quoting Agurs , 427 U.S. at 113, 96 S.Ct. 2392 ).

No physical evidence linked Floyd to either murder; in each instance, his claimed guilt was supported solely by his confession and other incriminating statements. At trial, Floyd attacked the validity of his confessions, maintaining his statements were untrue and coerced through the detective’s conduct, and maintained a third-party-guilt defense.

The evidence of the fingerprint-comparison results from both crime scenes links a third-party to the scene in support of that defense. Further, the exclusion of Floyd’s fingerprints from the whiskey bottle discovered in Hines' kitchen with one of the glasses undermines the detective’s testimony that the credibility of Floyd’s confession is conveyed through the evidence of glasses of whiskey in Hines' residence. Again, the fingerprint-comparison results bolster the lack of physical evidence linking Floyd to the crime scenes, undermine the State’s theory and the detective’s testimony, and strengthen Floyd’s assertions he falsely confessed.

As demonstrated supra , Floyd need not prove "that[,] after discounting the inculpatory evidence in light of the undisclosed evidence, there would not have been enough left to convict". Kyles , 514 U.S. at 434–35, 115 S.Ct. 1555. Even if the withheld evidence had been admitted at trial, the trial judge may have determined, even after considering the withheld evidence, that Floyd’s confession was sufficient to prove his guilt. Nonetheless, the fingerprint-comparison results and Clegg statement "would have entitled [the judge] to find" Floyd’s confession unreliable, and, therefore, affects confidence in the verdict. Id . at 454, 115 S.Ct. 1555. That, of course, is all that is required for this prong of a successful Brady claim.

Further, the materiality of the fingerprint evidence is strengthened through scrutiny of the evidence related to Robinson’s murder. During the joint trial, Floyd presented physical evidence from the Robinson scene of, inter alia : blood from the scene, seminal fluid discovered in his hotel room and body, and hair discovered near his body. The physical evidence excluded Floyd and suggested a black perpetrator with type-A blood; and Floyd was found not guilty of Robinson’s murder. Floyd , 435 So.2d at 996. In short, it is difficult, if not impossible, to conclude the fingerprint-comparison results excluding Floyd from the Hines scene are immaterial, in the light of the evidence excluding him from the Robinson scene and the simultaneous acquittal.

Additionally, Clegg’s pre-trial statement, regarding Hines' preference for black males, further undermines the State’s theory, including presented at trial, that Floyd was guilty of murdering Hines after having sexual relations with him. Conversely, it supports a possibility Hines was killed by a black male and corroborates the physical evidence (a black person’s hair) discovered at the Hines scene.

Considering the minimal inculpatory evidence presented by the State, Floyd clearly demonstrates the requisite "reasonable likelihood" his Brady material could have "affected the judgment" of his guilt in Hines' murder. Wearry , 136 S.Ct. at 1006 (quoting Giglio , 405 U.S. at 154, 92 S.Ct. 763 ). In the light of the absence of physical evidence linking Floyd to the Hines scene, the State’s theory Hines was murdered by a welcome male companion, and Floyd’s third-party-guilt defense, the fingerprint-comparison results, and Clegg’s statement "could reasonably be taken to put the whole case in such a different light as to undermine confidence in the verdict". Strickler , 527 U.S. at 290, 119 S.Ct. 1936 (quoting Kyles , 514 U.S. at 435, 115 S.Ct. 1555 ).

In sum, the State withheld favorable, material evidence, in violation of Brady , 373 U.S. 83, 83 S.Ct. 1194. But, our merits inquiry does not end here; for the following reasons, Floyd’s being granted relief also passes muster under AEDPA’s unreasonableness standard.

C.

A state prisoner seeking federal habeas relief pursuant to 28 U.S.C. § 2254 carries the heavy burden of demonstrating entitlement to that relief. Avila v. Quarterman , 560 F.3d 299, 304 (5th Cir. 2009) ; Lockett v. Anderson , 230 F.3d 695, 707 (5th Cir. 2000) ; Orman v. Cain , 228 F.3d 616, 619 (5th Cir. 2000). Prior to Floyd’s seeking such relief, state-court post-conviction relief was denied by both the Criminal District Court for the Parish of Orleans, and the Supreme Court of Louisiana. Floyd , 62 So.3d at 57. In granting relief, the district court concluded: the State withheld material evidence in violation of Brady , and the state-court contrary decisions were an unreasonable application of clearly-established federal law. Floyd , 2017 WL 1837676, at *16. In reviewing de novo the district court’s granting relief, we "apply[ ] the same standards to the state court’s decision[s] as did the district court". Lewis , 701 F.3d at 787 (quoting Busby , 359 F.3d at 713 ).

When reviewing, as here, the reasonableness of an unexplained state-court decision, our court applies the "look-through" presumption to examine the last reasoned state-court decision, with the presumption that all later unexplained (unreasoned) decisions "rest upon the same ground". Hittson v. Chatman , ––– U.S. ––––, 135 S.Ct. 2126, 2127, 192 L.Ed.2d 887 (2015) (quoting Ylst v. Nunnemaker , 501 U.S. 797, 803, 111 S.Ct. 2590, 115 L.Ed.2d 706 (1991) ). But, as discussed supra , in this instance the two state-court denials are unexplained. Therefore, because there is no reasoned state-court opinion, our court must hypothesize the reasons or theories that could have supported the denial of relief. Id. (citing Richter , 562 U.S. at 98, 131 S.Ct. 770.)

