BATCHELDER, J., delivered the opinion of the court, in which GILMAN, J., joined. MOORE, J. (pp. 643-53), delivered a separate dissenting opinion.
OPINION
ALICE M. BATCHELDER, Circuit Judge.Petitioner John Fautenberry (“Fauten-berry”), a prisoner in the state of Ohio awaiting execution, appeals the district *620court’s denial of his petition for writ of habeas corpus. Fautenberry raises eight issues on appeal. Finding no merit in any of them, we AFFIRM the judgment of the district court.
I.
In November 1990, Fautenberry, who had recently quit his job as a cross-country truck driver, met Donald Nutley at a truck stop outside Portland, Oregon, and the two men went target shooting together. After they had finished and were leaving the range, Fautenberry shot Nutley in the head and stole $10,000 from him. Fauten-berry then drove to Cincinnati, Ohio, where he stayed with his sister for a short time before traveling to Connecticut to visit an old friend. In February 1991, while en route back to Cincinnati, Fauten-berry — out of money and in need of gasoline to continue his travels — stopped at a truck stop in New Jersey. There he met Gary Farmer, who, after learning of Fau-tenberry’s need for money, offered to buy Fautenberry breakfast and give him money in exchange for sex. Fautenberry got into the cab of Farmer’s truck, shot Farmer in the head, and took his wallet. Fau-tenberry then returned to his sister’s residence in Cincinnati.
On February 17, 1991, after another brief stay in Cincinnati, Fautenberry again left his sister’s residence, this time on foot, in search of money. Fautenberry walked down Highway 125, in the eastern suburbs of Cincinnati, stopped at the on-ramp to Interstate 275, and began hitchhiking. Joseph Daron offered to give Fautenberry a ride. Daron intended to travel only ten miles north to his home in Milford, Ohio, but, upon learning that Fautenberry wanted to go north to Columbus, Ohio, he drove Fautenberry an extra ten miles and dropped him near the intersection of Interstate 275 and Interstate 71, which goes directly to Columbus. As he exited Dar-on’s vehicle, Fautenberry reached back into the car and shot Daron twice in the chest. Fautenberry then drove Daron’s car south to Cincinnati, and threw Daron’s body into a wooded area on the north bank of the Ohio River, where it was eventually found more than a month later by the local authorities. Fautenberry took Daron’s car, wallet, briefcase, wristwatch, and Bible, and returned to Oregon.
Fautenberry arrived in Portland on February 24, 1991, and spent the next few days at the Oregon coast with some old friends and acquaintances, including a woman named Christine Guthrie. Guthrie accompanied Fautenberry back to Portland from the coast, and along the way, they stopped on an old logging road. Fau-tenberry escorted Guthrie to a secluded portion of the woods, shot her three times in the back of the head, and stole her bank card. A few days later, after withdrawing cash from her bank account, Fautenberry traveled to Juneau, Alaska, where he began working aboard a fishing boat. On March 13, 1991, Fautenberry met Jefferson Diffee at a local bar, and the two men went to Diffee’s apartment. While there, Fautenberry beat Diffee, handcuffed him, and stabbed him seventeen times, which resulted in his death. The local police discovered Fautenberry’s fingerprints at the scene of the crime, and on March 16, 1991, they arrested him for the murder of Diffee. The police then searched Fauten-berry’s storage locker and hotel room, where they found Daron’s briefcase, wristwatch, and Bible.
On March 17, 1991, while in police custody, Fautenberry called Federal Bureau of Investigation (“FBI”) Agent Larry Ott and left a message indicating that he wanted to talk. Agent Ott went to the jail, informed Fautenberry of his Miranda rights (which Fautenberry subsequently waived), and re*621corded Fautenberry’s confession to the murders of Nutley, Farmer, Daron, and Guthrie. Fautenberry accurately described the wounds inflicted upon each victim, and indicated that robbery was the motive for each killing. A few days later, Fautenberry called his old girlfriend, Olivia Priest-Herndon, and told her that he was “only after ... money” and that he “did it[,] so [he] gotta pay the price now.” Fautenberry also confessed to Tom Nelson of the Portland Police Department, informing Nelson where the bodies of Nutley and Guthrie were located. In August 1991, Fautenberry pleaded guilty in an Alaskan state court to the murder of Jefferson Diffee, and the court sentenced him to 99 years’ imprisonment.
In September 1991, the Alaskan authorities transferred Fautenberry to Hamilton County, Ohio, the county in which Cincinnati is located, where a grand jury had returned a five-count indictment charging Fautenberry with two counts of aggravated murder (both pertaining to the death of Daron), aggravated robbery, theft of a motor vehicle, and theft of a credit card. The aggravated murder charges included two specifications, either of which would render Fautenberry eligible for the death penalty under Ohio law: (1) killing Daron while committing aggravated robbery; and (2) killing Daron as part of a course of conduct involving the purposeful killing of two or more persons. See Ohio Rev.Code § 2929.04(A). Fautenberry waived his right to a trial by jury and later proffered a no-contest plea to all counts and specifications in the indictment.
The prosecution presented the three-judge panel1 with evidence, including the murder weapon, various other pieces of physical evidence, and transcripts of Fau-tenberry’s confessions to Agent Ott, Officer Nelson, and Ms. Priest-Herndon. After reviewing this evidence, the court concluded beyond a reasonable doubt that Fautenberry was guilty of all counts and specifications in the indictment, and accepted his plea. In September 1992, the three-judge panel held a sentencing hearing. The defense presented its mitigating evidence, which included testimony from Fautenberry, Dr. Nancy Schmidt-goessling, and friends who knew Fauten-berry well. Those friends included Louise Corcoran (a long-time friend of Fautenberry’s family), Ms. Priest-Hern-don (Fautenberry’s former girlfriend with whom he had lived), and Mary Theresa Slayback (a friend with whom Fautenber-ry lived during his early twenties). After hearing all of this evidence, as well as the testimony of the six law-enforcement officers presented during the mitigation hearing by the prosecution, the three-judge panel imposed the death penalty, finding that, despite the defense’s “thorough job in presenting the[] mitigating factors,” it was beyond a reasonable doubt that the aggravating factors sufficiently outweighed the mitigating factors.
Fautenberry appealed to the state appellate court, which affirmed his conviction and sentence in 1994. See State v. Fautenberry, No. C-920734, 1994 WL 35023 (Ohio Ct.App. February 9, 1994). The Ohio Supreme Court also affirmed on direct appeal, see State v. Fautenberry, 72 Ohio St.3d 435, 650 N.E.2d 878 (1995), and the United States Supreme Court denied Fautenberry’s request for review, see Fautenberry v. Ohio, 516 U.S. 996, 116 S.Ct. *622534, 133 L.Ed.2d 439 (1995). In March 1996, Fautenberry filed a motion for reconsideration with the Ohio Supreme Court, arguing that he had received ineffective assistance of appellate counsel during his direct appeal to that court. That motion was summarily denied in May 1996. See State v. Fautenberry, 78 Ohio St.3d 320, 677 N.E.2d 1194 (1997). In July 1996, Fautenberry filed, pursuant to Ohio App. R. 26(B), an application for reopening (i.e., a motion for delayed reconsideration) with the state court of appeals, alleging ineffective assistance of appellate counsel during his direct appeal to that court. This application was denied because Fautenberry “failed to demonstrate that there [was] good cause for filing [ ]his application more than two years after th[e] court’s judgment was journalized.” The Ohio Supreme Court affirmed that decision. See State v. Fautenberry, 78 Ohio St.3d 320, 677 N.E.2d 1194 (1997). Also in July 1996, Fautenberry filed his petition for post-conviction relief. The trial court denied the petition, and the court of appeals affirmed that denial. See State v. Fautenberry, No. C-971017, 1998 WL 906395 (Ohio Ct.App. December 31, 1998). The Ohio Supreme Court then declined to review Fautenberry’s petition, see State v. Fautenberry, 85 Ohio St.3d 1477, 709 N.E.2d 849 (1999), and refused Fautenber-ry’s request for reconsideration, see State v. Fautenberry, 86 Ohio St.3d 1422, 711 N.E.2d 1015 (1999).
In May 2000, Fautenberry filed his petition for a writ of habeas corpus with the federal district court, asserting nineteen grounds for relief. The State moved to dismiss, contending that many of Fauten-berry’s claims had been procedurally defaulted. The district court partially granted the State’s motion and dismissed some of Fautenberry’s claims. In a later opinion and order, the district court addressed the substance of Fautenberry’s non-defaulted claims, found them to be without merit, and denied Fautenberry’s petition for habeas relief. The district court issued a certificate of appealability on nine of Fautenberry’s claims (two of which relate to his claim for ineffective assistance of trial counsel during the pretrial and plea hearing, and which Fautenberry has consolidated for purposes of this appeal). Fautenberry asserts eight claims on appeal, and we will address them separately.
II.
