During the rebuttal portion of her closing argument, the prosecutor told Grant's jury, "[T]he law says ... that before something can be mitigating it must reduce the moral culpability or blame of the defendant." R. Vol. 4, Trial Tr. 8, at 75 (emphasis added). And to ensure that no reasonable juror would have cause to doubt her, the prosecutor reinforced that this wasn't just what she was saying or what the court was saying, but what "the law sa[id]." Id. (emphasis added). I would find that no reasonable juror would have doubted her. Nor would a reasonable juror have doubted the prosecutor on the three other separate occasions when she told the jurors that the law prohibited them from considering evidence as mitigating unless it reduced Grant's culpability or blame.

The majority doesn't dispute that the prosecutor misstated the law on mitigating evidence. Nor does it dispute that she did so repeatedly. Instead the majority questions whether the jury believed those repeated misstatements. Yet I see no reason to think it wouldn't have. The prosecutor's numerous misstatements found explicit support in the jury instructions.1 Further, the trial court implicitly endorsed the prosecutor's misstatements by overruling the defense's objection to them. It's therefore clear that the jury felt legally precluded from considering all of Grant's mitigation evidence. Grant's death sentence therefore violates the Eighth and Fourteenth amendments, and I would reverse the district court's order denying Grant's habeas petition.

Analysis

The majority all but concedes that the prosecutor's comments impermissibly narrowed the scope of evidence that the jury could treat as mitigating. But it nevertheless *961affirms Grant's death sentence because it opines that the OCCA reasonably concluded that the prosecutor's improper comments didn't actually mislead the jury. In doing so, the majority errs in two respects. First, the majority overlooks the fact that the OCCA misunderstood Grant's argument on direct appeal and therefore didn't actually adjudicate this claim on the merits. Thus, AEDPA's deferential standard of review doesn't apply. See Chadwick v. Janecka , 312 F.3d 597, 606 (3d Cir. 2002) (explaining that "if an examination of the opinions of the state courts shows that they misunderstood the nature of a properly exhausted claim and thus failed to adjudicate that claim on the merits, the deferential standards of review in AEDPA do not apply"); 28 U.S.C. § 2254(d) (requiring us to defer to state-court decision only when state court resolved particular claim "on the merits"). Second, even if the OCCA had resolved this claim on the merits, it would have been unreasonable for the OCCA to conclude that there was no reasonable likelihood the prosecutor's comments misled at least one juror.

The majority "acknowledge[s] that a plausible argument could be made here that the prosecutor's rebuttal arguments were 'improper.' " Maj. Op. 938 (quoting Harris v. State , 164 P.3d 1103, 1114 (Okla. Crim. App. 2007) ). But because the majority doesn't "definitively opine on the matter," id. , it has no reason to discuss just how "improper" those statements actually were, id. (quoting Harris , 164 P.3d at 1114 ). I take a different approach. First, I explain how the prosecutor's misstatements undermined the Constitution's promise that the death penalty will not be imposed without the most circumspect deliberation. Then, with that conclusion in mind, I explain the two overarching deficiencies in the majority's analysis: its deference to the OCCA and its conclusion that the OCCA reasonably resolved this issue in the state's favor.

I. Flaws in Grant's Sentencing Proceeding

"[I]n capital cases the fundamental respect for humanity underlying the Eighth Amendment requires consideration of the character and record of the individual offender ... as a constitutionally indispensable part of the process of inflicting the penalty of death." Woodson v. North Carolina , 428 U.S. 280, 304, 96 S.Ct. 2978, 49 L.Ed.2d 944 (1976) (plurality opinion) (citation omitted). Thus, a jury must "not be precluded from considering, as a mitigating factor , any aspect of a defendant's character or record and any of the circumstances of the offense that the defendant proffers as a basis for a sentence less than death." Lockett v. Ohio , 438 U.S. 586, 604, 98 S.Ct. 2954, 57 L.Ed.2d 973 (1978) (plurality opinion) (footnote omitted); see also Eddings v. Oklahoma , 455 U.S. 104, 110, 102 S.Ct. 869, 71 L.Ed.2d 1 (1982) (adopting Lockett plurality's rule). Critically, this is true even if the evidence that "the defendant proffers," Lockett , 438 U.S. at 604, 98 S.Ct. 2954, doesn't provide a legal excuse for the defendant's crime, see Eddings , 455 U.S. at 112-14, 117, 102 S.Ct. 869 (reversing death sentence where sentencing judge stated that he was legally precluded from considering defendant's violent background; explaining that although sentencer "may determine the weight to be given relevant mitigating evidence," it "may not give it no weight by excluding such evidence from [its] consideration" (emphasis added)).

Clearly established Supreme Court precedent therefore requires that Grant's jury felt free to at least consider all the mitigating evidence that Grant presented. Of course the jury could have properly decided to give that evidence little weight.

*962See Eddings , 455 U.S. at 114-15, 102 S.Ct. 869. But Grant's death sentence cannot stand if there's a "reasonable likelihood that the jury would have found itself foreclosed from [even] considering" his mitigating evidence. Johnson v. Texas , 509 U.S. 350, 367-68, 113 S.Ct. 2658, 125 L.Ed.2d 290 (1993). And although this standard requires more than the mere "possibility of such an inhibition," it doesn't require Grant to show "that the jury was more likely than not to have been impermissibly inhibited" from considering all of Grant's mitigating evidence. Boyde v. California , 494 U.S. 370, 380, 110 S.Ct. 1190, 108 L.Ed.2d 316 (1990).

Finally, in determining whether the jury felt free to consider all of Grant's mitigating evidence, we review the totality of the jury instructions and closing arguments. See id. at 383-86, 110 S.Ct. 1190 (considering challenged instruction in context of closing arguments and other jury instructions); Penry v. Lynaugh (Penry I) , 492 U.S. 302, 325-26, 109 S.Ct. 2934, 106 L.Ed.2d 256 (1989) (considering challenged instruction in context of closing arguments), abrogated on other grounds by Atkins v. Virginia , 536 U.S. 304, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002) ; Hanson v. Sherrod , 797 F.3d 810, 852 (10th Cir. 2015) (considering challenged instruction in context of other instructions and closing arguments).

A. Instruction 12

Like the majority, I start with the instruction that defined mitigating circumstances for the jury. Instruction 12 stated, in relevant part, "Mitigating circumstances are those which, in fairness, sympathy, and mercy, may extenuate or reduce the degree of moral culpability or blame. The determination of what circumstances are mitigating is for you to resolve under the facts and circumstances of this case." O.R. 2349.

Instruction 12 is ambiguous at best. The word "may" might arguably broaden the instruction. Id. But at least some jurors very likely read Instruction 12 to prevent the jury from "consider[ing], as a matter of law ," Eddings , 455 U.S. at 114, 102 S.Ct. 869, "aspects of [Grant's] character and record" that didn't reduce his moral culpability or blame, Lockett , 438 U.S. at 605, 98 S.Ct. 2954. This definition of mitigating circumstances was impermissibly narrow. See Lockett , 438 U.S. at 604, 98 S.Ct. 2954 (holding that jury must be free to consider "any aspect of a defendant's character or record and any of the circumstances of the offense that the defendant proffers as a basis for a sentence less than death" (emphasis added)).

Indeed, even the OCCA has recognized that Instruction 12 can be problematic when the prosecution takes certain liberties in its closing arguments. See Harris v. State , 164 P.3d 1103, 1113-14 (Okla. Crim. App. 2007). In fact, after Grant's 2005 trial, the OCCA recommended that the Oklahoma Uniform Jury Instruction Committee amend Instruction 12 because prosecutors had been using the instruction to "egregious[ly] misstate[ ] ... the law on mitigating evidence" by "argu[ing] that evidence of a defendant's history, characteristics or propensities should not be considered as mitigating simply because it does not go to his moral culpability or extenuate his guilt." Id. at 1113-14. But the OCCA refrained from declaring that Instruction 12 was unconstitutional in the absence of such improper arguments.2 See id. at 1113.

*963B. The Prosecution's Improper Assertions

Our starting point is thus an instruction that-at minimum-flirts with the impermissible. But as Penry I , Boyde , and Hanson instruct us, we must consider the totality of the jury instructions and closing arguments. Doing so here only reinforces the unavoidable conclusion that the jury likely believed it couldn't consider Grant's evidence as mitigating. That's because, during closing arguments, the prosecution here did exactly what "troubled" the OCCA in Harris : the prosecution used Instruction 12 to "argue that evidence of [Grant's] history, characteristics [and] propensities should not be considered as mitigating simply because" that evidence didn't "go to his moral culpability or extenuate his guilt." 164 P.3d at 1114.

Grant produced two categories of mitigating evidence during the sentencing phase. First, he proffered significant evidence showing that he was schizophrenic. Second, he offered testimony from family members establishing that he experienced an abhorrent childhood that was riddled with parental drug abuse, extreme poverty, and violence. In its closing argument-in-chief during Grant's penalty phase, the prosecution focused on attacking the sufficiency of the evidence that Grant offered. In other words, the prosecution tried to convince the jury that Grant wasn't actually schizophrenic and that his childhood wasn't that bad. Grant doesn't dispute that this was appropriate; nor do I. The jury isn't required to believe all the testimony it hears, and the prosecution is free to implore it not to.

Then, during Grant's closing arguments, his attorneys asked for the jury's sympathy. They were frank about the fact that none of the mitigating evidence Grant presented would actually reduce his moral culpability or guilt. As one of Grant's attorneys told the jury, "I want to be clear that none of the mitigating evidence that we ... will ask you to consider excuses ... what happened in this case. And we *964understand that and we don't ask you to excuse what happened in this case. It's an explanation." R., vol. 4, Trial Tr. 8, at 44. Both of Grant's attorneys repeated this theme throughout their closing arguments. For example, one attorney explained, "We didn't get up in [the guilt] stage and try to tell you ... Grant is not guilty because he is schizophrenic, but it is an explanation. It's a reason that he ended up where he ended up." Id. at 53. Then, after reviewing the evidence of Grant's terrible upbringing, Grant's attorney again explained, "I'm not t[r]ying to say that it's okay that [Grant] committed a crime because he had tough environments, I'm just trying to explain to you how ... Grant got to this point in his life." Id. at 58.

Then Grant's other attorney took over. She began discussing Grant's life and then told the jury,

I'm going to stop in the story here and talk about that idea of fault for a minute because what we're talking about doesn't have to do with fault .... So don't let [the prosecution] tell you or don't presume on your own that we're telling you that ... Grant's life is an excuse for what he did. It is not an excuse for what he did. We're not saying that. It never will be an excuse. What it is is appropriate information to consider in determining punishment.

Id. at 64-65 (emphasis added). The defense's theme is clear and reasonable. Grant's attorneys recognized that he committed a horrible crime. And they recognized that neither his mental illness nor his childhood reduced his culpability or his blame. But these are undisputedly relevant mitigating factors, so the defense urged the jury to consider them and show Grant mercy.

Given the defense's theme, I question whether the jury would have read Instruction 12-even in isolation-as "provid[ing] a vehicle for the jury to give mitigating effect to" Grant's evidence. Penry I , 492 U.S. at 324, 109 S.Ct. 2934. That's because Instruction 12 defined mitigating circumstances as "those which, in fairness, sympathy, and mercy, may extenuate or reduce the degree of moral culpability or blame." O.R. 2349. And in light of the defense's repeated concessions that Grant's mental illness and childhood did neither, it's difficult to see how the jury could have read Instruction 12 as allowing it to consider this evidence as mitigating.

But even if some jurors might have read Instruction 12 in isolation to allow the jury to consider evidence that didn't reduce Grant's culpability or blame, the prosecution's rebuttal arguments made it nearly certain that the jurors would conclude they couldn't do so.

For instance, shortly into her rebuttal, the prosecutor picked up where Grant's attorneys left off and told the jury the following:

You know, what I noted about the argument of defense counsel is they spent a long time trying to describe to you how what they have offered in mitigation is not an excuse, how what they're trying to tell you about the life and times of ... Grant is not an excuse for the behavior that he committed at the La Quinta Inn in July of 2001. And that is exactly what the law says . Because one thing that I noticed that they did not talk about in their closing argument is what the definition of a mitigating circumstance is. Because the law tells you what that means . It tells you that in order-first of all, you have two choices: Do you believe that the mitigating circumstance has been proven because you don't have to believe everything that you have heard from the witnesses who testified. You can choose what parts you want to believe and disregard those *965parts that you believe perhaps people were exaggerating about or somehow t[r]ying to make things sound a whole lot worse than they actually were. But let's assume for the sake of argument that everything that you were told was correct, that not any person made up or fudged a little bit in what they were telling you about. What does it say mitigating circumstances are? What does that mean when we say that something may mitigate the murder of these two women, the lives that he took? It says that mitigating circumstances are those which reduce the moral culpability or blame of the defendant . That those things, in order to be mitigating, must reduce his moral culpability or blame.

R. Vol. 4, Trial Tr. 8, at 73-74 (emphases added).

At this point, the defense objected, the parties conferred at the bench, and the trial court overruled the objection. Then the prosecutor continued:

It's not Sandra Elliott [the prosecutor] telling you that this will make something mitigating, that's what the law says . And we all talked about during voir dire that we would be discussing the law that the [c]ourt's going to be giving you. And the law says, not Sandra Elliott, not what the defense attorneys say, but what the [c]ourt tells you and what the law says is that before something can be mitigating it must reduce the moral culpability or blame of the defendant.

Id. at 75 (emphases added).

Finally, after discussing the specifics of the evidence Grant presented, the prosecutor repeated, "So while [the defense] may say to you that [they are] not offering this as an excuse for ... Grant's behavior, you have to look at whether or not it reduces his moral culpability or blame. That is what the law says that you must do ." Id. at 79 (emphasis added).

The prosecutor thus made clear that the jurors could fully exercise their discretion to decide whether, as a factual matter, Grant's evidence proved the circumstances he asserted as mitigating. But she told them that they couldn't stop there. Instead, they then had to decide whether the factors that Grant proved reduced his culpability or blame. If not, the prosecutor explained, the law prohibited the jurors from considering those factors as mitigating.

Using mandatory language like "must," the prosecutor's comments improperly conveyed to the jury that each factor Grant asserted as mitigating had to impact moral culpability or blame to be a mitigating factor. Id. at 75; accord id. at 79. And even more problematically, she added clear qualifying language indicating that this was a prerequisite to ultimately considering a factor as mitigating at all. See id. ("[B ]efore something can be mitigating it must reduce the moral culpability or blame of the defendant." (emphasis added)). With these comments, the prosecutor explicitly tethered the legal definition of "mitigating circumstance" to a circumstance that reduced Grant's moral culpability or blame. And she repeatedly and explicitly made clear that this wasn't just her own guidance to the jury; it was what "the law" compelled. Id. at 75; accord id. at 73; id. at 79.

Put differently, my objection is that the prosecution used an ambiguous jury instruction to inject an additional, unconstitutional step into the mitigation analysis. Normally, the jury should undertake a two-step process. First, it should consider whether the evidence actually supports the defendant's asserted mitigating factor. Cf. Eddings , 455 U.S. at 113, 102 S.Ct. 869 (explaining that sentencer may "evaluate the evidence in mitigation and find it wanting as a matter of fact"). Next, it *966should consider what weight to afford that mitigating factor when balanced against the aggravating factors. Cf. id. at 114-15, 102 S.Ct. 869 ("The sentencer ... may determine the weight to be given relevant mitigating evidence ...."). But the prosecution's comments here suggest an intermediate step: if the jury determines that the evidence supports the asserted mitigating factor, then it must also determine whether that factor is actually mitigating-that is, whether it "reduce[s] the moral culpability or blame of the defendant." R. vol. 4, Trial Tr. 8, at 75. If so, then-and only then-can the jury reach the final step of the inquiry and determine what weight to afford the mitigating factor.

The result is the precise scheme that the Court rejected in Eddings . In that case, the sentencing judge concluded that " 'in following the law,' he could not 'consider the fact of [the defendant]'s violent background.' " Eddings , 455 U.S. at 112-13, 102 S.Ct. 869 (citation omitted). The judge's error, the Supreme Court later explained, was not that he "evaluate[d] the evidence in mitigation and [found] it wanting as a matter of fact; rather he found that as a matter of law he was unable even to consider the evidence." Id. at 113, 102 S.Ct. 869. This is precisely what the prosecution told the jury it must do here: if the jury did not believe the evidence reduced Grant's culpability or blame, then the "the law sa[id]" the jury could not find the evidence to be mitigating. R. Vol. 4, Trial Tr. 8, at 75.

The majority disagrees. In doing so, it relies heavily on comparisons to Hanson , 797 F.3d 810. But the error here is materially-albeit subtly-different from the error asserted in that case.

We've explained that the prosecution may "comment[ ] on the weight that should be accorded to the mitigating factors" but may not "suggest that the jury [is] not permitted to consider the factors." Fox v. Ward , 200 F.3d 1286, 1300 (10th Cir. 2000) ; see also Eddings , 455 U.S. at 114-15, 102 S.Ct. 869 ("The sentencer, and the Court of Criminal Appeals on review, may determine the weight to be given relevant mitigating evidence. But they may not give it no weight by excluding such evidence from their consideration."). In Hanson , the prosecution did the former. It encouraged "the jury to consider whether any of the mitigating circumstances 'really extenuate[d] or reduce[d] [the defendant's] degree of culpability or blame in this case.' " 797 F.3d at 851 (citation omitted). And it said, "Consider all the mitigating circumstances and you'll consider what weight [to give them]." Id. at 852 (alteration in original) (citation omitted). But it doesn't appear that the prosecution in Hanson suggested-as the prosecutor did here-that, "as a matter of law ," the jury couldn't conclude the evidence was mitigating if it didn't reduce the defendant's culpability or blame. Eddings , 455 U.S. at 113, 102 S.Ct. 869. Thus, while the prosecution in Hanson permissibly "commented on the weight" that the jury should accord the mitigating factors, the prosecution here impermissibly asserted "that the jury was not permitted to consider [Grant's] factors." Fox , 200 F.3d at 1300.

II. Grounds for Habeas Relief

It's thus abundantly clear that the prosecutor's comments here were "an egregious misstatement of the law on mitigating evidence." Harris , 164 P.3d at 1114. And the majority doesn't seriously dispute this. See Maj. Op. 938 ("To be sure, we acknowledge that a plausible argument could be made here that the prosecutor's rebuttal arguments were 'improper'-that is, one could plausibly contend that those remarks resemble the kind of 'misuse' of *967the moral-culpability text that concerned the OCCA in Harris ." (quoting Harris , 164 P.3d at 1114 )). Rather, the majority questions the effect of the prosecution's "improper" misstatements on the jury. Id. (quoting Harris , 164 P.3d at 1114 ). And it concludes that the OCCA could have reasonably decided that there is no reasonable likelihood the jury would have credited those misstatements.

The majority errs in two respects. First, the OCCA didn't address this argument "on the merits," 28 U.S.C. § 2254(d), so AEDPA's deferential standard of review doesn't apply. And second, it beggars belief to conclude that there's no reasonable likelihood the jury followed the prosecution's commands. So even if the OCCA reached this conclusion, it did so unreasonably.

A. Proper Standard of Review

1. Grant's Failure to Argue for De Novo Review

Preliminarily, the majority concludes that we must defer to the OCCA because Grant doesn't argue for de novo review. But "the correct standard of review under AEDPA is not waivable." Gardner v. Galetka , 568 F.3d 862, 879 (10th Cir. 2009). And despite the majority's assertion to the contrary, we made clear in Gardner that this principle applies to both habeas petitioners and respondents: our discussion in that case broadly (and repeatedly) referred to the "parties." Id. ("Other courts of appeal ... all have concluded that the standard of review under AEDPA cannot be waived by the parties ." (emphasis added)); accord id. ("It is one thing to allow parties to forfeit claims, defenses, or lines of argument; it would be quite another to allow parties to stipulate or bind us to application of an incorrect legal standard, contrary to the congressional purpose." (emphases added)). Moreover, as support for our decision, we relied on a Sixth Circuit case specifically holding that a habeas petitioner can't waive de novo review of an exhausted claim that the state court didn't adjudicate on the merits. See id. (citing Brown v. Smith , 551 F.3d 424, 428 n.2 (6th Cir. 2008) ); Brown , 551 F.3d at 428-30, 428 n.2.

Despite Gardner 's broad language, the majority suggests it doesn't apply to the specific facts before us here. According to the majority, congressional interests in federalism, comity, and finality require us to defer to the state court regardless of whether the state argues we should. But those same concerns, the majority indicates, forbid us from reviewing the state court de novo unless the petitioner argues we should.

For several reasons, I find the majority's attempts to distinguish Gardner unavailing. First, Gardner simply applies to the habeas context the more general rule "that 'the court, not the parties, must determine the standard of review, and therefore, it cannot be waived.' " United States v. Fonseca , 744 F.3d 674, 682 (10th Cir. 2014) (quoting Worth v. Tyer , 276 F.3d 249, 262 n.4 (7th Cir. 2001) ). Thus, although federalism, comity, and finality might provide additional reasons why we shouldn't let the state waive deferential review under AEDPA, Gardner doesn't in any way suggest we would depart from Fonseca 's more general rule in the habeas context absent these additional policy considerations.

Second, the majority doesn't explain how we would offend principles of federalism or comity by reviewing de novo an argument that the OCCA didn't address on the merits, even if the petitioner hasn't asked us to apply that less deferential standard of review. Federalism and comity require us-within reason-to respect how a state court chooses to resolve an issue. See *968§ 2254(d) ; Woodford v. Garceau , 538 U.S. 202, 206, 123 S.Ct. 1398, 155 L.Ed.2d 363 (2003). But this concern simply isn't implicated when the state court doesn't resolve that issue on the merits. Thus, there's no reason federalism or comity should reenter the equation just because a petitioner fails to argue for de novo review.

Third, as much as AEDPA seeks to protect federalism and comity, it also seeks to ensure that defendants aren't convicted or sentenced in violation of federal law. See Murdoch v. Castro , 609 F.3d 983, 1001 (9th Cir. 2010) (Kozinski, J., dissenting) (explaining that deferring to state-court decision that doesn't resolve particular federal claim "upsets the 'delicate balance' struck by AEDPA between vindicating the rights of criminal defendants and upholding the authority of state courts as the primary forum for adjudicating these rights" (quoting Williams v. Taylor , 529 U.S. 420, 436, 120 S.Ct. 1479, 146 L.Ed.2d 435 (2000) )). Thus, although AEDPA cautions us to respect a state court's decision, it also requires that at least one court-state or federal-hears the defendant's exhausted claims. See id. (noting that although state court has "primary responsibility" to adjudicate federal habeas claims, federal court must adjudicate those claims on merits if state court forgoes this responsibility). In short, if federalism and comity require us to defer to a state court's decision regardless of whether the state asks us to, then AEDPA's interest in the protection of federal rights likewise requires us to grant de novo review of claims not addressed on the merits regardless of whether the petitioner asks us to.

Finally, the case that the majority relies on- Eizember v. Trammell , 803 F.3d 1129 (10th Cir. 2015) -doesn't establish otherwise. In Eizember , we concluded that a habeas petitioner waived an argument that he could overcome AEDPA deference; we did not, as the majority asserts here, "rebuff[ ]" an "effort by the dissent" in that case to establish that AEDPA deference didn't apply at all. Maj. Op. 931 n.20; see also Eizember , 803 F.3d at 1140-41. That is, the dissent in Eizember never suggested, as I do here, that the state court failed to adjudicate the petitioner's claim on the merits, thus triggering de novo review. See Eizember , 803 F.3d at 1160-61 (Briscoe, J., concurring in part and dissenting in part). And this distinction explains why we didn't acknowledge Gardner in Eizember , let alone suggest that we were overruling it or our more general rule that parties can't waive the appropriate standard of review. See id. at 1140-41 (majority opinion). Nor would we have had any reason to do so. Gardner explicitly left open the path the majority followed in Eizember : it distinguished between "lines of [substantive] argument" (which are subject to traditional principles of waiver and forfeiture) and AEDPA's standard of review (which is not). Gardner , 568 F.3d at 879. Eizember concerned only the former. In reaching a different conclusion about the latter, the majority here charts new ground.

2. The OCCA's Misunderstanding of Grant's Argument

In arguing to the OCCA on direct appeal that the state improperly precluded the jury from considering his mitigating evidence, Grant framed his argument more or less as I do above. The state's key error, he explained, was that it "argued to the jury not to even consider his proffered evidence as mitigating under the law given to them." Aplt. Br. at 79, Grant v. State , 205 P.3d 1 (Okla. Crim. App. 2009). Thus, he posited, "[t]hough none of [Grant's] ten mitigating circumstances excused his crime or reduced his moral or legal culpability or blame, they may have given a reasonable juror a reason to exercise sympathy and mercy upon ... Grant and *969spare his life." Id. But, "in light of the prosecutor's persistent emphasis upon what 'the law sa[id],' [Grant's] jury could not ... give effect to ... Grant's mitigating [factors]." Id. (citation omitted).

The OCCA misunderstood this argument as merely "claim[ing] the prosecutor misstated the law by telling the jurors that the evidence [Grant] had presented as 'mitigating' did nothing to justify a sentence less than death." Grant , 205 P.3d at 21. It then rejected this argument, explaining, "While there is no restriction whatsoever on what information might be considered mitigating, no juror is bound to accept it as such, and the State is free to try to persuade the jury to that end." Id.

The majority accepts this faulty characterization of Grant's argument. See Maj. Op. 930-31-. But a look to Grant's briefing on direct appeal reveals that Grant never asserted that the state erred by arguing the mitigating evidence did nothing to justify a sentence less than death. Instead, Grant quite clearly argued the state improperly told the jury that unless his evidence reduced his culpability or blame, then the evidence, as a matter of law, wasn't mitigating even if the jury thought it might justify a sentence less than death.3

I won't build on this strawman. And I certainly won't defer to it. "[I]f an examination of the opinions of the state courts shows that they misunderstood the nature of a properly exhausted claim and thus failed to adjudicate that claim on the merits, the deferential standards of review in AEDPA do not apply." Chadwick , 312 F.3d at 606 ; see also Campbell v. Bradshaw , 674 F.3d 578, 596 (6th Cir. 2012) (reviewing habeas claim de novo because "[i]t appear[ed] that, when considering the trial court's ruling, the Ohio Supreme Court misconstrued [the defendant's] argument"); DeBerry v. Portuondo , 403 F.3d 57, 71-72 (2d Cir. 2005) (explaining that "habeas claims may be reviewed without AEDPA deference" when "a properly preserved claim, recognized as such, was misconstrued by the state court and hence not decided 'on the merits' " (quoting *970§ 2254(d) ); Davis v. Sec'y for Dep't of Corr. , 341 F.3d 1310, 1313 (11th Cir. 2003) (explaining that "controversy [fell] outside of § 2254(d)(1)'s [deference] requirement" because state court misconstrued defendant's claim and thus "failed to resolve [it]" on the merits); Henderson v. Cockrell , 333 F.3d 592, 600-601 (5th Cir. 2003) (holding that "AEDPA's standards of review [were] inapplicable" because state court misconstrued defendant's claim).

Our inquiry should thus begin and end by asking "whether there is a reasonable likelihood that the jury ... applied [Instruction 12] in a way that prevent[ed] the consideration of constitutionally relevant evidence." Boyde , 494 U.S. at 380, 110 S.Ct. 1190. In light of the prosecution's insistence that the jury not consider Grant's mitigating evidence, the answer to this question is an easy "yes." Thus, applying de novo review, I would reverse the district court's order denying Grant's habeas petition.

B. The Other Instructions and Comments

Alternatively, I would reverse under § 2254(d)'s deferential standard of review because the OCCA would have been unreasonable to reject this claim on the merits, had it hypothetically done so. The majority says that portions of the other jury instructions and the prosecution's closing argument-in-chief could have reasonably led the OCCA to "conclude[e] that there was no reasonable likelihood that the jury was precluded by the prosecution's closing arguments from considering all of ... Grant's mitigation evidence-including the evidence that did not extenuate or reduce his moral culpability or blame." Maj. Op. 945. I agree that the other jury instructions and the remainder of the prosecution's closing argument are relevant considerations. See Hanson , 797 F.3d at 851-52. But I can't agree that either has ameliorative power here.