AEDPA’s standards control the review of the state-court decision where, as here, the petition was filed after its effective date. Williams v. Taylor , 529 U.S. 362, 402, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000). Contrary to the criticism by the dissent at 246 and 247–48, our having found Brady error does not color or prejudge our application of AEDPA. To the contrary, our bifurcated approach in this instance provides greater clarity and precision because it permits a far cleaner and more focused AEDPA analysis, against the backdrop of the Brady analysis.

Under AEDPA, federal habeas applications centered on claims "adjudicated on the merits in State court proceedings" are denied unless the adjudication: (1) "resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States"; or (2) "resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding". 28 U.S.C. § 2254(d). Because Brady claims involve mixed questions of law and fact, § 2254(d)(1), instead of subpart (d)(2), is applied. DiLosa v. Cain , 279 F.3d 259, 262 n.2 (5th Cir. 2002) (citing Trevino v. Johnson , 168 F.3d 173, 184 (5th Cir. 1999) ).

The Criminal District Court for the Parish of Orleans denied, without reasons, Floyd’s petition from the bench; similarly, the Supreme Court of Louisiana provided no explanation for its denial. Floyd , 62 So.3d 57 (denial of Floyd’s writ application in a 4-3 vote without assigning reasons). The only state-court reasoning available on review is the dissent from the state-supreme-court denial, with the dissent’s stating Floyd was entitled to a new trial because the fingerprint evidence "undermine[s] confidence in the outcome of Floyd’s trial". Id . at 60.

In any event, "[ 28 U.S.C.] § 2254(d) does not require a state court to give reasons before its decision can be deemed to have been adjudicated on the merits". Richter , 562 U.S. at 100, 131 S.Ct. 770 (internal quotation omitted). "When a federal claim has been presented to a state court and the state court has denied relief, it may be presumed that the state court adjudicated the claim on the merits in the absence of any indication[,] or state-law procedural principles[,] to the contrary". Id . at 99, 131 S.Ct. 770. Therefore, where, as here, the state-court denial has no explanation, we review the "ultimate decision" for reasonableness. Charles v.Thaler , 629 F.3d 494, 501 (5th Cir. 2011) (quoting Neal v. Puckett, 286 F.3d 230, 246 (5th Cir. 2002) (en banc) ).

The state court’s "adjudication of the claim result[s] in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States", 28 U.S.C. § 2254(d)(1), when: it "reaches a legal conclusion in direct conflict with a prior decision of the Supreme Court or ... it reaches a different conclusion than the Supreme Court based on materially indistinguishable facts", Miller v. Dretke , 404 F.3d 908, 913 (5th Cir. 2005).

Accordingly, although Floyd’s Brady claim is meritorious, he is not entitled to relief under AEDPA based simply on our holding the state-court decision incorrectly applied the clearly-established Brady standard. Cullen v. Pinholster , 563 U.S. 170, 202–03, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011) ; Williams , 529 U.S. at 409, 120 S.Ct. 1495 ("a federal habeas court may not issue the writ simply because [it] concludes in its independent judgment that the relevant state-court decision applied clearly established federal law erroneously or incorrectly"). In other words, AEDPA reasonableness vel non is distinct from our analysis of legal accuracy. E.g. , Richter , 562 U.S. at 88, 131 S.Ct. 770. A state court’s deciding the post-conviction-relief "claim lacks merit precludes federal habeas relief so long as fairminded jurists could disagree" that the state court’s denial was inconsistent with Supreme Court precedent. Id. at 101, 131 S.Ct. 770 (quoting Yarborough v. Alvarado , 541 U.S. 652, 664, 124 S.Ct. 2140, 158 L.Ed.2d 938 (2004) ). Re-stated, our review is limited to "whether the state court’s application of clearly established [ Brady ] law was objectively unreasonable". Williams , 529 U.S. at 409, 120 S.Ct. 1495.

Because the state courts provided no explanation for their denial of post-conviction relief, we must hypothesize the reasons that supported, or could have supported, the denial consistent with Supreme Court precedent. Richter , 562 U.S. at 98, 102, 131 S.Ct. 770. The decision is an "unreasonable application" under 28 U.S.C. § 2254(d) only if, after this hypothetical inquiry, we determine there was no reasonable basis for it. Id. at 98, 101, 131 S.Ct. 770.

Our analysis, of course, corresponds to the specificity of the Brady rule. E.g. , Yarborough , 541 U.S. at 664, 124 S.Ct. 2140. In other words, because the Court’s Brady requirements are specific, the range of reasonable judgment is more narrow, and "[a]pplications of the rule may be plainly correct or incorrect". See id. ;see also Richter, 562 U.S. at 101, 131 S.Ct. 770. Again, the Court defines Brady violations as: (1) prosecutorial suppression of evidence; (2) favorable to the defense; and (3) material to either guilt or innocence. Brady , 373 U.S. at 87, 83 S.Ct. 1194. Accordingly, AEDPA’s requisite hypothetical analysis is whether any reasonable theory exists to support the state courts' conclusion that Floyd failed to show each of Brady ’s three specific elements.