“We review de novo the district court’s denial of ... [a] petition for a writ of habeas corpus.” Clinkscale v. Carter, 375 F.3d 430, 435 (6th Cir.2004). Fautenberry filed his habeas petition after the enactment of the Antiterrorism and Effective Death Penalty Act of 1996 (“AED-PA”); we will therefore review his claims under the standards set forth in that statute. See Lindh v. Murphy, 521 U.S. 320, 336, 117 S.Ct. 2059, 138 L.Ed.2d 481 (1997). AEDPA permits a federal court to grant a writ of habeas corpus only where the state court’s adjudication “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.” 28 U.S.C. § 2254(d)(1). “A federal habeas court may issue the writ under the ‘contrary to’ clause if the state court applies a rule different from the governing law set forth in [Supreme Court] cases, or if it decides a case differently than [the Supreme Court has] done on a set of materially indistinguishable facts.” Bell v. Cone, 535 U.S. 685, 694, 122 S.Ct. 1843, 152 L.Ed.2d 914 (2002). “The [federal] court may grant relief under the ‘unreasonable application’ clause if the state court correctly identifies the governing legal principle from [Supreme Court] decisions but unreasonably applies it to the facts of the *623particular case.” Id. “An unreasonable application of federal law is different from an incorrect application of federal law.” Woodford v. Visciotti, 537 U.S. 19, 25, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002) (quotations omitted). We may not issue the writ “simply because [we] conclude[] in [our] independent judgment that the relevant state-court decision applied clearly established federal law erroneously or incorrectly.” Williams v. Taylor, 529 U.S. 362, 411, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000). Instead, we must find that “the state court’s application of clearly established federal law [was] objectively unreasonable.” Cone, 535 U.S. at 694, 122 S.Ct. 1843. When identifying the relevant, controlling federal law for purposes of AED-PA analysis, we look to “the holding, as opposed to the dicta, of [the Supreme] Court’s decisions as of the time of the relevant state-court decisions.” Williams, 529 U.S. at 412, 120 S.Ct. 1495.
A. Ineffective Assistance of Trial Counsel During the Penalty Phase of the Proceedings
Fautenberry’s primary argument — to which he devoted most of his time at oral argument — is that his trial counsel rendered ineffective assistance during the penalty phase of his proceedings. An ineffective-assistance claim has two components: (1) counsel’s performance must have been deficient, and (2) counsel’s deficient performance must have prejudiced the defendant. Strickland v. Washington, 466 U.S. 668, 687, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). “[T]he proper standard for attorney performance is that of reasonably effective assistance,” id., as measured by “prevailing professional norms.” Rompilla v. Beard, 545 U.S. 374, 380, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005) (quotation marks and citations omitted). To establish deficient performance, the ha-beas petitioner must show that “counsel’s representation fell below an objective standard of reasonableness.” Strickland, 466 U.S. at 687-88, 104 S.Ct. 2052. When evaluating counsel’s performance, we “must indulge a strong presumption that counsel’s conduct falls within the wide range of reasonable professional assistance.” Darden v. Wainwright, 477 U.S. 168, 186, 106 S.Ct. 2464, 91 L.Ed.2d 144 (1986).
“A convicted defendant making a claim of ineffective assistance must identify the acts or omissions of counsel that are alleged not to have been the result of reasonable professional judgment.” Strickland, 466 U.S. at 690, 104 S.Ct. 2052. Fautenberry contends that his trial counsel rendered deficient performance by: (1) failing to communicate meaningfully with him, (2) failing to conduct a “reasonable” investigation for mitigating evidence, and (3) failing to retain “reasonable and necessary” experts. The state court rejected this claim, relying primarily on Fautenber-ry’s refusal to cooperate. We address Fautenberry’s three allegations in turn.
First, Fautenberry contends that his counsel failed to communicate meaningfully with him. The Sixth Amendment, however, protects the criminal defendant’s right to “adversarial process”; that is, to “have counsel[,] acting in the role of advocate^] ... require the prosecution’s case to survive the crucible of meaningful adversarial testing.” United States v. Cronic, 466 U.S. 648, 656, 104 S.Ct. 2039, 80 L.Ed.2d 657 (1984) (quotation marks and citations omitted).
[T]he appropriate inquiry focuses on the adversarial process, not on the accused’s relationship with his lawyer as such. If counsel is a reasonably effective advocate, he meets constitutional standards irrespective of his client’s evaluation of his performance. It is for this reason *624that we attach no weight to either respondent’s expression of satisfaction with counsel’s performance at the time of his trial, or to his later expression of dissatisfaction.
Id. at 657 n. 21, 104 S.Ct. 2039 (citations omitted). Thus, at its root, the ineffective-assistance analysis is based on “an objective standard of reasonableness.” Strickland, 466 U.S. at 688, 104 S.Ct. 2052.
The record shows that any communication problems between Fautenberry and his attorneys were the result of Fauten-berry’s own determined refusal to communicate, not a lack of availability of, or effort by, counsel. The record contains abundant evidence indicating that Fautenberry independently resolved not to communicate or cooperate with anyone, including his lawyers. Fautenberry’s former girlfriend, Ms. Priest-Herndon, testified that Fautenberry told her of his tactical decision not to communicate with his attorneys because he “thought that was his best way to beat this thing.” As the district court aptly noted, Fautenberry “has not cited, and the court is not aware of, any Supreme Court case holding that the Sixth Amendment requires counsel to persuade a recalcitrant client to do that which he ... refused to do.” We find nothing objectively unreasonable in these attorneys’ inability to communicate meaningfully with Fauten-berry, who steadfastly refused to communicate with them. See Strickland, 466 U.S. at 691, 104 S.Ct. 2052 (“The reasonableness of counsel’s actions may be determined or substantially influenced by the [accused’s] own statements or actions.”).
The evidence proffered to demonstrate that counsel had failed to communicate with Fautenberry is wholly unpersuasive. Fautenberry relies heavily upon statements in the medical notes recorded during his pretrial incarceration, in which he said, “I just want to be treated like a human being,” and “I[ am] as good as dead now.” When considered in their proper context, these statements, contrary to Fautenberry’s contention, do not demonstrate a dysfunctional attorney-client relationship. Instead, they display Fautenber-ry’s despondency because of the situation at hand — he had been incarcerated for six months awaiting trial for a murder that he had already (on at least three occasions) confessed to committing. Fautenberry also relies on the affidavit of Dr. Susan D. Shorr, a mitigation specialist who assisted his attorneys during their preparation for the sentencing hearing, in which she opined that Fautenberry became uncooperative and noncommunicative only “[a]s a result of the breakdown of the relationship between ... Fautenberry and his attorneys.” This unsubstantiated statement contradicts Ms. Priest-Herndon’s testimony that Fautenberry had told her of his independent, tactical decision not to communicate with his lawyers, but it substantiates all other accounts that Fautenberry, in fact, refused to communicate. Fauten-berry’s attempts to blame his counsel for his own recalcitrance and unwillingness to communicate do not demonstrate that counsel were deficient in failing to communicate meaningfully with Fautenberry.
Second, Fautenberry argues that his attorneys rendered ineffective assistance by failing to conduct a “reasonable” mitigation investigation that would have discovered his organic brain damage.2 Fautenberry contends that further investigation would have revealed his past head *625injuries and alerted his attorneys to the likelihood that he had permanent brain damage. We conclude, contrary to Fau-tenberry’s assertion, that counsel did in fact conduct a reasonable mitigation investigation.
Defense counsel interviewed and procured extensive testimony from lay witnesses, including Louise Corcoran (a close family friend) and Ms. Priest-Herndon (Fautenberry’s former girlfriend). These women testified regarding Fautenberry’s unstable family environment and emotionally abusive upbringing. Defense counsel also garnered the assistance of numerous experts, including Dr. Nancy Schmidt-goessling and Dr. James Tanley. In fact, counsel hired Dr. Tanley for the express purpose of conducting a neuropsychological examination — the most effective means possible of determining whether Fauten-berry had a brain impairment. But Fau-tenberry refused to submit to the examination. Counsels’ inability to discover or establish organic brain damage is directly attributable to Fautenberry’s refusal to cooperate, rather than any insufficiency in the investigation. Had Fautenberry not impeded Dr. Tanley’s examination, the doctor presumably would have discovered, verified, and revealed any such brain damage. We will not find counsel deficient simply because they did not succeed in discovering his brain damage or pursue unspecified, alternate avenues (which may or may not have revealed the brain damage). In light of the “strong presumption that counsel’s conduct falls within the wide range of reasonable professional assistance,” Darden, 477 U.S. at 186, 106 S.Ct. 2464, Fautenberry’s attorneys conducted a sufficient and reasonable mitigation investigation under the circumstances Fauten-berry created.
Third, Fautenberry contends that his counsel failed to retain “reasonable and necessary” experts, asserting that neither Dr. Schmidtgoessling nor Dr. Tanley provided him with a reasonable level of assistance. We find this argument baseless. Dr. Tanley provided as much assistance as was possible under the circumstances: He attempted to examine Fautenberry to determine whether he suffered from a brain impairment, but Fautenberry refused to participate in the examination. Thus, any deficiencies in Dr. Tanley’s assistance were the result of Fautenberry’s admitted refusal to cooperate, rather than any shortcoming of Dr. Tanley or any flaw in his work. We conclude as well that any alleged defect in Dr. Schmidtgoessling’s assistance cannot be attributed to counsel.