Before I address the majority's specific arguments, I pause to clarify the standard that the majority applies when it defers to the OCCA. Following § 2254(d)(1), the majority asks whether the OCCA unreasonably applied clearly established Supreme Court law-i.e., Lockett and its progeny. And recall that the Boyde inquiry is "whether there is a reasonable likelihood that the jury ... applied the challenged instruction in a way that prevent[ed] the consideration of constitutionally relevant evidence." 494 U.S. at 380, 110 S.Ct. 1190. Combining these standards, we ask whether the OCCA reasonably concluded that there is no "reasonable likelihood that the jury ... applied the challenged instruction in a way that prevents the consideration of constitutionally relevant evidence." Id.

It's important not to lose sight of what this inquiry involves. We don't ask whether the OCCA could have reasonably concluded that the jury might have known it could consider all of Grant's mitigating evidence. I agree that the OCCA could have reasonably reached such a conclusion. But I cannot agree that the OCCA could have reasonably concluded that it's unlikely that even a single juror felt precluded from considering Grant's evidence. And this is the applicable standard. See id.

1. Suggestions that the Jury Could Decide Whether the Evidence was Mitigating

The majority makes much of the fact that both the jury instructions and the prosecution emphasized that the jury could ultimately decide if Grant's evidence was actually mitigating. But this wouldn't have corrected a juror's misunderstanding of what "mitigating" means. That's because the jury received an erroneous definition of "mitigating." Thus, simply reminding *971the jury that it could decide whether the evidence met that definition didn't cure the error.

Accordingly, the portion of Instruction 12 telling the jury that "[t]he determination of what circumstances [were] mitigating [was] for [the jury] to resolve under the facts and circumstances in this case," gave Grant no relief from the prosecution's erroneous definition of "mitigating." Nor did the portion of Instruction 13 that told the jury that it could "decide that other mitigating circumstances exist[ed]." O.R. 2351. And this is also true for the prosecution's comments that whether a fact is mitigating was "for [the jury] to consider," R. vol. 4, Trial Tr. 8, at 31, and that it was "up to [the jury] to determine whether or not these mitigators-whether or not these circumstances somehow mitigate[d] what ... Grant did," id. at 32.

The majority is quick to point out that in Hanson we concluded that the same instructions and similar statements during closing argument alleviated the asserted error. But as I explain above, this case isn't Hanson . The prosecution in Hanson didn't give the jury an erroneous definition of "mitigating." See 797 F.3d at 851-52. The prosecution in that case certainly urged the jury not to give the defendant's evidence any mitigating weight unless it reduced the defendant's culpability or blame. See id. at 851. But-unlike the prosecutor in this case-the prosecution in Hanson didn't tell the jury that the law precluded it from doing so. Thus, in the absence of such of a misstatement from the prosecutor, the OCCA could have reasonably concluded in Hanson that the jury read the phrase "[t]he determination of what circumstances are mitigating is for you to resolve," O.R. 2349, to mean that the jury had discretion to consider anything to be mitigating. But these instructions have no similar effect here, where the prosecutor instead urged the jurors not to consider Grant's evidence as mitigating at all. Cf. Eddings , 455 U.S. at 115, 102 S.Ct. 869 (explaining that capital sentencer may not "give [defendant's mitigating evidence] no weight by excluding such evidence from [its] consideration").

2. Instruction 13's List of Mitigating Evidence

I do recognize that other parts of Instruction 13 may have led some jurors to correctly infer that they could consider all of Grant's mitigating evidence. For instance, Instruction 13 stated, "Evidence has been introduced as to the following mitigating circumstances," and then provided a summary of the various evidence that Grant introduced. O.R. 2350. Certainly a juror might have inferred from this list that the jury could legally consider these factors as mitigating even if it concluded that the factors didn't reduce Grant's moral culpability or blame. But Boyde doesn't require us to be certain that the jury felt it couldn't consider a relevant mitigating factor. See 494 U.S. at 380, 110 S.Ct. 1190. Indeed, it doesn't even require the defendant to "establish that the jury was more likely than not to have been impermissibly inhibited." Id. It merely requires a "reasonable likelihood." Id.

Even considering Instruction 13, Grant easily meets this burden. Instruction 13 might have led some jurors to disregard Instruction 12 and the prosecutor's statements and apply the proper law. But other jurors might have simply (1) read Instruction 13 as summarizing the evidence that Grant presented and (2) concluded that the jury nevertheless had to decide whether these factors reduced Grant's culpability or blame. At most, Instruction 13 gives skeptical jurors grounds to infer that the prosecutor misstated the law. But nothing about Instruction 13 makes it unlikely that *972at least one juror would have believed the prosecutor's misstatements.

Further, to the extent that a juror could read Instruction 13 to say that the jury could consider all of Grant's mitigating evidence, regardless of whether it reduced his moral culpability or blame, Instruction 13 conflicts with Instruction 12, which suggests the opposite. And this conflict "inserted 'an element of capriciousness' into the sentencing decision, 'making the jurors' power to avoid the death penalty dependent on their willingness' to elevate [Instruction 13] over [Instruction 12]." Penry v. Johnson(Penry II) , 532 U.S. 782, 800, 121 S.Ct. 1910, 150 L.Ed.2d 9 (2001) (quoting Roberts v. Louisiana , 428 U.S. 325, 335, 96 S.Ct. 3001, 49 L.Ed.2d 974 (1976) (plurality opinion)).

3. The Lack of Objection to Grant's Evidence

Finally, the majority speculates that the jury might have attached some significance to the fact that Grant was allowed to present his mitigating evidence at all. Essentially, the majority surmises, the jury might have expected the prosecution to object to Grant's evidence if the evidence was indeed legally irrelevant to the jury's inquiry-and for the trial court to sustain that objection. Ergo, these perceptive jurors would have concluded that the evidence was not legally irrelevant.

I simply can't subscribe to the speculation that all 12 members of a layperson jury engaged in such spontaneous consideration of the rules regarding the admissibility of irrelevant evidence. Frankly, I'm not even sure that 12 lawyers would make this inference. But even if the jury contemplated such an approach, it doesn't necessarily follow that just because the evidence was admissible, the jury would consider it as mitigating regardless of whether it thought the evidence reduced Grant's culpability or blame. Some jurors might have believed that they had leeway to determine that Grant's evidence did reduce his moral culpability or blame (for example, if they concluded that Grant committed the crimes during a schizophrenic delusion). But then, if these jurors decided that the evidence didn't reduce Grant's moral culpability or blame, they wouldn't have considered whether that evidence nevertheless justified sentencing Grant to a sentence less than death, as Lockett and its progeny demand they be allowed to do. See 438 U.S. at 604, 98 S.Ct. 2954. In any event, we need not say for certain what the jury might have thought as long as there's a reasonable likelihood a juror could have been misled. There undoubtedly is here.

C. The Jury's Natural Impression

The dispute here can be boiled down to one question: Is it reasonably likely that at least one juror believed the prosecutor when she purported to tell the jury what "the law sa[id]?" R. vol. 4, Trial Tr. 8, at 75. The only reasonable answer is "yes." We've warned in the past that we must be "especially aware of the imprimatur of legitimacy that a prosecutor's comments may have in the eyes of the jury." Le v. Mullin , 311 F.3d 1002, 1018 (10th Cir. 2002) ; see also Berger v. United States , 295 U.S. 78, 88, 55 S.Ct. 629, 79 L.Ed. 1314 (1935) (explaining that prosecution's "interest ... in a criminal prosecution is not that it shall win a case, but that justice shall be done"; noting that "the average jury ... has confidence that these obligations ... will be faithfully observed"; and concluding that prosecution's "improper suggestions" are therefore "apt to carry much weight against the accused").

And we cannot ignore that the trial court overruled Grant's objection when the prosecutor said that Grant's evidence, "in *973order to be mitigating," had to "reduce his moral culpability or blame." Id. at 74, 55 S.Ct. 629. Far from it, we've explained that when a trial court overrules defense counsel's objections in a case like this, "[t]he official imprimatur thereby placed upon the prosecution's misstatements of law obviously amplifie[s] their potential prejudicial effect on the jury." Mahorney v. Wallman , 917 F.2d 469, 473 (10th Cir. 1990).

In sum, the jury had no reason to doubt the prosecutor when she said, "[T]he law tells you ... that mitigating circumstances are those which reduce the moral culpability or blame of the defendant." R. vol. 4, Trial Tr. 8, at 73-74. It had no reason to doubt her when she said, "[Grant's evidence] in order to be mitigating, must reduce his moral culpability or blame." Id. at 74. It had no reason to doubt her when she said, "[What] the law says, not Sandra Elliott [the prosecutor], not what the defense attorneys say, but what the [c]ourt tells you and what the law says is that before something can be mitigating it must reduce the moral culpability or blame of the defendant." Id. at 75. And it had no reason to doubt her when she said, "[Y]ou have to look at whether or not it reduces his moral culpability or blame. That is what the law says that you must do." Id. at 79.

On the contrary, all three sources that the jury would have looked to for an accurate statement of the law-Instruction 12, the prosecution, and the trial court itself-gave the jury good reason to believe that it couldn't consider Grant's mitigating evidence if that evidence didn't reduce Grant's culpability or blame, as the defense conceded it didn't. The only thing that would have led the jurors to believe otherwise is an inference that they could have made from Instruction 13. It's thus clear that "[i]n light of the prosecutor's argument, and in the absence of appropriate jury instructions, a reasonable juror could well have believed that there was no vehicle for expressing the view that [Grant] did not deserve to be sentenced to death based upon his mitigating evidence." Penry I , 492 U.S. at 326, 109 S.Ct. 2934. And to the extent that the OCCA concluded otherwise on the merits-and again, I'm confident that it did not-that conclusion was unreasonable.

III. Harmless Error

Because the majority doesn't believe that AEDPA allows us to reach the trial court's error, it has no opportunity to consider whether this error was harmless. I would find that it plainly is not.

An error is only reversible on habeas review if it "had substantial and injurious effect or influence in determining the jury's verdict." Brecht v. Abrahamson , 507 U.S. 619, 637, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993) (quoting Kotteakos v. United States , 328 U.S. 750, 776, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946) ). We've elaborated that this means we must grant habeas relief if we find ourselves "in 'grave doubt' about the effect of the error on the jury's verdict." Welch v. Workman , 639 F.3d 980, 992 (10th Cir. 2011) (quoting O'Neal v. McAninch , 513 U.S. 432, 435, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995) ); see also McAninch , 513 U.S. at 435, 115 S.Ct. 992 ("By 'grave doubt' we mean that, in the judge's mind, the matter is so evenly balanced that he feels himself in virtual equipoise as to the harmlessness of the error.").

I find myself in grave doubt here, especially considering the theme of the defense's closing argument. Essentially, the defense (1) admitted to the jury that the factors it asserted as mitigating didn't reduce Grant's culpability or blame but (2) argued those factors nevertheless warranted the jury's sympathy. For the prosecution *974to rebut this by telling the jury it couldn't actually consider the evidence as mitigating unless it reduced Grant's culpability or blame gutted the defense's argument and left the jury with no room to decide whether the factors that Grant identified as mitigating-and the powerful evidence he offered to prove those factors-warranted a sentence other than death. Because wholly denying Grant his opportunity to present his case in mitigation had a substantial and injurious effect on the jury's determination that the death penalty was warranted in this case, the error wasn't harmless under Brecht .

* * *

Grant's crimes were abhorrent. But even the worst offenders have an absolute right to ask for mercy. It is disturbingly clear to me that Grant never had that opportunity. I would not allow Grant's execution to proceed without giving Grant an opportunity to explain to a jury why he doesn't deserve to die. I would thus reverse the district court's order denying Grant's habeas petition.

*886Petitioner Donald Anthony Grant, an Oklahoma state prisoner on death row, appeals from the district court's denial of his 28 U.S.C. § 2254 habeas petition. Additionally, Mr. Grant filed a motion to expand the certificate of appealability ("COA"). Exercising jurisdiction under 28 U.S.C. § 1291, we affirm the district court's denial of Mr. Grant's § 2254 petition and deny Mr. Grant's motion to expand the COA.

I. BACKGROUND

A. Facts

"The OCCA outlined the facts of Mr. [Grant's] crimes, and '[w]e presume that the factual findings of the state court are correct' unless Mr. [Grant] presents clear and convincing evidence otherwise." Clayton Lockett v. Trammel , 711 F.3d 1218, 1222 (10th Cir. 2013) (quoting Fairchild v. Workman , 579 F.3d 1134, 1137 (10th Cir. 2009) ); see also 28 U.S.C. § 2254(e)(1) ("[A] determination of a factual issue made by a State court shall be presumed to be correct. The applicant shall have the burden of rebutting the presumption of correctness by clear and convincing evidence.").

The OCCA provided the following factual summary:

The essential facts of the crimes are not disputed. On July 18, 2001, [Mr. Grant] entered a LaQuinta Inn in Del City, ostensibly to fill out an employment application. In reality, [Mr. Grant] had planned to rob the hotel in order to obtain money to post bond for a girlfriend, Shlonda Gatewood (who was in the Oklahoma County Jail at the time), and was prepared to kill any witnesses to the crime. [Mr. Grant] may have been motivated to strike this particular business because another girlfriend of his, Cheryl Tubbs, had been fired from employment there a few months before; in any event, [Mr. Grant] was familiar with the layout of the property and the location of video surveillance equipment.

When [Mr. Grant] saw the hotel manager, Brenda McElyea, he approached her with a pistol in his hand and ordered her to walk to a storage room, where he fatally shot her once in the head, and slashed her neck and back with a box knife to make sure the knife was sharp enough to use on his next victim. [Mr. Grant] then left the storage room and approached another employee, Suzette Smith, in the break room. [Mr. Grant] ordered Smith at gunpoint to give him the money from the hotel register, which she did. [Mr. Grant] then ordered Smith to walk back to the manager's office, where he shot her three times in the face. Smith continued to struggle to escape, so [Mr. Grant] brutally beat her and cut her numerous times with his knife. He hit Smith in the head with his pistol, attempted to break her neck, and threw a computer monitor on her head *887in an effort to stop her struggling. Eventually, Smith succumbed to her wounds and died in the office. Before leaving the office, [Mr. Grant] took personal property from Smith's purse.

[Mr. Grant] then left the hotel and walked to a nearby discount store, where he abandoned his pistol and some traveler's checks he had taken in the robbery. He then called a cab to take him to the home of Cheryl Tubbs. Later that day, [Mr. Grant] used money from the robbery to pay Shlonda Gatewood's bond, which was about $200. [Mr. Grant] and Gatewood then used a stolen car to drive from Oklahoma City to New York City, where [Mr. Grant] had family. About a month after the murders, [Mr. Grant] was arrested in New York and returned to Oklahoma.

Grant v. State , 205 P.3d 1, 7 (Okla. Crim. App. 2009) (numeric paragraph notations and footnote omitted).

B. Procedural History

In August 2001, Mr. Grant was charged with two counts of first degree murder and two counts of robbery with a firearm for the murders of Brenda McElyea and Suzette Smith during the robbery of the La Quinta Inn in Del City, Oklahoma. With respect to the murder counts, the State sought the death penalty. It charged several aggravating circumstances to support such a sanction:

(1) that the defendant knowingly created a great risk of death to more than one person; (2) that the murders were committed for the purpose of avoiding arrest or prosecution; (3) that the murders were committed by a person serving a sentence of imprisonment on conviction of a felony; and (4) that a probability existed that the defendant would pose a continuing threat to society. As to one of the murder counts (Count 2) [relating to Ms. Smith], the State also alleged that the murder was heinous, atrocious, or cruel.

Grant , 205 P.3d at 6 n.1.

In November 2001, Mr. Grant's counsel moved for a determination of his competency to stand trial. The parties litigated Mr. Grant's competency for the next four years, culminating in a February 2005 trial, at which a jury found Mr. Grant competent to stand trial.

Mr. Grant's eight-day jury trial began on November 14, 2005. The jury found Mr. Grant guilty on all counts. "As to each of the murder counts, the jury found the existence of all aggravating circumstances alleged, and recommended punishment of death on each count." Id. at 6-7. Mr. Grant filed a direct appeal and an application for an evidentiary hearing with the Oklahoma Court of Criminal Appeals ("OCCA"). The OCCA affirmed Mr. Grant's conviction and death sentence and denied his request for an evidentiary hearing. In 2008 Mr. Grant filed an application for post-conviction relief with the OCCA. The OCCA again denied relief.

In October 2012, Mr. Grant filed the instant 28 U.S.C. § 2254 petition with the United States District Court for the Western District of Oklahoma. Mr. Grant raised numerous propositions of error, five of which are relevant to the present appeal. First, he argued that he was denied procedural due process because the trial court failed to hold a second competency hearing in response to Mr. Grant's alleged manifestations of incompetence leading up to and during trial. Second, he raised several ineffective-assistance-of-counsel claims relating to trial counsel's failures to investigate and present evidence regarding his competence and other mitigating circumstances. Third, Mr. Grant challenged the constitutionality of a jury instruction and related prosecutorial statements concerning *888mitigation evidence. Fourth, Mr. Grant raised a constitutional challenge to the peremptory strike of a potential juror on the basis of race. Finally, Mr. Grant argued that he was prejudiced by cumulative error. The district court denied Mr. Grant's petition and granted a COA on the single issue of procedural competency.

Mr. Grant filed a timely appeal. In our December 12, 2014 Case Management Order, we granted a COA on Mr. Grant's additional claims concerning (1) ineffective assistance of trial counsel, (2) the challenged jury instruction and related prosecutorial statements, (3) the peremptory strike of a minority (i.e., African-American) juror, and (4) cumulative error. On December 29, 2014, Mr. Grant filed a motion to expand the COA to include one additional issue. This motion is still pending before us.

II. STANDARD OF REVIEW

The Antiterrorism and Effective Death Penalty Act of 1996 ("AEDPA") "circumscribes our review of federal habeas claims that were adjudicated on the merits in state-court proceedings." Hooks v. Workman (" Victor Hooks II "), 689 F.3d 1148, 1163 (10th Cir. 2012). "Under AEDPA, a petitioner is entitled to federal habeas relief on a claim only if he can establish that the state court's adjudication of the claim on the merits (1) 'resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law'; 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.' " Littlejohn v. Trammell ("Littlejohn I "), 704 F.3d 817, 824 (10th Cir. 2013) (quoting 28 U.S.C. § 2254(d)(1), (2) ); see Kernan v. Cuero , --- U.S. ----, 138 S.Ct. 4, 5, 199 L.Ed.2d 236 (2017) (per curiam); Byrd v. Workman , 645 F.3d 1159, 1166 (10th Cir. 2011).

The AEDPA standard is "highly deferential [to] state-court rulings [and] demands that state-court decisions be given the benefit of the doubt." Woodford v. Visciotti , 537 U.S. 19, 24, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002) (per curiam) (citation omitted) (quoting Lindhv. Murphy , 521 U.S. 320, 333 n.7, 117 S.Ct. 2059, 138 L.Ed.2d 481 (1997) ); accord Littlejohn I , 704 F.3d at 824 ; Victor Hooks II , 689 F.3d at 1163. "A habeas petitioner meets this demanding standard only when he shows that the state court's decision was 'so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.' " Dunn v. Madison , --- U.S. ----, 138 S.Ct. 9, 11, 199 L.Ed.2d 243 (2017) (per curiam) (quoting Harrington v. Richter , 562 U.S. 86, 103, 131 S.Ct. 770, 178 L.Ed.2d 624 (2011) ). As the Court explained in Harrington , "If this [AEDPA] standard is difficult to meet, that is because it was meant to be .... It preserves authority to issue the writ in cases where there is no possibility fairminded jurists could disagree that the state court's decision conflicts with this Court's precedents. It goes no further." 562 U.S. at 102, 131 S.Ct. 770 (citations omitted).

"In applying the legal inquiry under § 2254(d)(1), we ask at the threshold 'whether there exists clearly established federal law, an inquiry that focuses exclusively on holdings of the Supreme Court.' " Littlejohn I , 704 F.3d at 825 (quoting Victor Hooks II , 689 F.3d at 1163 ); see Cuero , 138 S.Ct. at 8 ("[W]e still are unable to find in Supreme Court precedent that 'clearly established federal law' demanding specific performance as a remedy. To the contrary, no 'holdin[g] of this Court' requires the remedy of specific performance *889under the circumstances present here." (alteration in original) (quoting Harrington , 562 U.S. at 100, 131 S.Ct. 770 ). "The absence of clearly established federal law is dispositive under § 2254(d)(1)" and results in the denial of habeas relief. Victor Hooks II , 689 F.3d at 1163 (quoting House v. Hatch , 527 F.3d 1010, 1018 (10th Cir. 2008) ).

If clearly established federal law exists, a state-court decision is contrary to it only if the 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." Id. (alteration in original) (quoting Bell v. Cone , 535 U.S. 685, 694, 122 S.Ct. 1843, 152 L.Ed.2d 914 (2002) ). "A state court decision unreasonably applies federal law if it 'identifies the correct governing legal principle from [Supreme Court] decisions but unreasonably applies the principle to the facts of the prisoner's case.' " Littlejohn I , 704 F.3d at 825 (alteration in original) (quoting Bland v. Sirmons , 459 F.3d 999, 1009 (10th Cir. 2006) ).

Finally, "[h]abeas relief is also warranted if the state court's adjudication of a claim on the merits '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.' " Victor Hooks II , 689 F.3d at 1163 (quoting 28 U.S.C. § 2254(d)(2) ). "We will not conclude that a state court's determination of the facts is unreasonable unless the court plainly and materially misstated the record or the petitioner shows that reasonable minds could not disagree that the finding was in error." Michael Smith v. Duckworth , 824 F.3d 1233, 1250 (10th Cir. 2016) (citing Ryder ex rel. Ryder v. Warrior , 810 F.3d 724, 739 (10th Cir. 2016) ), cert. denied , --- U.S. ----, 137 S.Ct. 1333, 197 L.Ed.2d 526, reh'g denied , --- U.S. ----, 137 S. Ct. 2153, 198 L.Ed.2d 225 (2017).

"[W]e review the district court's legal analysis of the state court decision de novo ," Byrd , 645 F.3d at 1165 (alteration in original) (quoting Bland , 459 F.3d at 1009 ), and "the factual findings of the state court are [presumed] correct unless the petitioner rebuts that presumption by 'clear and convincing evidence,' " id. (quoting 28 U.S.C. § 2254(e)(1) ). Moreover, our review "is limited to the record that was before the state court that adjudicated the claim on the merits." Victor Hooks II , 689 F.3d at 1163 (alteration in original) (quoting Cullen v. Pinholster , 563 U.S. 170, 181, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011) ).

"Our standard of review changes if there has been no state-court adjudication on the merits of the petitioner's claim." Byrd , 645 F.3d at 1166. That is, "[t]he [deferential] § 2254(d) standard does not apply to issues not decided on the merits by the state court." Welch v. Workman , 639 F.3d 980, 992 (10th Cir. 2011) (emphasis added) (quoting Bland , 459 F.3d at 1010 ). For such claims, "we exercise our 'independent judgment' and 'review the federal district court's conclusions of law de novo,' " and its factual findings for clear error. Victor Hooks II , 689 F.3d at 1163-64 (quoting McCracken v. Gibson , 268 F.3d 970, 975 (10th Cir. 2001) ). And, even in the setting where we lack a state court merits determination, "[a]ny state-court findings of fact that bear upon the claim are entitled to a presumption of correctness rebuttable only by 'clear and convincing evidence.' " Id . (quoting 28 U.S.C. § 2254(e)(1) ).

"Finally, we may not consider claims that have been 'defaulted in state court on adequate and independent state procedural grounds' " absent the petitioner's demonstration of " 'cause for the default *890and actual prejudice as a result of the alleged violation of federal law, or [that] failure to consider the claims will result in a fundamental miscarriage of justice.' " Byrd , 645 F.3d at 1167 (quoting Matthews v. Workman , 577 F.3d 1175, 1195 (10th Cir. 2009) ); see also Davila v. Davis , --- U.S. ----, 137 S.Ct. 2058, 2064, 198 L.Ed.2d 603 (2017) ("[A] state prisoner must exhaust available state remedies before presenting his claim to a federal habeas court. The exhaustion requirement is designed to avoid the 'unseemly' result of a federal court 'upset[ting] a state court conviction without' first according the state courts an 'opportunity to ... correct a constitutional violation[.]' " (alteration in original) (citation omitted) (quoting Rose v. Lundy , 455 U.S. 509, 518, 102 S.Ct. 1198, 71 L.Ed.2d 379 (1982) )).

We now turn to assessing Mr. Grant's claims.

III. MERITS

A. Procedural Due Process Competency Claim

Mr. Grant argues that the trial court violated his procedural due process rights by allowing his trial to proceed while he was incompetent. The district court deemed this claim to be procedurally barred, finding that Mr. Grant failed to exhaust the claim in state court. Because we agree with the district court that Mr. Grant failed to exhaust his procedural due process competency argument before the OCCA, we uphold this aspect of the district court's ruling.

1. Legal Framework

"A state prisoner generally must exhaust available state-court remedies before a federal court can consider a habeas corpus petition." Bland , 459 F.3d at 1011 ; see 28 U.S.C. § 2254(b)(1)(A) ; accord Thacker v. Workman , 678 F.3d 820, 838-39 (10th Cir. 2012) ; see also Ellis v. Raemisch , 872 F.3d 1064, 1076 (10th Cir. 2017) ("More specifically, AEDPA prohibits federal courts from granting habeas relief to state prisoners who have not exhausted available state remedies."). "Exhaustion requires that the claim be 'fairly presented' to the state court, which 'means that the petitioner has raised the "substance" of the federal claim in state court.' " Fairchild , 579 F.3d at 1151 (quoting Bland , 459 F.3d at 1011 ); accord Jeremy Williams v. Trammell , 782 F.3d 1184, 1210 (10th Cir. 2015), cert. denied sub nom. Williams v. Warrior , --- U.S. ----, 136 S.Ct. 806, 193 L.Ed.2d 726 (2016). Put another way, "a federal habeas petitioner [must] provide the state courts with a 'fair opportunity' to apply controlling legal principles to the facts bearing upon his constitutional claim." Anderson v. Harless , 459 U.S. 4, 6, 103 S.Ct. 276, 74 L.Ed.2d 3 (1982) ; accord Picard v. Connor , 404 U.S. 270, 277, 92 S.Ct. 509, 30 L.Ed.2d 438 (1971).

"[T]he crucial inquiry is whether the 'substance' of the petitioner's claim has been presented to the state courts in a manner sufficient to put the courts on notice of the federal constitutional claim." Prendergast v. Clements , 699 F.3d 1182, 1184 (10th Cir. 2012) (citing Picard , 404 U.S. at 278, 92 S.Ct. 509 ). "A petitioner need not invoke 'talismanic language' or cite 'book and verse on the federal constitution.' " Id. (quoting Nichols v. Sullivan , 867 F.2d 1250, 1252 (10th Cir. 1989) ); accord Picard , 404 U.S. at 278, 92 S.Ct. 509. But, a " '[f]air presentation' requires more than presenting 'all the facts necessary to support the federal claim' to the state court." Bland , 459 F.3d at 1011 (emphasis added) (quoting Anderson , 459 U.S. at 6, 103 S.Ct. 276 ).

*891Nor is citing the relevant legal principles, absent the relevant facts, sufficient to fairly present a claim. See Picard , 404 U.S. at 277, 92 S.Ct. 509 (finding no fair presentation where state court had no "opportunity to apply controlling legal principles to the facts bearing upon [the federal] claim"); Anderson , 459 U.S. at 6, 103 S.Ct. 276 ("[A] federal habeas petitioner [must] provide the state courts with a 'fair opportunity' to apply controlling legal principles to the facts bearing upon his constitutional claim."); Jeremy Williams , 782 F.3d at 1210 (noting that the substance of the claim "includes not only the constitutional guarantee at issue, but also the underlying facts that entitle a petitioner to relief").