1.

For the suppression prong, it is undisputed the State failed to disclose the fingerprint-comparison results and Clegg statement. Floyd , 2017 WL 1837676, at *9–11. The State’s theory the evidence was effectively disclosed through the lab report, list of evidence, and Clegg’s identity, is in direct conflict with Banks . 540 U.S. at 695–96, 124 S.Ct. 1256. Consequently, the court could not have reasonably relied on that theory to find the evidence was not suppressed. Similarly, the State’s assertions the evidence was not suppressed because Floyd had an opportunity to conduct his own analysis and interview Clegg is counter to the Court’s Banks decision. Id . Therefore, the state court was presented with no reasonable theory for concluding the State did not withhold the fingerprint-comparison results and the Clegg statement; nor were we presented with any; nor do we perceive any.

2.

As for Brady ’s favorability prong, it would be an unreasonable application of Brady and its progeny to conclude that the withheld evidence was not favorable. It was favorable because it supported Floyd’s third-party-guilt defense, and impeached Detective Dillmann’s testimony that the "two highball glasses filled with a liquid on each side of [Hines'] bed" corroborated the details of Floyd’s confession. (As noted repeatedly, the detective erroneously stated the glasses were found in the bedroom; instead, one was found in the bedroom and one was found in the kitchen, where the whiskey bottle was also located.)

"[T]he character of a piece of evidence as favorable will often turn on the context of the existing or potential evidentiary record." Kyles , 514 U.S. at 439, 115 S.Ct. 1555. Supreme Court precedent defines evidence tending to strengthen a defense as favorable evidence under Brady . Cone v. Bell , 556 U.S. 449, 470, 129 S.Ct. 1769, 173 L.Ed.2d 701 (2009). And, again, the Court has held evidence impeaching a prosecution witness is favorable Brady evidence. Bagley , 473 U.S. at 676, 105 S.Ct. 3375. Any reason to support a conclusion the evidence was not favorable to Floyd is contrary to Court precedent, and, therefore, an unreasonable application of clearly-established federal law. Re-stated, there was no reasonable theory for the state courts' concluding the evidence was not favorable to Floyd under Brady .

Along that line, and as the dissent maintains at 248, the state court could have concluded that the withheld fingerprint-comparison results from the Hines scene do not impeach Detective Dillmann’s testimony because he did not testify that the whiskey bottle, from which the prints were lifted, corroborated Floyd’s confession. But that conclusion would be an unreasonable application of Supreme Court law.

First, the Court has been clear that favorability depends on context. The detective testified that the whiskey glasses found at the Hines scene—one of which was actually found in the kitchen, as was the whiskey bottle—corroborated Floyd’s statement that the two were drinking together. Evidence that a third person—neither Floyd nor Hines—touched the whiskey bottle undermines Detective Dillmann’s testimony that the confession was credible based on Floyd’s statement that he and Hines were drinking together. Second, although the detective did not specifically reference the whiskey bottle, to conclude that that negates the favorability of the fingerprint-comparison results "confuses the weight of the evidence with its favorable tendency". Kyles , 514 U.S. at 450, 115 S.Ct. 1555.

The dissent also asserts at 248–49 that the state court could reasonably have concluded that the fingerprints lifted from the whiskey bottle were only neutral evidence. We disagree. The presence of a third party’s fingerprints at a crime scene does not itself prove Floyd was not present; but, it is evidence that a third party, not Floyd, touched an item that was singled out for dusting by investigators and linked to the commission of the crime through Detective Dillmann’s testimony. See id . (holding that a list of cars at the crime scene that did not include the defendant’s car "would obviously have helped" the defendant in countering investigator’s assumption, argued to the jury, that the killer had driven to the scene and left his car there). Furthermore, although the fingerprint-comparison results do not conclusively establish that Floyd was not present at the Hines scene, any such contention would again confuse weight with favorability, and also misapply the relevant standard for materiality. See id. at 434, 115 S.Ct. 1555 ("[A] showing of materiality does not require demonstration by a preponderance that disclosure of the suppressed evidence would have resulted ultimately in the defendant’s acquittal....").

In addition, it would be unreasonable to conclude that the Clegg statement is not favorable. Under clearly-established Supreme Court precedent, evidence that could have been used to impeach a witness's testimony is favorable. Strickler , 527 U.S. at 281–82, 119 S.Ct. 1936. Clegg’s statement, that Hines' sexual preference was for black males, could have been used to impeach Detective Dillmann’s testimony that he "had learned that Mr. Hines' sexual preferences was not to any one race". (The dissent at 250 asserts Detective Dillmann’s testimony "suggests he relied on more than just one person" for his determinations regarding Hines' sexual preferences. Nonetheless, regarding Hines' sexual preferences, the detective’s report, in the record for this habeas proceeding, states only that "Mr. Clegg stated that to his knowledge the victim was homosexual and frequently had sexual relations with both black and white males".) Any assertion that Clegg’s knowledge of Hines' sexual preferences may not have been exhaustive again would go to weight, not favorability.

3.