Fautenberry argues that Dr. Schmidt-goessling misdiagnosed his mental condition when she concluded that he did not suffer from organic brain damage. Even if we assume that Dr. Schmidtgoessling did misdiagnose Fautenberry,3 “[a] licensed practitioner is generally held to be competent, unless counsel has good reason to believe to the contrary.” Lundgren v. Mitchell, 440 F.3d 754, 772 (6th Cir.2006). Fautenberry has not shown that counsel had “good reason” to believe that Dr. Schmidtgoessling was incompetent, and we conclude that it was objectively reasonable for counsel to rely upon the doctor’s opinions and conclusions. See Campbell v. Coyle, 260 F.3d 531, 555 (6th Cir.2001) (holding, in a case where there was “no evidence that [the expert] was incompetent!] or that [the petitioner’s] lawyers had any reason to question [the expert’s] professional qualifications,” that “it was *626objectively reasonable for ... trial counsel to rely upon [the expert’s] diagnosis”). Under these circumstances, any inadequacies in Dr. Schmidtgoessling’s expert assistance — assuming there were any — cannot be the basis for a meritorious ineffective-assistance claim. Accordingly, we find no deficiency in counsels’ performance.
Even if Fautenberry could show that his counsels’ performance was deficient, he has not established that he was prejudiced by that performance. The prejudice prong requires the petitioner to show that “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Darden, 477 U.S. at 184, 106 S.Ct. 2464. “When a [petitioner] challenges a death sentence such as the one at issue in this case, the question is whether there is a reasonable probability that, absent the errors, the sentencer ... would have concluded that the balance of aggravating and mitigating circumstances did not warrant death.” Strickland, 466 U.S. at 695, 104 S.Ct. 2052. We thus “reweigh the evidence in aggravation against the totality of available mitigating evidence,” which, in this case, includes the mitigation evidence that was omitted because of counsel’s alleged deficiencies. See Wiggins v. Smith, 539 U.S. 510, 534, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003). A petitioner does not establish the prejudice element where he shows only that his counsel failed to present “cumulative” mitigation evidence, that is, evidence already presented to the jury. Broom v. Mitchell, 441 F.3d 392, 410 (6th Cir.2006). Rather, “to establish prejudice, the new evidence that a habeas petitioner presents must differ in a substantial way— in strength and subject matter — from the evidence actually presented at sentencing.” Clark v. Mitchell, 425 F.3d 270, 286 (6th Cir.2005).
Fautenberry argues that he was prejudiced because, due to counsel’s alleged deficiencies, the jury did not hear evidence about (1) his personal struggle with, and his family’s history of, depression, (2) the connection between his abusive childhood and the commission of these murders, (3) his head injuries and resulting organic brain damage, and (4) the sexual aspects of the murders he committed. But, Corcoran and Priest-Herndon testified extensively regarding Fautenberry’s rough upbringing, his difficult family life, his poor relationships with the father figures in his life, his mother’s constant struggle with depression, and his longing for acceptance. After hearing this testimony, as well as all the other mitigating evidence, the three-judge panel concluded that the mitigating factors included: (1) Fautenberry’s “past history”; (2) his “abuse[ ] as a child”; (3) the “rage of his childhood”; (4) his abuse of drugs and alcohol; and (5) his “low self-esteem and rejection.” The omitted mitigation evidence of Fautenberry’s depression, his family’s history of depression, and his abusive childhood mirrors the evidence actually presented at the sentencing hearing, and therefore its omission cannot be held to have prejudiced his mitigation defense. See Broom, 441 F.3d at 410. Furthermore, the three-judge panel had already heard some evidence about the sexual nature of the murders. At the guilt phase, the prosecutor had informed the court that prior to Fautenberry’s murder of Farmer in New Jersey, Fautenberry agreed to have sex with him in exchange for money. See Gillard v. Mitchell, 445 F.3d 883, 896 (6th Cir.2006) (stating that the sentencer was “privy” to evidence introduced during the guilt phase of trial and that counsel need not reintroduce it during the sentencing hearing). Thus, the court was aware that at least one of Fautenberry’s murders contained a sexual element, and, to the *627extent that Fautenberry relies upon this already-disclosed evidence, we find that it is cumulative and insufficient to establish prejudice.
Inclusion of the non-cumulative evidence (i.e., evidence of organic brain damage and the sexual nature of some of the murders) — which purportedly would have explained his impulsive and violent nature, inability to tolerate frustration, and sexual confusion — does not create “a reasonable probability that ... the sentencer ... would have concluded that the balance of aggravating and mitigating circumstances did not warrant death.” See Strickland, 466 U.S. at 695, 104 S.Ct. 2052. Furthermore, we question whether evidence of the sexual nature of the murders is even mitigating. The three-judge panel might just as well have viewed Fautenberry’s apparent strategy of preying on gay victims as an aggravating factor. Nor do the side-effects of organic brain damage significantly mitigate Fautenberry’s actions. According to Fautenberry’s medical reports, “[bjrain impairment of the sort ... apparent in ... Fautenberry’s case can ... cause serious problems in such areas of day-to-day functioning as impulse control[,] modulation of affect[,] planningf,] problem-solving[,] and the capacity to tolerate frustration.” It is highly unlikely that this sort of evidence would have altered the three-judge panel’s decision to impose the death sentence for Fautenber-ry’s murder of Daron, which they found was “contemplated and calculating” — a conclusion that is not at all mitigated or reduced by the traits associated with or the side effects of organic brain disorder. In short, when we aggregate all the mitigating evidence, including Fautenberry’s brain disorder and the sexual nature of the crimes (to the extent that those are even mitigating), and reweigh this evidence against the aggravating factors, we find that “these mitigating factors pale before the simple fact that [Fautenberry’s] actions were plotted, vicious, persistent[,] and utterly callous,” which was the finding of the three-judge panel. Accordingly, even if trial counsel rendered deficient performance, we find that those deficiencies did not prejudice Fautenberry.4
B. Trial Counsel’s Alleged Conflict of Interest
Fautenberry next contends that one of his trial counsel labored under a conflict of interest, which rendered his assistance per se ineffective and violated Fautenberry’s Sixth Amendment rights. Fautenberry argues that attorney Michael Walton had a conflict of interest because he was a trustee for Anderson Township, the township in which Daron’s body was found; as a trustee, Walton had a fiduciary duty to the citizens of that township to ensure that the criminal laws were upheld; and Walton’s duty to the citizens of that *628township conflicted with his duty to Fau-tenberry. The state trial court addressed this claim, contained in Fautenberry’s petition for post-conviction relief, and found the following facts: (1) “No evidentiary documents demonstrate that Anderson Township or the Anderson Township Trustees had an interest in the outcome of [Fautenberry’s] trial”; and (2) “No eviden-tiary documents demonstrate that service as a Township Trustee hampered [Fauten-berry’s] attorneys in any way.” The state appellate court determined that, “absent evidence of an actual conflict, there is no presumption of prejudice arising from the mere fact that defense counsel also serves in some capacity as a public official.” Fautenberry, 1998 WL 906395, at *5. The court then rejected Fautenberry’s claim, finding that Fautenberry had “not presented any evidence to support the conclusion that Walton’s position on the board [of trustees] in any way influenced his ability to defend Fautenberry at trial.” Id.
A habeas petitioner asserting an ineffective-assistance claim generally must show that his counsel’s performance was deficient and that the deficiency resulted in prejudice. Strickland, 466 U.S. at 687, 104 S.Ct. 2052. But a habeas petitioner can establish an ineffective-assistance claim without having to show prejudice if he demonstrates that his counsel labored under an “actual conflict” of interest. See Cuyler v. Sullivan, 446 U.S. 335, 348, 100 S.Ct. 1708, 64 L.Ed.2d 333 (1980); Strickland, 466 U.S. at 692, 104 S.Ct. 2052 (“Prejudice is presumed only if [the petitioner] demonstrates that counsel actively represented conflicting interests and that an actual conflict of interest adversely affected his lawyer’s performance.”) (quotations omitted). “An ‘actual conflict,’ for Sixth Amendment purposes, is a conflict of interest that adversely affects counsel’s performance.” Mickens v. Taylor, 535 U.S. 162, 171 n. 5, 122 S.Ct. 1237, 152 L.Ed.2d 291 (2002).
Fautenberry has not established an “actual conflict.” He does not challenge the state court’s factual findings that “[n]o ev-identiary documents demonstrate that ... the Anderson Township Trustees had an interest in the outcome of [his] trial” or that “[n]o evidentiary documents demonstrate that service as a Township Trustee hampered [his] attorneys in any way.” See 28 U.S.C. § 2254(e)(1) (providing that “a determination of a factual issue made by a[s]tate court shall be presumed to be correct” unless the habeas petitioner “re-butís] the presumption of correctness by clear and convincing evidence”). After independently reviewing the record, we find that Fautenberry has failed to demonstrate that his “counsel actively represented conflicting interests” or that any alleged conflict “adversely affected his lawyer’s performance.” See Strickland, 466 U.S. at 692, 104 S.Ct. 2052. Accordingly, we find this claim baseless.
C. The Prosecution Withheld Material Exculpatory Evidence in Violation of Brady
Fautenberry next argues that the prosecution withheld material exculpatory evidence in violation of Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963). The State argues that this claim has been procedurally defaulted. A claim is procedurally defaulted and is thus unreviewable by a federal habeas court where the “habeas petitioner fail[ed] to obtain consideration of [that] claim by a state court ... due to [his] failure to raise that claim before the state courts while state-court remedies [were] still available.” Seymour v. Walker, 224 F.3d 542, 549-50 (6th Cir.2000). Fautenberry admits that he did not present his Brady claim to the state courts, but argues that the prosecu*629tion’s failure to disclose the evidence is both the cause of and the prejudice resulting from the default. See Wainwright v. Sykes, 433 U.S. 72, 87, 97 S.Ct. 2497, 53 L.Ed.2d 594 (1977) (discussing “cause” and “prejudice”). A habeas petitioner shows “cause” where he demonstrates that he failed to raise a constitutional issue because it was “reasonably unknown to him” at the time. Amadeo v. Zant, 486 U.S. 214, 222, 108 S.Ct. 1771, 100 L.Ed.2d 249 (1988). The district court found that most of the challenged Brady evidence had not been disclosed to Fautenberry during his state-court proceedings, and the State does not directly dispute the “cause” element. Thus we will assume without deciding that Fautenberry satisfied that element.