Furthermore, a "petitioner cannot assert entirely different arguments [in his or her request for habeas relief] from those raised before the state court." Bland , 459 F.3d at 1011. That is, there is no fair presentation if the claim before the state court was only "somewhat similar" to the claim pressed in the habeas petition. Duncan v. Henry , 513 U.S. 364, 366, 115 S.Ct. 887, 130 L.Ed.2d 865 (1995) ; see also Bland , 459 F.3d at 1012 (finding failure to exhaust "[b]ecause presentation of a 'somewhat similar' claim is insufficient to 'fairly present' a federal claim"). Indeed, "mere similarity of claims is insufficient to exhaust." Id. And the assertion of a general claim before the state court is insufficient to exhaust a more specific claim asserted for habeas relief. See Gray v. Netherland , 518 U.S. 152, 163, 116 S.Ct. 2074, 135 L.Ed.2d 457 (1996) ("[I]t is not enough to make a general appeal to a constitutional guarantee as broad as due process to present the 'substance' of such a claim to a state court."); see also Thomas v. Gibson , 218 F.3d 1213, 1221 n.6 (10th Cir. 2000) (holding petitioner's "generalized" state-court ineffective-assistance claim was insufficient to exhaust his later, more specific federal habeas claim).

Indeed, in order to be fairly presented, the state-court claim must be the "substantial equivalent" of its federal habeas counterpart. Picard , 404 U.S. at 278, 92 S.Ct. 509. There is no such substantial equivalency where the claim raised in habeas proceedings is "in a significantly different and stronger posture than it was when the state courts considered it." Jones v. Hess , 681 F.2d 688, 694 (10th Cir. 1982). To satisfy exhaustion, then, the habeas petition's focus-as well as the alleged error that it identifies-cannot depart significantly from what the petitioner had presented to the state court. See, e.g. , Bland , 459 F.3d at 1012 (noting that the habeas "challenge to the actions of the prosecution differs significantly from" the state-court "challenge to the instructions given by the court," even where both concerned the propriety of a given jury instruction). Nor is it acceptable for the habeas petitioner to "shift" the "basis for [his or her] argument" away from what was previously raised in state court. Gardner v. Galetka , 568 F.3d 862, 872 (10th Cir. 2009) (claims were not "substantially the same" where petitioner's state-court ineffective-assistance claim was predicated on counsel's inaccurate description of petitioner's injury, but where his habeas claim was grounded on counsel's failure to undertake a thorough investigation of the murder weapon); Smallwood v. Gibson , 191 F.3d 1257, 1267 (10th Cir. 1999) (holding a claim for ineffective assistance of counsel not exhausted where petitioner "based [his state-court claim] on different reasons" and on different "bases [than those] upon which his current ineffective assistance of counsel claims rely").

There are consequences for failing to properly present a claim. "Generally, a federal court should dismiss unexhausted *892claims without prejudice so that the petitioner can pursue available state-court remedies." Bland , 459 F.3d at 1012 ; see 28 U.S.C. § 2254(b)(1)(A). "However, dismissal without prejudice for failure to exhaust state remedies is not appropriate if the state court would now find the claims procedurally barred on independent and adequate state procedural grounds." Smallwood , 191 F.3d at 1267 (citing Coleman v. Thompson , 501 U.S. 722, 735 n.1, 111 S.Ct. 2546, 115 L.Ed.2d 640 (1991) ). Where the relevant state courts "would now find those claims procedurally barred, there is a procedural default for the purposes of federal habeas review." Bland , 459 F.3d at 1012 (quoting Dulin v. Cook , 957 F.2d 758, 759 (10th Cir. 1992) ); see also Moore v. Schoeman, 288 F.3d 1231, 1233 n.3 (10th Cir. 2002) (" 'Anticipatory procedural bar' occurs when the federal courts apply procedural bar to an unexhausted claim that would be procedurally barred under state law if the petitioner returned to state court to exhaust it.") (citing Hain v. Gibson , 287 F.3d 1224, 1240 (10th Cir. 2002) ). A petitioner may overcome the procedural bar only if he can "demonstrate cause for the default and actual prejudice as a result of the alleged violation of federal law, or demonstrate that failure to consider the claims will result in a fundamental miscarriage of justice." Coleman , 501 U.S. at 750, 111 S.Ct. 2546.

2. Analysis

We conclude after thorough review that Mr. Grant presented only a substantive due process competency argument to the OCCA on direct appeal. In short, he failed to present the substance of his procedural due process competency argument to the state court. Mr. Grant's argument to the contrary depends on the inherent similarities between the two types of competency challenges. We conclude, however, that Mr. Grant's presentation of a "somewhat similar" claim, Duncan , 513 U.S. at 366, 115 S.Ct. 887, on direct appeal1 was insufficient to have put the state courts on notice of the procedural competency claim he now urges, Jones , 681 F.2d at 694.

a

We begin the analysis by delineating the differences between claims of procedural due process competency and claims of substantive due process competency.

"[C]ompetency claims can raise issues of both substantive and procedural due process." Walker v. Attorney Gen. , 167 F.3d 1339, 1343 (10th Cir. 1999). Although "sometimes there is overlap," procedural competency and substantive competency are distinct claims. Barnett v. Hargett , 174 F.3d 1128, 1133 (10th Cir. 1999). "A procedural [due process] competency claim is based upon a trial court's alleged failure to hold a competency hearing, or an adequate competency hearing, while a substantive competency claim is founded on the allegation that an individual was tried and convicted while, in fact, incompetent." Allen v. Mullin , 368 F.3d 1220, 1239 (10th Cir. 2004) (quoting McGregor v. Gibson , 248 F.3d 946, 952 (10th Cir. 2001) (en banc)).

"The distinction between substantive and procedural claims is significant because courts have evaluated these claims under differing evidentiary standards." Walker , 167 F.3d at 1344. To make out a procedural competency claim, a defendant must demonstrate that "a reasonable judge should have had a bona fide doubt as to [the defendant's] competence *893at the time of trial," McGregor , 248 F.3d at 954, but the claim does "not require proof of actual incompetency," Allen , 368 F.3d at 1239. Further, procedural competency imposes on the trial court a continuing duty to monitor the defendant's behavior. See Drope v. Missouri , 420 U.S. 162, 181, 95 S.Ct. 896, 43 L.Ed.2d 103 (1975) ("Even when a defendant is competent at the commencement of his trial, a trial court must always be alert to circumstances suggesting a change that would render the accused unable to meet the standards of competence to stand trial.").

"A substantive competency claim, on the other hand, requires the higher standard of proof of incompetency by a preponderance of the evidence." Allen , 368 F.3d at 1239 (emphasis added) (citing Cooper v. Oklahoma , 517 U.S. 348, 368-69, 116 S.Ct. 1373, 134 L.Ed.2d 498 (1996) ). A petitioner alleging a substantive claim must demonstrate that he actually lacked a "sufficient present ability to consult with his lawyer with a reasonable degree of rational understanding [and] a rational as well as a factual understanding of the proceedings against him." Dusky v. United States , 362 U.S. 402, 402, 80 S.Ct. 788, 4 L.Ed.2d 824 (1960). Thus, a petitioner alleging a substantive competency claim must show that he was convicted during a period of incompetency, McGregor , 248 F.3d at 953, whereas a procedural competency petitioner "states a procedural competency claim by alleging the trial court failed to hold a competency hearing after the defendant's mental competency was put in issue." Valdez v. Ward , 219 F.3d 1222, 1239 (10th Cir. 2000).

Moreover, the two claims rest on different evidentiary bases. In evaluating a procedural competency claim, "[o]ur review is limited to the evidence that was made available to the state trial court." Lay v. Royal , 860 F.3d 1307, 1314 (10th Cir. 2017) (citing Allen , 368 F.3d at 1239 ). However, post-conviction evidence can often be relevant to establishing substantive incompetency. See, e.g. , Nguyen v. Reynolds , 131 F.3d 1340, 1345-47 (10th Cir. 1997) (considering post-conviction behavior in prison in the context of a substantive competency claim). In further juxtaposition, competency claims based on substantive due process are subject neither to waiver, nor to procedural bar, whereas their procedural counterparts are susceptible to both. See id. at 1346 ; Barnett , 174 F.3d at 1133.

Further, because procedural competency and substantive competency guard against distinct harms, it should come as no surprise that their corresponding remedies are also different. Since the error asserted in a procedural claim is the court's failure to provide adequate procedures-i.e., the failure to conduct a competency hearing-a defendant who prevails on a procedural competency claim is entitled to the procedures (i.e., a competency hearing) that he should have received in the first instance. See McGregor , 248 F.3d at 952 (noting that "[a] procedural competency claim is based upon a trial court's alleged failure to hold a competency hearing, or an adequate competency hearing"); id. at 962 ("Our conclusion that McGregor's procedural due process rights were violated does not end the analysis. We next consider whether a retrospective competency hearing can be held."); see also Barnett , 174 F.3d at 1133-34 (noting that a petitioner pursuing a procedural competency claim alleges that he was deprived of the right "to an adequate state procedure to insure that he is in fact competent to stand trial," which he should have received in the first place); see also United States v. Grist , 299 Fed.Appx. 770, 775, 778 (10th Cir. 2008) (unpublished) (finding no error in the magistrate judge's assessment that *894no relief was due to the petitioner, as he "ha[d] been afforded ... the only relief to which [he] would be entitled for a procedural due process competency violation: a retrospective competency determination").2

Thus, the issuance of the habeas writ is not mandated in situations where the procedural competency claimant is successful; it is resorted to only where a retrospective competency hearing would not be feasible. Compare McGregor , 248 F.3d at 962 ("[W]e conclude that a meaningful retrospective competency determination can not be made in this case. As such, McGregor's due process rights can not adequately be protected by remanding to the state court for such a determination .... Accordingly, we GRANT McGregor's request for habeas corpus relief."), and Pate v. Robinson , 383 U.S. 375, 387, 86 S.Ct. 836, 15 L.Ed.2d 815 (1966) (rejecting the option of remand for a "limited hearing as to Robinson's mental competence at the time he was tried"), with Bryan v. State , 935 P.2d 338, 347 n.4 (Okla. Crim. App. 1997) ("The [ Robinson ] Court did not, however, rule that retrospective hearings may not be held where such hearings are feasible, and has never so held despite the opportunity in subsequent cases."), and Walker , 167 F.3d at 1347 n.4 ("In these circumstances, we are not persuaded Mr. Walker was deprived of due process by the retrospective competency hearing."). Such a hearing would be infeasible where-due to the passage of time, the unavailability of contemporaneous medical evidence, the lack of defendant's statements on the trial record, or the absence of eyewitnesses to the defendant's behavior during trial-the examination would no longer be meaningful. See McGregor , 248 F.3d at 962 (citing Clayton v. Gibson , 199 F.3d 1162, 1169 (10th Cir. 1999) ).

As noted, the right at issue in a substantive competency claim is the right not to be tried while incompetent; therefore, in the habeas context, the remedy must involve the issuance of the writ because the conviction cannot constitutionally stand. See, e.g. , id. at 952 (noting that "a substantive competency claim is founded on the allegation that an individual was tried and convicted while, in fact, incompetent"); see also Godinez v. Moran , 509 U.S. 389, 396, 113 S.Ct. 2680, 125 L.Ed.2d 321 (1993) ("A criminal defendant may not be tried unless he is competent.").

It falls upon us to determine whether Mr. Grant actually presented a procedural competency claim to the OCCA on direct appeal, in addition to a substantive competency claim; we have undertaken analogous inquiries. See, e.g. , Walker , 167 F.3d at 1343 (disagreeing with the district court's characterization of competency claim as procedural, rather than substantive); Barnett , 174 F.3d at 1134 (holding that the district court erred in construing the claim as substantive when it was in fact procedural). In so doing, we parse the pleadings of Mr. Grant's direct-appeal brief and proceed with "care[ ] not to collapse the distinction between procedural and substantive due process [claims]." McGregor , 248 F.3d at 953.

b

Mr. Grant argues that he presented a procedural competency claim on direct appeal: more specifically, he says that he cited the relevant caselaw, set forth the *895relevant facts, and "correctly argued ... that the trial judge independently had a duty, based on all the information available to him prior to and during trial, to halt the trial for additional competency proceedings." Aplt.'s Opening Br. at 30. We are not persuaded.

Our contrary view is supported by an independent examination of Mr. Grant's direct-appeal briefing: this review makes clear that the unwavering focus of Mr. Grant's arguments was on establishing his actual incompetence at the time of trial-that is, on mounting a substantive competency claim, and not a procedural one. We also find validation for our conclusion by juxtaposing Mr. Grant's state-court claim with the one filed for habeas relief; this side-by-side view reveals that Mr. Grant's original competency claim bears little resemblance to the one he now advances.

We are thus left with a firm conviction that Mr. Grant's direct-appeal competency claim was "based ... on different reasons," and grounded on different legal "bases [than those] upon which his current ... claim[ ] rely." Smallwood , 191 F.3d at 1267. We accordingly hold that Mr. Grant did not fairly present a procedural competency claim to the OCCA. Furthermore, for reasons explicated infra , this claim is subject to an anticipatory procedural bar. Therefore, we are precluded from considering it.

i

Several aspects of Mr. Grant's direct-appeal brief contradict his argument that he fairly presented a procedural competency claim to the OCCA. Indeed, we discern that the focus of Mr. Grant's direct-appeal briefing-and therefore, the substance of the claim before the OCCA-was on whether Mr. Grant was in fact incompetent at the time of his November 2005 trial.

At the outset, we note that the heading to Proposition I-the section under which Mr. Grant contends he raised his procedural competency claim-did nothing to put the OCCA on notice of any such claim. In fact, the heading neatly set out the standard-not for a procedural claim-but for proving a violation of substantive due process. Compare Aplt.'s Direct Appeal Opening Br. at 3 ("Mr. Grant Was Incompetent When He Stood Trial, In Violation of Due Process"), with Allen , 368 F.3d at 1239 ("[A] substantive competency claim is founded on the allegation that an individual was tried and convicted while, in fact, incompetent.").3 We also observe that Mr. Grant himself characterized Proposition I as setting out "Appellant['s] complain[t] that he was tried while incompetent." Aplt.'s Direct Appeal App. for Evidentiary Hr'g. on Sixth and Fourteenth Amendment Claims (filed Oct. 11, 2007), at 1.4

*896Nor would the body of the direct-appeal brief have put the OCCA on notice that the competency claim was anything other than substantive in nature. Proposition I was devoted to establishing Mr. Grant's actual incompetency: the unalloyed thrust of the facts presented there was that Mr. Grant was incompetent in November 2005, when he stood trial. The argument opened by asserting that the reason for the four-year delay in Mr. Grant's trial was due to his incompetence. It proceeded to lay out the lengthy chronology of Mr. Grant's history with his lawyers, the court, and the psychological experts they appointed to assess his competency to stand trial, who had found him to be incompetent. It is evident from the brief that Mr. Grant's quarrel was with the fact of his conviction while incompetent.

Consistent with the requirements of a substantive claim, the brief alleged facts showing that Mr. Grant "lack[ed] the capacity to understand the nature and object of the proceedings against him, to consult with counsel, and to assist in preparing his defense." Drope , 420 U.S. at 171, 95 S.Ct. 896. Specifically, it noted Mr. Grant's "understanding of the charges and criminal proceedings was 'fleeting at best,' " Aplt.'s Direct Appeal Opening Br. at 4 (quoting from the underlying state court record). Furthermore, it recounted how Mr. Grant "was agitated and [the] presence [of his counsel] aggravated him." Id. at 5. And the brief observed that Mr. Grant's lawyer was "concerned because" Mr. Grant was "making statements that may hurt him in the future." Id. (quoting from the underlying state court record).

Conversely, Mr. Grant's direct-appeal briefing does not similarly bear the trappings of a procedural competency claim, which would have alleged that a "reasonable judge should have had a bona fide doubt as to [petitioner's] competence at the time of trial." McGregor , 248 F.3d at 954. For one, the direct-appeal brief never mentioned the established "bona fide doubt" standard of proof. For another, the scope of the information presented on direct appeal exceeded the bounds of evidence relevant to a procedural claim. For example, in order to demonstrate that Mr. Grant was not taking his medications, Mr. Grant's direct-appeal brief requested that the OCCA consider Mr. Grant's medical records for the period leading up to trial. See Aplt.'s Direct Appeal Opening Br. at 11. However, these medical records were neither part of the trial record; nor were they at any point before the trial judge.5 As such, they were irrelevant to a procedural due process claim, because "[o]ur review [as to such a claim] is limited to the evidence that was made available to the state trial court." Lay , 860 F.3d at 1314 (citing Allen , 368 F.3d at 1239 ); see also McGregor , 248 F.3d at 954 ("We view the evidence in the record objectively, from *897the standpoint of a reasonable judge presiding over petitioner's case at the time of trial .") (emphases added). Therefore, Mr. Grant's reliance on evidence that was never before the trial court, nor part of the record, is inconsistent with his assertion now that he was presenting a procedural due process claim to the OCCA.

Quite significant, moreover, is the stark lack of fit between the remedy requested from the OCCA on direct appeal and the procedural competency claim Mr. Grant purports to have raised. Nowhere in Mr. Grant's direct-appeal brief did he mention the possible relief of a retrospective competency hearing-though ordinarily, if feasible, this would be the relief available to him if he had prevailed. See McGregor , 248 F.3d at 962 ("Our conclusion that McGregor's procedural due process rights were violated does not end the analysis. We next consider whether a retrospective competency hearing can be held."); see also Grist , 299 Fed.Appx. at 775, 778 ; Bryan , 935 P.2d at 347 n.4 (noting that, in the procedural competency context, the OCCA "has remanded the issue for a determination of feasibility and a [retrospective] hearing"); Boltz v. State , 806 P.2d 1117, 1121 (Okla. Crim. App. 1991) ("It is the opinion of this Court that if a defendant's competency at the time of trial can be meaningfully determined at a subsequent time on the basis of credible and competent evidence, then error committed by a district court in failing to hold a hearing at the proper time can be cured. If the procedural defect is thereby cured, any due process attack based upon the defect must fail."). Nor did Mr. Grant ever allege that the trial court should have halted the trial to perform a competency hearing. See Valdez , 219 F.3d at 1239 ("[A] petitioner states a procedural competency claim by alleging that the trial court failed to hold a competency hearing."). The notable absence of a request for a (retrospective) competency hearing belies Mr. Grant's assertion that he sought to vindicate a procedural error on direct appeal.

Indeed, Mr. Grant demonstrated in his direct-appeal brief that he knew how to ask for a hearing when he believed one was necessary to vindicate his rights. Specifically, he alleged such an entitlement just a mere moment after his competency arguments, in the context of Proposition II, which set out his Sixth Amendment self-representation claim. See Aplt.'s Direct Appeal Opening Br. at 12-13 ("It was thus incumbent upon the trial court to hold a hearing on the issue [of Mr. Grant's desire to dismiss his attorneys], at which the trial court should have evaluated [Mr. Grant's] competency to waive [representation by counsel] and warned him of the dangers of self-representation."). The fact that Mr. Grant did not ask for a hearing in Proposition I-which he now identifies as the source of his procedural due process competency claim-strongly indicates that he did not believe at that time that this remedy was material to the claim he was pursuing there. And that belief would have been correct-if, as we conclude-Mr. Grant was presenting there a substantive competency claim. Put another way, given Mr. Grant's demonstrated ability to challenge the court's failure to hold a hearing in the self-representation context, his silence regarding a hearing in the competency context is deafening and strongly suggests that the kind of competency claim that he actually was pursuing was not one that would have been satisfied by a hearing. See Duncan , 513 U.S. at 366, 115 S.Ct. 887 (noting that failure to raise a specific due process claim "is especially pronounced in that respondent did specifically raise a due process objection before the state court based on a different claim").

Instead of requesting an additional hearing, Mr. Grant solely sought reversal; as *898discussed, this remedy is clearly appropriate in the context of a claim for substantive competency, but not a remedy at least of first resort in the setting of a procedural competency claim. See, e.g. , McGregor , 248 F.3d at 962 ; Bryan , 935 P.2d at 347 n.4. And the concluding paragraph of Proposition I is particularly illuminating: there, Mr. Grant sought reversal based on the fact that he "was not competent when he was tried .... in violation of his Fourteenth Amendment due process right to be competent." Aplt.'s Direct Appeal Opening Br. at 11. This is paradigmatic language of a substantive due process competency claim.

To be sure, throughout Proposition I, there are references to facts that could have been relevant to a procedural due process competency claim. But given the "blurred ... distinctions" between substantive and procedural competency claims, Walker , 167 F.3d at 1344, this is entirely unremarkable. For instance, direct-appeal counsel quoted statements Mr. Grant made to the trial court, Mr. Grant's testimony during the penalty phase, and a letter Mr. Grant wrote to the prosecutor. Direct-appeal counsel asserted that these statements were "full of ramblings and delusions," and that despite this, Mr. Grant's "competence to stand trial was never revisited, even though ... he [showed] signs that his competence may have slipped." Aplt.'s Direct Appeal Opening Br. at 6-8. Yet the mere presentation of potentially-relevant facts is not enough. See Bland , 459 F.3d at 1011 (" '[F]air presentation' requires more than presenting ... 'all the facts necessary to support the federal claim' " to the state court) (quoting Anderson , 459 U.S. at 6, 103 S.Ct. 276 ).

Similarly, in Proposition I, Mr. Grant also cited to legal principles that could have been relevant to a procedural due process competency claim. For instance, he cited Drope , 420 U.S. at 180, 95 S.Ct. 896, for the proposition that "[e]ven if the defendant is competent when the trial begins, the trial court must always be alert to circumstances suggesting a change that would render the accused unable to meet the standard of competence to stand trial." Aplt.'s Direct Appeal Opening Br. at 11. And he cited Robinson , 383 U.S. at 385-86, 86 S.Ct. 836, for the proposition that "[e]vidence of a defendant's irrational behavior, his demeanor at trial, and any prior medical opinion on competence to stand trial are all relevant in determining whether further inquiry is required." Id.

However, recitations of caselaw-tethered neither to factual allegations nor directed argumentation-also fall short of fairly presenting a legal claim. See Gray , 518 U.S. at 163, 116 S.Ct. 2074 ("[I]t is not enough to make a general appeal to a constitutional guarantee as broad as due process to present the 'substance' of such a claim to a state court."); see also Thomas , 218 F.3d at 1221 n.6 (holding petitioner's "generalized" state-court ineffective-assistance claim was insufficient to exhaust his later, more specific federal habeas claim). Indeed, Mr. Grant's citations to Robinson and Drope came at the tail-end of Proposition I and were not connected, in any readily-discernible way, to the relevant factual allegations discussed some five pages before. See Burnett v. Hargett , 139 F.3d 911, 1998 WL 88153, at *5 (10th Cir. 1998) (unpublished) (finding lack of exhaustion where petitioner belatedly made "stray citations to a pair of potentially relevant federal cases"). Because Mr. Grant failed to tie a legal theory grounded in procedural due process to the salient facts, we hold that he failed to fairly present such a claim to the OCCA.

ii

Finally, our conclusion is further bolstered by a brief comparison of the argument *899that Mr. Grant presented on direct appeal with the argument that he has presented in his habeas petition. The claim on direct appeal is not the "substantial equivalent" of the one asserted before us. Picard , 404 U.S. at 278, 92 S.Ct. 509. "[T]he basis for [the competency] argument ... shifted" from Mr. Grant's alleged incompetency during trial to the trial court's failure to monitor his condition. Gardner , 568 F.3d at 872.

Whereas the direct-appeal briefing made only fleeting references to the trial court's obligation to inquire into Mr. Grant's competency, the habeas petition is replete with assertions that the trial judge knew, was aware of, or otherwise ignored signs that ought to have instilled in him a "bona fide doubt regarding Mr. Grant's competency to stand trial." Aplt.'s Habeas Pet. at 11 (dated Jan. 25, 2011). Moreover, the petition explicitly identifies the procedural defect as the trial judge's "neglect [of] duty ... to have a hearing on Mr. Grant's competency." Id. at 23. This specific allegation of error-the trial court's "failure to adopt a procedure to ensure Mr. Grant was competent," id. at 12-13-is nowhere to be seen in Mr. Grant's direct-appeal brief before the OCCA. The juxtaposition between that brief and Mr. Grant's habeas petition strongly underscores the deficiencies of the former as an ostensible presentation of a procedural due process competency claim. The direct-appeal brief did not offer a fair presentation of such a claim.

c

In sum, in light of the foregoing, we cannot say that Mr. Grant "provide[d] the state court with a 'fair opportunity' to apply controlling legal principles to the facts bearing upon his" procedural due process competency claim. Anderson , 459 U.S. at 6, 103 S.Ct. 276.6

Mr. Grant asks us to reach a contrary conclusion based in part on the contents of the State's direct-appeal response brief. Specifically, Mr. Grant contends that, though the State argues now for a lack-of-exhaustion determination, on direct appeal it sounded a different "tune" in that its response brief "acknowledged the procedural due process element of the claim" and responded to that element in its briefing. Aplt.'s Reply Br. at 3. Mr. Grant contends that the State's argument against a procedural due process claim is tantamount to a "concession" that Mr. Grant actually presented such a claim to the OCCA. Id. However, the State made no such concession, and certainly not the kind of explicit one that would be necessary under the law to remove the exhaustion-fair-presentation "issue from consideration." Fairchild , 579 F.3d at 1148 n.7 ("[T]he State also has not explicitly argued exhaustion but that fact does not remove the issue from consideration."); see 28 U.S.C. § 2254(b)(3) ("A State shall not be deemed to have waived the exhaustion requirement or be estopped from reliance upon the requirement unless the State, through counsel, expressly waives the requirement.").

Therefore, our focus properly remains fixed on whether Mr. Grant satisfied his burden to fairly present the argument to the OCCA. See Picard , 404 U.S. at 276, 92 S.Ct. 509 ("[W]e have required a state prisoner to present the state courts with the same claim he urges upon the federal courts.") (emphasis added). And Mr. Grant cannot rely on the hypervigilence or extraordinary *900circumspection of others to demonstrate his satisfaction of this burden. See Baldwin v. Reese , 541 U.S. 27, 32, 124 S.Ct. 1347, 158 L.Ed.2d 64 (2004) (holding that "ordinarily a state prisoner does not 'fairly present' a claim to a state court if that court must read beyond a petition or a brief (or a similar document) that does not alert it to the presence of a federal claim in order to find material ... that does so"). In Baldwin , the Supreme Court rejected the proposition that a petitioner could establish fair presentation as to the Oregon Supreme Court simply by relying on the fact that a "lower state trial court" had picked up on his federal constitutional violation, and that the Oregon Supreme Court "had had 'the opportunity to read ... the lower [Oregon] court['s] decision.' " Id. at 30, 124 S.Ct. 1347. A similar logic applies here. That the State's attorney elected to respond to an ostensible procedural due process competency claim-whether out of hypervigilence, prophylactic intent, or simple misunderstanding-does not relieve Mr. Grant of his burden of demonstrating that he actually fairly presented such a claim to the OCCA. And we conclude that he has not carried that burden.7

Nor does Mr. Grant's reliance on Sanders v. United States , 373 U.S. 1, 16, 83 S.Ct. 1068, 10 L.Ed.2d 148 (1963), give us pause. Mr. Grant points to the district court's characterization of this question as a "close" one, see R., Vol. I, at 1578, and argues that, under Sanders , we should resolve close calls in his favor and thus conclude that he fairly presented a procedural due process competency argument to the OCCA, Aplt.'s Opening Br. at 32. In Sanders , the Supreme Court explained: "Should doubts arise in particular cases as to whether two grounds are different or the same, they should be resolved in favor of the applicant." 373 U.S. at 16, 83 S.Ct. 1068. In Mr. Grant's view, because it is a close call whether the competency argument in his direct-appeal brief was based on substantive due process alone or substantive and procedural due process, we should resolve the question in favor of a determination that Mr. Grant fairly presented a procedural due process claim to the OCCA. We disagree.

Even assuming arguendo that the fair-presentation issue was close, Mr. Grant's reliance on Sanders would be unavailing. Sanders arises in a distinct procedural context; it is not a fair-presentation case. The language Mr. Grant relies on appears in the Court's discussion of the principles that should determine whether prior and successive habeas petitions, or such motions under 28 U.S.C. § 2255, present the same ground for federal relief. See Sanders , 373 U.S. at 16, 83 S.Ct. 1068. In other words, this language was offering guidance in resolving a distinct problem, and Mr. Grant has not explained why it should apply in the fair-presentation/exhaustion context, and we are not aware of any reason that it should, especially given the unique federalism concerns at stake here. See Picard , 404 U.S. at 275, 92 S.Ct. 509 (noting that "it would be unseemly in our dual system of government for a federal ... court to upset a state court conviction without an opportunity to the state courts to correct a constitutional violation" (quoting Darr v. Burford , 339 U.S. 200, 204, 70 S.Ct. 587, 94 L.Ed. 761 (1950) )).