For the final prong, we consider whether any reasonable theory could have supported a conclusion that the withheld evidence was collectively immaterial. Kyles , 514 U.S. at 436, 115 S.Ct. 1555. The materiality of Brady evidence is not considered in the light of the probability of acquittal. Bagley , 473 U.S. at 680, 105 S.Ct. 3375. Rather, evidence is understandably material under Brady where it simply undermines "confidence in the outcome", id. at 682, 105 S.Ct. 3375 (quoting Strickland , 466 U.S. at 694, 104 S.Ct. 2052 ), and is more likely material when the State presents a weaker case for guilt, e.g. , Smith v. Cain , 565 U.S. 73, 76, 132 S.Ct. 627, 181 L.Ed.2d 571 (2012) (eyewitness "testimony was the only evidence linking [the petitioner] to the crime", and, therefore, the undisclosed statements contradicting this testimony were "plainly material"). As noted, the evidence is collectively material if there is any "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.

Floyd was indicted for the second-degree murder of Hines and Robinson. In the joint trial, Floyd’s incriminating statements (confession and threat to Edwards) were the only evidence presented to support his guilt for Hines' murder. And, that evidence was contradicted by the suppressed evidence at issue, analogous to the evidence at issue in Cain . 565 U.S. at 76, 132 S.Ct. 627.

The fingerprint-comparison results undermine Floyd’s confessions to each murder, and impeach Detective Dillmann’s testimony for the Hines murder that the "glasses filled with a liquid" (in fact, discovered in Hines' bedroom and kitchen) corroborated Floyd’s confession. The fingerprint-comparison evidence contradicts the physical evidence purported to corroborate Floyd’s confessions to each murder, such as the glasses containing whiskey being on each side of Robinson’s bed, undermining "confidence in the verdict". Kyles , 514 U.S. at 435, 115 S.Ct. 1555. Likewise, the Clegg statement impeaches the detective’s testimony that Hines' sexual preference was for black and white males, and further challenges the credibility of Floyd’s confession. In the light of the entire case, the fingerprint-comparison results and the Clegg statement significantly impact the only evidence supporting Floyd’s guilt (his incriminating statements, including, most especially, his confession), rendering it material under Brady . Id. In other words, the fingerprint-comparison results and the Clegg statement create a "reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different". Id. at 433, 115 S.Ct. 1555 (quoting Bagley , 473 U.S. at 682, 105 S.Ct. 3375 ).

Any conclusion to the contrary would be an unreasonable application of Supreme Court law. The state court could have concluded that neither the fingerprint-comparison results nor the Clegg statement conclusively prove Floyd did not commit the Hines murder. But that would constitute an unreasonable application of the Supreme Court’s holding that "a showing of materiality does not require demonstration by a preponderance that disclosure of the suppressed evidence would have resulted ultimately in the defendant’s acquittal". Id. at 434, 115 S.Ct. 1555.

The state court could also have concluded that, despite the withheld evidence, the trial judge could still have convicted Floyd on the basis of his incriminating statements to Edwards. But that, too, would be an unreasonable application of Supreme Court law. "[M]ateriality ... is not a sufficiency of evidence test." Id.

Floyd "need not demonstrate that after discounting the inculpatory evidence in light of the undisclosed evidence, there would not have been enough left to convict". Id . at 434–35, 115 S.Ct. 1555. Where the proof on which a conviction was based was thin to begin with, the Supreme Court has been clear that withheld evidence undermining that proof is material. See Wearry , 136 S.Ct. at 1006 ; Cain , 565 U.S. at 76, 132 S.Ct. 627 ; Agurs , 427 U.S. at 113, 96 S.Ct. 2392. In short, while the trial judge could have convicted Floyd of the Hines murder on the basis of Floyd’s incriminating statement to Edwards, or could have continued to credit his confession, there can be "no confidence that [the trial judge] would have done so", and that is all that Brady requires. Wearry , 136 S.Ct. at 1007 (quoting Cain , 565 U.S. at 76, 132 S.Ct. 627 ).

Materiality of the suppressed Hines evidence is further demonstrated by the simultaneous acquittal at the bench trial for Robinson’s murder. After considering the exculpatory physical evidence from the Robinson scene, the trial judge found Floyd not guilty of that murder. Floyd , 435 So.2d at 994 (1983). ("[Floyd] was found not guilty of the murder of Robinson (evidence showed that Robinson’s assailant had been a black man with Type A blood; Floyd is white with Type B blood") ). Because the trial judge determined the physical evidence rendered Floyd’s incriminating statements, including his confession, insufficient to support his guilt for Robinson’s murder, there is a "reasonable probability" that, had the similarly favorable physical evidence from the Hines scene been disclosed, "the result of the proceeding would have been different". Cain , 565 U.S. at 75, 132 S.Ct. 627 (quoting Cone , 556 U.S. at 469–70, 129 S.Ct. 1769 ). Re-stated, there is a " ‘reasonable probability’ that the [trial judge] would have been [similarly] persuaded by the undisclosed evidence" undermining Floyd’s Hines confession. Id. at 77, 132 S.Ct. 627 (Thomas, J., dissenting) (citing Bagley , 473 U.S. at 682, 105 S.Ct. 3375 ).