The question of whether we may excuse Fautenberry’s procedural default, therefore, turns on the issue of prejudice. “Prejudice, for purposes of procedural default analysis, requires a showing that the default of the claim not merely created a possibility of prejudice to the defendant, but that it worked to his actual and substantial disadvantage, infecting his entire trial with errors of constitutional dimensions.” Jamison v. Collins, 291 F.3d 380, 388 (6th Cir.2002) (citing United States v. Frady, 456 U.S. 152, 170-71, 102 S.Ct. 1584, 71 L.Ed.2d 816 (1982)). Procedural default analysis on the issue of prejudice mirrors Brady materiality analysis, see id., so in determining whether Fautenberry has procedurally defaulted his Brady claim, we will follow the Supreme Court’s example and proceed under a Brady materiality analysis. See Strickler v. Greene, 527 U.S. 263, 282, 119 S.Ct. 1936, 144 L.Ed.2d 286 (1999). Evidence is deemed material for purposes of Brady “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.” United States v. Bagley, 473 U.S. 667, 682, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985). In a case such as this, involving a guilty or no-contest plea, the habeas petitioner may establish prejudice by showing that there is a reasonable probability that, but for the non-disclosure of evidence, “he would not have [entered. his plea], and would have insisted on going to trial.” See Hill v. Lockhart, 474 U.S. 52, 59, 106 S.Ct. 366, 88 L.Ed.2d 203 (1985). Alternatively, he can show that the findings and decision by the three-judge panel “would have been different.” Bagley, 473 U.S. at 682, 105 S.Ct. 3375. Our analysis must consider the totality of the undisclosed evidence, not each item in isolation. Id.
Fautenberry contends that the prosecution failed to produce five categories of exculpatory evidence: (1) evidence suggesting that FBI Agent Ott violated Fau-tenberry’s Fifth Amendment right to counsel by contravening the rule announced in Edwards v. Arizona, 451 U.S. 477, 484-85, 101 S.Ct. 1880, 68 L.Ed.2d 378 (1981) (upon requesting counsel, an accused may not be subjected to further questioning until counsel is provided, unless the accused reinitiates the questioning); (2) evidence indicating that the State of Ohio lacked jurisdiction and that venue was improper in Hamilton County, Ohio; (3) evidence demonstrating that victim Joseph Daron had verbal arguments with his girlfriend and another friend just prior to his disappearance; (4) evidence relating to the sexual nature of the murders; and (5) evidence indicating that Fautenberry was depressed and suicidal just prior to the murders. For the following reasons, we find that Fautenberry has failed to establish that any of this evidence is material for purposes of Brady.
First, the undisclosed documents concerning Fautenberry’s interaction with *630Agent Ott do not establish an Edwards violation. Fautenberry was arrested on May 16, 1991, and sometime during the following day, while in police custody, he invoked his Fifth Amendment right to counsel. According to Edwards and its progeny, an accused in police custody who has invoked his Fifth Amendment right to counsel is protected from further police questioning so long as “the accused himself [does not] initi'ate[ ] further communication, exchanges, or conversations with the police.” Edwards, 451 U.S. at 484-85, 101 S.Ct. 1880; see also Van Hook v. Anderson, 488 F.3d 411, 415 (6th Cir.2007) (en banc). Here, the evidence shows that it was Fautenberry himself who initiated communications with Agent Ott, after invoking his right to counsel. Sometime in the evening of May 17, 1991, a few hours after asking to speak with a lawyer, Fau-tenberry called Agent Ott and left a message indicating that he wanted to talk. At this point, Fautenberry had clearly initiated further communications with the police, and Agent Ott did not violate Fauten-berry’s Fifth Amendment right to counsel by questioning him at that time. The alleged Brady evidence shows only that Agent Ott returned Fautenberry’s call a few hours later and left a message stating that Fautenberry should call him if he still wanted to talk. After waiting for two hours without hearing from Fautenberry, Agent Ott called the jail, made arrangements to visit Fautenberry, spoke with Fautenberry (apparently without any objection or refusal on the part of Fauten-berry), and received Fautenberry’s confession to the four murders.
Fautenberry contends that this newly discovered evidence establishes an Edwards violation. We disagree. Fautenberry’s mere failure to return Agent Ott’s call and “confirm” his desire to speak does not negate Fautenberry’s pri- or, unambiguous initiation of further communication. To be sure, had Fautenberry reinvoked his Fifth Amendment right to counsel after initiating communication with Agent Ott, he would have been protected from further questioning, but Fau-tenberry does not assert that he did so and the alleged Brady evidence does not demonstrate that he did so. In order to invoke one’s Fifth Amendment right to counsel, the “suspect must unambiguously request counsel,” meaning that “he must articulate his desire to have counsel present sufficiently clearly that a reasonable police officer in the circumstances would understand the statement to be a request for an attorney.” Davis v. United States, 512 U.S. 452, 459, 114 S.Ct. 2350, 129 L.Ed.2d 362 (1994) (holding that an ambiguous mention of an attorney is not a request for counsel). The alleged Brady evidence does not indicate that Fautenber-ry renewed his request for an attorney, unambiguously or otherwise, and is therefore insufficient to establish an Edwards violation or to require the suppression of his confession to Agent Ott. Moreover, all of the events that transpired during this time were known to Fautenberry and there is no basis for assuming that the non-disclosure of this evidence affected his decision to enter his no-contest plea. See Bagley, 473 U.S. at 682, 105 S.Ct. 3375.
Second, the undisclosed evidence does not establish that Ohio lacked jurisdiction or that Hamilton County was not the proper venue for prosecution. The state court found that Fautenberry entered Daron’s car at “the on-ramp to [Interstate] 275 off of Highway 125,” see Fautenberry, 1994 WL 35023, at *2, which is located to the east of Cineinnáti, near the border of Hamilton County, Ohio, and Clermont County, Ohio. The state court concluded that Daron drove north on Interstate 275 to a destination “some ten *631miles north of Milford, Ohio,” see id., and dropped Fautenberry at a “restaurant near the junction of Interstate 71,” see Fautenberry, 650 N.E.2d at 880. The state court determined that Fautenberry shot Daron at this location, see id., which is in Hamilton County, Ohio; and that Fautenberry dumped Daron’s body on the north bank of the Ohio River, near the intersection of Highway 52 and Interstate 275, see Fautenberry, 1994 WL 35023, at *2, which is also in Hamilton County, Ohio. The body was later found in this general vicinity. It is therefore clear, based on the facts as found by the state court, that Ohio had jurisdiction over this murder, see Ohio Rev.Code §§ 2901.11(A)(1), 2901.11(B) (stating that a “person is subject to criminal prosecution and punishment in [Ohio] if ... [t]he person commits an offense under the laws of [Ohio], any element of which takes place in [Ohio],” and that the elements of a homicide offense “inelude[ ] the act that causes death”), and that Hamilton County was the proper venue. See Ohio Rev.Code § 2901.12(A) (noting that venue is proper “in the territory of which the offense or any element of the offense was committed”). None of the proffered Brady material rebuts any of these factual findings by clear and convincing evidence; thus we must presume that these factual findings are correct. See 28 U.S.C. § 2254(e)(1).
Fautenberry argues that two pieces of evidence indicate that the murder occurred in the Commonwealth of Kentucky. First, the prosecutor, in his recitation of the facts during the plea hearing, stated that the murder occurred at “a restaurant and motel parking lot just past the junction of I-71 and 1-275.” Fautenberry emphasizes that Interstate 275 — a highway that circles Cincinnati — and Interstate 71 intersect twice, once in the State of Ohio and once in the Commonwealth of Kentucky. This argument, of course, entirely ignores the overwhelming evidence indicating that Daron picked up Fautenberry in an eastern suburb of Cincinnati, Ohio, and drove north to Columbus, Ohio. This one arguably ambiguous statement by the prosecutor is woefully insufficient to rebut the abundant evidence indicating that this murder occurred near the intersection of Interstate 275 and Interstate 71 in Hamilton County, Ohio. Second, a psychiatric report, which contains a second-hand account of Fautenberry’s description of the murder, states that Fautenberry “drove to Kentucky!,]' pulled overt,] • • • [and] shot the man in the chest a couple of times.” This hearsay statement is contradicted by all the evidence in the record and does not amount to clear and convincing evidence to rebut the state court’s factual findings.
The alleged Brady evidence consists of FBI intra-department communications indicating uncertainty about jurisdiction and venue, and instructing the officers to obtain more evidence on these issues. These documents do not establish that jurisdiction or venue were proper elsewhere; at best, they call these issues into question. Even if we were to assume, as alleged by Fautenberry, that this evidence creates a genuine ambiguity as to the location of the murder, both jurisdiction and venue would nevertheless be proper in Hamilton County, Ohio. Ohio law provides that “[w]hen the offense involves the death of a person, and it cannot reasonably be determined in which jurisdiction the offense was committed, the offender may be tried in the jurisdiction in which the dead person’s body or any part of the dead person’s body was found.” Ohio Rev.Code § 2901.12(J). It is undisputed that Daron’s body was found on the north bank of the Ohio River in Hamilton County, Ohio. Therefore, even if the evidence as to the location of the murder were unclear, both jurisdiction and venue would lie in Hamilton County. We *632conclude that this evidence was not material because it did not establish an error in jurisdiction or venue.