Indeed, Picard -which is a fair-presentation case-sheds useful light on Sanders . Picard describes Sanders as a "ready example" of one of those "instances in which *901'the ultimate question for disposition' will be the same despite variations in the legal theory or factual allegations urged in its support." Id. at 277, 92 S.Ct. 509 (quoting United States ex rel. Kemp v. Pate , 359 F.2d 749, 751 (7th Cir. 1966) ). And Picard then explains that the Sanders example was not applicable there because the claim that the petitioner pursued in federal court in Picard was not the "substantial equivalent" of the claim presented in state court. Id. at 278, 92 S.Ct. 509. This holds true here as well: the question for resolution in procedural and substantive competency claims is not the same, and it is clear to us that the competency claim that Mr. Grant presented in his direct-appeal brief to the OCCA is not a "substantial equivalent" of the claim he presented in his habeas petition. Id. In sum, Mr. Grant's reliance on Sanders is misplaced.

Thus, we reject Mr. Grant's arguments opposing our conclusion that he failed to present a procedural due process competency claim to the OCCA.

d

All that said, we deem Mr. Grant's procedural due process competency claim unexhausted and-with an eye toward Oklahoma law-this claim is seemingly subject to an anticipatory procedural bar. Specifically, if Mr. Grant attempted to pursue this procedural competency claim in state court, that court would deem the claim procedurally barred under Oklahoma law because Mr. Grant could have raised it on direct appeal. Under Oklahoma's Uniform Post-Conviction Procedure Act, "only claims which '[w]ere not and could not have been raised' on direct appeal will be considered [in post-conviction proceedings]." Conover v. State , 942 P.2d 229, 230-31 (Okla. Crim. App. 1997) (quoting OKLA. STAT. TIT . 22, § 1089(c)(1) ); see also James v. Gibson , 211 F.3d 543, 550 (10th Cir. 2000) (citing cases in which the OCCA applied Section 1089(c)(1) to competency claims not raised on direct appeal); Walker v. State , 933 P.2d 327, 338-39 (Okla. Crim. App. 1997), superceded by statute on other grounds as recognized in Davis v. State , 123 P.3d 243, 245 (Okla. Crim. App. 2005) (holding capital petitioner's competency claim procedurally barred because he failed to raise the issue on direct appeal).8 While it is true that "[t]o preclude federal habeas review, a state procedural bar must be adequate," Mr. Grant makes no arguments relating to the adequacy of Oklahoma's procedural default rule, much less mount a challenge to the propriety of applying it here.9 Spears v. Mullin , 343 F.3d 1215, 1251 (10th Cir. 2003).

Therefore, we have no reason to question the rule's application to him. See id. at 1252 ("[B]ecause Powell does not address his alleged procedural default, let alone challenge the adequacy of Oklahoma's procedural rules, we conclude that Oklahoma's procedural bar is adequate to preclude our *902habeas review of these particular ineffective-trial-representation claims."). And so we hold that Mr. Grant's claim is subject to an anticipatory procedural bar. Cf. Thacker , 678 F.3d at 841 ("Were Thacker to now return to state court to attempt to exhaust a claim that trial counsel was ineffective in advising him to enter a blind plea and in failing to file a motion to withdraw the guilty plea, by filing a fourth application for post-conviction relief, it would be procedurally barred under Oklahoma law because Thacker failed to assert it in any of his applications for post-conviction relief."); Cummings v. Sirmons , 506 F.3d 1211, 1222-23 (10th Cir. 2007) ("Although the claim is technically unexhausted, it is beyond dispute that, were Cummings to attempt to now present the claim to the Oklahoma state courts in a second application for post-conviction relief, it would be deemed procedurally barred.").

Furthermore, Mr. Grant makes no effort to overcome this bar by arguing cause and prejudice, or a fundamental miscarriage of justice. Consequently, we hold that we are precluded from considering Mr. Grant's procedural due process competency claim. See Coleman , 501 U.S. at 750, 111 S.Ct. 2546 ; see also Thacker , 678 F.3d at 841-42 ("The only way for Thacker to circumvent this anticipatory procedural bar is by making either of two alternate showings: he may demonstrate 'cause and prejudice' for his failure to raise the claim in his initial application for post-conviction relief, or he may show that failure to review his claim will result in a 'fundamental miscarriage of justice.' " (quoting Anderson , 476 F.3d at 1140 )).10

B. Ineffective Assistance of Counsel Claims

Mr. Grant argues that his trial counsel rendered ineffective assistance resulting in an "unfair trial and an unreliable death sentence in violation of the Sixth, Eighth, and Fourteenth Amendments." Aplt.'s Opening Br. at 50 (capitalization altered). We granted COAs regarding the following issues: whether trial counsel rendered ineffective assistance by failing to (1) monitor Mr. Grant's competency, (2) investigate and present evidence of the effects of Mr. Grant's frontal-lobe damage (i.e., organic *903brain damage), and (3) investigate and present evidence of (a) Mr. Grant's purported delusional belief system and (b) pertinent aspects of Mr. Grant's childhood. Because Mr. Grant has not shown that the OCCA's resolution of his ineffective-assistance claims is contrary to or an unreasonable application of clearly established federal law, or premised on an unreasonable determination of fact, we affirm the district court's denial of habeas relief regarding Mr. Grant's ineffective-assistance claims.

1. Legal Framework

We review claims of "ineffective assistance of counsel under the familiar framework laid out in Strickland [v. Washington , 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984) ]." Byrd , 645 F.3d at 1167. Under Strickland , a petitioner "must show both that his counsel's performance 'fell below an objective standard of reasonableness' and that 'the deficient performance prejudiced the defense.' " Id. (emphasis omitted) (quoting Strickland , 466 U.S. at 687-88, 104 S.Ct. 2052 ). "These two prongs may be addressed in any order, and failure to satisfy either is dispositive." Victor Hooks II , 689 F.3d at 1186 ; see Littlejohn v. Royal ("Littlejohn II "), 875 F.3d 548, 552 (10th Cir. 2017) ("These two prongs may be addressed in any order; indeed, in Strickland , the Supreme Court emphasized that 'if it is easier to dispose of an ineffectiveness claim on the ground of lack of sufficient prejudice, ... that course should be followed.' " (omission in original) (quoting Strickland , 466 U.S. at 697, 104 S.Ct. 2052 )).

"[O]ur review of counsel's performance under the first prong of Strickland is a 'highly deferential' one." Byrd , 645 F.3d at 1168 (quoting Danny Hooks v. Workman , 606 F.3d 715, 723 (10th Cir. 2010) ). "Every effort must be made to evaluate the conduct from counsel's perspective at the time." Littlejohn I , 704 F.3d at 859 (quoting United States v. Challoner , 583 F.3d 745, 749 (10th Cir. 2009) ). "[C]ounsel is strongly presumed to have rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment." Victor Hooks II , 689 F.3d at 1187 (quoting Byrd , 645 F.3d at 1168 ). And the "petitioner 'bears a heavy burden' when it comes to overcoming that presumption." Byrd , 645 F.3d at 1168 (quoting Fox v. Ward , 200 F.3d 1286, 1295 (10th Cir. 2000) ). "To be deficient, the performance must be outside the wide range of professionally competent assistance. In other words, it must have been completely unreasonable, not merely wrong." Danny Hooks , 606 F.3d at 723.

"A state prisoner in the § 2254 context faces an even greater challenge." Victor Hooks II , 689 F.3d at 1187 (citing Byrd , 645 F.3d at 1168 ). "[W]hen assessing a state prisoner's ineffective-assistance-of-counsel claims on habeas review, '[w]e defer to the state court's determination that counsel's performance was not deficient and, further, defer to the attorney's decision in how to best represent a client.' " Id. (alterations in original) (quoting Byrd , 645 F.3d at 1168 ). "Thus our review of ineffective-assistance claims in habeas applications under § 2254 is 'doubly deferential.' " Id. ; Harrington , 562 U.S. at 105, 131 S.Ct. 770 ("The standards created by Strickland and § 2254(d) are both 'highly deferential,' and when the two apply in tandem, review is 'doubly' so ...." (citations omitted) (quoting first Strickland , 466 U.S. at 689, 104 S.Ct. 2052, and then Knowles v. Mirzayance , 556 U.S. 111, 123, 129 S.Ct. 1411, 173 L.Ed.2d 251 (2009) )).

"Federal habeas courts must guard against the danger of equating unreasonableness under Strickland with unreasonableness under § 2254(d). When *904§ 2254(d) applies, the question is not whether counsel's actions were reasonable. The question is whether there is any reasonable argument that counsel satisfied Strickland 's deferential standard." Harrington , 562 U.S. at 105, 131 S.Ct. 770 (emphasis added). And "because the Strickland standard is a general standard, a state court has ... more latitude to reasonably determine that a defendant has not satisfied that standard." Byrd , 645 F.3d at 1168 (emphasis added) (quoting Knowles , 556 U.S. at 123, 129 S.Ct. 1411 ); accord Frost v. Pryor , 749 F.3d 1212, 1223 (10th Cir. 2014) ; see also Harrington , 562 U.S. at 105, 131 S.Ct. 770 ("The Strickland standard is a general one, so the range of reasonable applications is substantial."); Acosta v. Raemisch , 877 F.3d 918, 925 (10th Cir. 2017) (noting that under AEDPA "our inquiry is informed by the specificity of the governing rule").

Despite our strong presumption that counsel rendered constitutionally reasonable assistance, "we have recognized a need to apply ... closer scrutiny when reviewing attorney performance during the sentencing phase of a capital case." Cooks , 165 F.3d at 1294 ; see also Osborn v. Shillinger , 861 F.2d 612, 626 n.12 (10th Cir. 1988) ("[T]he minimized state interest in finality when resentencing alone is the remedy, combined with the acute interest of a defendant facing death, justify a court's closer scrutiny of attorney performance at the sentencing phase."). "We judge counsel's performance by reference to 'prevailing professional norms,' which in capital cases include the ABA Guidelines for the Appointment and Performance of Counsel in Death Penalty Cases ('ABA Guidelines')." Victor Hooks II , 689 F.3d at 1201 (quoting Young v. Sirmons , 551 F.3d 942, 957 (10th Cir. 2008) ). "Among the topics defense counsel should investigate and consider presenting include medical history, educational history, employment and training history, family and social history, prior adult and juvenile correctional experiences, and religious and cultural influences." Young , 551 F.3d at 957.

"Counsel has a duty to conduct a 'thorough investigation-in particular, of mental health evidence-in preparation for the sentencing phase of a capital trial.' " Victor Hooks II , 689 F.3d at 1201 (quoting Michael Wilson v. Sirmons (" Michael Wilson I "), 536 F.3d 1064, 1083 (10th Cir. 2008) ); accord Littlejohn I , 704 F.3d at 860. "[D]rawing on a trilogy of Supreme Court cases- [Terry]Williams v. Taylor , 529 U.S. 362, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000), Wiggins v. Smith , 539 U.S. 510, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003), and Rompilla v. Beard , 545 U.S. 374, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005) -involving ineffective assistance at capital-sentencing proceedings[,]" Littlejohn I , 704 F.3d at 860, we divined the following three principles:

First, the question is not whether counsel did something ; counsel must conduct a full investigation and pursue reasonable leads when they become evident. Second, to determine what is reasonable investigation, courts must look first to the ABA guidelines, which serve as reference points for what is acceptable preparation for the mitigation phase of a capital case. Finally, because of the crucial mitigating role that evidence of a poor upbringing or mental health problems can have in the sentencing phase, defense counsel must pursue this avenue of investigation with due diligence. Our own Circuit has emphasized this guiding principle. In Smith v. Mullin , 379 F.3d 919, 942 (10th Cir. 2004), we held that it was "patently unreasonable" for trial counsel to fail to present evidence of Smith's borderline mental retardation, brain damage, and troubled childhood, *905and stated that this type of mitigating evidence "is exactly the sort of evidence that garners the most sympathy from jurors."

Michael Wilson I , 536 F.3d at 1084-85 (citations omitted); accord Littlejohn I , 704 F.3d at 860.

"Under the prejudice prong [of Strickland ], a petitioner must demonstrate 'a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different.' " Littlejohn II , 875 F.3d at 552 (quoting Strickland , 466 U.S. at 694, 104 S.Ct. 2052 ); accord Victor Hooks II , 689 F.3d at 1202. "When a petitioner alleges ineffective assistance of counsel stemming from a failure to investigate mitigating evidence at a capital-sentencing proceeding, 'we evaluate the totality of the evidence' " that AEDPA permits us to consider. Jeremy Williams , 782 F.3d at 1215 ; accord Littlejohn II , 875 F.3d at 552-53 ; see Cullen , 563 U.S. at 181, 131 S.Ct. 1388 ("We now hold that review under § 2254(d)(1) is limited to the record that was before the state court that adjudicated the claim on the merits.").

More specifically, "we 'reweigh the evidence in aggravation against the totality of available mitigating evidence,' considering 'the strength of the State's case and the number of aggravating factors the jury found to exist, as well as the mitigating evidence the defense did offer and any additional mitigating evidence it could have offered[.']" Littlejohn II , 875 F.3d at 553 (quoting first Hooks , 689 F.3d at 1202, and then Knighton v. Mullin , 293 F.3d 1165, 1178 (10th Cir. 2002) ). Furthermore, "we must consider not just the mitigation evidence that Defendant claims was wrongfully omitted, but also what the prosecution's response to that evidence would have been." Michael Wilson v. Trammell (" Michael Wilson II "), 706 F.3d 1286, 1306 (10th Cir. 2013) ; accord Littlejohn II , 875 F.3d at 553 ; see Michael Wilson II , 706 F.3d at 1305 ("To resolve whether there was prejudice, we do not consider omitted mitigation evidence in a vacuum.").

"If there is a reasonable probability that at least one juror would have struck a different balance, ... prejudice is shown." Littlejohn I , 704 F.3d at 861 (quoting Victor Hooks II , 689 F.3d at 1202 ). Put another way, in the capital-sentencing context, if the petitioner demonstrates that there is a reasonable probability "that at least one juror would have refused to impose the death penalty," the petitioner has successfully shown prejudice under Strickland . Victor Hooks II , 689 F.3d at 1202 (quoting Michael Wilson I , 536 F.3d at 1124 (Hartz, J., concurring)); accord Littlejohn II , 875 F.3d at 553.

2. Misstatement of Strickland Standard

We first address Mr. Grant's overarching claim that the OCCA's rejection of his ineffective-assistance claims is contrary to clearly established federal law because the OCCA applied the incorrect legal framework-that is, the OCCA failed to apply Strickland 's well-established rubric. When setting forth "the legal framework for evaluating [Mr. Grant's] ineffective-assistance claims," the OCCA stated that "[Mr. Grant] must demonstrate that trial counsel's performance was so deficient as to have rendered [him], in essence, without counsel." Grant , 205 P.3d at 22. This statement of law, in Mr. Grant's view, "placed an extra burden on him which was contrary to law." Aplt.'s Opening Br. at 83. Mr. Grant thus argues that we must apply de novo review to his ineffective-assistance claims. We disagree.

On habeas review, we properly eschew the role of strict English teacher, *906finely dissecting every sentence of a state court's ruling to ensure all is in good order. Cf. Renico v. Lett , 559 U.S. 766, 773, 130 S.Ct. 1855, 176 L.Ed.2d 678 (2010) (noting that "AEDPA imposes a 'highly deferential standard for evaluating state-court rulings[ ]' " (quoting Lindh v. Murphy , 521 U.S. 320, 333 n.7, 117 (S.Ct. 2059, 138 L.Ed.2d 481 1997) )); Michael Williams , 529 U.S. at 411, 120 S.Ct. 1495 ("Congress specifically used the word 'unreasonable,' and not a term like 'erroneous' or 'incorrect.' "). Rather, our focus is on the reasonableness of the state court's decision -viz. , whether that decision is contrary to or an unreasonable application of clearly established federal law or based on an unreasonable determination of the facts.

Consequently, our inquiry relates to the overall substance of the state court's analysis and the conclusion it thereafter makes. The Supreme Court has emphasized § 2254's focus on the decision of the state court: "Avoiding [ § 2254's] pitfalls does not require [a state court's] citation of [Supreme Court] cases-indeed, it does not even require [a state court's] awareness of [Supreme Court] cases, so long as neither the reasoning nor the result of the state-court decision contradicts them." Early v. Packer , 537 U.S. 3, 8, 123 S.Ct. 362, 154 L.Ed.2d 263 (2002) (emphases added).

Viewed through this proper prism, there is no occasion here for us to apply de novo review based on the OCCA's language in a single sentence. Admittedly, that language-especially, the "rendered without counsel" phrase-deviates from the proper formulation of the Strickland standard. Cf. United States v. Cronic , 466 U.S. 648, 659, 104 S.Ct. 2039, 80 L.Ed.2d 657 (1984) ("[I]f counsel entirely fails to subject the prosecution's case to meaningful adversarial testing, then there has been a denial of Sixth Amendment rights that makes the adversary process itself presumptively unreliable."). But that is not material. The true question presented by Mr. Grant's challenge is simply whether the overall substance of the OCCA's analysis, as well as the result it reached, reflects that the court understood and decided the ineffective-assistance issue under the proper Strickland framework. And we answer that question with no difficulty in the affirmative.

Virtually in the same breath-indeed, in the same paragraph on the same page as the language that Mr. Grant identified-the OCCA articulated the proper rubric of Strickland : it stated with respect to the performance prong-"We assess counsel's performance for reasonableness in light of prevailing professional norms"-and regarding the prejudice prong-"Appellant must also demonstrate that the allegedly deficient performance caused prejudice." Grant , 205 P.3d at 22. And the OCCA never repeated in its Strickland discussion the "rendered without counsel" linguistic formulation. Therefore, we have no difficulty concluding that the OCCA's analysis reflects that it understood and resolved the ineffective-assistance issue under the proper Strickland framework. We consequently reject Mr. Grant's call for the application of de novo review based on the (admittedly inaccurate) wording of a single sentence.

3. Failure to Monitor Competency

Mr. Grant argues that he is entitled to habeas relief based on the OCCA's rejection of his ineffective-assistance claim asserting that trial counsel's alleged unreasonable failure to monitor his competency-or, more precisely, his purported decline into incompetency-prejudiced him. The OCCA summarized the nature of Mr. Grant's argument:

*907[Mr. Grant] submits that in the months between the competency trial and the trial on guilt and punishment, his competency may well have deteriorated. He points to his statements at various pretrial and in camera hearings, pro se writings, and his testimony in the punishment stage of the trial in an attempt to support this claim. He also submits extra-record evidence to support a related claim, based on the Sixth Amendment right to counsel, that trial counsel was deficient for not challenging his competency at the time of trial . Specifically, he presents (1) an expert's retrospective opinion, based on evaluation of various materials, that [Mr. Grant] was not competent to stand trial in November 2005; and (2) documentary evidence suggesting that in mid-2005, Mr. Grant was not diligent about taking medications prescribed to treat his mental illness.

Grant , 205 P.3d at 8 (emphasis added).

Regarding his pretrial statements, Mr. Grant highlights certain comments that he made during a hearing in May 2005, when he waived a possible conflict of interest involving one of his attorneys. After telling the court that he "underst[ood] conflict of interest" and was prepared to "fire all staff and represent myself once I feel that honestly [sic] matters are being taken out of proportion meaning I'm being plotted against with the DA," Mr. Grant offered his "theory" in response to the court's assurances that no such plotting was taking place: "My theory plays my whole background. That's for one. My way of life is I'm going to leave this planet earth. That's my theory. My theory I stand on it and it don't have nothing to do with this. My theory is my theory, you see what I'm saying." Mot. Hr'g Tr. at 7-8 (dated May 2, 2005). However, in response to follow-up questioning, Mr. Grant indicated that he was not accusing the potentially conflicted attorney of plotting with the District Attorney, and his other attorney advised the court-without objection by Mr. Grant-that when Mr. Grant was speaking about such plotting, he was simply "expressing concern that lawyers sometimes do that." Id. at 9.

Mr. Grant also draws our attention to two letters that he wrote in September 2005 to the judge and prosecutor, respectively. The first letter-which Mr. Grant calls the "Eye" letter-refers to "electrons," "eye," "God," and "Allah," among other things, R., Vol. I, at 782-84, and the author of Mr. Grant's retrospective competency hearing-Dr. Antoinette McGarrahan, PhD-subsequently suggested that it "revealed incoherent ramblings and religious and grandiose delusions," id. at 771. The second letter (directed to the prosecutor) amounted to a confession to the charged crimes which Mr. Grant wrote with the apparent hope of securing the release of another inmate whom Mr. Grant described as "a good man from the heart." Id. at 786.

The OCCA's ineffective-assistance analysis implicated Strickland 's first prong- viz. , the performance prong. Considering the extra-record materials that Mr. Grant specifically marshaled with respect to this claim, the court held that the materials were "insufficient to overcome the presumption that trial counsel had a sound basis for believing [Mr. Grant] was competent at the time of trial." Grant , 205 P.3d at 10. And the OCCA offered the following comments regarding the underlying issue of competency:

[W]e find no reason to second-guess the judgment of those parties most familiar with [Mr. Grant's] history of mental problems before and during the trial-defense counsel, the trial court, and the defense experts retained at that time. The record supports a conclusion that *908[Mr. Grant] was competent at the time of his trial.

Id.

In resolving this claim, we first address below Mr. Grant's contentions that certain specific legal and factual errors are embedded in the OCCA's analysis. Concluding that these contentions are without merit, we then turn to the substantive Strickland question. We determine that Mr. Grant's ineffective-assistance claim based on trial counsel's alleged failure to monitor his purported decline into incompetency fails under Strickland 's second prong-that is, on the issue of prejudice. Accordingly, we affirm the district court's denial of habeas relief as to this claim.

a

First, Mr. Grant argues that the OCCA's rejection of his failure-to-monitor claim was contrary to clearly established federal law because it rested on a determination of competence that was legally flawed. The OCCA rejected Mr. Grant's failure-to-monitor claim in significant part because it determined that Mr. Grant was in fact competent when tried. Mr. Grant argues that the OCCA's competency determination was legally flawed because it addressed only one prong of the two-pronged test for competency. We disagree.

"The [two-pronged] test for incompetence is ... well settled. A defendant may not be put to trial unless he 'has [ (1) ] sufficient present ability to consult with his lawyer with a reasonable degree of rational understanding ... [and (2) ] a rational as well as factual understanding of the proceedings against him.' " Cooper v. Oklahoma , 517 U.S. 348, 354, 116 S.Ct. 1373, 134 L.Ed.2d 498 (1996) (quoting Dusky v. United States , 362 U.S. 402, 402, 80 S.Ct. 788, 4 L.Ed.2d 824 (1960) (per curiam)); see also Drope , 420 U.S. at 171, 95 S.Ct. 896 ("It has long been accepted that a person whose mental condition is such that he lacks the capacity to understand the nature and object of the proceedings against him, to consult with counsel, and to assist in preparing his defense may not be subjected to a trial." (emphasis added)).

Mr. Grant argues that the OCCA's competence determination is legally flawed because it addressed only the second prong of the competency standard-i.e., whether Mr. Grant understood the proceedings-and, therefore, the OCCA's resultant rejection of his failure-to-monitor claim is contrary to clearly established federal law. Our review of the OCCA's direct-appeal opinion, however, reveals that the OCCA understood and also sufficiently considered the first prong of the competence standard-i.e., whether Mr. Grant was able to assist counsel in preparing his defense.

The OCCA clearly recognized that the competence standard had two components and explicitly set them out. See Grant , 205 P.3d at 8 ("Under Oklahoma law, a person is competent to stand trial if he has the present ability to understand the nature of the charges and proceedings brought against him and to rationally assist in his own defense. These standards are consistent with federal constitutional requirements." (emphasis added) (citations omitted, including to the Supreme Court's seminal decisions in Cooper and Drope )).

Furthermore, the OCCA's reasoning also reflects its awareness of the first prong of the standard and application of it. For example, the court reasoned that Mr. Grant was able to make decisions regarding his defense; during the pre-trial phase, for instance, the trial court "had several discussions with [Mr. Grant] about various matters" and "these exchanges show[ed that Mr. Grant] had a rather keen understanding *909of the legal process, and ... was able to make important decisions." Id. at 9.

In short, we are confident that all fairminded jurists would not agree that the OCCA misunderstood the appropriate federal standard for competency or misapplied it. Its decision in this regard was not contrary to or an unreasonable application of clearly established federal law.

b

Mr. Grant next argues that the OCCA's rejection of his failure-to-monitor claim and its related rejection of Mr. Grant's request for an evidentiary hearing were premised on an unreasonable determination of several facts. Specifically, he claims that the OCCA unreasonably determined that (1) Mr. Grant "had a rather keen understanding of the legal process ... [and] was able to make important decisions," Grant , 205 P.3d at 9 ; (2) Mr. Grant had an awareness and understanding of the "ramifications" of the "admission" in his confession letter, id. ; and (3) Mr. Grant's purported "delusions" were actually "related to an unconventional philosophy, or religion of sorts," id. at 9 n.6.

We conclude that Mr. Grant has not preserved these three arguments for appellate review because he failed to raise them in his habeas petition. See Owens v. Trammell , 792 F.3d 1234, 1246 (10th Cir. 2015) ("Because the argument was not raised in his habeas petition, it is waived on appeal."); Stouffer v. Trammell , 738 F.3d 1205, 1222 n.13 (10th Cir. 2013) ("We do not generally consider issues that were not raised before the district court as part of the habeas petition."); Parker v. Scott , 394 F.3d 1302, 1327 (10th Cir. 2005) (deeming waived certain ineffective-assistance claims where petitioner "fail[ed] to assert them in his district court habeas petition"); see also Hancock v. Trammell , 798 F.3d 1002, 1021-22 ("But in the habeas petition, Mr. Hancock did not present this allegation as a separate basis for habeas relief. As a result, this issue has been forfeited." (footnote omitted)). Accordingly, we do not reach the merits of them.

The third alleged unreasonable factual determination warrants a brief discussion.11 Mr. Grant argues that the OCCA unreasonably determined that the " 'delusions' found by experts in their diagnoses of schizophrenia were not delusions, but 'related to an unconventional philosophy, or religion of sort.' " Aplt.'s Opening Br. at 92 (quoting Grant , 205 P.3d at 9 n.6 ). Mr. Grant asserts that this factual determination was unreasonable in light of the OCCA's contrasting determination that "experts tended to agree that [Mr. Grant] had ... a form of schizophrenia." Id. (quoting Grant , 205 P.3d at 8 ). As Mr. Grant sees things, these statements reflect "internally inconsistent determinations" and are, therefore, unreasonable. Id.

The State contends that Mr. Grant waived this argument. In the State's view, *910Mr. Grant "never argued that the OCCA's findings were unreasonable because it made inconsistent findings." Aplee.'s Br. at 51 n.9 (emphasis added). We conclude that Mr. Grant has failed to preserve this argument for appellate review. To be sure, Mr. Grant did contend that the OCCA made an unreasonable determination of facts bearing on his delusions: "The OCCA also made an unreasonable determination of the facts in concluding comments Mr. Grant made in writings, colloquies with the court, and in his trial testimony are not 'delusions that sprang [from Mr. Grant's] own mind [because] they related to an unconventional philosophy, or religion of sorts.' " R., Vol. I, at 561 (quoting Grant , 205 P.3d at 9 n.6 ). And Mr. Grant bolstered this assertion by noting that the OCCA's finding was contrary to the evidentiary findings of others : "The wealth of evidence from all the experts indicate that his grandiose delusions, which have a religious element, are a significant symptom of his mental illness and not merely an unconventional philosophy." Id. In other words, Mr. Grant contrasted the OCCA's factual findings with the allegedly abundant contrary evidentiary findings of others and, on this basis, declared the OCCA's findings unreasonable.

However, at no point did Mr. Grant contend that the OCCA's findings were unreasonable because they were at odds with themselves -i.e., internally inconsistent. More specifically, in seeking to establish their unreasonableness, Mr. Grant did not compare one set of OCCA findings with another. But this logically is what Mr. Grant would have done if he were attempting to demonstrate that the unreasonableness of the OCCA's findings was based on their internal inconsistency. And, not surprisingly, this is precisely the line of argument that Mr. Grant pursues on appeal.