In the light of the withheld evidence undermining the only evidence supporting Floyd’s guilt for Hines' murder, and the trial judge’s simultaneously acquitting Floyd of Robinson’s murder after considering similar physical evidence excluding Floyd from the Robinson scene, there is no sound theory, considering the record as a whole, to support the conclusion that the evidence of the fingerprint-comparison results and the Clegg statement were not reasonably likely to affect Floyd’s trial for Hines' murder. Id . ; Kyles , 514 U.S. at 435, 115 S.Ct. 1555. Accordingly, any theory supporting the conclusion that the withheld, favorable evidence was immaterial is an unreasonable application of Brady ’s materiality standard.

In sum, "fairminded jurists could [not ] disagree" that the state-court denial of post-conviction relief was contrary to Supreme Court precedent. Richter , 562 U.S. at 88, 131 S.Ct. 770. Re-stated, "the state court’s application of clearly established [ Brady ] law was objectively unreasonable". Williams , 529 U.S. at 409, 120 S.Ct. 1495.III.

For the foregoing reasons, the judgment is AFFIRMED.

JERRY E. SMITH, Circuit Judge, dissenting:

For the first time ever, this court finds a meritorious claim of actual innocence under McQuiggin v. Perkins , 569 U.S. 383, 133 S.Ct. 1924, 185 L.Ed.2d 1019 (2013). But, given the panel majority’s errant analysis under Brady v. Maryland , 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), I would reverse and deny habeas corpus relief. I therefore respectfully dissent from the cogent and well-intended majority opinion.

"A state court’s determination that a claim lacks merit precludes federal habeas relief so long as ‘fairminded jurists could disagree’ on the correctness of the state court’s decision."¹ That "is [a] difficult [standard] to meet ... [and] it was meant to be."² Meeting that standard can become even more unlikely where, as here, a claim is adjudicated on the merits but lacks a written opinion elucidating the state court’s reasons. Floyd "can satisfy the ‘unreasonable application’ prong of [ 28 U.S.C.] § 2254(d)(1)only by showing that ‘there was no reasonable basis ’ for the [Louisiana] Supreme Court’s decision."³ "[A] habeas court must determine what arguments or theories supported or, as here, could have supported , the state court’s decision."⁴

1 Harrington v. Richter , 562 U.S. 86, 101, 131 S.Ct. 770, 178 L.Ed.2d 624 (2011) (quoting Yarborough v. Alvarado , 541 U.S. 652, 664, 124 S.Ct. 2140, 158 L.Ed.2d 938 (2004) ).

2 Id. at 102, 131 S.Ct. 770.

3 Cullen v. Pinholster , 563 U.S. 170, 188, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011) (quoting Richter , 562 U.S. at 98, 131 S.Ct. 770 ) (emphasis added).

4 Richter , 562 U.S. at 102, 131 S.Ct. 770 (emphasis added).

Though the majority recites the appropriate standards, its Brady methodology fails to apply them rigorously. Instead, it allows its analysis to become colored by the gateway question of whether Floyd proved actual innocence under Perkins . This is one of the rare occasions where we must cope with the tension between a meritorious gateway actual-innocence claim and the strong deference AEDPA accords to a state court’s resolution of the underlying constitutional claim—the latter being the only type of claim that can justify relief.⁵

5 See Perkins , 569 U.S. at 392, 133 S.Ct. 1924 ("We have not resolved whether a prisoner may be entitled to habeas relief based on a freestanding claim of actual innocence."); United States v. Fields , 761 F.3d 443, 479 (5th Cir. 2014) ("[O]ur caselaw does not recognize freestanding actual innocence claims.").

To understand why it is possible to find a petitioner, such as Floyd, "actually innocent" while simultaneously denying him habeas relief, it is important to recognize exactly what an actual-innocence claim is. First, it is a gateway claim. Neither this circuit nor the Supreme Court has recognized a freestanding claim of innocence. Instead, a petitioner can assert actual innocence only to overcome a procedural bar, such as limitations.⁶ After establishing actual innocence, the petitioner must still prove a meritorious constitutional violation while overcoming § 2254 ’s mandated deference. Without a meritorious constitutional violation, an actual-innocence claim is meaningless.

6 Perkins , 569 U.S. at 386, 133 S.Ct. 1924 (holding that "actual innocence, if proved, serves as a gateway through which a petitioner may pass whether the impediment is a procedural bar ... or, as in this case, expiration of the statute of limitations").

Second, the postures in which we review the actual-innocence claim and the underlying constitutional claim are different. Because an actual-innocence claim is a gateway claim asserted to overcome some procedural barricade, it is a claim that has not been reviewed by a state court and thus is accorded no AEDPA deference. A federal court independently determines whether the Perkins standard is met. Conversely, the Brady claims here were adjudicated on the merits by the Louisiana Supreme Court and thus are accorded AEDPA deference. We cannot independently determine whether the Brady standard is met. Instead, we must add an additional layer and decide whether " ‘there was no reasonable basis ’ for the [Louisiana] Supreme Court’s decision."⁷

7 Pinholster , 563 U.S. at 188, 131 S.Ct. 1388 (quoting Richter , 562 U.S. at 98, 131 S.Ct. 770 ) (emphasis added).

Finally, but importantly, when reviewing Floyd’s Brady claims, we cannot consider much of the new evidence presented in the actual-innocence analysis. Under Perkins , we can take into account old and new (reliable) evidence alike. To determine materiality under Brady , however, we can consider only the evidence presented at trial and the suppressed evidence. Thus, new and arguably strong evidence favoring Floyd, such as the fact that he was shown photos of the crime scene, cannot, as a matter of law, color our review of the alleged Brady violations.