The remainder of the alleged Brady evidence — evidence regarding Daroris arguments prior to his disappearance, the sexual nature of the murders, and Fautenberry’s depression and suicidal inclinations — is not material for Brady purposes. In fact, the evidence is virtually insignificant in light of the overwhelming evidence both of guilt (i.e., the confessions to Agent Ott, Officer Nelson, and Ms. Priest-Herndon and the physical evidence connecting Fautenberry to Daroris murder) and the sentencing specifications (i.e., the three-judge panel’s finding that the “mitigating factors pale before the simple fact that [Fautenberry’s] actions were plotted, vicious, persistent!,] and utterly callous”). Considering as we must the cumulative effect of all the alleged Brady evidence, we conclude that Fautenberry has failed to establish a “reasonable probability” that the disclosure of this evidence would have altered either his decision to enter a no-contest plea or the three-judge panel’s sentence of death. See Bagley, 473 U.S. at 682, 105 S.Ct. 3375. Because this evidence is not material under Brady, Fautenberry cannot show prejudice to excuse his procedural default. See Jamison, 291 F.3d at 388. And because Fautenberry cannot establish prejudice to excuse his procedurally defaulted Brady claim, he is not entitled to habeas relief on that basis.5
D. Ineffective Assistance of Trial Counsel During the Pretrial Preparation and Plea Hearing
Fautenberry argues that his trial counsel rendered ineffective assistance during their pretrial preparation and at the plea hearing. In his habeas petition, Fautenber-ry presents three subparts to this ineffective-assistance claim: (A) counsel failed to engage in an adequate investigation, (B) counsel provided him with erroneous information regarding the implications of pleading no contest to the charges against him, and (C) counsel failed to hold the prosecution to its burden of proof at the plea hearing. The district court determined that each subpart of this ineffective-assistance claim had been procedurally defaulted. The court granted Fautenberry a certificate of appealability only on Subparts A and C, so we do not address the allegations asserted under Subpart B.
The State argues that we should affirm the district court’s conclusion that Fautenberry procedurally defaulted this claim. A habeas petitioner procedurally defaults a claim where “a state procedural rule ... prevents the state courts from reaching the merits of the petitioner’s claim.” Seymour, 224 F.3d at 549-50. Federal courts must consider four factors when determining whether a habeas petitioner has procedurally defaulted a claim. Gonzales v. Elo, 233 F.3d 348, 353 (6th Cir.2000); see also Maupin v. Smith, 785 F.2d 135, 138 (6th Cir.1986).
*633First, the court must determine that there is a state procedural rule that is applicable to the petitioner’s claim and that the petitioner failed to comply with the rule. Second, the court must decide whether the state courts actually enforced the state procedural sanction. Third, the court must decide whether the state procedural forfeiture is an “adequate and independent” state ground on which the state can rely to foreclose review of a federal constitutional claim.
Jacobs v. Mohr, 265 F.3d 407, 417 (6th Cir.2001) (quoting Maupin, 785 F.2d at 138) (alterations omitted). “Once the court determines that a state procedural rule was not complied with and that the rule was an adequate and independent state ground,” the court must move to the fourth factor. Maupin, 785 F.2d at 138. The fourth factor allows a petitioner to avoid or excuse procedural default if he demonstrates “that there was cause for him to not follow the procedural rule and that he was actually prejudiced by the alleged constitutional error.” Id. (quotation omitted).
Under Subpart A of this claim, Fauten-berry contends that his counsels’ performance was deficient because: (1) they did not interview a sufficient number of the prosecution’s potential witnesses, (2) they did not object to venue in Hamilton County, and (3) they did not file a motion to suppress Fautenberry’s confession to Agent Ott. Fautenberry presented this claim to the state court in his petition for post-conviction relief, but he alleged only that counsel were ineffective because they failed to object to venue; he did not challenge counsels’ failure to interview a sufficient number of witnesses or to file a suppression motion. Because Fautenber-ry undeniably did not present these allegations to the state courts, we find that the district court correctly held that they were procedurally defaulted. See Seymour, 224 F.3d at 549-50.
The state appellate court addressed the only allegation presented to it, namely, that counsel did not object to venue, and held, pursuant to State v. Cole, 2 Ohio St.3d 112, 443 N.E.2d 169 (1982), that “this is a claim which could and should have been raised by Fautenberry on direct appeal and is, therefore, barred by the doctrine of res judicata.” Fautenberry, 1998 WL 906395, at *3. In Cole, the Ohio Supreme Court held that res judicata is a proper basis upon which to dismiss an ineffective-assistance claim in a petition for post-conviction relief where a defendant who is represented by new counsel on direct appeal fails to raise that claim and the basis for that claim “could fairly be determined without examining evidence outside the record.” 443 N.E.2d at 171. We have in the past recognized that Ohio’s application of res judicata pursuant to Cole is an actually enforced, adequate and independent state ground upon which the Ohio state courts consistently refuse to review the merits of a defendant’s claims. See Byrd v. Collins, 209 F.3d 486, 521-22 (6th Cir.2000) (“Ohio state courts consistently invoke Cole and apply res judicata when a defendant, who is represented by new counsel on direct appeal, fails to raise at that stage of the litigation an ineffective assistance of trial counsel claim appearing on the face of the record.”); see also Mapes v. Coyle, 171 F.3d 408, 421 (6th Cir.1999) (rejecting the habeas petitioner’s attempt to “demonstrate Ohio’s wavering commitment to its procedural default rules”). We therefore conclude that Fau-tenberry has procedurally defaulted the allegation that counsel failed to object to venue. In short, then, Fautenberry has procedurally defaulted all the claims of *634deficiency raised in Subpart A.6
Even if we were to conclude that Fau-tenberry did not procedurally default Sub-part A of this ineffective-assistance claim, we would find it to be without merit. First, Fautenberry argues that his counsel did not interview a sufficient number of the prosecution’s potential witnesses. Noting that his attorneys billed most of their investigation time in one-hour increments, he surmises that they were not interviewing any of the out-of-state witnesses. This argument is based on sheer speculation; it does not account for the reasonable inference that counsel interviewed some of the witnesses (including out-of-state witnesses) via phone. Fauten-berry does not indicate how many witnesses were actually interviewed or how many more should have been interviewed. Fautenberry has the burden of establishing his counsel’s deficiency, and this speculative argument is insufficient to support an ineffective-assistance claim. Second, Fautenberry cannot establish that he was prejudiced by counsel’s failure to raise the venue issue because the evidence overwhelmingly indicated that the murder occurred in Hamilton County and, to the extent that the evidence was less than conclusive on this issue, venue was proper in the jurisdiction where Daron’s body was found, which was Hamilton County. See Ohio Rev.Code § 2901.12(J). Finally, Fautenberry cannot establish prejudice from counsel’s failure to file a motion to suppress the statements made to Agent Ott because he cannot prove that conversation was improper, as there is no evidence of an Edwards violation. See Edwards, 451 U.S. at 484-85, 101 S.Ct. 1880.
In Subpart C of this claim, Fautenberry alleges that his trial counsel should not have stipulated to the prosecution’s evidence at the plea hearing. Fautenberry readily admits that he did not raise that portion of his claim on direct appeal or during post-conviction proceedings, but contends that he preserved this issue by raising it in his motion for reconsideration with the Ohio Supreme Court and his application for reopening with the Ohio Court of Appeals. Both Fautenberry’s motion for reconsideration with the Ohio Supreme Court and his application for reopening with the court of appeals alleged ineffective assistance of appellate counsel. This claim, on the other hand, alleges ineffective assistance of trial counsel. The district court correctly concluded that the allegations in Fautenberry’s motions for reconsideration and reopening, which argued only ineffective assistance of appellate counsel, did not fairly present his ineffective assistance of trial counsel claim to the state court, and that Fautenberry had defaulted Subpart C of this claim.
E. Waiver of Right to a Jury Trial During the Penalty Phase of the Proceedings
Fautenberry argues here that although he waived his right to a jury trial during the guilt phase of his proceedings, he did not waive his right to a jury trial during the penalty phase. The substance of this argument is entirely different from the “waiver of jury trial” argument he presented to the state trial court in his petition for post-conviction relief, in which he alleged that: (1) his attorneys failed to gain his trust and thus failed to provide him with the necessary information re*635garding his constitutional rights; (2) his attorneys provided him with incorrect information about the waiver; and (3) he was psychologically and mentally unable to waive his right to a jury trial. The state trial court made the following findings of fact: (1) “[Fautenberry] was competent when the jury waiver occurred”; (2) “[Fautenberry] acknowledged discussing the [waiver] with both attorneys”; and (3) “[t]he court went over the law regarding the waiver and the proceedings before a three[-]judge panel with petitioner[,] and petitioner acknowledged that he fully understood what he was doing.” The state trial court concluded that this claim “could have been raised at trial or on direct appeal” and therefore was “barred by res judicata.” The state appellate court, however, did not consider whether this claim was proeedurally barred, but addressed the merits of the claim and rejected it because “the record on review show[ed] that Fautenberry was engaged in a colloquy by the judge[ ] and indicated squarely that he understood that he was waiving his right to a jury trial and that no promises had been made to him.” Fautenberry, 1998 WL 906395, at *7.