We do not understand the State-as Mr. Grant does-to be "essentially assert[ing] that by not using the word 'inconsistent' ... Mr. Grant waived the argument." Aplt.'s Reply Br. at 18. That is because both the totality of the relevant language and structure of Mr. Grant's arguments in his habeas petition make patent to us that he did not advance this internal-inconsistency ground for declaring the OCCA's factual findings unreasonable in his habeas petition. Accordingly, he has failed to preserve this contention for appellate review.

c

Having rejected Mr. Grant's specific contentions of legal and factual error, we turn to the merits of his ineffective-assistance claim based on trial counsel's alleged failure to monitor his competency. We conclude that Mr. Grant cannot prevail on this ineffective-assistance claim under Strickland 's (second) prejudice prong. That is, Mr. Grant cannot demonstrate that there is a reasonable probability that the result of the proceeding would have been different but for counsel's failure to monitor Mr. Grant's competency-or, as Mr. Grant views it, his "slide into incompetency." R., Vol. I, at 582.

It is undisputed-as the district court found-that the OCCA did not "expressly address [ Strickland 's ] second prong."12 Id. at 1584. The parties joust about whether we should apply AEDPA deference or de novo review in our consideration of the prejudice prong. Advocating for de novo review, Mr. Grant has the better of this argument-but only up to a point. Because the OCCA did not-by the plain terms of its ruling-reach the prejudice question, we resolve this overarching question de novo. See, e.g. , Wiggins , 539 U.S. at 534, 123 S.Ct. 2527 ("In this case, our review is not circumscribed by a state court conclusion with respect to prejudice, as neither of the state courts below reached this prong of the Strickland analysis."); Rompilla , 545 U.S. at 390, 125 S.Ct. 2456 (noting that "[b]ecause the state courts found the representation adequate, they never reached the issue of prejudice, and so we examine this element of the Strickland claim de novo" (citation omitted)); Victor Hooks II , 689 F.3d at 1188 ("[I]n those instances where the OCCA did not address the performance prong of Strickland and we elect to do so, our review is de novo."). But cf. McBride v. Superintendent, SCI Houtzdale , 687 F.3d 92, 100 n.10 (3d Cir. 2012) (noting some "possible tension between" language in Harrington regarding the adjudication of habeas claims and the approach of Wiggins and its progeny where a portion of a Strickland claim is not reached by a state court, and collecting cases).

*911However, in resolving the merits of Strickland 's first prong (i.e., performance), the OCCA made a related, but distinct and independent, merits determination regarding Mr. Grant's substantive competency. See Grant , 205 P.3d at 10. We are obliged to take this ruling into account in our analysis of Strickland 's prejudice prong. And it seemingly can be determinative regarding whether Mr. Grant can prevail under that prong. For instance, commenting on what constitutionally effective counsel would have done for him, Mr. Grant states:

Reasonably effective counsel, with a client whose competency was as mercurial as Grant's, would have investigated, checked medication records, interviewed the mental health professionals at [the Oklahoma County Detention Center] who were treating him daily, challenged Grant's mental capacity to make a written confession, and heeded the signs of decompensation noted by their own expert.

Aplt.'s Opening Br. at 64. But, even assuming arguendo the performance of Mr. Grant's trial counsel was constitutionally deficient for failing to take such steps-and, significantly, for failing to seek a second competency trial-if Mr. Grant was actually competent , their unconstitutional performance would not have prejudiced him. More specifically, any motion that Mr. Grant's purportedly effective counsel would have filed for a second competency proceeding would have been properly denied.

AEDPA's deferential standards are appropriately applied to our review of the OCCA's distinct and independent merits determination of Mr. Grant's competency. And this is true even though-as Mr. Grant urges-we conduct a de novo review of the overarching and related question of whether Mr. Grant was prejudiced by trial counsel's assumed unconstitutional performance. Compare Spears , 343 F.3d at 1249-50 (using de novo review of petitioner's ineffective-assistance claim involving counsel's failure to object to a flight-from-the-crime instruction, but seemingly *912applying a deferential reasonableness standard of AEDPA to the OCCA's related determination of a mixed question of law and fact, specifically, that "the State had presented sufficient evidence to support giving the flight instruction"), with Victor Hooks v. Ward (" Victor Hooks I "), 184 F.3d 1206, 1223 (10th Cir. 1999) (applying de novo review where "the Oklahoma courts never considered Hooks' federal constitutional claim with regard to his requested instructions on lesser included offenses" but according AEDPA's presumption of correctness where "the Oklahoma Court of Criminal Appeals made some factual determinations that may bear on this issue " (emphasis added)).

The question then becomes what AEDPA standards govern this question: viz. , AEDPA's standards pertaining to issues of fact (notably, § 2254(e)(1) )13 or those relating to law-dependent mixed questions of fact and law (i.e., § 2254(d)(1) ). Precedent from the Supreme Court and our court at least strongly suggests that, in determining that Mr. Grant was substantively competent, the OCCA resolved "a factual issue" that "shall be presumed to be correct"; Mr. Grant would thus bear the burden of rebutting that presumption "by clear and convincing evidence." 28 U.S.C. § 2254(e)(1) ; see Thompson v. Keohane , 516 U.S. 99, 113, 116 S.Ct. 457, 133 L.Ed.2d 383 (1995) (noting that certain "practical considerations ... prompted the Court to type questions like ... competency" as "factual issues"); Demosthenes v. Baal , 495 U.S. 731, 735, 110 S.Ct. 2223, 109 L.Ed.2d 762 (1990) (per curiam) ("We have held that a state court's conclusion regarding a defendant's competency is entitled to such a presumption [i.e., of correctness]."); Spitzweiser-Wittgenstein v. Newton , 978 F.2d 1195, 1197 (10th Cir. 1992) ("Competency is a question of fact subject to the rebuttable presumption of correctness established in § 2254."); 1 Randy Hertz & James S. Liebman, FEDERAL HABEAS CORPUS PRACTICE & PROCEDURE § 20.3[d] at 1156 n.55 (noting that "[t]he lower federal courts are divided on the question of whether competency to stand trial is a factual issue entitled to a presumption of correctness or a mixed question of law and fact," and citing Tenth Circuit cases in the former camp).

The Supreme Court "typed" the competency issue as a factual one, even though it is not comprised solely of "simple historical fact[s]," Miller v. Fenton , 474 U.S. 104, 113-14, 106 S.Ct. 445, 88 L.Ed.2d 405 (1985), but, rather, is determined by applying legal standards to the subsidiary facts, see Thompson , 516 U.S. at 111, 116 S.Ct. 457 ("[T]he Court has classified as 'factual issues' within § 2254(d)'s compass [certain] questions extending beyond the determination of 'what happened.' This category notably includes: competency to stand trial ...."). The Thompson Court explained the "practical considerations" underlying this choice:

While these issues encompass more than "basic, primary, or historical facts," their *913resolution depends heavily on the trial court's appraisal of witness credibility and demeanor. This Court has reasoned that a trial court is better positioned to make decisions of this genre, and has therefore accorded the judgment of the jurist-observer "presumptive weight."

Id. at 100 (citation omitted) (citing Wainwright v. Witt , 469 U.S. 412, 429, 105 S.Ct. 844, 83 L.Ed.2d 841 (1985), and quoting Miller , 474 U.S. at 114, 106 S.Ct. 445 ).14

We recognize that the key cases cited supra - Thompson , Demosthenes , and its seminal Tenth Circuit progeny, Spitzweiser-Wittgenstein -were decided under a pre-AEDPA version of § 2254 that contained the presumption-of-correctness language that subsequently found a home in subsection (e)(1) of § 2254, with AEDPA's 1996 enactment. However, the presumption-of-correctness language before and after AEDPA is intended to effectuate federalism principles by "giv[ing] great weight to the considered conclusions of a coequal state judiciary," Miller , 474 U.S. at 112, 106 S.Ct. 445. Compare Thompson , 516 U.S. at 108, 116 S.Ct. 457 (quoting Miller , in discussing the federalism principles underlying the presumption of correctness in the pre-AEDPA regime), with Sharpe v. Bell , 593 F.3d 372, 379 (4th Cir. 2010) ("AEDPA in general and Section 2254(e) in particular were designed 'to further the principles of comity, finality, and federalism.' Section 2254(e)(1) plainly seeks to conserve judicial resources and reflects Congress's view that there is no reason for a do-over in federal court when it comes to facts already resolved by state tribunals. That section also reflects Congress's respect for principles of federalism, recognizing that a decision to set aside state court factual findings intrudes on the state's interest in administering its criminal law." (citation omitted) (quoting Michael Williams , 529 U.S. at 436, 120 S.Ct. 1479 )). Furthermore, we have no reason to believe that the "practical considerations" that Thompson cited, 516 U.S. at 113, 116 S.Ct. 457, do not remain in full effect in the post-AEDPA era. Cf. United States v. Mackovich , 209 F.3d 1227, 1232 (10th Cir. 2000) ("Competency to stand trial is a factual determination that can be set aside only if it is clearly erroneous." (quoting United States v. Boigegrain , 155 F.3d 1181, 1189 (10th Cir. 1998) )). Therefore, there is a strong suggestion in controlling precedent that competency should be treated as a factual issue in the habeas context-even in the post-AEDPA setting. If so, it would be subject to § 2254(e)(1)'s presumption of correctness.

*914As it turns out, however, we need not definitively determine whether subsection (e)(1)'s standard applies here. That is because whether this standard applies or, alternatively, AEDPA's standard pertaining to law-dependent mixed questions-specifically, the standard of subsection (d)(1), see Michael Williams , 529 U.S. at 398, 120 S.Ct. 1495 ; Gilson v. Sirmons , 520 F.3d 1196, 1233-34 (10th Cir. 2008) (holding that whether there is sufficient evidence to warrant giving a lesser-included-offense instruction is "a mixed question of law and fact and is thus reviewable under § 2254(d)(1)"); Cook v. McKune , 323 F.3d 825, 829-30 (10th Cir. 2003) (discussing Michael Williams and noting that § 2254(d)(1)"applies to errors of law and mixed questions of fact and law")-Mr. Grant cannot prevail. In the former scenario, § (e)(1)'s "standard is demanding but not insatiable." Miller-El v. Dretke (" Miller-El II "), 545 U.S. 231, 240, 125 S.Ct. 2317, 162 L.Ed.2d 196 (2005) ; accord Danny Hooks , 606 F.3d at 721 ; House , 527 F.3d at 1019. But Mr. Grant would be hard-pressed to satisfy it here because he makes no effort to take up the cudgel by making specific arguments under § 2254(e)(1)'s framework to rebut the presumption of correctness. In the latter scenario, Mr. Grant cannot demonstrate that the OCCA's determination of competency was contrary to or an unreasonable application of clearly established federal law. See, e.g. , Dusky , 362 U.S. at 402, 80 S.Ct. 788 ; Drope , 420 U.S. at 171, 95 S.Ct. 896. At the very least, "it is possible fairminded jurists could disagree" regarding whether the OCCA's competency determination conflicts with Supreme Court precedent. Harrington , 562 U.S. at 102, 131 S.Ct. 770.

Relevant to both scenarios, specifically, the OCCA considered the totality of the evidence, including the supplemental materials that Mr. Grant submitted with respect to his purported incompetency, and found "no reason to second-guess the judgment of those parties most familiar with [Mr. Grant's] history of mental problems before and during the trial-defense counsel, the trial court, and the defense experts retained at that time." Grant , 205 P.3d at 10. This approach was not unreasonable. And, as the OCCA explained, these sources uniformly did not express doubt regarding Mr. Grant's competency to stand trial. See id. at 9.

Notably, the OCCA stated:

[Mr. Grant's] two-lawyer defense team was experienced and zealous, considering that the overwhelming evidence against their client limited their options. At no time did either of them express doubts about their client's competency during the trial. Their chief mental-health expert, Dr. Grundy, who spent many hours interviewing [Mr. Grant] over the lengthy course of the prosecution, attended at least part of the trial and testified for the defense. Yet, there is no indication that Dr. Grundy had doubts about [Mr. Grant's] competency, either at that time or on reflection afterward.

Id.

The OCCA moreover called into question the adequacy of Mr. Grant's supplemental medical records that supposedly "suggest[ed] that in mid-2005, [Mr. Grant] was not diligent about taking medications prescribed to treat his mental illness." Id. at 8. Mr. Grant contends that these "psychiatric medications .... [were] required to keep him competent." Aplt.'s Opening Br. at 65. And Dr. McGarrahan's report underscored the point: "Most all of the mental health professionals who evaluated Mr. Grant, including court-ordered experts and State's experts agreed that Mr. Grant's competency was basically dependent *915upon ... his use of psychotropic medications." R., Vol. I, at 774.

Regarding the medication records, however, the OCCA stated:

The jail records regarding [Mr. Grant's] medication history do not warrant a different result. [Mr. Grant] claims these logs show that he was, at times, non-compliant in taking his prescribed medication. However, the affidavit accompanying these logs indicates that complete records for crucial time periods-particularly, most of September 2005, all of October 2005, and most of November 2005, when the trial was held-are missing or incomplete.

Grant , 205 P.3d at 10. In his appellate briefing, Mr. Grant does not dispute the OCCA's finding-based on the court's review of the submitted affidavit-that his medication records for a "key time period" leading up to trial were incomplete. Aplt.'s Opening Br. at 65. He simply responds that "[h]ad counsel been monitoring Grant's medication contemporaneously, the medication picture for October would be complete." Id. But that comment does nothing to undermine the OCCA's factual judgment regarding the inadequacy of his medication records to establish the alleged harmful pattern of Mr. Grant not taking his medication-or, relatedly, to call into question the OCCA's ultimate finding that Mr. Grant was competent.

Similarly, Mr. Grant also attacks vigorously the OCCA's reasoning regarding the competency implications of his decision to testify and to send his two September 2005 letters. But we discern nothing in these arguments that would undermine the presumption of correctness that would attach to any OCCA factual finding that Mr. Grant was competent, and at the very least, "it is possible fairminded jurists could disagree" about whether the OCCA's reasoning here regarding competency conflicts with Supreme Court precedent. Harrington , 562 U.S. at 102, 131 S.Ct. 770.

In this regard, the OCCA rejected Mr. Grant's argument that his decision to testify, despite trial counsel's contrary advice, militates in favor of a conclusion that he was not competent. See Grant , 205 P.3d at 9. Dr. McGarrahan had concluded that Mr. Grant was incompetent during his trial and sentencing, in part because "Mr. Grant testified in the punishment stage against counsel's advice." Id. The OCCA, however, generally refused to adopt Dr. McGarrahan's opinion because it was contrary to the opinions of the trial court, defense counsel, and Dr. Grundy-who each observed Mr. Grant leading up to and during trial and believed him to be competent. See id. Furthermore, the OCCA specifically reasoned that, insofar as Dr. McGarrahan's opinion rested on the fact that Mr. Grant testified over the contrary advice of counsel, it really implicated whether Mr. Grant's decision-making was wise , not necessarily whether he was competent to make such decisions. See id. More specifically, the OCCA explained that the wisdom, or lack of wisdom, of Mr. Grant's decisions is not dispositive of the competency inquiry: "Just as we do not judge counsel's effectiveness solely by the success of their strategies, we refuse to judge a defendant's competency solely by the wisdom of his own choices." Id.

The OCCA also did not shy away from addressing whether Mr. Grant's September 2005 letters evinced that he was incompetent. Relevant to the so-called EYE letter, the OCCA reasoned:

[Mr. Grant] points to several cryptic comments in his writings , in his colloquies with the court, and in his trial testimony, as evidence that he did not understand the nature of the proceedings. But these comments were not delusions that sprang from [Mr. Grant's]

*916own mind. They related to an unconventional philosophy, or religion of sorts, that [Mr. Grant] adhered to, similar in some respects to the Black Muslim or Nation of Islam movements, and known variously as "The Nation of Gods and Earths" or "The Five Percenters." This set of beliefs is not uncommon among inmates in the Northeastern United States, where [Mr. Grant] had grown up and been incarcerated.

Id. at 9 n.6. Moreover, in the same vein as its comments regarding Mr. Grant's decision to testify in the sentencing proceeding, the OCCA offered the following remarks regarding Mr. Grant's letter to the prosecutor:

The letter [Mr. Grant] addressed to the prosecutor shortly before trial, wherein he detailed his commission of the crimes, may not have been the most prudent course of action, but it does not show that he was unable to grasp the ramifications of such an admission. To the contrary, the letter indicates that [Mr. Grant] was fully aware of what he was doing.

Id. at 9. Notably, in the letter, Mr. Grant acknowledges that he may pay a "price" for confessing, insists that he is not "crazy" and explains that he is making the statement because he is "tired" and "want[s] to help someone." R., Vol. I, at 786-87.

Based on the foregoing, we conclude that Mr. Grant has not rebutted the presumption of correctness that attaches to the OCCA's arguably factual competency finding; or , alternatively, has not demonstrated that the OCCA's competency determination was contrary to or an unreasonable application of clearly established federal law regarding substantive competency. Consequently, we must accept this competency determination in our de novo consideration of whether Mr. Grant can prevail under Strickland 's prejudice prong.15 And we conclude that this competency *917determination sounds the death knell for Mr. Grant's ineffective-assistance claim because it undercuts his ability to satisfy this prong.

Specifically, Mr. Grant cannot establish that, but for his trial counsel's failure to monitor his alleged descent into incompetency, the result of his proceeding would have been different. For example, even if counsel had responded to Mr. Grant's seemingly strange behavior and requested a second competency trial, they would not have been successful because Mr. Grant was in fact competent. In other words, given Mr. Grant's competency, there is no reasonable probability that the court would have ruled favorably on such a motion. In sum, we conclude that Mr. Grant's ineffective-assistance claim based on the failure to monitor fails under Strickland 's prejudice prong. The district court reached a similar alternative holding. See R., Vol. I, 1584-85 ("[G]iven the OCCA's conclusion that petitioner was competent at the time of his trial, even assuming that his counsel's performance in allegedly failing to monitor petitioner's competence was deficient, petitioner cannot show that he was prejudiced by his counsel's failure or that fairminded jurists could not disagree that petitioner was prejudiced."). We uphold the district court's denial of habeas relief on this claim.

4. Failure to Investigate and Present Evidence of Organic Brain Damage

Mr. Grant's habeas petition asserts that his trial counsel was constitutionally ineffective for failing to investigate and present evidence of his frontal-lobe damage at the sentencing phase of his trial. Trial counsel's mitigation case focused on evidence of Mr. Grant's schizophrenia, and his dysfunctional family background. Trial counsel presented "red flags" suggestive of organic brain damage, but never offered definitive evidence that Mr. Grant suffered from organic brain damage.

To support his ineffective-assistance claim on direct appeal, Mr. Grant presented a neuropsychological evaluation, performed by clinical psychologist Dr. Michael M. Gelbort, PhD, in which Dr. Gelbort diagnosed Mr. Grant with organic brain damage-specifically, damage to the frontal lobes. Based on Dr. Gelbort's evaluation, Mr. Grant argued on direct appeal that trial counsel's failure to investigate and present evidence of organic brain damage deprived him of his constitutional right to effective assistance of counsel. The OCCA rejected this argument, concluding that trial counsel's performance was not constitutionally deficient and that Mr. Grant was not prejudiced by trial counsel's failure to investigate and present evidence of his organic brain damage.

Mr. Grant raised this claim in his habeas petition in the district court and, applying AEDPA deference, the district court found no error in the OCCA's denial of Mr. Grant's claim. Mr. Grant argues before us that the OCCA's rejection of his claim is contrary to and an unreasonable application of clearly established federal law and is premised on unreasonable factual determinations. To aid our analysis, we first pause to explicate the OCCA's adjudication of this particular claim. We then address Mr. Grant's arguments and ultimately conclude *918that he has not shown that the OCCA's denial of his claim was unreasonable or otherwise erroneous under AEDPA's standards.

a

In analyzing Mr. Grant's claim, the OCCA considered both the additional evidence that Mr. Grant argued trial counsel should have discovered and presented at the penalty phase of his trial-that is, as relevant here, Dr. Gelbort's report-as well as the mitigation evidence that was in fact presented to the jury. The OCCA characterized Dr. Gelbort's report as "linking some of [Mr. Grant's] mental deficits to an organic brain disorder, and concluding that these deficits appeared very early in [Mr. Grant's] life." Grant , 205 P.3d at 23. Next, the OCCA reviewed trial counsel's mitigation strategy: "Because the evidence of [Mr. Grant's] guilt was overwhelming, defense counsel focused on punishment." Id. The OCCA found that trial counsel "spent considerable time presenting [mitigation] evidence to the jury." Id. Specifically, trial counsel's "strategy was to present expert evidence on [Mr. Grant's] mental illness, and testimony from family members about his disadvantaged and dysfunctional childhood." Id. The OCCA stated that trial counsel's mitigation strategies "were by no means antagonistic" and concluded that "jurors might have found the circumstances surrounding [Mr. Grant's] formative years to have created, or at least aggravated, his mental problems." Id.

Based on the mitigation evidence before the jury, the OCCA reasoned that "the fact that [Mr. Grant] had some sort of mental illness was never in serious dispute" and, as a result, even though evidence of "organic brain disorder might have shed light on one potential cause of [Mr. Grant's] mental illness," there was not a reasonable probability that Dr. Gelbort's report "would have affected the jury's sentencing decision." Id. In particular, the OCCA opined that "the affidavits [Mr. Grant] submit[ted] on appeal do not present anything qualitatively different from what was presented at trial on these issues." Id. The OCCA further found that Mr. Grant failed to "overcome the strong presumption that his trial counsel performed competently," and denied his claim. Id.

b

Mr. Grant first argues that "the OCCA mischaracterized [his] claim as a failure of trial counsel to call several witnesses who could have testified to [Mr.] Grant's mental illness generically." Aplt.'s Opening Br. at 91 (emphasis omitted) (citing Grant , 205 P.3d at 23 ). Presumably, Mr. Grant is referring to the following statement of the OCCA: "The final complaint is defense counsel's failure to call several witnesses who could have testified about [Mr. Grant's] mental illness." Grant , 205 P.3d at 23. Mr. Grant asserts that this characterization was unreasonable because trial counsel's "deficient performance did not arise from a strategic decision not to call witnesses but from a failure to thoroughly investigate mitigating evidence in order to make reasonable strategic decisions about what witnesses to call." Aplt.'s Opening Br. at 91 (emphasis added).

As a threshold matter, it appears that Mr. Grant neglected to raise this argument in his habeas petition. Consequently, we could decline to consider it. See, e.g. , Parker , 394 F.3d at 1327. However, even if we saw fit to do so, see, e.g. , Abernathy v. Wandes , 713 F.3d 538, 552 (10th Cir. 2013) ("[T]he decision regarding what issues are appropriate to entertain on appeal in instances of lack of preservation is discretionary."), we would be hard-pressed to *919conclude that the OCCA mischaracterized Mr. Grant's claim.

Reading the statement at issue in context, it appears to be merely a shorthand reference to Mr. Grant's claim, rather than a mischaracterization. The OCCA's opinion explicitly introduced the claim as one involving "trial counsel's failure to investigate aspects of [Mr. Grant's] mental health." Grant , 205 P.3d at 23 (emphasis added). And the concepts of deficient investigation and deficient presentation are closely intertwined. Indeed, the OCCA reviewed the evidence that Mr. Grant alleges trial counsel would have discovered from a reasonable investigation and concluded that Mr. Grant was not prejudiced by the absence of this evidence before the sentencing jury. See id. In light of the foregoing, we are not persuaded that this stray line from the OCCA's opinion constitutes a misapprehension of Mr. Grant's claim.

Furthermore, as previously noted, our focus under AEDPA's deferential standard is on the reasonableness of a state court's decision -viz. , whether that decision is contrary to or an unreasonable application of clearly established federal law or based on an unreasonable determination of the facts-not on the unalloyed rectitude of each line of text of a state court's opinion. And Mr. Grant has not explained how this isolated line resulted in the OCCA performing an unreasonable analysis in contravention of AEDPA's standards. Consequently, on this basis alone, we would reject Mr. Grant's argument.

c

Mr. Grant argues that the OCCA's finding that he suffered no prejudice from trial counsel's failure to investigate and present evidence of his organic brain damage constitutes an unreasonable application of clearly established federal law.16 He asserts that "[p]rejudice has been specifically found by the Supreme Court from trial counsel's deficient performance in not investigating a client's cognitive deficits." Aplt.'s Opening Br. at 92 (citing Michael Williams , 529 U.S. at 396, 120 S.Ct. 1495 ). To bolster his point, Mr. Grant challenges the OCCA's subsidiary conclusion that Mr. Grant's particular organic-brain-damage evidence was not qualitatively different, in *920terms of mitigating effect, from the evidence of schizophrenia that trial counsel had presented to the jury.17

In line with Mr. Grant's arguments, we elect to focus our deferential review under AEDPA on the OCCA's prejudice ruling-viz. , its determination that there is not a reasonable probability that, but for trial counsel's failure to investigate and present organic-brain-damage evidence, the jury's sentencing verdict would have been different. See, e.g. , Strickland , 466 U.S. at 697, 104 S.Ct. 2052 ( "If it is easier to dispose of an ineffectiveness claim on the ground of lack of sufficient prejudice, which we expect will often be so, that course should be followed."); see also Littlejohn II , 875 F.3d at 552 (noting that "[t]hese two prongs" of the Strickland standard "may be addressed in any order"). And, viewed through AEDPA's prism, we conclude that Mr. Grant has not carried his burden to demonstrate prejudice under Strickland .

i

"Evidence of organic brain damage is something that we and other courts, including the Supreme Court, have found to have a powerful mitigating effect." Victor Hooks II , 689 F.3d at 1205 (citing Rompilla , 545 U.S. at 392, 125 S.Ct. 2456 ; Michael Wilson I , 536 F.3d at 1094 ; Roderick Smith v. Mullin , 379 F.3d 919, 942-43 (10th Cir. 2004) ); see Littlejohn I , 704 F.3d at 864 ("Evidence of organic mental deficits ranks among the most powerful types of mitigation evidence available."). "And for good reason-the involuntary physical alteration of brain structures, with its attendant effects on behavior, tends to diminish moral culpability, altering the causal relationship between impulse and action." Victor Hooks II , 689 F.3d at 1205.

In Littlejohn II , we had occasion to elaborate on this "proposition"-and, thereby, clarify the mitigating role that organic-brain-damage evidence plays in the capital-sentencing context. See *921Littlejohn II , 875 F.3d at 559. Specifically, we said:

[T]his proposition only has explanatory power with respect to our caselaw when appropriately qualified in two salient respects. First , it does not mean that all evidence of organic brain damage has the same potency in the Strickland prejudice analysis and will ineluctably result in a determination of prejudice. Our caselaw requires us to examine the precise nature of the alleged organic brain damage. In this regard, in several instances, we have concluded that evidence alleged to show organic brain damage, or related mental-health evidence, would have had little, if any, impact on the jury's decision-making process. ...

....

Second , we have concluded, in some instances, that organic-brain-damage evidence would have been just as likely-if not more likely-to have had an aggravating effect rather than a mitigating effect on a sentencing jury.

Id. at 559-60 (citations omitted).

Put more simply, with respect to the first point, we must carefully consider in our analysis the "precise nature" of the organic-brain-damage evidence at issue and recognize that, though this category of evidence generally packs a powerful mitigating punch, particular versions of it may be "qualitatively weak in their mitigating effects on jurors." Id. at 559, 566. And, regarding the second point, because omitted mitigation evidence may have the effect of a "double-edged sword"-cutting both in favor of mitigation and in favor of aggravation-we must remain cognizant of the possibility that the evidence actually would have hurt more than helped the petitioner's mitigation case.18 See Davis v. Exec. Dir. of Dep't of Corr. , 100 F.3d 750, 762 (10th Cir. 1996) (noting that courts must carefully review omitted mitigation evidence to determine if it truly mitigates or, instead, has the possibility of being a "two-edged sword" (quoting Davis v. People , 871 P.2d 769, 774 (Colo. 1994) )); see Gilson , 520 F.3d at 1250 (noting, as to organic-brain-damage evidence, that "the presentation of this evidence would likely have weighed against [petitioner] by erasing any lingering doubts that may have existed as to his role in [the victim's] murder, and by confirming the jury's conclusion that he represented a continuing threat, even if confined in prison for life"); see also Littlejohn II , 875 F.3d at 564 (noting that "the introduction of Mr. Littlejohn's organic-brain-damage evidence at resentencing likely would have been the impetus for developments harmful to his case"). In sum, the identification of organic-brain-damage evidence that counsel allegedly omitted unreasonably from the sentencing phase marks the beginning-not the end-of our prejudice analysis under Strickland .

ii

(1)

On direct appeal, Mr. Grant submitted the neuropsychological evaluation performed by clinical psychologist Dr. Gelbort.