The majority accords insufficient AEDPA deference to the state court. At the outset, the majority seems fundamentally to misunderstand the process by which we review state-court decisions under AEDPA. The majority bifurcates its evaluation of Brady , such that the first section seems to be a de novo review of whether Brady violations occurred, and the second section is a perfunctory nod to AEDPA.

That bifurcation is at odds with how we evaluate previously adjudicated claims. We do not first independently evaluate the claim. Instead, the entire analysis is driven from the perspective of whether the state court reasonably applied settled Supreme Court law. By bifurcating the discussion, the majority implicitly turns the analysis into one in which it weighs whether its review of the asserted claims is more reasonable than the state’s review of the asserted claims. Or, at the very least, by first independently reviewing the evidence without considering the hypothetical reasons underpinning the state court’s determination, the majority risked further coloring its own analysis as to whether there was a reasonable basis for the state court’s decision.⁸

8 Cf. Richter , 562 U.S. at 102, 131 S.Ct. 770 ("The Court of Appeals appears to have treated the unreasonableness question as a test of its confidence in the result it would reach under de novo review: Because the Court of Appeals had little doubt that Richter’s Strickland claim had merit, the Court of Appeals concluded the state court must have been unreasonable in rejecting it.").

Contrary to the majority’s insistence, the Supreme Court has never blessed such bifurcation. To the contrary, it has said, "The Ninth Circuit requires federal habeas courts to review the state court decision de novo before applying the AEDPA standard of review. We disagree with this approach ." Lockyer v. Andrade , 538 U.S. 63, 71, 123 S.Ct. 1166, 155 L.Ed.2d 144 (2003) (emphasis added).⁹ Moreover, it makes sense to prohibit bifurcated analyses, given that "even clear error [does] not suffice" to show an "unreasonable application." White v. Woodall , ––– U.S. ––––, 134 S.Ct. 1697, 1702, 188 L.Ed.2d 698 (2014) (internal citation and quotation marks omitted). Instead, " ‘the state court’s ruling on the claim being presented in federal court [must be] so lacking in justification that there [is] an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.’ " Id. (quoting Richter , 562 U.S. at 103, 131 S.Ct. 770 ).

9 See also Richter , 562 U.S. at 102, 131 S.Ct. 770 (explaining that first "a habeas court must determine what arguments or theories ... could have supported[ ] the state court’s decision; and then it must ask whether it is possible fairminded jurists could disagree" (emphasis added) ); Williams v. Taylor , 529 U.S. 362, 410, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000) ("an unreasonable application of federal law is different from an incorrect application of federal law"); id . at 411, 120 S.Ct. 1495 ("Under § 2254(d)(1) ’s ‘unreasonable application’ clause, then, a federal habeas court may not issue the writ simply because that court concludes in its independent judgment that the relevant state-court decision applied clearly established federal law erroneously or incorrectly.").

When its decision is viewed in the proper light, the state court plainly had a reasonable basis for denying relief under Brady . To prove a Brady violation, the petitioner must show that the evidence was withheld, favorable, and material.¹⁰ I agree in full with the majority’s analysis in regard to suppression. Thus, I address only the other two Brady prongs, favorability and materiality.

10 Strickler v. Greene , 527 U.S. 263, 281–82, 119 S.Ct. 1936, 144 L.Ed.2d 286 (1999) ; see also United States v. Sipe , 388 F.3d 471, 477–78 (5th Cir. 2004).

Floyd says that the following evidence is Brady material: analysis of fingerprints found on a whiskey bottle in Hines's kitchen; analyses of fingerprints lifted from two drinking glasses in Robinson’s hotel room, on the passenger side of Robinson’s car, and on a glass, cup, and whiskey bottle in Robinson’s car; and John Clegg’s statement concerning Hines's sexual preferences. The majority classifies the fingerprint analysis from the whiskey bottle as favorable because the analysis could be used to impeach state witness Detective Dillmann. Of note, neither Floyd nor the district court ever contended that the fingerprint analyses could constitute impeachment evidence. Those analyses, however, could reasonably be viewed as not impeaching Dillmann.

The majority avers that the analysis impeaches Dillmann because he testified that the presence of glasses corroborated Floyd’s confession, in which Floyd stated, "We were both drinking." Dillmann, however, never mentioned the whiskey bottle or even whiskey. Instead, he testified only that "there were two highball glasses filled with a liquid on each side of the bed." And, the whiskey bottle was not found at the murder scene¹¹ but in the kitchen.

11 Hines was murdered in his bedroom. No testimony or evidence was provided that indicated he or the murderer ever entered the kitchen. The majority says that "Detective Dillmann testified about the importance of evidence discovered in Hines' kitchen." As with its discussion of the whiskey bottle, the majority again fails to address Dillmann’s testimony with precision.
Dillmann never even mentioned the kitchen. The one time the word "kitchen" was used during his examination, it was by Floyd’s attorney asking whether Floyd’s confession contained any specific details about the layout of the apartment, such as where the bedroom and kitchen were located. Dillmann did not even reply because the court interrupted and asked the attorney to allow Dillmann to finish his testimony on a previous line of questioning.