In his habeas petition, Fautenberry argues that he did not waive his right to a jury trial on the penalty phase of his prosecution. This claim that he did not knowingly waive his right to be sentenced by a jury is materially different from the claim he raised in the state court, which challenged his competence and knowledge in connection with the waiver of his right to a trial by jury and did not distinguish between the waiver of his right to a jury at the guilt phase and the waiver of his right to a jury at the penalty phase. The argument Fautenberry raises in these habeas proceedings, on the other hand, effectively concedes that he waived his right to a jury trial but contends that he did not knowingly or voluntarily waive his right to be sentenced by a jury.
Before us, the State argues that Fautenberry’s claim that he did not waive his right to a jury trial on the penalty phase has been proeedurally defaulted because the state court applied res judicata and refused to address the merits. “In determining whether state courts have relied on a procedural rule to bar review of a claim, we look to the last reasoned opinion of the state courts.... ” Mason v. Mitchell, 320 F.3d 604, 635 (6th Cir.2003). Because the state appellate court’s decision is the “last reasoned opinion of the state courts,” we must look to that decision. That decision, however, did not mention res judicata but addressed the merits of the waiver claim that Fautenberry raised in his post-conviction proceedings. But that decision did not address at all the claim that Fautenberry makes in his habe-as petition and in this appeal, because Fautenberry never presented that claim to the state post-conviction courts. We therefore conclude that although the State’s reason for claiming procedural-default argument is incorrect, this claim is nonetheless defaulted. See Seymour, 224 F.3d at 549-50.
Even if we were to reach the substance of this claim, we would find it to be without merit. Fautenberry has not presented, and we have not discovered, any Supreme Court precedent establishing that a capital defendant has a constitutional right to be sentenced by a jury in state court. Fautenberry argues, relying on Hicks v. Oklahoma, 447 U.S. 343, 100 S.Ct. 2227, 65 L.Ed.2d 175 (1980), that Ohio statutory law creates a right to be sentenced by a jury and that the Fourteenth Amendment protects that right. Fauten-berry’s reliance on Hicks is misplaced. In Hicks, it was undisputed that the defendant had a statutory right to be sentenced *636by the jury; the issue in that case was whether the state court violated the defendant’s due process rights by restricting the jury’s discretion through a habitual-offender statute that was later declared to be unconstitutional. Id. at 345-46, 100 S.Ct. 2227. Here, however, Fautenberry did not have a statutory right to be sentenced by a jury. The applicable Ohio statute states:
[I]f the offender is found guilty of both the [aggravated murder] charge and one or more of the specifications, the penalty to be imposed on the offender ... shall be determined by one of the following:
(a) By the panel of three judges that tried the offender upon the offender’s waiver of the right to trial by jury;
(b) By the trial jury and the trial judge, if the offender was tried by jury.
Ohio Rev.Code § 2929.03(C)(2) (1981).7 See also Ohio R.Crim. Pro. 11(c)(3).
Fautenberry clearly and expressly waived his right to a jury trial. His waiver stated: I, John Fautenberry, ... hereby knowingly, intelligently!,] and voluntarily waive and relinquish my right to a trial by Jury, and elect to be tried by a Judge of the Court in which the said cause be pending.” The trial court explained Fau-tenberry’s rights and asked him if it was his desire to knowingly, intelligently, and voluntarily relinquish his right to a jury trial; he responded in the affirmative. The trial court also informed Fautenberry that if his waiver were accepted and if he pleaded guilty to the charges against him, he would be sentenced by a three-judge panel (rather than a jury). We find, as did the state court, that Fautenberry knowingly and voluntarily waived his right to a jury trial. We conclude further that he did not have a statutory right (let alone a constitutionally protected right) to be sentenced by a jury. This case, therefore, is distinguishable from Hicks, and we find no basis upon which to grant habeas relief.
F. Knowing and Voluntary Nature of the No-Contest Plea
Fautenberry next argues that he did not knowingly or voluntarily enter his no-contest plea. The state trial court, in resolving Fautenberry’s petition for post-conviction relief, made the factual finding that “the plea was properly accepted,” and arrived at the legal conclusion that the plea was knowingly, intelligently, and voluntarily entered. The state appellate court, in affirming the trial court’s decision, evaluated Fautenberry’s three evi-dentiary bases for this claim: (1) the affidavit of mitigation specialist, Dr. Shorr, which stated that defense counsel failed to “maintain a positive, working relationship with ... Fautenberry”; (2) documents concerning Fautenberry’s psychological condition prior to his plea; and (3) Fautenberry’s own affidavit stating that his attorneys did not adequately apprise him of the consequences of his plea. Fautenberry, 1998 WL 906395, at *6. The court concluded that Dr. Schorr’s opinion was insufficient to rebut the abundant evidence in the record demonstrating that his plea was knowingly and voluntarily entered. Id. at *7. The court found the documents concerning his psychological condition to be unpersuasive because he “was twice found competent to stand trial.” Id. And the court refused to give much weight to Fau-tenberry’s self-serving affidavit. Id.
A guilty or no-contest plea involves a waiver of many substantial constitutional rights, see Boykin v. Alabama, 395 U.S. 238, 243, 89 S.Ct. 1709, 23 L.Ed.2d 274 (1969), and a court may accept a guilty or *637no-contest plea only where it is a “voluntary!,] • • • knowing, intelligent act[ ] done with sufficient awareness of the relevant circumstances and likely consequences,” see Brady v. United States, 397 U.S. 742, 748, 90 S.Ct. 1463, 25 L.Ed.2d 747 (1970). On appeal, Fautenberry asserts that his plea was not knowing and voluntary for four reasons: (1) his trial counsel incorrectly informed him that if he pleaded no contest, the three-judge panel would not learn that he committed other murders, the three-judge panel would perceive his no-eontest plea as mitigating, and he would preserve his right to appeal the denial of his pretrial motions; (2) he was unaware of exculpatory evidence impermissibly withheld by the prosecution (i.e., the alleged Brady evidence); (3) he suffered from serious mental illnesses at the time of his plea; and (4) during the plea colloquy, the court “was disorganized!,] ... failed to [engage] in a meaningful dialogue,” and failed to correct the misinformation provided by trial counsel.
Each asserted basis for this claim is without merit. First, aside from Fauten-berry’s unsubstantiated, self-serving affidavit, there is no evidence in the record indicating that trial counsel incorrectly advised him of the implications of entering a no-contest plea. During the plea colloquy, trial counsel indicated that they “explained to [Fautenberry] in detail what the plea of no contest means” and that “he understood] ... that by pleading no contest ... he [was] in essence giving up [a] substantial number of rights, particularly those that may be available to him at the appellate level if he were to go to trial.” We agree with the state courts that Fautenberry’s own self-serving affidavit is insufficient to rebut the contrary evidence in the record or to undermine the knowing and intelligent nature of his plea. Second, we have already concluded, in the context of the Brady claim, that there is no reasonable likelihood that the disclosure of the undisclosed evidence would have altered Fautenberry’s decision to enter his plea because that evidence was not material to his defense. Third, none of the psychological evidence indicates that Fautenberry was mentally incapable of understanding, appreciating, and waiving his constitutional rights, and he does not challenge the state court’s finding that he was twice found competent to stand trial. Fourth, and most importantly, the state court’s plea colloquy was methodical and thorough, not “disorganized” or “failing to result in a meaningful dialogue” as alleged by Fautenberry. And, at the plea hearing, Fautenberry indicated that he did not have any questions about his rights. In sum, we reject this claim as unfounded, and instead agree with the state court that Fautenberry knowingly and voluntarily waived his rights during the entry of his guilty plea.
G. Admission of Victim Impact Evidence
Fautenberry argues that the state court violated his Eighth Amendment rights by admitting victim-impact evidence. Fautenberry specifically challenges the admission of statements from Daron’s fi’iends and family recommending that he receive the maximum available sentence: (1) Daron’s ex-wife stated that Fautenberry “should receive the maximum possible sentence”; (2) Daron’s father stated that Fautenberry is “an animal with no conscience” and that “the maximum possible sentence is the only appropriate punishment”; and (3) Daron’s supervisor at work stated that Fautenberry committed “an extremely brutal offense” and he “should receive a maximum penalty.” The Ohio Supreme Court addressed this argument on direct appeal and found “error in the admission of ... the victim-impact *638statements [that] relate to sentencing recommendations.” Fautenberry, 650 N.E.2d at 882. The court nevertheless was “not persuaded that such error warrant[ed] reversal” because there was no indication that the three judges who sentenced Fau-tenberry “contemplated or relied upon the victim-impact evidence which was available to them.” Id.