*922In summary, Dr. Gelbort's report indicated that Mr. Grant has a "frontal lobe syndrome," causing him to suffer from certain cognitive impairments. Aplt.'s Direct Appeal Appl. for Evidentiary Hr'g, Ex. G-2. He found that Mr. Grant's "impairments ... predate the [offense conduct]," "are omnipresent[,] and continuously affect his behavior in a negative way." Id. at Ex. G-3. More specifically, Dr. Gelbort found that Mr. Grant is "never as able as a normal individual to think logically, adaptively, and coherently," and that "[h]is capacity for normal reasoning" is impaired. Id. As a result, "his behavior [was] less likely to conform to normal standards at the time of the crime." Id.

Mr. Grant argues that this evidence could have explained to the jury that his frontal lobe damage "caused [the] neuro-cognitive deficiencies that are linked to his lifetime of impulsive, aggressive, and irrational behaviors." Aplt.'s Opening Br. at 87. Moreover, he argues that Dr. Gelbort's evidence could have "proved the brain damage to which trial counsel alluded, [and] also explained .... in mitigating terms the self-destructive impulsive behaviors that carried over into his devastating confession letter and testimony, and explained how his impulse to sabotage his case was impossible for him to control." Id. at 61-62.

We conclude that the OCCA could have reasonably concluded that the organic-brain-damage evidence from Dr. Gelbort "would have been qualitatively weak in [its] mitigating effects on jurors." Littlejohn II , 875 F.3d at 566. Dr. Gelbort spoke only in general terms about the presence of "frontal-lobe damage" and Mr. Grant's inability to think "logically, adaptively, and coherently." See Aplt.'s Direct Appeal Appl. for Evidentiary Hr'g, Ex. G-2, G-3. Contrary to Mr. Grant's suggestion, Dr. Gelbort's testimony never indicated that Mr. Grant's brain defects caused his behavior to be "impulsive" or "aggressive" in a way that would meaningfully explain his involvement in the double murders for which he suffered convictions.

Indeed, evidence of impulse-control impairments would have been of modest explanatory power in this particular case, where the overwhelming evidence in the record indicates that Mr. Grant's commission of the offenses at issue was not the result of impulse; on the contrary, Mr. Grant's criminal episode was a planned, organized, and methodical one designed to secure money for his girlfriend's bail. See, e.g. , R., Vol. IV, Trial Tr. VI, at 194 ("Q: So was it part of your plan, before you went there, that you were going to get the videotape [from the security camera]? [Mr. Grant:] Most definitely."); id. at 194-95 ("Q: And so before you ever entered the La Quinta Inn that day you had a knife and a gun and you knew that you were going to kill whoever was there; is that correct? [Mr. Grant:] Most definitely.").

As a result, in this case, any evidence of Mr. Grant's inability to control his impulses would have done little to connect the dots between his brain damage and the offense conduct. Cf. Hooks , 689 F.3d at 1204 ("Counsel in capital cases must explain to the jury why a defendant may have acted as he did-must connect the dots between, on the one hand, a defendant's mental problems, life circumstances, and personal history and, on the other, his commission of the crime in question."). Furthermore, with regard to Dr. Gelbort's opinion that Mr. Grant's "behavior [was] less likely to conform to normal standards at the time of the crime," Aplt.'s Direct Appeal Appl. for Evidentiary Hr'g. at Ex. G-3, the OCCA could have reasonably concluded that "such evidence[, which] tends to depict [Mr. Grant] as unstable and unable to control his actions[,] ... could have *923[had] an overall aggravating, rather than mitigating, effect," and, therefore, its omission was not prejudicial. Littlejohn II , 875 F.3d at 562.

Moreover, the OCCA also could have reasonably concluded that the potency of Dr. Gelbort's organic-brain-damage evidence would have been significantly weakened by the fact that he never indicated that the negative manifestations of Mr. Grant's organic brain damage-for instance, his inability to conform to societal norms-were treatable with medication or other such means. See id. at 565 ("[T]he mitigating effect of Mr. Littlejohn's evidence of organic brain damage would likely have been diminished by the lack of reliable treatment options for Mr. Littlejohn's attention deficit and impulse-control disorders."); cf. Littlejohn I , 704 F.3d at 865 n.24 (evidence of organic brain damage coupled with evidence of available medical treatments could give a sentencing jury "some assurance that" the petitioner's "criminal, violent past would not be prologue"); Michael Wilson I , 536 F.3d at 1094 ("Diagnoses of specific mental illnesses such as schizophrenia or bipolar, which are associated with abnormalities of the brain and can be treated with appropriate medication, are likely to [be] regarded by a jury as more mitigating than generalized personality disorders, which are diagnosed on the basis of reported behavior, are generally inseparable from personal identity, and are often untreatable through medical or neurological means ." (emphases added)); cf. also Hooks , 689 F.3d at 1205 (citing Michael Williams , 529 U.S. at 398, 120 S.Ct. 1495 ).

In sum, we conclude that the OCCA could have reasonably concluded that the "mitigating effects on the jurors" of the particular organic-brain-damage evidence identified by Mr. Grant "would have been qualitatively weak." Littlejohn II , 875 F.3d at 566.

(2)

Furthermore, this qualitatively-weak evidence would not have been considered by the sentencing jury in a vacuum; Strickland and its progeny lead us to examine the role that the evidence would have played in Mr. Grant's overall mitigation case. As the Supreme Court put it,

In making this determination, a court hearing an ineffectiveness claim must consider the totality of the evidence before the judge or jury. Some of the factual findings will have been unaffected by the errors, and factual findings that were affected will have been affected in different ways. Some errors will have had a pervasive effect on the inferences to be drawn from the evidence, altering the entire evidentiary picture, and some will have had an isolated, trivial effect.

Strickland , 466 U.S. at 695-96, 104 S.Ct. 2052 ; accord Jeremy Williams , 782 F.3d at 1215 ; Michael Wilson II , 706 F.3d at 1305.

The OCCA observed that trial counsel "spent considerable time presenting [mitigation] evidence to the jury." Grant , 205 P.3d at 23. Trial counsel's mitigation case included expert evidence of Mr. Grant's schizophrenia presented by Dr. Grundy; testimony from family members and Dr. Art Williams, PhD, about his disadvantaged and dysfunctional childhood; and various references to certain "red flag" indicators of organic brain damage, though no definitive evidence of organic brain damage was presented.

Dr. Grundy testified that he diagnosed Mr. Grant with schizophrenia and that this illness "significantly impaired his competency," but that medication and a structured environment "helped his symptoms *924go into partial remission ... to a great extent." R., Vol. IV, Trial Tr. VI, at 231. He also noted that Mr. Grant was subject to certain "stressors" including that he grew up in "a high crime neighborhood" with "poor parental supervision," and possibly "parental neglect." Id. , Trial Tr. VII, at 19. Dr. Grundy explained that when these "stressors are present ... mental illness worsens." Id. at 20.

Dr. Grundy did not testify as to organic brain damage because he was not qualified to "examine ... and assess" the "potential for organic brain damage." Id. at 16, 18. However, he testified to certain "red flag" indicators of organic brain damage including that Mr. Grant was deprived of oxygen at birth. Id . at 16, 28. Moreover, Mr. Grant's mother testified that she drank alcohol heavily during her pregnancy with Mr. Grant and that he was born "blue" and without a pulse. Id. , Vol. IV, Trial Tr. VI, at 133-34. And the jury was given the following mitigation instruction further reinforcing the potentiality for organic brain damage: "There are indications of brain damage existing at or before Donald Grant's birth[, including] his mother's heavy consumption of alcohol during her pregnancy with Donald Grant and the loss of oxygen to him during delivery." Id. , Trial Tr. VIII, at 34-35.

Dr. Williams testified regarding Mr. Grant's "[v]ery pathological" childhood. Id. , Trial Tr. VII, at 181. Dr. Williams explained that Mr. Grant's father was an alcoholic and his mother was addicted to crack cocaine, and that, as a result, "[h]e didn't get any of the core values ... the discipline accountability [values] in terms of a family system and support system. ... [H]e didn't get any positive rol[e] models early in life in relationship of support." Id. at 181-82.

In light of the foregoing, we may certainly conclude that this is patently not a situation where Mr. Grant's trial counsel just "did something " in the mitigation case. Michael Wilson I , 536 F.3d at 1084. Far from it. Consequently, with the totality of this evidence before it, we think that the OCCA's rejection of Mr. Grant's prejudice showing under Strickland cannot be found to be contrary to or an unreasonable application of clearly established federal law. At a minimum, " 'fairminded jurists could disagree' on the correctness of the [OCCA's] decision." Harrington , 562 U.S. at 101, 131 S.Ct. 770 (quoting Yarborough , 541 U.S. at 664, 124 S.Ct. 2140 ).

Significantly, the omitted evidence of organic brain damage in this case would have merely supplemented the same mitigation theory as was already presented in the record. The jury heard evidence, inter alia , of Mr. Grant's schizophrenia diagnosis, and red flags of organic brain damage, as part of trial counsel's strategy to show that Mr. Grant was relatively less morally culpable for his crimes. The omitted evidence of Dr. Gelbort's evaluation and related testimony would have followed this same theme: Mr. Grant is less morally culpable for his crimes because of a mental impairment beyond his control. More specifically, trial counsel's mitigation case was targeted specifically to explain and lessen Mr. Grant's culpability for his offense conduct by underscoring a clinically-diagnosable mental-health condition afflicting Mr. Grant, that is, schizophrenia. And to this end, counsel put substantial evidence before the sentencing jury regarding Mr. Grant's mental illness.

Thus, evidence of organic brain damage would have only supplemented, rather than introduced, this mitigation theory to the sentencing jury. Indeed, the further evidence of organic brain damage could have been in tension with the mitigation case and had a doubled-edged effect. Whereas Dr. Grundy testified, as to *925schizophrenia, that-though it "significantly impaired his competency"-medication and a structured environment "helped his symptoms go into partial remission ... to a great extent," R., Vol. IV, Trial Tr. VI, at 231, there was not similar testimony from Dr. Gelborn regarding the possibility of successful treatment options for Mr. Grant's organic brain damage. See, e.g. , Littlejohn II , 875 F.3d at 565 ("[T]he mitigating effect of Mr. Littlejohn's evidence of organic brain damage would likely have been diminished by the lack of reliable treatment options for Mr. Littlejohn's attention deficit and impulse-control disorders."); cf. Michael Wilson I , 536 F.3d at 1094 ("Diagnoses of specific mental illnesses such as schizophrenia or bipolar [disorder], which are associated with abnormalities of the brain and can be treated with appropriate medication, are likely to [be] regarded by a jury as more mitigating than generalized personality disorders, which are diagnosed on the basis of reported behavior, are generally inseparable from personal identity, and are often untreatable through medical or neurological means ." (emphases added)). In sum, for Mr. Grant, evidence of his organic brain damage-absent any indication that the symptoms could be treated-could well have caused some damage to the substantial mitigation case trial counsel had already presented to the jury.

Mr. Grant points to a number of cases in which the Supreme Court and this one have found prejudice stemming from counsel's failure to investigate and present evidence of organic brain damage. In several of these cases-unlike the circumstances here-trial counsel's only mitigation theory was far afield from that supported by (omitted) evidence of organic brain damage. For instance, in Sears v. Upton , 561 U.S. 945, 947, 949, 130 S.Ct. 3259, 177 L.Ed.2d 1025 (2010), the Supreme Court found prejudice where trial counsel's mitigation case "portray[ed] the adverse impact of [the defendant's possible] execution on his family and loved ones," and omitted evidence of the defendant's "frontal lobe abnormalities" and dysfunctional childhood, which impaired the defendant's planning, sequencing, and impulse control. Similarly, in Rompilla , 545 U.S. at 378, 392, 125 S.Ct. 2456, the Court found prejudice where trial counsel presented testimony from family members that the defendant "was innocent and a good man," and omitted evidence that the defendant suffered from organic brain damage and fetal alcohol syndrome.

And, in Wiggins , the Court found prejudice where counsel argued only that the defendant had a "clean record," with no prior convictions, despite the fact that "[t]he mitigating evidence that counsel failed to discover and present in this case [was] powerful," including evidence of petitioner's "diminished mental capacities." 539 U.S. at 515, 534-35, 123 S.Ct. 2527. Finally, in Anderson v. Sirmons , 476 F.3d 1131, 1146-47 (10th Cir. 2007), we found prejudice where counsel presented evidence that the defendant "was a kind, hard-working, normal man who could be of some help to his daughter if his life were spared," and omitted evidence, inter alia , that the defendant had brain damage and was "borderline mentally defective."

In each of these cases, the good-guy and beloved-family-member mitigating evidence presented by counsel was significantly different from the omitted-and potentially more powerful-evidence of organic brain damage, which could have served to explain and lessen the defendants' moral culpability for their offense conduct. Not so here. Evidence of organic brain damage would have only supplemented, rather than introduced, the mitigation theory of Mr. Grant's counsel to the sentencing jury.

*926Furthermore, a proper Strickland prejudice analysis would also necessarily take into account the State's potential case in aggravation. See, e.g. , Michael Wilson II , 706 F.3d at 1306 (noting that, in making the prejudice determination under Strickland , we must evaluate the strength of the omitted evidence in light of "what the prosecution's response to that evidence would have been"). In support of its contention that Mr. Grant deserved the death penalty, the State argued, inter alia , that Mr. Grant posed a continuing threat to society. At least under facts akin to these, "we have characterized a petitioner's potential for continued dangerousness, even if incarcerated, as 'perhaps [the] most important aggravating circumstance' that juries consider in weighing the death penalty." Littlejohn II , 875 F.3d at 564 (quoting John Grant v. Trammell , 727 F.3d 1006, 1017 (10th Cir. 2013) ); see also Littlejohn I , 704 F.3d at 865 (holding that "[t]he potential prejudice flowing from th[e] omission" of organic-brain-damage evidence could have been "heightened" where "a considerable portion of the State's case in aggravation relate[d] to the continuing-threat aggravator"). Thus, in considering the State's potential case in aggravation, the jury would likely have used evidence of Mr. Grant's organic brain damage and its possibly untreatable symptoms as support for a conclusion that Mr. Grant would pose a continuing threat to society, notwithstanding his incarceration. The OCCA thus could have reasonably concluded that, in this instance, "organic-brain-damage evidence would have been just as likely-if not more likely-to have had an aggravating effect rather than a mitigating effect on a sentencing jury." Littlejohn II , 875 F.3d at 560.

* * *

In sum, we are not persuaded that the OCCA was unreasonable in concluding that this additional mental-health explanation for Mr. Grant's offense conduct-offered in the form of organic-brain-damage evidence from Dr. Gelbort-would have created a reasonable probability that a juror would have voted differently at Mr. Grant's sentencing. See Grant , 205 P.3d at 23 ("[W]hile the presence or absence of organic brain disorder might have shed light on one potential cause of [Mr. Grant's] mental illness, the fact that [Mr. Grant] had some sort of mental illness was never in serious dispute. In our view, the affidavits [Mr. Grant] submits on appeal do not present anything qualitatively different from what was presented at trial on these issues."); cf. [ Lois] Smith , 235 F.3d at 1282 ("Although the evidence pertaining to Smith's organic brain damage would have been proper mitigating evidence and may have helped explain the crime to some degree .... we are not persuaded it is reasonably probable that the introduction of the organic brain damage evidence would have led the jury to choose a life sentence rather than a death sentence."). At a minimum, " 'fairminded jurists could disagree' on the correctness of the [OCCA's] decision." Harrington , 562 U.S. at 101, 131 S.Ct. 770 (quoting Yarborough , 541 U.S. at 664, 124 S.Ct. 2140 ). We cannot conclude that the OCCA's Strickland prejudice determination concerning the omitted organic-brain-damage evidence is contrary to or an unreasonable application of clearly established federal law. This determination dooms the organic-brain-damage aspect of Mr. Grant's Strickland claim because he necessarily cannot satisfy his two-part burden. We accordingly affirm the district court's denial of habeas relief as to this claim.

5. Failure to Investigate and Present Evidence of Delusional Belief System

Mr. Grant argues that he was denied effective assistance of counsel because *927trial counsel "fail[ed] to investigate and present evidence that [Mr. Grant's] delusions were," in fact, delusions and "not related to any recognized religion." Aplt.'s Opening Br. at 77 (capitalization altered). The district court found that this claim was unexhausted. As we have noted, exhaustion in the state courts is a prerequisite for habeas review in federal court. See, e.g. , Davila , 137 S.Ct. at 2064 ; Picard , 404 U.S. at 275, 92 S.Ct. 509.

Mr. Grant argues that the district court erred because he presented this claim to the OCCA on direct appeal and in his application for post-conviction relief by presenting certain affidavits to the OCCA at each of these phases. Specifically, he contends that, on direct appeal, he presented this claim through the affidavits of Anna Wright and Natasha Briggs, both of whom worked in the medical unit at the prison where Mr. Grant was incarcerated. Additionally, he contends that he presented this argument on post-conviction through the affidavit of Stacey Hemphill, Mr. Grant's cellmate.

We have comprehensively explicated in Part III.A.1, supra , the relevant principles governing exhaustion and, more specifically, the fair presentation of federal claims in state court. We will not repeat that discussion here. Suffice it to say that presenting the relevant facts to the state court is not enough to constitute the fair presentation of a claim. See Bland , 459 F.3d at 1011. To exhaust his claim in state court, Mr. Grant needed to also present the argument that he was denied effective assistance of counsel because trial counsel failed to investigate and present evidence that his delusions were, in fact, delusions, and not expressions of religion. In his briefing to this court, Mr. Grant has identified no such language in his direct-appeal or post-conviction briefing.

Nor could we find any on our independent review of Mr. Grant's briefing before the OCCA. His direct-appeal briefing makes no such argument. And his post-conviction brief likewise offers no argument on these matters. Mr. Grant does refer to Mr. Hemphill's affidavit in his post-conviction brief, but Mr. Grant does so only in relation to his argument that he was tried while purportedly incompetent. Mr. Grant does not link Mr. Hemphill's statement to an argument that trial counsel failed to investigate and present evidence that his seemingly delusional statements reflected actual delusions, rather than religious beliefs.

Accordingly, Mr. Grant failed to exhaust this claim in state court. This argument would appear to be subject to an anticipatory procedural bar because the Oklahoma courts would not entertain it if Mr. Grant were to return to those courts either because they would find that Mr. Grant could have raised the claim on direct-appeal or subsequently in his post-conviction proceeding. See, e.g. , Thacker , 678 F.3d at 841 ; Conover , 942 P.2d at 230. Mr. Grant makes no attempt to argue that such a bar would not apply here were we to determine that this claim is unexhausted. And he has made no showing of cause and prejudice or a fundamental miscarriage of justice to excuse his default. See, e.g. , Davila , 137 S.Ct. at 2064-65 ; Bland , 459 F.3d at 1012. Therefore, we have no occasion to reach the merits of this claim; our review is precluded.

6. Failure to Investigate and Present Evidence of Pertinent Aspects of Mr. Grant's Childhood

Mr. Grant received a COA to challenge trial counsel's failure to investigate and present evidence of pertinent aspects of his childhood. Mr. Grant argues that trial counsel failed to "uncover family historians *928who observed unusual behaviors in Donald as a child." Aplt.'s Opening Br. at 78. Specifically, Mr. Grant points to affidavits submitted with his post-conviction application that as a child he "acted crazy," and was "unpredictable[,] impulsive," and "mentally ill." Id. (quoting Aplt.'s App. to Post-Conviction Appl., Ex. 7, 8). He asserts that his "case in mitigation would have been significantly stronger if counsel had thoroughly interviewed [his] mother and siblings, and interviewed other family members [that] post-conviction counsel easily found." Id. For the reasons set forth below, we uphold this aspect of the district court's denial of Mr. Grant's petition.

a

To make sense of Mr. Grant's request for relief, we pause briefly to review the procedural history of this particular claim.

On direct appeal, Mr. Grant argued that trial counsel rendered ineffective assistance, inter alia , by omitting certain evidence of his dysfunctional childhood from its mitigation case. For support, he submitted "three affidavits from friends or family which discuss his disadvantaged childhood, and offer occasional examples of his strange behavior in the years preceding the instant crimes." Grant , 205 P.3d at 23. Specifically, he submitted the affidavit of his step-father, Ronald Williams, who knew Mr. Grant as a child, as well as Mr. Grant's former girlfriend Cheryl Tubbs and her nephew Keith Tubbs, who each averred that they had second-hand knowledge of Mr. Grant's childhood. The OCCA found that these "affidavits [did] not present anything qualitatively different from what was presented at trial on these issues" and concluded that there was "no reasonable probability that these additional witnesses would have affected the jury's sentencing decision," and accordingly denied Mr. Grant relief under Strickland . Id.

In his post-conviction application, Mr. Grant argued that trial and direct-appeal counsel rendered ineffective assistance by "failing to investigate and present to the jury [additional] compelling mitigation evidence." Aplt.'s Post-Conviction Appl. at 35 (capitalization altered). Specifically, Mr. Grant identified several of his family members who could have testified to pertinent aspects of his childhood. He submitted affidavits of his mother, Mary Williams, his sister, Juzzell Robinson, and his younger brother, Lennox Grant. These three family members each testified at sentencing, but Mr. Grant argued that they had additional information that went undiscovered by trial and direct-appeal counsel. He also submitted affidavits from two uncles, John Robinson and Isaiah Robinson, and his cousin, Louis Robinson. Mr. Grant explained that these family members would have been willing to testify but were never asked to do so. In its denial of Mr. Grant's post-conviction application, the OCCA found that Mr. Grant's claim "reformulate[d] an argument presented and addressed on direct appeal," and was therefore "barred [from post-conviction review] under the doctrine of res judicata ." Grant II , No. PCD-2006-615, slip op. at 7.

b

Mr. Grant argues that our merits review of this claim should include the materials that he submitted in post-conviction proceedings. See Aplt.'s Opening Br. at 93 ("The post-conviction material should be considered as part of the merits review of Grant's claim."). Mr. Grant asserts that, under Cone v. Bell , a state court's refusal to "review the merits of a petitioner's claim on the ground that it has done so already"-that is, res judicata -"creates no bar to federal habeas review." Id. (quoting Cone v. Bell , 556 U.S. 449, 466, 129 S.Ct. 1769, 173 L.Ed.2d 701 (2009) ). Therefore, *929he urges us to consider his post-conviction materials as part of our merits review of his claim.

In Cone , the state post-conviction court refused to consider a petitioner's claim based on its erroneous finding that the claim had been previously adjudicated in state court. Cone , 556 U.S. at 466, 129 S.Ct. 1769 ("That conclusion rested on a false premise: ... [the petitioner] had not presented his Brady claim in earlier proceedings and, consequently, the state courts had not passed on it."). The Court held that the state court's refusal to review the merits of the petitioner's claim on this mistaken ground-i.e., that it had done so already-created no bar to federal habeas review. See id. at 467, 129 S.Ct. 1769. Because the claim had never been reviewed on the merits in state court, it was not subject to AEDPA review, and the Court thus proceeded to review the claim de novo. Id. As applied here, Mr. Grant seems to argue that Cone would permit us to consider his post-conviction materials de novo in resolving his ineffective-assistance claim, given that the OCCA never considered them on the merits due to its res judicata ruling.

But Mr. Grant is mistaken. Assuming arguendo that Cone is instructive here at all, it leads in a different direction. It is true that, under Cone , the OCCA's refusal to consider his claim in the post-conviction context based on res judicata created no bar to habeas review of the claim in the federal courts. However, unlike in Cone , the OCCA has already adjudicated this claim on the merits on direct appeal. In other words, the OCCA was not mistaken in concluding that it had previously considered the claim. Indeed, Mr. Grant does not dispute this; more specifically, he does not contest the OCCA's finding that his post-conviction claim was merely a reformulation of his direct-appeal claim and thus barred by res judicata . Consequently, as applied here, Cone would simply instruct that the OCCA's refusal to consider the claim post-conviction on res judicata grounds creates no barrier to our review of its resolution of this claim on direct appeal . And because the OCCA reached the merits on direct appeal, unlike in Cone , we are bound by AEDPA and, notably, its prohibition against the consideration of materials that were not part of the state-court record when the state court ruled. See, e.g. , Cullen , 563 U.S. at 181, 131 S.Ct. 1388 ("We now hold that review under § 2254(d)(1) is limited to the record that was before the state court that adjudicated the claim on the merits."). Thus, we are not permitted to consider the post-conviction materials that Mr. Grant tenders here.

Stripped of those materials and the possibility of de novo review, Mr. Grant offers little to advance his cause. Specifically, Mr. Grant has not shown-or, for that matter, even argued here-that, under AEDPA, the OCCA's rejection of this particular ineffective-assistance claim is contrary to or an unreasonable application of clearly established federal law, or premised on an unreasonable determination of the facts. Rather, he asserts only that his "case in mitigation would have been significantly stronger if counsel had thoroughly interviewed [his] mother and siblings, and interviewed other family members [that] post-conviction counsel easily found." Aplt.'s Opening Br. at 78. Mr. Grant may be correct, but this argument falls patently short of satisfying the AEDPA standard.

To satisfy AEDPA, Mr. Grant needed to go further-for example, arguing that the OCCA's rejection of his ineffective-assistance claim is contrary to or an unreasonable application of clearly established federal law in light of the "significantly *930stronger" mitigation case that trial counsel could have-but did not-present. See, e.g. , Littlejohn I , 704 F.3d at 824. This Mr. Grant has not done. Therefore, under AEDPA, he is not entitled to relief.

* * *

In sum, Mr. Grant has not persuaded us that the OCCA's rejection of his ineffective-assistance claim here-viz. , that trial counsel was constitutionally ineffective in failing to investigate and present further evidence of his troubled and dysfunctional childhood-was contrary to or an unreasonable application of clearly established federal law, or an unreasonable determination of the facts. Accordingly, we uphold the district court's resolution of this portion of Mr. Grant's petition.19

C. Jury Instruction and Closing Statements on Mitigation Evidence

Mr. Grant argues that one of the sentencing-phase jury instructions, Instruction 12, standing alone and in conjunction *931with the State's closing arguments, unconstitutionally limited the jury's consideration of evidence presented in mitigation of his death sentence. Instruction 12 provides in pertinent part: "Mitigating circumstances are those which, in fairness, sympathy, and mercy, may extenuate or reduce the degree of moral culpability or blame." O.R. 2349 (Instr. 12). In Mr. Grant's view, this text of Instruction 12 (i.e., the "moral-culpability text") caused the jury to ignore otherwise proper mitigating evidence, and the prosecution exploited this allegedly infirm instruction in its closing arguments.

On direct appeal, the OCCA reviewed and rejected on the merits Mr. Grant's dual challenge to the moral-culpability text of Instruction 12 and the related prosecutorial statements. Mr. Grant argued that "the prosecutor focused on only part of the definition of mitigating evidence, and thus unfairly limited the jurors' consideration of the evidence [he] had offered as mitigating." Grant , 205 P.3d at 20. And he relatedly contended that "the prosecutor misstated the law by telling the jurors that the evidence he had presented as 'mitigating' did nothing to justify a sentence less than death." Id. However, the OCCA ruled that "[t]he jurors in this case were properly instructed that anything could be considered mitigating." Id. at 21. It further reasoned that

[Mr. Grant] confuses what kind of information may be offered as mitigating evidence, with whether that information successfully serves its intended purpose. While there is no restriction whatsoever on what information might be considered mitigating, no juror is bound to accept it as such, and the State is free to try to persuade the jury to that end. The prosecutor's arguments did not misstate the law on this point.

Id. Accordingly, the OCCA denied Mr. Grant's dual challenge. Because it did so on the merits, we must accord AEDPA deference to the OCCA's decision.20 See, *932e.g. , Victor Hooks II , 689 F.3d at 1163. For the reasons set forth below, we conclude that Mr. Grant cannot prevail on any of his arguments here.