The majority does not address these details with enough precision,¹² so let me emphasize this: The unidentified fingerprints were found on the whiskey bottle, not the highball glasses, and Dillmann never mentioned the "whiskey bottle" or "whiskey" generally. Reviewing with the proper AEDPA deference, it is easy to see that the presence of an unidentified third party’s partial prints on a whiskey bottle located in the kitchen could reasonably be interpreted as not impeaching Dillmann’s testimony that the presence of glasses in the bedroom (the murder scene) corroborated Floyd’s confession that he and Hines shared a drink.¹³

12 See, e.g. , "Rather, Detective Dillmann testified the evidence of the glasses of whiskey discovered in Hines' apartment ... corroborated ‘perfectly’ the descriptions in Floyd’s confession, and supported its credibility." (emphasis added).

13 The majority responds by claiming that Dillmann provided "erroneous" testimony, given that he said there were two glasses in the bedroom. The majority points to a tech report that says the tech dusted a glass in the bedroom and a glass in the kitchen. First, the majority has decided, because it fits its narrative, to credit the tech over Dillmann. That is curious because, as the majority admits, the photograph from the kitchen depicts two bottles of whiskey but no whiskey glass (or glasses of any sort). Thus, the glass was not "with the whiskey bottle" as the majority states. Second, it is possible for Dillmann and the tech report both to be accurate, as maybe there were a glass in the kitchen and two glasses in the bedroom. Third, even assuming Dillmann mischaracterized where the glasses were, that does not undermine the fact that there is no evidence connecting the kitchen to the murder scene, and Dillmann still never testified about "whiskey."

The majority also contends that the fingerprint analysis is "favorable because it supported Floyd’s third-party-guilt defense." Though the majority is correct that evidence strengthening a defense can be favorable under Brady , the majority again fails to view the issue through the proper lens.

We must review whether it would be reasonable for the Louisiana courts to conclude that the presence of an unidentified third party’s partial prints on a whiskey bottle not directly connected to the murder scene does not strengthen Floyd’s third-party defense. Without a stronger connection between the item containing the fingerprints and the crime, it is not unreasonable for the Louisiana courts to conclude the evidence did not strengthen the defense¹⁴ and thus was only neutral evidence of innocence or guilt.¹⁵

14 See, e.g. , Lines v. Terrell , No. CIV. A. 07-3532, 2009 WL 2870162, at *15 (E.D. La.), report and recommendation adopted , No. CIV. A. 07-3532, 2009 WL 2929334 (E.D. La. 2009) ("While evidence regarding the lack of petitioner’s fingerprints might have been helpful to the defense, that is not the standard for required disclosure. Brady is not violated simply because potentially helpful information is withheld.... [T]he negative fingerprint analysis would not show that petitioner never handled the evidence, but rather only that there were no fingerprints proving that he had done so. That information is not exculpatory and does not put the whole case in such a different light as to undermine confidence in the verdict.").

15 See Black’s Law Dictionary 675 (10th ed. 2014) (defining exculpatory evidence as "[e]vidence tending to establish a criminal defendant’s innocence."); United States v. Ruiz , 536 U.S. 622, 628, 122 S.Ct. 2450, 153 L.Ed.2d 586 (2002) ("[E]xculpatory evidence is evidence the suppression of which would ‘undermine the confidence in the verdict.’ " (quoting Kyles v. Whitley , 514 U.S. 419, 435, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995) ) ); United States v. Bagley , 473 U.S. 667, 676, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985) ("Such evidence is evidence favorable to an accused, so that, if disclosed and used effectively, it may make the difference between conviction and acquittal." (internal quotation marks and citations omitted) ).

Again, under a de novo review we may disagree with that conclusion, but AEDPA deference requires us to test for any reasonable explanation. And it is plausible to characterize the fingerprint analysis "as neutral evidence." Sipe , 388 F.3d at 487. Review of the fingerprint analysis rightly ends here, on the favorability prong.

Regarding the analyses of the fingerprints from the Robinson crime scene, all of the prints on one glass in the hotel room belonged to Robinson, while all others belonged to an unidentified person. Unlike the prints discovered at the Hines crime scene, some but not all of the prints at the Robinson crime scene were on items potentially connected to the murder. The prints on the drinking glasses in the hotel room (the murder scene) certainly could serve as exculpatory evidence—for the Robinson murder. Some may believe that additional evidence exculpating Floyd of the Robinson murder could potentially favor exculpation from the Hines murder. But it is also reasonable to believe that evidence exculpating Floyd of one murder—a murder that he was previously acquitted of because there was already evidence presented in the joint case exculpating him of that murder—does not tend to show innocence of the other murder.

And, the prints from the vehicle suffer largely the same fate as the prints at the Hines crime scene. The vehicle has never been directly connected to the crime, and it would not be unreasonable for there to be numerous third-party prints (including those of Robinson’s friend whom he drove home earlier in the evening) within a vehicle.¹⁶ Thus, the prints from the vehicle could easily be classified as neutral, and, after we accord the proper AEDPA deference, so too could the prints on glasses found at the Robinson crime scene.

16 Accord Sosa v. Dretke , 133 Fed.Appx. 114, 121–22 (5th Cir. 2005) (explaining that the presence of other fingerprints in putative getaway car was not exculpatory because it "merely shows ... that others had been in the car at some point in time").