In Booth v. Maryland, 482 U.S. 496, 509, 107 S.Ct. 2529, 96 L.Ed.2d 440 (1987), the Supreme Court held that the introduction of victim-impact evidence “at the sentencing phase of a capital murder trial violates the Eighth Amendment.” Id. Just a few years later, however, the Court retreated from this holding, declaring that “if the State chooses to permit the admission of victim impact evidence ..., the Eighth Amendment erects no per se bar.” Payne v. Tennessee, 501 U.S. 808, 827, 111 S.Ct. 2597, 115 L.Ed.2d 720 (1991). The Payne Court noted that it overturned only that part of Booth that disallowed “evidence ... relating to the victim and the impact of the victim’s death on the victim’s family.” Id. at 830 n. 2, 111 S.Ct. 2597. The Court did not disturb that portion of Booth that forbids “a victim’s family members’ characterization and opinions about the crime, the defendant, and the appropriate sentence.” Id.; see also Welch v. Sirmons, 451 F.3d 675, 703 (10th Cir.2006) (recognizing that many circuits have found that this portion of Booth’s holding “survived the holding in Payne and remains valid”). We agree with the Ohio Supreme Court that the state trial court erred in admitting this evidence.
The issue, then, is whether the Ohio Supreme Court’s reasoning that this error did not “warrant reversal” was contrary to or involved an unreasonable application of Supreme Court precedent. Fau-tenberry argues that the Ohio Supreme Court’s decision was an unreasonable application of Booth because it employed a sort of “harmless error” analysis, whereas Booth did not require the defendant to demonstrate “actual prejudice.”
We find helpful the Tenth Circuit’s decision in Hain v. Gibson, 287 F.3d 1224, 1239-40 (10th Cir.2002), and our unpublished decision in Brewer v. Anderson, 47 Fed.Appx. 284, 287-88 (6th Cir.2002) (unpublished case). In Hain, the Tenth Circuit noted that “[t]he decision in Booth does not expressly indicate whether the Court believed such errors to be trial errors subject to harmless error review, or structural error[s] requiring automatic reversal.” Hain, 287 F.3d at 1239 n. 11. The Hain court thus concluded that the state court in its case did not “unreasonably appl[y] Booth in concluding that such errors [were] subject to harmless error review.” Id. In Brewer, we reviewed, as we do here, the admission of victim-impact evidence before a three-judge panel. The Ohio Supreme Court, both in Brewer and in the present case, relied on State v. Post, 32 Ohio St.3d 380, 513 N.E.2d 754, 759 (1987), which states: “Absent an indication that the panel was influenced by or considered the victim impact evidence in arriving at its sentencing decision, the admission of the victim impact statement ... did not constitute prejudicial error.” Id. And both here and in Brewer, the Ohio Supreme Court found no indication that the three-judge panel considered or contemplated the victim-impact evidence available to them.8 We concluded in Brewer, 47 Fed. *639Appx. at 288, that the state court’s application of Booth was not unreasonable and we reach the same conclusion here.
“An unreasonable application of federal law is different from an incorrect application of federal law,” Woodford, 537 U.S. at 25, 123 S.Ct. 357 (quotation marks omitted); in order to grant habeas relief under the “unreasonable application” clause, we must determine that “the state court’s application of clearly established federal law [was] objectively unreasonable,” Cone, 535 U.S. at 694, 122 S.Ct. 1843. Contrary to Fautenberry’s argument, Booth does not indicate whether the erroneous admission of victim-impact evidence warrants automatic reversal or whether such errors are subject to harmless-error review. See Rain, 287 F.3d at 1239 n. 11. Therefore, the Ohio Supreme Court’s decision to engage in a form of harmless-error analysis does not constitute an unreasonable application of Booth because Booth did not address that issue.
Moreover, we question whether Booth even applies here. Booth involved the improper admission of victim-impact evidence to a jury, whereas this case involves the improper admission of victim-impact evidence to a three-judge panel. The Court in Booth was greatly concerned that the victim-impact evidence might (1) “distract the sentencing jury from its constitutionally required task [of] determining whether the death penalty is appropriate in light of the background and record of the accused and the particular circumstances of the crime,” Booth, 482 U.S. at 507, 107 S.Ct. 2529, (2) “divert the jury’s attention away from the defendant’s background and record[ ] and the circumstances of the crime,” id. at 505, 107 S.Ct. 2529, or (3) “create an impermissible risk that the capital sentencing decision will be made in an arbitrary manner.” Id. Those considerations are severely diminished—if not entirely obviated—when the sentencer is a judge or a three-judge panel, rather than a lay jury. We conclude that Booth has minimal relevance when the victim-impact evidence is presented to a three-judge panel, see Brewer, 47 Fed.Appx. at 287-88 (affirming, as a reasonable application of Supreme Court precedent, the Ohio Supreme Court’s conclusion that “Booth does not apply to situations where a defendant is tried by a three-judge panel rather than a jury”), and hold that the Ohio Supreme Court did not unreasonably apply Supreme Court precedent.
H. Ineffective Assistance of Appellate Counsel
Fautenberry contends that he received ineffective assistance of counsel during his direct appeal to the state court of appeals. The district court found that this claim had been procedurally defaulted because Fau-tenberry did not present it to the state appellate court in a timely application for reopening, which is the proper procedure in Ohio for raising ineffective assistance of appellate counsel claims.
The State urges us to affirm the district court’s conclusion that this claim has been procedurally defaulted. Under Ohio law, a criminal defendant must raise his ineffective assistance of appellate counsel claim in an application for reopening (i.e., a motion for delayed reconsideration) filed “in the court of appeals where the alleged error took place.” State v. Mur-*640nahan, 63 Ohio St.3d 60, 584 N.E.2d 1204, 1209 (1992). Ohio App. R. 26(B)(1) states that an application for reopening “shall be filed in the court of appeals where the appeal was decided within ninety days from journalization of the appellate judgment unless the applicant shows good cause for filing at a later time.” The state court of appeals -rejected Fautenberry’s direct appeal in February 1994, and the Ohio Supreme Court denied his appeal in 1995- In January 1996, after the conclusion of Fautenberry’s direct appeal, the state court appointed new counsel to represent Fautenberry during his post-conviction proceedings. In March 1996, Fauten-berry, through his new counsel, filed a motion for reconsideration with the Ohio Supreme Court, alleging ineffective assistance of appellate counsel during his direct appeal to that court; the motion was denied in May 1996. In July 1996, Fauten-berry filed an application for reopening with the state court of appeals, asserting ineffective assistance of appellate counsel during his direct appeal to the court of appeals. The court of appeals denied that application because Fautenberry “failed to demonstrate that there [was] good cause for filing []his application more than two years after th[e] court’s judgment was journalized.” The court further reasoned that Fautenberry could have raised the issue of ineffective assistance of appellate counsel in his “previous application for reopening in the Supreme Court” and “provided no explanation as to why the application of res judicata would be unjust.” The Ohio Supreme Court affirmed this decision in April 1997 “for the same reasons articulated by the court of appeals.” Fautenberry, 677 N.E.2d at 1195.
We hold that Fautenberry has procedurally defaulted this claim. The state appellate court denied his direct appeal in February 1994, and Fautenberry waited more than two years to file his application for reopening with that court. Fautenberry demonstrated good cause for not filing his application prior to January 1996, because until that point he was represented by the same counsel on appeal and we conclude that it would be unreasonable to expect counsel to raise an ineffective assistance claim against himself. But Fautenberry did not show good cause for failing to file his application until July 1996, six months after the appointment of new counsel. Additionally, he did not explain why he failed to raise this claim in his motion for reconsideration filed with the Ohio Supreme Court in March 1996. We conclude that Fautenberry did not comply with the timeliness requirements in Ohio App. R. 26(B) and that those time constraints are an actually enforced, adequate and independent state ground upon which the Ohio courts consistently refuse to address ineffective assistance of appellate counsel claims. See Coleman v. Mitchell, 244 F.3d 533, 539-40 (6th Cir.2001) (finding that the petitioner procedurally defaulted his claim because he failed to comply with the requirements in Ohio App. R. 26(B)); Wickline v. Mitchell, 319 F.3d 813, 823 (6th Cir.2003) (finding that the petitioner’s ineffective assistance of appellate counsel claims were procedurally defaulted because he failed to comply with the rule set forth in Murnahan).
Fautenberry argues that the timeliness requirements in Ohio App. R. 26(B) are not “adequate and independent” state grounds upon which to find that his claims have been procedurally defaulted. Maupin, 785 F.2d at 138. “To be adequate, a state procedural rule must be firmly established and regularly followed.... ” Hutchison v. Bell, 303 F.3d 720, 737 (6th Cir.2002) (quotation marks omitted). Fauten-berry relies on our decision in Franklin v. Anderson, 434 F.3d 412, 418-21 (6th Cir.2006), to argue that an untimely applica*641tion for reopening pursuant to Ohio App. R. 26(B) is not a “firmly established and regularly followed” procedural rule, id. at 418, particularly emphasizing the Franklin court’s statement that “[a] review of the relevant case law reveals that the Ohio Supreme Court has been erratic in its handling of untimely Rule 26(B) applications in capital cases.” Id. at 420. Fau-tenberry wishes to elevate this statement to an all-encompassing, ever-applicable.legal proposition that will forever (or at least for a very long time) bar the federal courts from finding that an ineffective assistance of appellate counsel claim has been proee-durally defaulted where the state court refused to address the merits of that claim because of the time constraints in Ohio App. R. 26(B). But the “firmly established and regularly followed” inquiry cannot be made once and for all. Instead we must consider whether the “adequate and independent state procedural bar ... [was] ‘firmly established and regularly followed’ by the time as of which it [was] to be applied.” Ford v. Georgia, 498 U.S. 411, 424, 111 S.Ct. 850, 112 L.Ed.2d 935 (1991) (emphiasis added). Cf. Rogers v. Howes, 144 F.3d 990, 994 n. 5 (6th Cir.1998) (“[T]he question is not whether the state courts consistently apply the procedural bar in the present.”). Put differently, we ask “whether, at the time of the petitioner’s actions giving rise to the default, the petitioner could ... be deemed to have been apprised of the rule’s existence.” Hutchison, 303 F.3d at 737 (quotations and alterations omitted).