1. Legal Framework

"[T]he Eighth and the Fourteenth Amendments require that the sentencer, in all but the rarest kind of capital case, not be precluded from considering, as a mitigating factor , any aspect of a defendant's character or record and any of the circumstances of the offense that the defendant proffers as a basis for a sentence less than death." Lockett v. Ohio , 438 U.S. 586, 604, 98 S.Ct. 2954, 57 L.Ed.2d 973 (1978) (plurality opinion) (footnote omitted); accord Eddings v. Oklahoma , 455 U.S. 104, 110, 102 S.Ct. 869, 71 L.Ed.2d 1 (1982). " Lockett and its progeny stand only for the proposition that a State may not cut off in an absolute manner the presentation of mitigating evidence, either by statute or judicial instruction, or by limiting the inquiries to which it is relevant so severely that the evidence could never be part of the sentencing decision at all." Johnson v. Texas , 509 U.S. 350, 361-62, 113 S.Ct. 2658, 125 L.Ed.2d 290 (1993) (quoting McKoy v. North Carolina , 494 U.S. 433, 456, 110 S.Ct. 1227, 108 L.Ed.2d 369 (1990) (Kennedy, J., concurring in judgment)); see also Brecheen v. Reynolds , 41 F.3d 1343, 1361 n.13 (10th Cir. 1994) (summarizing the holdings of Lockett and its progeny and observing that "[t]hose cases all involved situations where the sentencer was, for a variety of reasons, prevented or precluded from considering relevant mitigating evidence"). Though this body of authority highlights that there are various ways in which the sentencer (here, the jury) might be precluded from considering all mitigating evidence, these ways do not all have the same potency. Notably, improper comments of the prosecution "are not to be judged as having the *933same force as an instruction from the court." Boyde v. California , 494 U.S. 370, 384-85, 110 S.Ct. 1190, 108 L.Ed.2d 316 (1990). As relevant here, the Court in Boyde elaborated on this point:

[A]rguments of counsel generally carry less weight with a jury than do instructions from the court. The former are usually billed in advance to the jury as matters of argument, not evidence, and are likely viewed as the statements of advocates; the latter, we have often recognized, are viewed as definitive and binding statements of the law. Arguments of counsel which misstate the law are subject to objection and to correction by the court. This is not to say that prosecutorial misrepresentations may never have a decisive effect on the jury, but only that they are not to be judged as having the same force as an instruction from the court. And the arguments of counsel, like the instructions of the court, must be judged in the context in which they are made.

Id. (citations omitted).

Regarding the importance of context, "we accept at the outset the well established proposition that a single instruction to a jury may not be judged in artificial isolation, but must be viewed in the context of the overall charge." Cupp v. Naughten , 414 U.S. 141, 146-47, 94 S.Ct. 396, 38 L.Ed.2d 368 (1973). Where a jury instruction is alleged to be "subject to an erroneous interpretation .... the proper inquiry in such a case is whether there is a reasonable likelihood that the jury has applied the challenged instruction in a way that prevents the consideration of constitutionally relevant evidence." Boyde , 494 U.S. at 380, 110 S.Ct. 1190 ; accord Hanson v. Sherrod , 797 F.3d 810, 850 (10th Cir. 2015). "Although a defendant need not establish that the jury was more likely than not to have been impermissibly inhibited by the instruction," the defendant must show more than "a possibility of such an inhibition." Boyde , 494 U.S. at 380, 110 S.Ct. 1190. The Court in Boyde further observed:

There is, of course, a strong policy in favor of accurate determination of the appropriate sentence in a capital case, but there is an equally strong policy against retrials years after the first trial where the claimed error amounts to no more than speculation. Jurors do not sit in solitary isolation booths parsing instructions for subtle shades of meaning in the same way that lawyers might. Differences among them in interpretation of instructions may be thrashed out in the deliberative process, with commonsense understanding of the instructions in the light of all that has taken place at the trial likely to prevail over technical hairsplitting.

Id. at 380-81 (footnote omitted).

2. Analysis

a

We first address Mr. Grant's argument that the moral-culpability text of Instruction 12 violated his constitutional rights and that the OCCA's approval of the instruction was thus contrary to or an unreasonable application of clearly established federal law. For support, Mr. Grant relies on an Oklahoma state case decided after his trial called Harris v. State -which was on the books at the time of his direct appeal-in which the OCCA expressed concern that prosecutors "consistent[ly] misus[e] ... the language in this instruction [i.e., the moral-culpability text identical to that found in Instruction 12]" to argue that mitigating evidence cannot be considered when it does not bear on moral culpability or blame, 164 P.3d 1103, 1114 (Okla. Crim. App. 2007). For this reason, the OCCA recommended reformation *934of this pattern instruction to "clarify" its meaning. Id. Specifically, the OCCA "refer[red] this issue to the Oklahoma Uniform Jury Instruction Committee (Criminal) for promulgation of a modified jury instruction defining mitigating circumstances in capital cases." Id. Nevertheless, the Harris court determined that the existing instruction, while imperfect, was constitutionally sound. See id. at 1113. In response to the OCCA's referral, a reformed instruction was created; it defines mitigating circumstances to include those "which in fairness, sympathy, or mercy may lead you as jurors individually or collectively to decide against imposing the death penalty. " See OUJI-CR 4-78 (Sup. 2008) (emphasis added).

Mr. Grant argues, based on the OCCA's criticism in Harris of the moral-culpability text of the instruction before it, that the OCCA acted contrary to (as well as unreasonably applied) clearly established federal law here when it held that the identical moral-culpability text-which is contained in Instruction 12-was not unconstitutional. He asserts that the "OCCA's actions [-i.e., endorsing reformation of the instruction in Harris -] speak far louder than its internally inconsistent and thus unreasonable endorsement of the instruction as constitutionally sound." Aplt.'s Opening Br. at 103. In this regard, Mr. Grant contends that in Mills v. Maryland , 486 U.S. 367, 382, 108 S.Ct. 1860, 100 L.Ed.2d 384 (1988), "[t]he Supreme Court [indicated that] the remedial measures state courts perform are at least some indicia the prior instructions were infirm." Id. at 101, 131 S.Ct. 770 (emphasis omitted). His argument thus seems to be that the OCCA's approval of the moral-culpability text of the instruction here is contrary to or an unreasonable application of Mills and, more generally, Lockett 's progeny.

We rejected this very argument, however, under similar circumstances in Hanson . In particular, we rebuffed the notion that Mills was "germane," in light of our examination of the "OCCA's explanation as to why it amended the instruction" in Harris . 797 F.3d at 851. That explanation belied the petitioner's contention that the OCCA's comments regarding the reformation of the instruction had amounted to a tacit acknowledgment that the pre-reformation instruction was constitutionally infirm. See id. We noted that the Harris court "emphasize[d] that the language of the current instruction [i.e., the pre-reformation instruction] itself [wa]s not legally inaccurate, inadequate, or unconstitutional," and explicitly stated that "[c]ases in which the current [instruction] ha[d] been used and applied are not subject to reversal on this basis." Id. at 850-51 (first and fourth alteration in original) (quoting Harris , 164 P.3d at 1114 ). Consequently, in Hanson , we reasoned that Mills was inapposite in discerning the import of Oklahoma's reformation of the mitigation instruction and, more specifically, its moral-culpability text. See id. at 851.

Significantly, in repudiating the petitioner's Eighth Amendment challenge to the instruction's moral-culpability text, we also observed that "some of the other instructions from [the petitioner's] trial concerning mitigating evidence broadened the scope of evidence the jury could consider." Id. at 851. First of all, within the instruction itself containing the moral-culpability text, there was language that informed the jury that it was its responsibility to determine what circumstances are mitigating under the facts before it. See id. We said that "[t]his statement broadened any potential limitations imposed by the first sentence of the instruction [i.e., the moral-culpability text]." Id. Further, we observed that there was another instruction given to the jury that "listed 11 specific mitigating *935circumstances for the jury to consider, some of which had nothing to do with [the petitioner's] moral culpability"; it listed circumstances such as his family and emotional history, his fatherhood of a young son, and his trait of being "a follower." Id. (quoting the instruction from the record). And that instruction concluded by counseling the jury as follows: "In addition, you may decide that other mitigating circumstances exist , and if so, you should consider those circumstances as well." Id. (emphasis added) (quoting the instruction from the record). In light of these additional instructions, our resolution of the petitioner's Eighth Amendment attack on the moral-culpability text of the instruction at issue was clear: "Viewing the challenged instruction in the context of all the instructions, we do not think the jury would have felt precluded from considering any mitigating evidence ...." Id. (emphasis added).

Hanson controls our resolution of Mr. Grant's challenge to the moral-culpability text of Instruction 12 here. For the reasons stated in Hanson , Mr. Grant's Mills -based argument is without merit. Furthermore, as explicated infra in connection with our resolution of Mr. Grant's prosecution-exploitation claim, the additional instructions relating to mitigating evidence that we concluded in Hanson "broadened the scope of evidence the jury could consider" also were present-in all material respects-in Mr. Grant's case. Id. Therefore, Hanson 's conclusion-through the broader lens of all of the instructions-that "the jury would [not] have felt precluded [by the moral-culpability text of the instruction] from considering any mitigating evidence," id. , governs here as well regarding Instruction 12's identical moral-culpability text. Indeed, Mr. Grant does not meaningfully dispute this conclusion. See Aplt.'s Opening Br. at 101 ("This Court recently held [in Hanson ] that, given other Oklahoma jury instructions, the instruction itself does not violate the Constitution."). But he has "respectfully persist[ed] in presenting his concerns about the instruction to, at minimum, preserve them for potential further review." Id. However, applying AEDPA deference to the OCCA's determination upholding the constitutionality of the moral-culpability text of Instruction 12, and adhering to Hanson 's reasoning and holding, we must reject Mr. Grant's contention here.

b

We turn now to the contention that Mr. Grant pursues with greater vigor: specifically, that the prosecution's arguments to the jury impermissibly exploited Instruction 12's moral-culpability text in a way that makes it reasonably likely that the jury believed that it was limited to only considering evidence in mitigation that extenuated or reduced Mr. Grant's moral culpability or blame. See Aplt.'s Opening Br. at 103 (noting that "the prosecutor's exploitation" of the moral-culpability text effected an Eighth Amendment violation). According to Mr. Grant, the prosecution improperly argued that certain "evidence did not qualify as 'mitigating' because it did not reduce his moral culpability or blame for the crime." Id. (citing R., Vol. IV, Trial Tr. VIII, at 73-75, 79-80).21 Mr. *936Grant contends that the prosecution's arguments violated his Eighth Amendment rights safeguarded by Lockett and its progeny by preventing the sentencing jury from "consider[ing] all mitigating evidence." Aplt.'s Opening Br. at 107.22 We disagree.

We begin our analysis by explicating the prosecution's arguments upon which Mr. Grant grounds his Eighth Amendment challenge. Then, putting those arguments in the broader context of the instructions before the jury and other arguments to the jury by the prosecution and the defense, we conclude that the OCCA did not act unreasonably when it concluded that there was no reasonable likelihood that the jury believed-based on the prosecution's arguments-that it was limited to only considering evidence in mitigation that had the effect of extenuating or reducing Mr. Grant's moral culpability or blame.

i

Mr. Grant's central concerns relate to the arguments made by one of the two prosecutors representing the State at trial-Sandra Elliott-during the prosecution's rebuttal closing. This closing followed *937the defense's closing argument, which repeatedly informed the jury that the defense's intention was not to "excuse what happened in this case," but rather to offer "an explanation" for it. R., Vol. IV, Trial Tr. VIII, at 44. As material here, the prosecution responded:

You know, what I noted about the argument of defense counsel is they spent a long time trying to describe to you how what they have offered in mitigation is not an excuse, how what they're trying to tell you about the life and times of Donald Grant is not an excuse for the behavior that he committed at the La Quinta Inn in July of 2001. And that is exactly what the law says. Because one thing that I noticed that they did not talk about in their closing argument is what the definition of a mitigating circumstance is. Because the law tells you what that means . It tells you that in order-first of all, you have two choices: Do you believe that the mitigating circumstance has been proven because you don't have to believe everything that you have heard from the witnesses who testified. You can choose what parts you want to believe and disregard those parts that you believe perhaps people were exaggerating about or somehow trying to make things sound a whole lot worse than they actually were. But let's assume for the sake of argument that everything that you were told was correct, that not any person made up or fudged a little bit in what they were telling you about. What does it say mitigating circumstances are? What does that mean when we say that something may mitigate the murder of these two women, the lives that he took? It says that mitigating circumstances are those which reduce the moral culpability or blame of the defendant. That those things, in order to be mitigating, must reduce his moral culpability or blame .

Id. (emphases added). After the defense unsuccessfully challenged this argument on the ground that it "would tend to mislead the jury into believing" that the defense's evidence did not qualify as "proper mitigation circumstances to consider," the prosecution continued:

It's not Sandra Elliott telling you that this will make something mitigating, that's what the law says . And we all talked about during voir dire that we would be discussing the law that the Court's going to be giving you. And the law says, not Sandra Elliott, not what the defense attorneys say, but what the Court tells you and what the law says is that before something can be mitigating it must reduce the moral culpability or blame of the defendant .

Id. at 75 (emphases added). Lastly, in one instance, the prosecution specifically employed the moral-culpability text of Instruction 12 to argue against one element of Mr. Grant's mitigation case, relating to his alleged schizophrenia. In this regard, Ms. Elliott stated:

Does it reduce his moral culpability, his moral blame for what he did? And I would submit to you that it does not in any way.

So while they may say to you that I'm not offering this as an excuse for Mr. Grant's behavior, you have to look at whether or not it reduces his moral culpability or blame. That is what the law says that you must do .

Id. at 79 (emphasis added).

Mr. Grant contends that the prosecution's rebuttal closing arguments had the unconstitutional effect of precluding the jury from considering that portion of his proferred mitigation evidence that did not extenuate or reduce his moral culpability or blame; as he reasons, the jury would have been operating under the mistaken *938belief-instilled by the prosecution-that this evidence was not legally proper mitigating evidence and thus should not be considered. However, applying AEDPA's deferential standards, we ultimately conclude that the OCCA was not unreasonable in reaching a contrary conclusion.

To be sure, we acknowledge that a plausible argument could be made here that the prosecutor's rebuttal arguments were "improper"-that is, one could plausibly contend that those remarks resemble the kind of "misuse" of the moral-culpability text that concerned the OCCA in Harris . 164 P.3d at 1114. We need not, however, definitively opine on the matter. The important point is that such a conclusion would not be determinative of the constitutional question before us: whether those prosecution arguments (improper or not) had the effect of violating Mr. Grant's Eighth Amendment rights under Lockett and its progeny. The test of constitutional error under Lockett is not (as relevant here) whether the prosecution's arguments were improper, but rather whether there is a reasonable likelihood that they had the effect of precluding the jury from considering mitigating evidence.

Put another way, even if we were to accept that the prosecution's rebuttal arguments here were improper, that would not necessarily mean that the OCCA was unreasonable in concluding that there was no Lockett error because there was no reasonable likelihood that the jury was precluded by those arguments from considering all of Mr. Grant's mitigating evidence-including the evidence that did not extenuate or reduce his moral culpability or blame. Indeed, considering the record as a whole-notably, the jury instructions and other unchallenged aspects of the prosecution's closing arguments, as well as the defense's closing arguments-we conclude that the OCCA was not unreasonable in ruling (in substance) that the prosecution's closing arguments did not violate Mr. Grant's Eighth Amendment rights established by Lockett and its progeny.

Significantly, based on the totality of the record, it would not have been unreasonable for the OCCA to conclude that, even if some of the prosecution's comments were improper, that it was not reasonably likely that the jury read the comments as doing anything more than vigorously-but permissibly-attacking the veracity, credibility, and weight of Mr. Grant's mitigating evidence, rather than barring or "cut[ting] off in an absolute manner," Johnson , 509 U.S. at 361, 113 S.Ct. 2658 (quoting McKoy , 494 U.S. at 456, 110 S.Ct. 1227 (Kennedy, J., concurring in judgment)), the jury's discretion to consider all of his mitigating evidence. Cf. Bland , 459 F.3d at 1026 (rejecting the petitioner's claim that he was deprived of a fair trial because "the prosecution improperly demeaned his mitigating evidence by suggesting that the jury ignore mitigating evidence and by recharacterizing the mitigating evidence as evidence in aggravation," after concluding that "[a]s long as the jury is properly instructed on the use of mitigating evidence, the prosecution is free to comment on the weight the jury should accord to it" and "[t]he prosecutors, while critical of [the petitioner's] mitigating evidence, never told the jury it could not consider [the petitioner's] mitigating evidence"); Fox v. Ward , 200 F.3d 1286, 1299-1300 (10th Cir. 2000) (rebuffing the petitioner's claim that he received "a fundamentally unfair trial," after determining that "the prosecutor merely commented on the weight that should be accorded to the mitigating factors" but "did not suggest that the jury was not permitted to consider the factors"). Indeed, in essence, this is precisely what the OCCA held. See Grant , 205 P.3d at 21 ("While there is no restriction whatsoever on what information might be considered *939mitigating, no juror is bound to accept it as such, and the State is free to try to persuade the jury to that end.").

Before turning to the specifics, we underscore that the question before us is not whether the OCCA's determination of this issue is wrong. Rather it is whether "it is possible fairminded jurists could disagree" about whether the OCCA's decision conflicts with Supreme Court precedent. Harrington , 562 U.S. at 102, 131 S.Ct. 770. And we conclude that such a disagreement is indeed possible.

ii

Our Hanson decision is also helpful in our resolution of Mr. Grant's prosecution-exploitation claim because there we rejected essentially the same challenge to an Oklahoma prosecutor's alleged exploitation of the moral-culpability text of the pattern mitigating-evidence jury instruction to preclude the jury's consideration the defense's proferred mitigating evidence. See Hanson , 797 F.3d at 851-52. Though the factual circumstances of Hanson are not entirely on all fours with this case, its mode of analysis is instructive and its substantive holding provides cogent support for the conclusion we reach here regarding Mr. Grant's claim.

The petitioner in Hanson argued that the prosecutor unconstitutionally exploited the moral-culpability text when "the prosecutor told the jury to consider whether any of the mitigating circumstances 'really extenuate or reduce [the petitioner's] degree of culpability or blame in this case.' " 797 F.3d at 851 (quoting the record). In rejecting this contention, the Hanson panel reasoned that the OCCA could reasonably conclude that it was not reasonably likely that the prosecutor's comment precluded the jury from considering mitigating evidence, in light of the jury instructions and the other unchallenged comments of the prosecution. See id. at 852 ("In light of all of the instructions and of the prosecutor's various comments, we find it hard to imagine that the jurors thought they were prohibited from considering any of the mitigating evidence they heard at the resentencing hearing."); see also id. ("We conclude that there is no reasonable likelihood that the jurors applied [the moral-culpability text of] Instruction No. 22 in a way that precluded them from considering mitigating evidence.").

(1)

The Hanson panel made only a shorthand reference to the jury instructions in the context of determining the prosecution-exploitation claim. Id. (saying "[i]n light of all of the instructions"). But the panel's extensive analysis of those instructions in disposing of the petitioner's challenge to the constitutionality of the instruction's moral-culpability text makes sufficiently clear to us the import of this reference. As noted supra , the Hanson panel concluded that "some of the other instructions from [the petitioner's] trial concerning mitigating evidence broadened the scope of evidence the jury could consider." Id. at 851. These instructions supported the panel's conclusion that the OCCA could have reasonably determined that the jury instruction containing the moral-culpability text was not itself unconstitutional. Id. The logical implications of the panel's analysis for its resolution of the prosecution-exploitation claim are patent. First, because the moral-culpability text itself was not unconstitutional-at least in the context of other, broadening instructions-the prosecutor's isolated references to that text, without more, did not effect a constitutional violation. And, second, by "broaden[ing] the scope of evidence the jury could consider," id. , the jury instructions *940militated against and served to counteract any allegedly impermissible efforts of the prosecution to limit the jury's consideration of mitigating evidence to only that evidence that extenuated or reduced moral culpability and blame. We are confident that the Hanson panel discerned these obvious implications and took them into account in rejecting the petitioner's prosecution-exploitation claim.

This mode of analysis based on jury instructions is significant here as well and contributes to our decision to reject Mr. Grant's claim. In this regard, we underscore at the outset that a jury is presumed to follow the trial court's instructions. See, e.g. , Weeks v. Angelone , 528 U.S. 225, 234, 120 S.Ct. 727, 145 L.Ed.2d 727 (2000) ("A jury is presumed to follow [the trial court's] instructions."); Louis Jones v. United States , 527 U.S. 373, 394, 119 S.Ct. 2090, 144 L.Ed.2d 370 (1999) ("The jurors are presumed to have followed these [trial court] instructions."). As we have noted supra , the instructions in Hanson that "broadened the scope of evidence the jury could consider," 797 F.3d at 851, also were present-in all material respects-in Mr. Grant's case. For instance, as in Hanson , id. at 849, 851, in the same instruction that included the moral-culpability text, there was language that vested the jury with the responsibility for determining what evidence was mitigating: "The determination of what circumstances are mitigating is for you to resolve under the facts and circumstances of this case." O.R. 2349 (Instr. 12).

Further, akin to Hanson , id. at 151, there was another instruction that informed the jury that "[e]vidence ha[d] been introduced as to [specified] ... mitigating circumstances " O.R. 2350 (Instr. 13) (emphasis added), and then listed ten items, "some of which had nothing to do with [the petitioner's] moral culpability," Hanson , 797 F.3d at 851. In other words, in this instruction-Instruction 13-the trial judge specifically characterized as "mitigating" factors that ordinarily would not be deemed to have extenuated or reduced Mr. Grant's moral culpability or blame. See Hanson , 797 F.3d at 851 (noting that some of the instructions that did not extenuate moral culpability or blame discussed the petitioner's family and emotional history and his fatherhood of a young son).

For instance, Instruction 13 listed the following:

• "A substantial portion of Donald Grant's childhood was spent in a violent and drug-infested neighborhood."

• "For extended periods of time, Donald Grant's mother was unable or unwilling to take care of him to the extent that he sometimes was deprived of food and nurturing."

• "Donald Grant's life will be of value to other persons besides himself."

• "Donald Grant's family and cultural history indicate that he did not receive what most families consider important for their children to have success in the world."

• "Donald Grant periodically became a ward of the government at a young age."

O.R. 2350-51.

Significantly, as evident from the foregoing quotation, many of the identified mitigating circumstances that did not extenuate or reduce moral culpability or blame-viz. , that fell outside of the scope of the moral-culpability text-related to Mr. Grant's difficult and turbulent upbringing. These factors closely reflected a major thrust of Mr. Grant's mitigation evidentiary presentation and related closing arguments. As his defense counsel observed in *941closing argument, the "horrible circumstances in [Mr. Grant's] upbringing" constituted one of "two categories" of circumstances it had stressed in its evidentiary presentation and would discuss in his oral argument (the other being that Mr. Grant "suffered from a very serious mental illness"). R., Vol. IV, Trial Tr. VIII, at 47-48. One could reasonably conclude that the jury might be inclined to view evidence that played such a marquee role in the defense's mitigation evidentiary presentation and oral argument as actually being legally permissible mitigation evidence, absent a clear court instruction to the contrary. After all, the court did not stop the defense from putting the evidence before the jury. See Boyde , 494 U.S. at 384-85, 110 S.Ct. 1190 ("All of the defense evidence presented at the penalty phase ... related to petitioner's background and character, and we think it unlikely that reasonable jurors would believe the court's instructions transformed all of this 'favorable testimony into a virtual charade.' ... Presentation of mitigating evidence alone, of course, does not guarantee that a jury will feel entitled to consider that evidence. But the introduction without objection of volumes of mitigating evidence certainly is relevant to deciding how a jury would understand an instruction which is at worst ambiguous." (emphasis added) (citation omitted) (quoting California v. Brown , 479 U.S. 538, 542, 107 S.Ct. 837, 93 L.Ed.2d 934 (1987) ); accord Buchanan v. Angelone , 522 U.S. 269, 278, 118 S.Ct. 757, 139 L.Ed.2d 702 (1998). There was no such clearly contrary instruction here. Indeed, the obverse was true: there was a clear introductory instruction from the court that identified these horrific-upbringing factors-which did not extenuate or reduce moral culpability or blame-as "mitigating circumstances." O.R. 2350.

And it would have been reasonable to conclude that Instruction 13's list would have tended to militate against and counteract any belief that the prosecution's rebuttal argument might have planted in the jury's mind that it was precluded from considering evidence in mitigation that did not extenuate or reduce moral culpability or blame. This is especially so because-as in Hanson , 797 F.3d at 851 -Instruction 13 ended with this admonition: "In addition, you may decide that other mitigating circumstances exist , and if so, you should consider those circumstances as well." O.R. 2351 (emphasis added). In other words, in the context of an instruction that included factors that did not extenuate or reduce moral culpability or blame, the court advised that the jury was free to determine that other circumstances also should be given mitigating effect. Thus, the jury might reasonably infer from this instruction that-among the other circumstances that it could deem to be mitigating-would be those like the ones listed in Instruction 13 that did not extenuate or reduce moral culpability or blame.

In sum, the Hanson panel relied in significant part on other unchallenged jury instructions in the record in concluding that the OCCA would not have been unreasonable in determining that the prosecution's closing argument did not have the unconstitutional effect of precluding the jury from considering the petitioner's proferred mitigating evidence that did not extenuate or reduce moral culpability or blame. The materially similar instructions in Mr. Grant's record lead us in the same direction.

In addition, the trial court here specifically admonished the jury that its instructions "contain all the law and rules you must follow," O.R. 2357 (Instr. 17), and during the course of the prosecution's arguments reminded the jury on two occasions that the prosecution's statements were "argument only" and "for purposes of *942persuasion." R., Vol. IV, Trial Tr. Vol. VIII, at 37; see id. at 34. One could reasonably conclude that these instructions also would have tended to make it less reasonably likely that the jurors would have "applied Instruction [12] in a way that precluded them from considering mitigating evidence," Hanson , 797 F.3d at 852, despite any prosecution arguments that may have had the effect (whether inadvertently or not) of militating to an appreciable extent in favor of such preclusion. See Boyde , 494 U.S. at 384, 110 S.Ct. 1190 ("[A]rguments of counsel generally carry less weight with a jury than do instructions from the court. The former are usually billed in advance to the jury as matters of argument, not evidence, and are likely viewed as the statements of advocates; the latter, we have often recognized, are viewed as definitive and binding statements of the law." (citation omitted); see also Fox , 200 F.3d at 1299 ("Only the court instructs the jury. The prosecutor merely argues to the jury.").

(2)

The Hanson panel's rejection of the petitioner's claim also took account of other unchallenged prosecution arguments that it deemed to have a corrective effect. Specifically, the panel noted that "the prosecutor made a number of other comments to the jury that encouraged them to consider any and all mitigating evidence they thought relevant." Hanson , 797 F.3d at 851. For example, apparently in his opening closing argument, the prosecutor referred to the instruction (which we have discussed supra ) that contained a list of mitigating circumstances-"some of which had nothing to do with [the petitioner's] moral culpability"-and told the jury to "[g]o back, talk about those. Think about those. Make a decision." Id. (quoting the trial record). "And in rebuttal argument, the prosecutor said, 'Consider all the mitigating circumstances and you'll consider what weight [to give them.]' " Id. at 851-52 (alteration in original) (quoting trial record). The panel thus concluded that "[t]hese statements demonstrate that the prosecutor encouraged the jury to consider all sorts of mitigating evidence," including the kind that did not extenuate or reduce moral culpability or blame. Id. at 852.

Following Hanson 's analytical methodology, we also have inquired into the other unchallenged prosecution comments to assess whether they would have made it less likely that a jury would have interpreted the arguably improper prosecution rebuttal arguments here as precluding them from considering mitigating evidence that did not extenuate or reduce moral culpability or blame. We believe that those other unchallenged prosecution comments would have made such an impermissible jury interpretation less likely. The prosecutor handling the opening closing argument (who commenced the round of oral arguments)-i.e., Suzanne Lister-spent the lion's share of her time casting doubt on the veracity, credibility, and weight of the evidence supporting the mitigating circumstances that the court identified in Instruction 13. Yet, as discussed supra , many of these circumstances cannot be deemed ones that extenuate or reduce moral culpability or blame. Nevertheless, as in Hanson (797 F.3d at 851 ), the prosecutor offered the following advice to the jury:

These [factors of Instruction 13] are for you to consider . You don't have to accept them. You can talk about them, you can talk amongst yourselves, you can talk about the testimony. ...