Even if the fingerprints on the glasses should have properly been deemed favorable, they would still fail the materiality prong. Throughout the joint trials, the defense undermined Floyd’s confession to the Robinson murder with numerous other pieces of evidence, such as the fact that though Floyd claimed he wiped himself with a tissue after receiving oral gratification from Robinson, that tissue actually contained semen that could not belong to either Floyd or Robinson.

So, ample evidence at trial indicated the presence of a third party and undermined the credibility of Floyd’s confession. A state court could thus deem any additional evidence to be cumulative and not material under Fifth Circuit precedent.¹⁷

17 See , e.g. , Sipe , 388 F.3d at 478 ("Thus, ‘when the undisclosed evidence is merely cumulative of other evidence [in the record], no Brady violation occurs.’ " (quoting Spence v. Johnson , 80 F.3d 989, 995 (5th Cir. 1996) ) ); Jackson v. Johnson , 194 F.3d 641, 650 (5th Cir. 1999) ("When Brady evidence would have only a cumulative or marginal impact on the jury’s credibility assessment, habeas relief is not in order because the evidence is not material....").

As for Clegg’s statement, I agree that it could only reasonably be labeled as favorable, because it could be used to weaken Dillmann’s testimony that during his "follow-up investigation, initially after the homicide," he spoke "with several people ... [and] had learned that Mr. Hines' sexual preferences was not to any one race. He was involved with both black and white males, and he was very indiscriminate...."¹⁸ Dillmann interviewed Clegg and reported that Clegg stated Hines was indiscriminate in his tastes. Thus, Clegg’s contradictory statement—that Hines had only ever pointed out black men the few times Clegg and Hines went to gay bars together—would serve as impeachment evidence.

18 Of note, Dillmann’s testimony suggests he relied on more than just one person for his belief that Hines had indiscriminate preferences.

That statement, however, fails the final prong of Brady —materiality. As the majority notes, under that prong we consider "the cumulative effect of all [suppressed] evidence." Sipe , 388 F.3d at 478. But, "[w]e include in this cumulative materiality analysis only the evidence that survived Brady’ s other prongs...." Id . at 491. As the only piece of evidence to clear the first two prongs, the Clegg statement is correctly evaluated by itself.¹⁹

19 As previously explained, the Hines fingerprint analysis fails the favorability prong. The Robinson fingerprint analyses also fails it, or at the very least is cumulative of evidence presented at trial.

The state court could have reasonably concluded that Clegg’s statement was not material. "[E]vidence is material only 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. "A ‘reasonable probability’ is a probability sufficient to undermine confidence in the outcome." Id. "[T]he question is whether ‘the favorable evidence could reasonably be taken to put the whole case in such a different light as to undermine confidence in the verdict.’ "²⁰ " ‘The materiality of Brady material depends almost entirely on the value of the evidence relative to the other evidence mustered by the state.’ "²¹

20 Strickler , 527 U.S. at 290, 119 S.Ct. 1936 (quoting Kyles , 514 U.S. at 435, 115 S.Ct. 1555 ).

21 Sipe , 388 F.3d at 478 (quoting Smith v. Black , 904 F.2d 950, 967 (5th Cir. 1990), vacated on other grounds , 503 U.S. 930, 112 S.Ct. 1463, 117 L.Ed.2d 609 (1992) ).

Clegg admitted he had limited knowledge of Hines's sexual preferences.²² The state court could conclude that Clegg’s statement does not significantly dispel the possibility that Hines was open to relations with a white male nor that a white male could have committed the murder. At least Thomas Bloodworth, another good friend of Hines's, testified he had never seen Hines "socially in the company of a black person" other than one friend who had moved away.

22 Clegg stated in his affidavit, "I was never, in fact, aware of the frequency of [Hines's] sexual relations with anyone."

Regardless, learning that Clegg (who had moved out of the state ten years before and had been back only for visits)²³ had, in the few instances they were at gay bars together, only heard Hines point out specific black men as attractive, can easily be regarded as not throwing the case into a whole new light or undermining confidence in the verdict. That is especially true in comparison to the value of the opposing evidence—Floyd’s separate confessions to the police and to bar owner Steven Edwards. Thus, when we properly and fully apply AEDPA deference, the state court reasonably could have discounted Clegg’s statement.²⁴ In sum, we are bound by AEDPA and Brady . Under AEDPA, we accord strong deference to the state court and test for any reasonable basis on which its decision could rest. Under Brady , we look only at evidence presented at trial and any allegedly suppressed evidence—but no more. For these reasons, the district court erred, and I respectfully dissent.

23 In fact, Clegg’s statement implies that he and Hines had not visited a gay bar together in ten years. That further illustrates why it would be reasonable for a state court to determine that an opinion based on ambiguous statements made ten years before the murder are not material.

24 Even assuming the majority is correct—that we can only cherry-pick certain sentences from Clegg’s affidavit instead of analyzing its reliability as a whole to determine whether the differing statement "put the whole case in such a different light as to undermine confidence in the verdict"—the fact that one friend believed Hines had a penchant only for black men does not inarguably "undermine confidence in the verdict." Strickler , 527 U.S. at 290, 119 S.Ct. 1936 (internal citations and quotation marks omitted).
--------