We find that, as of the time that Fauten-berry should have filed his application for reopening (i.e., when he acquired new counsel in January 1996), the time constraints in Ohio App. R. 26(B) were firmly established and regularly followed, and he was or should have been apprised of the rule’s existence. The Franklin decision is not to the contrary. To begin with, the facts in that case are materially different from those here. In Franklin, the defendant had filed his application for reopening with the court of appeals one year after the Ohio Supreme Court issued its decision in Mumahan and just prior to the effective date of Ohio App. R. 26(B). The Ohio procedural rule in effect at the time Mur-nahan was decided required that such an application be filed within ten days from the entry of the opinion for which reconsideration was sought; Mumahan itself did not set any time frame within which such an application was required to be filed, but opined that courts should take a more lenient approach with regard to the time for filing. Ohio App. R. 26(B) was then created to provide the time frame for these applications. As we observed in Franklin, at the time Franklin filed his application for delayed reconsideration, the law regarding the time period within which ineffective assistance of appellate counsel claims were required to be filed was unclear. But we specifically noted in Franklin that “[f]or several years following the enactment of amended Rule 26(B) [in July 1993], the Ohio Supreme Court regularly enforced the rule’s timeliness requirements.” Franklin, 434 F.3d at 420 (citing a string of Ohio Supreme Court cases).
Fautenberry’s circumstances are significantly different from those in Franklin. Fautenberry did not obtain new counsel (and thus cannot be held responsible for failing to file an application for reopening) until January 1996, which is almost four years after Murnahan was decided, and two-and-a-half years after Ohio App. R. 26(B) became effective.. By that time, the 90-day time limit in Ohio App. R. 26(B) was clearly established and Fautenberry was or should have been apprised of its existence. Nevertheless, after obtaining his new counsel, Fautenberry waited six months — during which he filed a motion *642for reconsideration with the Ohio Supreme Court in which he could have raised this claim but did not—before filing his application for reopening. We conclude that the holding of Franklin—in which the court stated that “Rule 26(B) is not an adequate and independent state rule that can preclude consideration of Franklin’s ineffective assistance of appellate counsel claim,” id. at 421—is inapplicable here. Fauten-berry has procedurally defaulted his ineffective assistance of appellate counsel claim.
But even if this claim was not procedurally defaulted, it must fail because it is meritless. A defendant is entitled to effective assistance of counsel during his first appeal of right, Evitts v. Lucey, 469 U.S. 387, 396, 105 S.Ct. 830, 83 L.Ed.2d 821 (1985), but effective assistance does not require counsel to raise every nonfrivo-lous argument on appeal, Jones v. Barnes, 463 U.S. 745, 751-52, 103 S.Ct. 3308, 77 L.Ed.2d 987 (1983). “[0]nly when ignored issues are clearly stronger than those presented, will the presumption of effective assistance of [appellate] counsel be overcome.” Monzo v. Edwards, 281 F.3d 568, 579 (6th Cir.2002).
Fautenberry argues that his appellate counsel rendered ineffective assistance because they raised only seven issues on direct appeal, and, more specifically, they failed to present four allegedly meritorious claims on appeal: (1) the trial court’s failure to review evidence as required by Ohio Crim. R. 11(C)(3); (2) ineffective assistance of trial counsel; (3) Ohio’s discriminatory method for selecting grand-jury forepersons; and (4) the fact that Fauten-berry was forced to wear shackles in the presence of the trial court. The mere fact that appellate counsel confined their appeal to seven issues does not establish that counsel were ineffective; it is often best to filter out less meritorious issues so that counsel can emphasize those that present the best opportunity for relief on appeal. See Jones, 463 U.S. at 752, 103 S.Ct. 3308 (“There can hardly be any question about the importance of having the appellate advocate examine the record with a view to selecting the most promising issues for review.”). And the four “ignored issues” were not “clearly stronger than those presented.” The first of the four appears to be baseless. Contrary to Fau-tenberry’s argument, the trial court did in fact review evidence as required by Ohio Crim. R. 11(C)(3); most notably, the prosecutor introduced and the trial court reviewed the transcripts of Fau-tenberry’s confessions to Agent Ott, Officer Nelson, and Ms. Priest-Herndon. We have already concluded that the second omitted issue—i.e., ineffective assistance of trial counsel—is baseless. The third omitted claim, which alleges the impropriety of Ohio’s selection of grand-jury forepersons, was not apparent from the record on direct appeal, and thus we do not fault appellate counsel for failing to raise that claim. Moreover, the case upon which Fautenberry relies to challenge Ohio’s grand-jury foreperson selection process, Campbell v. Louisiana, 523 U.S. 392, 398, 118 S.Ct. 1419, 140 L.Ed.2d 551 (1998), was not decided until 1998, long after Fautenber-r/s direct appeal had concluded. Finally, the fourth omitted claim—that Fau-tenberry was prejudiced by the wearing of shackles before the trial court—would not have had any merit on direct appeal because under Ohio law the appellate court “presum[es] that in a bench trial in a criminal case[,] the court consid-eres] only the relevant, material, and competent evidence in arriving at its judgment unless it affirmatively appears *643to the contrary.” Post, 513 N.E.2d at 759.
III.
After carefully reviewing the record and evaluating Fautenberry’s arguments on appeal, we conclude that the district court properly denied Fautenberry’s petition for habeas relief. Accordingly, we AFFIRM the judgment of the district court.
. Under Ohio law, a capital defendant who waives his right to trial by jury and elects to be tried by the court, is actually tried by a three-judge panel. See Ohio Rev.Code § 2945.06. Furthermore, even on a no-contest plea, the prosecution must produce evidence to prove aggravated murder with the specified aggravating circumstances. See Ohio R.Crim. Pro. 11(c)(3).
. The State disputes whether Fautenberry actually suffers from a functional brain impairment. Because it is not dispositive to our analysis of this ineffective-assistance claim, we will assume without deciding that Fauten-berry does in fact have some sort of organic brain damage.
. We have assumed for the sake of argument that Fautenberry does have some sort of organic brain damage. See n. 2, supra.
. At one point in his brief, Fautenberry implies that the state court’s denial of his ineffective-assistance claim was contrary to the Supreme Court’s decision in Rompilla v. Beard, 545 U.S. 374, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005). This argument is wholly meritless. Federal law for purposes of AED-PA is defined as "the holding, as opposed to the dicta, of [the Supreme] Court’s decisions as of the time of the relevant state-court decisions.” Williams, 529 U.S. at 412, 120 S.Ct. 1495. First, Rompilla was decided in 2005, long after Fautenberry had concluded his proceedings in state court. Second, the holding in Rompilla is inapposite here. Contrary to Fautenberry’s argument, the Court in Rompil-la did not hold that counsel, even though faced with his client’s uncooperative and obstructive behavior, rendered deficient performance by failing to conduct additional mitigation investigation. The court made no such statement, holding instead that "the lawyers were deficient in failing to examine the court file on [the petitioner's] prior conviction.” Rompilla, 545 U.S. at 383, 125 S.Ct. 2456.
. Fautenberry's Brady claim with regard to most of this evidence fails for an additional reason: The Brady rule applies only to evidence that was "known to the prosecution but unknown to the defense.” United States v. Agurs, 427 U.S. 97, 103, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976) (emphasis added); see also United States v. Mullins, 22 F.3d 1365, 1371 (6th Cir.1994). Regardless of his guilt in the commission of the murders, Fautenber-ry knew the details about his interaction with Agent Ott and his own state of mind (i.e., his depression and suicidal tendencies). Furthermore, by pleading no contest, Fautenberry admitted the facts charged in the indictment, so he also knew the location where he shot Joseph Daron and any pertinent sexual aspects of the murders. With the lone exception of the evidence pertaining to Joseph Dar-oris relationships, Fautenberry or his counsel had knowledge of all this "undisclosed evidence.”
. Fautenberry argues that because most of the evidence supporting the allegations in Sub-part A of this ineffective-assistance claim was the alleged Brady evidence, the prosecution’s failure to provide him with this evidence constitutes the cause and prejudice to excuse his procedural default. Because, as already noted, Fautenberry’s Brady claim lacks merit, that claim cannot constitute cause and prejudice to excuse procedural default.
. The quoted passage was in effect at the time of Fautenberiy’s trial and sentencing in 1992. The current statute, amended as of 1995, is identical.
. As we noted in Brewer, the Ohio Supreme Court's finding that there was no indication that the three-judge panel relied on the victim-impact evidence was a “factual finding that is presumed to be correct under the AEDPA.” See Brewer, 47 Fed.Appx. at 288. Fautenberry has failed to challenge this factual finding on appeal, much less to establish *639that it was clearly erroneous, and our review of the record persuades us that the finding is correct. See Cooey v. Coyle, 289 F.3d 882, 910-11 (6th Cir.2002) (affirming the district court's giving deference to the Ohio Supreme Court’s factual finding that there was "no affirmative indication that the victim impact statements were considered in sentencing [the petitioner] to death”).