....

The defendant has alleged the following mitigating circumstances [in Instruction 13]: And I want to talk about them individually. And it's up to you to deter mine whether or not these mitigators - whether or not these circumstances somehow mitigate what Donald Anthony Grant did *943....

R., Vol. IV, Trial Tr. VIII, at 31-32 (emphases added).

At no point during her opening closing remarks did Ms. Lister assert that the jury was not free under the law to consider all of the mitigating factors that the court identified in Instruction 13 on the ground that some of them did not extenuate or reduce moral culpability or blame. In other words, the first voice that the jury heard during closing arguments identified several circumstances-which the court had characterized as mitigating-that did not have the effect of extenuating or reducing moral culpability or blame and, yet, this voice never questioned whether those circumstances qualified under the law as mitigating evidence. The jury might logically infer from this presentation that the evidence actually did legally qualify as mitigating evidence, and that the question before them was the one that Ms. Lister hammered on: whether there was sufficiently accurate, credible, and weighty evidence to support a jury finding as to these alleged mitigating circumstances. Cf. Ayers v. Belmontes , 549 U.S. 7, 16-17, 127 S.Ct. 469, 166 L.Ed.2d 334 (2006) ("It is improbable the jurors believed that the parties were engaging in an exercise in futility when [the defendant] presented (and both counsel later discussed) his mitigating evidence in open court. Arguments by the prosecution and the defense assumed the evidence was relevant."). Such an inference that all of the Instruction 13 mitigating evidence was legally proper (even if otherwise insufficient) would have tended to undermine a subsequent suggestion by the prosecution (Ms. Elliott) in the rebuttal closing argument that the jury was legally precluded by the moral-culpability text from considering any evidence in mitigation that did not extenuate or reduce moral culpability or blame.

To be sure, unlike Hanson , there were no further statements from the prosecution-i.e., Ms. Elliott-in rebuttal closing that could reasonably suggest that "the prosecutor encouraged the jury to consider all sorts of mitigating evidence." Hanson , 797 F.3d at 852. Instead, on more than one occasion, the prosecution admittedly argued in rebuttal closing that "what the law says is that before something can be mitigating it must reduce the moral culpability or blame of the defendant." R., Vol. IV, Trial Tr. at 75; see id. at 79-80 (admonishing the jury that after it determines the evidentiary sufficiency and veracity of the proferred mitigation evidence, it must "ask ... does it reduce [Mr. Grant's] moral blame for what happened at the La Quinta Inn in July of 2001").

Therefore, we acknowledge that a plausible argument could be made under Hanson 's analytical methodology- due to the comparatively greater strength of the permissible prosecution mitigation-related statements in Hanson -that Hanson is a stronger case than this one for concluding that it was not reasonably likely that the jury was precluded from considering mitigating evidence. In this vein, Mr. Grant argues that, unlike the prosecutor in Hanson who "extensive[ly] 'encourage[d]' " the jurors to consider the mitigating circumstances, the prosecutors in his case made more "generic" comments that never made up for the combined unconstitutional effect of the jury instruction and the prosecutors' limiting statements regarding mitigation in rebuttal closing. Aplt.'s Opening Br. at 106. And Mr. Grant further suggests that any ameliorative statements the prosecutors made in his case were less effective than those in Hanson because they were made before the allegedly improper statements;

*944according to Mr. Grant, in his case, the prosecutors' "principal cabining [i.e., limiting] of mitigation evidence [took place] in their second closing [and] ... was among the last things the jury heard before their deliberations." Id.23

However, whether Hanson actually is a stronger case is immaterial. As explicated further infra , Hanson is not the measuring stick under AEDPA for assessing whether the OCCA acted unreasonably in resolving this prosecution-exploitation claim; Supreme Court law is.24 And Hanson certainly does not purport to establish the floor or an essential set of circumstances necessary for the State to prevail on such a claim. Hanson simply provides a useful analytical framework for resolving this case under its own unique facts because in Hanson we rejected essentially the same prosecution-exploitation challenge that Mr. Grant raises here.

In any event, even if we accept for purposes of argument that the legally permissible prosecution statements in this case were more "generic" than those in Hanson , that does not mean that, coupled with the jury instructions, those statements did not adequately highlight for the jury its singular responsibility to consider all evidence proferred in mitigation-including evidence that did not extenuate or reduce moral culpability or blame. And the OCCA could reasonably conclude that they did so. Furthermore, any fair comparison of the strengths of the two cases would have to take into account on the other side of the ledger the fact that there were instructions present here that were not mentioned in Hanson that expressly informed the jury that it should not treat the lawyers' arguments as expressing the governing law of the case. See, e.g. , O.R. 2357 (court admonishing the jury that its instructions "contain all the law and rules you must follow"); R., Vol. IV, Trial Tr. Vol. VIII, at 37 (court reminding the jury that the prosecution's closing-argument statements were "argument only" and "for purposes of persuasion"). Presuming that jurors follow the court's instructions, see *945Weeks , 528 U.S. at 234, 120 S.Ct. 727, those instructions would have aided the jury, in a way that may not have been not possible in Hanson , in avoiding any improper prosecution efforts to limit the scope of mitigation evidence to those things that extenuated or reduced moral culpability.

Moreover, Mr. Grant's argument regarding the timing of the allegedly harmful statements-viz. , they should be deemed more prejudicial than those in Hanson because they were delivered closer to the time the jury began its deliberations-relies on nothing more than speculation and conjecture about the impact of the timing of the statements on the jurors' decision-making. One could just as well argue that the ameliorative statements of the prosecution (i.e., Ms. Lister) in the opening closing arguments-which were the first thing that the jury heard in that phase of the trial-would have had a more powerful impact on the jury than the statements in rebuttal closing. Compare Richard B. Klein, TRIAL COMMUNICATION SKILLS 15:4 (2d ed.) Westlaw (database updated Nov. 2017) ("[I]t is not possible to establish a fixed rule saying that the best tactic will always be to apply the primacy or the recency theory ...."), and Spencer H. Silvergate, Closing Argument , 25 TRIAL ADVOCATE QUARTERLY 28, 29 (2005) ("The laws of primacy and recency tell us that people best remember what they hear first and last. Therefore, the strongest points should be made at the beginning and end of the closing."), with Bill Kanasky, Jr., The Primacy and Recency: The Secret Weapons of Opening Statements , 33 TRIAL ADVOCATE QUARTERLY 26, 29 (2014) ("The recency effect is far less powerful [than certain primacy effects], as it is a simple enhancement of short-term memory due to recent exposure to information. In other words, it is easy to remember information that is presented an hour ago compared to information from a week ago."). Mr. Grant cites no legal authority to support his position that the later comments in rebuttal closing in his case prejudiced him more than the ones in Hanson . For the foregoing reasons, we reject both of these arguments based on the comparative strength of Hanson .

At bottom, we must determine whether the OCCA would have been unreasonable in concluding that there was no reasonable likelihood that the jury was precluded by the prosecution's closing arguments from considering all of Mr. Grant's mitigation evidence-including the evidence that did not extenuate or reduce his moral culpability or blame. Considering the record as a whole-notably, the jury instructions, other unchallenged aspects of the prosecution's closing arguments and the closing arguments of the defense-we conclude that OCCA would not have been unreasonable. In this regard, we hold that the OCCA's decision was not contrary to or an unreasonable application of Lockett and its progeny. At the very least, "it is possible fairminded jurists could disagree" about whether the OCCA's decision conflicts with Supreme Court precedent. Harrington , 562 U.S. at 102, 131 S.Ct. 770.

iii

Regarding Supreme Court precedent, we do not lose sight of the fact that the focus of our deferential review under AEDPA is the clearly established law of the Supreme Court. In that regard, we are hard-pressed to conclude that the OCCA's determination of Mr. Grant's prosecution-exploitation claim was contrary to or an unreasonable application of Lockett and its progeny because Mr. Grant cannot point to even one case where the Supreme Court has approved of a grant of habeas relief under circumstances like those here. Put another way, we are reluctant to conclude that the OCCA's approach was unreasonable *946when there is no relevant guidance from the Supreme Court.

In arguing for a contrary result, Mr. Grant highlights three Supreme Court cases in which the Court examined a jury instruction and related prosecution statements under Lockett , and ultimately remanded for resentencing. However, we discern nothing in these cases to alter our conclusion.

First, he cites Penry v. Lynaugh , 492 U.S. 302, 109 S.Ct. 2934, 106 L.Ed.2d 256 (1989)abrogated on other grounds by Atkins v. Virginia , 536 U.S. 304, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002), in asserting that "[t]he Supreme Court has recognized similar [prosecution] arguments exacerbate faulty instructions." Aplt.'s Opening Br. at 107 (emphasis added). In Penry , the Court concluded that the court's jury instructions were inherently flawed and, therefore, the prosecutor's argument "stressing that the jurors had taken an oath to follow the law, and that they must follow the instructions they were given" served to compound the error of the instructions. 492 U.S. at 325, 109 S.Ct. 2934. The Court concluded that "[i]n light of the prosecutor's argument, and in the absence of appropriate jury instructions , a reasonable juror could well have believed that there was no vehicle for expressing the view that Penry did not deserve to be sentenced to death based upon his mitigating evidence." Id. at 326, 109 S.Ct. 2934 (emphasis added). In contrast, following Hanson 's lead, we have already concluded supra that the OCCA would not have been unreasonable in concluding that the moral-culpability text of Instruction 12 was not unconstitutional under Lockett and its progeny, at least when read in the context of other instructions in the record. In other words, unlike in Penry , it is not established here that the instruction at issue was inherently flawed or defective as a matter of law.

Moreover, even accepting that some of the prosecutors' comments here urged the jury to follow the law, as they did in Penry , if the jury had done so, it not only would have followed the moral-culpability text of Instruction 12, but notably also Instruction 13, which "set[ ] out the mitigators that the defendant allege[d the jury] should consider with regards to determining whether or not this is mitigation," R., Vol. IV, Trial Tr. VIII, at 31. And, as previously noted, many of those mitigating circumstances listed in Instruction 13 did not involve matters that properly could be viewed as extenuating or reducing moral culpability or blame. Therefore, the OCCA could have reasonably viewed the jury here-unlike in Penry -as having had a "vehicle for expressing the view that [Mr. Grant] did not deserve to be sentenced to death based upon his mitigating evidence," Penry , 492 U.S. at 326, 109 S.Ct. 2934 -including his evidence that did not extenuate or reduce moral culpability or blame. In any event, there is at least a "possibility fairminded jurists could disagree" about whether the OCCA would have been correct to adopt this position. Harrington , 562 U.S. at 102, 131 S.Ct. 770. Accordingly, Mr. Grant's reliance on Penry is unavailing.

Our exposition of our reasoning with respect to Penry permits us to demonstrate in short order the flaw in Mr. Grant's reliance on the second case, LaRoyce Smith v. Texas , 543 U.S. 37, 125 S.Ct. 400, 160 L.Ed.2d 303 (2004). As in Penry - and contrary to this case-the Court had declared that the instruction at issue in Smith itself was "constitutionally inadequate." 543 U.S. at 48, 125 S.Ct. 400. And, like Penry , but not this case, the prosecutor's remarks exacerbated the *947prejudicial nature of the legally infirm instruction because the prosecutor "similarly reminded the jury that each and every one of them had promised to 'follow the law' and return a 'Yes' answer to the special issues so long as the State met its burden of proof." Id. at 48 n.5, 125 S.Ct. 400 (quoting petitioner's certiorari petition). For the reasons we explicated above in addressing Penry , LaRoyce Smith lends Mr. Grant no succor.

Lastly, Mr. Grant comes up the shortest in his reliance on the third case- Caldwell v. Mississippi , 472 U.S. 320, 323, 105 S.Ct. 2633, 86 L.Ed.2d 231 (1985). He cites Caldwell for the proposition that "the type of misconduct at issue is one that invades specific Constitutional rights" and under a "strict scrutiny" standard of review applicable to such invasions, the "OCCA's ruling was ... contrary to law." Aplt.'s Opening Br. at 107. However, it is patent that Caldwell stands on a distant shore from this case. There, the Supreme Court considered the constitutionality of "a prosecutor urg[ing] the jury not to view itself as determining whether the defendant would die, because a death sentence would be reviewed for correctness by the State Supreme Court." 472 U.S. at 323, 105 S.Ct. 2633 (emphasis added). The Court in Caldwell held that "it is constitutionally impermissible to rest a death sentence on a determination made by a sentencer who has been led to believe that the responsibility for determining the appropriateness of the defendant's death rests elsewhere." Id. at 328-29, 105 S.Ct. 2633. Mr. Grant has not identified any such statements by the prosecutors in this case regarding the jury's role and the role of the reviewing court. Put directly, there is nothing in Caldwell that avails Mr. Grant.

In sum, we are not convinced-with our focus properly set on Supreme Court caselaw, and viewing the record as a whole-that the OCCA's conclusion that it was not reasonably likely that the prosecution's comments in closing argument unconstitutionally restricted the evidence the jurors could consider in mitigation was contrary to or an unreasonable application of Lockett and its progeny.

c

Finally, Mr. Grant argues that the OCCA's "unreasonable assessment ... is likely traceable to factual error and confusion revealed in the OCCA opinion."25 Aplt.'s Opening Br. at 109. Specifically, Mr. Grant points to footnote 34, in which the OCCA purported to refer to the jury *948instruction at issue-Instruction 12-but, by mistake , described the post- Harris reformed version of this standard mitigation instruction, which was never given in Mr. Grant's case. This misstatement, however, is merely a scrivener's error. In footnote 33, the OCCA explicitly stated that the Grant trial court applied the pre- Harris (i.e., pre-reformation) mitigation jury instruction-which contained the moral-culpability text-thus indicating an understanding of the pre- and post- Harris instruction distinction. As such, in our view, footnote 34's reference to this reformed instruction is a mere mistake, lacking in decisional or legal significance. It certainly does not suggest that the OCCA's ruling rested on "an unreasonable determination of the facts in light of the evidence." 28 U.S.C. § 2254(d)(2). Mr. Grant's argument is without merit.

* * *

In sum, Mr. Grant has not demonstrated that the OCCA's rejection of his challenge to Instruction 12-and, more specifically, that instruction's moral-culpability text-and the prosecution's statements relating to the instruction is contrary to or an unreasonable application of clearly established federal law, or premised on an unreasonable determination of the facts. Therefore, we uphold the district court's denial of habeas relief on this claim.

D. Batson Claim

Mr. Grant argues that he is entitled to habeas relief because the trial court permitted a prosecutor to use a peremptory strike to exclude a potential juror on the basis of race. Mr. Grant raised this claim before the OCCA on direct appeal, and the OCCA rejected the claim on the merits. Therefore, we ordinarily would accord AEDPA deference to the OCCA's resolution of this claim. Mr. Grant contends, however, the OCCA's decision was contrary to the controlling Supreme Court case- Batson v. Kentucky , 476 U.S. 79, 106 S.Ct. 1712, 90 L.Ed.2d 69 (1986) -and its progeny, because "the OCCA rejected, or at the very least devalued, the use of comparative juror analysis." Aplt.'s Opening Br. at 116. Therefore, reasons Mr. Grant, de novo review is the proper standard for assessing the merits of his Batson claim. More specifically, Mr. Grant asserts that, applying a de novo standard of review, we should conclude that "the State's discriminatory strike ... deprived Grant [of] due process as recognized in Batson ." Id. at 118, 106 S.Ct. 1712. Notably, Mr. Grant does not explicitly contend that the OCCA's Batson ruling is also infirm under AEDPA deference-that is, that we also should conclude that, under this deferential standard, the OCCA committed Batson error.

We determine that the OCCA's decision-and, more specifically, its treatment of comparative-juror analysis-is not contrary to, or an unreasonable application of Batson and its progeny. Therefore, AEDPA supplies the appropriate standard of review. And, given Mr. Grant's failure to make an explicit Batson argument under AEDPA, we could very well end our analysis there. But, even were we to reach the merits of the Batson claim under AEDPA, Mr. Grant could not prevail. We would conclude that the OCCA's Batson determination neither contravenes the legal nor the factual standards of AEDPA.

1. Legal Framework

In Batson , the Supreme Court held that "the Fourteenth Amendment's Equal Protection Clause prohibits the prosecution's use of peremptory challenges to exclude potential jurors on the basis of their race." House , 527 F.3d at 1020 (quoting Saiz v. Ortiz , 392 F.3d 1166, 1171 (10th Cir. 2004) ). The Batson Court provided a *949three-step analysis for determining whether the prosecution impermissibly used a peremptory challenge:

First, a defendant must make a prima facie showing that a peremptory challenge has been exercised on the basis of race [. S]econd, if that showing has been made, the prosecution must offer a race-neutral basis for striking the juror in question [. T]hird, in light of the parties' submissions, the trial court must determine whether the defendant has shown purposeful discrimination.

Miller-El II , 545 U.S. at 277, 125 S.Ct. 2317 (alterations omitted) (quoting Miller-El v. Cockrell ("Miller-El I "), 537 U.S. 322, 328-29, 123 S.Ct. 1029, 154 L.Ed.2d 931 (2003) ). Notably, in Miller-El II , the Court underscored " Batson 's explanation that a defendant may rely on 'all relevant circumstances' to raise an inference of purposeful discrimination." Id. at 240, 125 S.Ct. 2317 (quoting Batson , 476 U.S. at 96-97, 106 S.Ct. 1712 ). Consequently, the Court has "made it clear that in considering a Batson objection, or in reviewing a ruling claimed to be Batson error, all of the circumstances that bear upon the issue of racial animosity must be consulted ." Snyder v. Louisiana , 552 U.S. 472, 478, 128 S.Ct. 1203, 170 L.Ed.2d 175 (2008) (emphasis added); accord Foster v. Chatman , --- U.S. ----, 136 S.Ct. 1737, 1748, 195 L.Ed.2d 1 (2016).

"[I]f, at step three, the court finds the proffered ground to be pretextual, it may determine that the strike was purposeful discrimination." Black v. Workman , 682 F.3d 880, 894 (10th Cir. 2012). That "step turns on factual determinations," to which a reviewing court ordinarily accords deference. Foster , 136 S.Ct. at 1747 (quoting Snyder , 552 U.S. at 477, 128 S.Ct. 1203 ); see United States v. Atkins , 843 F.3d 625, 633 (6th Cir. 2016) (describing "the ultimate issue" as "factual" where "[d]efendant is squarely challenging the district court's finding at Batson step three that the government lacked intent to discriminate").

More specifically, in the AEDPA context, the deferential analytical rubric of § 2254(d)(2) comes into play and, to grant relief, "a federal habeas court must find the state-court conclusion 'an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.' " Rice v. Collins , 546 U.S. 333, 338, 126 S.Ct. 969, 163 L.Ed.2d 824 (2006) ; accord Black , 682 F.3d at 880. And § 2254(e)(1)'s "presumption of correctness" applies to state court findings relating to the ultimate factual question at Batson 's step three-i.e., purposeful discrimination; consequently, such findings can only be rebutted by "clear and convincing evidence." 28 U.S.C. § 2254(e)(1) ; see Black , 682 F.3d at 898 ("The question before us is whether the prosecution's failure to strike a white juror despite his not disclosing a prior misdemeanor conviction would establish by clear and convincing evidence that the prosecutor's challenge to the prospective African-American juror was racially motivated." (emphasis added)); Moody v. Quarterman , 476 F.3d 260, 269 (5th Cir. 2007) ("[T]he district court's task was not to assess whether it agreed with the state court's ruling, but to determine whether the state court's finding [at Batson step three] was entitled to the presumption of correctness and to decide whether that determination was unreasonable in light of the evidence presented."); see also Akins v. Easterling , 648 F.3d 380, 392 (6th Cir. 2011) (noting that "[t]he question at step three of Batson -whether the defendant has shown intentional discrimination on the basis of race-is a question of fact" and applying AEDPA's presumption of correctness to such findings);

*950Weaver v. Bowersox , 241 F.3d 1024, 1030 (8th Cir. 2001) (applying AEDPA's presumption of correctness to a state court's factual findings pertaining to its Batson step three determination); cf. Hernandez v. New York , 500 U.S. 352, 367, 111 S.Ct. 1859, 114 L.Ed.2d 395 (1991) (observing in the direct-appeal context that "[w]hether a prosecutor intended to discriminate on the basis of race in challenging potential jurors is, as Batson recognized, a question of historical fact."); Madison v. Comm'r, Ala. Dep't of Corr. , 761 F.3d 1240, 1251 n.11 (11th Cir. 2014) (noting that "the state court made no factual determination at Batson 's third step that could be entitled to a presumption of correctness").

2. Background and Adjudication on Direct Appeal

Mr. Grant's Batson claim concerns potential juror Valerie Jamerson, an African-American woman. When asked to give a race-neutral explanation for removing Ms. Jamerson, the prosecution stated that Ms. Jamerson "was one of the jurors who said she could not look at photographs and did not want to look at photographs that might be graphic." R., Vol. IV, Trial Tr. III, at 113. Mr. Grant challenged the prosecution's justification on direct appeal, "pointing out that several non-minority panelists who were not challenged by the State had expressed similar discomfort at having to consider photographs that were assured to be gruesome." Grant , 205 P.3d at 15. The OCCA rejected this claim of Batson error:

[R]acially-motivated discrimination is not established simply because panelists of different races provide similar responses, and one is excused while the other is not. Rather, all the attendant circumstances are relevant to whether the strike was racially motivated. [ Miller-El II , 545 U.S. at 240, 125 S.Ct. 2317.] [Mr. Grant] argues that statistical analysis-e.g. the percentage of minorities removed by the State from the panel-is relevant to whether racial discrimination was at work. But statistical analysis does not advance [Mr. Grant]'s argument here. The State used only two of its nine peremptory challenges to remove minority panelists, and left more minorities on the panel than it removed.

Id. (emphases added) (citations omitted). The OCCA thus concluded, "[t]he prosecutor's explanation for striking Panelist J. [i.e., Ms. Jamerson] was sufficiently race-neutral," and denied Mr. Grant's Batson claim. Id.

3. Analysis

a

Mr. Grant argues that the OCCA's rejection of his Batson claim is contrary to clearly established federal law. Specifically, he argues that the OCCA "rejected, or at the very least devalued, the use of comparative juror analysis," and that this "approach was contrary to [the] clearly established law of Miller-El[II] ." Aplt.'s Opening Br. at 116. More specifically, he challenges the OCCA's statement that "racially-motivated discrimination is not established simply because panelists of different races provide similar responses, and one is excused while the other is not." Id. at 114-15 (quoting Grant , 205 P.3d at 15 ). Mr. Grant reads this statement to mean that the OCCA "dismiss[ed] comparative juror analysis as a means of detecting discrimination." Aplt.'s Opening Br. at 114. "This approach," Mr. Grant asserts, "was both contrary to law and unreasonable." Id. at 115. We disagree with Mr. Grant's analysis.

At the outset, it is important to clarify what Batson and its progeny do not expressly hold. This line of cases does not explicitly state that comparative juror evidence, standing alone, must be accorded *951determinative effect with regard to the question of racially-motivated discrimination-i.e., Batson 's third step. In other words, these cases do not explicitly provide that, if a racial minority juror and a non-minority juror voice a similar response during voir dire and the prosecutor excludes only the minority juror, that this is conclusive proof of the prosecutor's discriminatory motive.

Notably, Miller-El II -the case that Mr. Grant principally relies on-does not articulate this proposition. Rather, it is reasonably read as standing for the patently logical view that such evidence is probative of discriminatory intent and, indeed, may be persuasive evidence thereof. In this regard, the Court in Miller-El II stated, "If a prosecutor's proffered reason for striking a black panelist applies just as well to an otherwise-similar nonblack who is permitted to serve, that is evidence tending to prove purposeful discrimination to be considered at Batson 's third step." Miller-El II , 545 U.S. at 241, 125 S.Ct. 2317 (emphasis added); see also Black , 682 F.3d at 897 ("Whenever the prosecutor's explanation for striking a minority juror would also apply to a white juror who was not struck, the explanation loses some credibility."). Thus, by its plain terms, Miller-El II establishes only that such comparative juror analysis "tend[s] to prove purposeful discrimination," not that it is definitive proof of such discrimination. 545 U.S. at 241, 125 S.Ct. 2317 (emphasis added).

Indeed, Miller-El II and subsequent cases of the Court underscore that "in considering a Batson objection, or in reviewing a ruling claimed to be Batson error, all of the circumstances that bear upon the issue of racial animosity must be consulted ." Snyder , 552 U.S. at 478, 128 S.Ct. 1203 (emphases added); accord Foster , 136 S.Ct. at 1748 ; see Miller-El II , 545 U.S. at 240, 125 S.Ct. 2317 (highlighting " Batson 's explanation that a defendant may rely on 'all relevant circumstances' to raise an inference of purposeful discrimination" (quoting Batson , 476 U.S. at 96-97, 106 S.Ct. 1712 )).

That is not to say that, under certain circumstances, the Court has not expressly found that comparative juror evidence was "compelling." Foster , 136 S.Ct. at 1754. However, even in Foster -a case where the evidence fit this description-the Court recognized that, in conducting its third-step, discriminatory-intent analysis, "that is not all." Id. In addition to comparative juror evidence, the Court noted the prosecutor's "shifting explanations" for dismissing the minority jurors and the "persistent focus on race in [notes relating to voir dire that the Court attributed to the prosecution team]." Id. Concluding its analysis, the Foster Court stated, "Considering all of the circumstantial evidence that 'bear[s] upon the issue of racial animosity,' we are left with the firm conviction that the strikes of [the minority jurors] were 'motivated in substantial part by discriminatory intent.' " Id. (quoting Snyder , 552 U.S. at 478, 485, 128 S.Ct. 1203 ). Thus, in Foster , even the "compelling" comparative juror evidence was not dispositive on the question of racially-discriminatory intent at Batson 's third step.

In light of the foregoing, we cannot conclude that the OCCA's ruling is contrary to the clearly established federal law of Batson and its progeny. There is at least a "possibility fairminded jurists could disagree," Harrington , 562 U.S. at 102, 131 S.Ct. 770, about whether the OCCA "rejected" or "devalued" the role that the Supreme Court has indicated that comparative juror analysis plays in the Batson discrimination rubric, Aplt.'s Opening Br. at 116, by merely concluding that "racially-motivated discrimination is not established simply because panelists of different races *952provide similar responses, and one is excused while the other is not," Grant , 205 P.3d at 15 (emphasis added). Indeed, such jurists are more likely to agree that the OCCA's ruling does not deviate from the clear pronouncements of the Batson line of cases and, more specifically, Miller-El II , which the OCCA explicitly cited-in that the OCCA explicitly recognized that "all the attendant circumstances are relevant to whether the strike was racially motivated." Id. (emphasis added).

Because the OCCA's Batson ruling, as it pertains to comparative juror evidence, is not contrary to Batson and its progeny, we must reject Mr. Grant's call for us to apply de novo review on this basis to his Batson claim. Significantly, Mr. Grant does not explicitly contend that the OCCA's Batson ruling is also infirm under AEDPA deference-viz. , he does not argue that we should conclude under AEDPA that the OCCA committed Batson error. Therefore, we could end our analysis here. However, even were we to reach the merits of the Batson claim under AEDPA, Mr. Grant could not prevail. Specifically, as we have previously discussed, the OCCA's Batson determination regarding Ms. Jamerson was neither contrary to nor an unreasonable application of Batson . And, most relevant here, it also was not a decision "based on an unreasonable determination of the facts in light of the evidence." 28 U.S.C. § 2254(d)(2), and there is no evidence that would clearly and convincingly rebut the presumption of correctness that attaches to the OCCA's factual findings regarding whether the prosecutor acted with purposeful discrimination in excluding Ms. Jamerson from the jury.

b

Like Mr. Grant's briefing, we focus on the third step of Batson -that is, whether the prosecutor engaged in purposeful discrimination in striking Ms. Jamerson. Significantly, Mr. Grant's discussion of his Batson claim centers solely on comparative juror evidence. In other words, he does not identify other circumstances from which the OCCA could have found a Batson violation at step three. Therefore, we similarly focus on this evidence.

The comparative juror evidence in the record clearly indicates that the OCCA was not unreasonable in finding that there was no showing of purposeful discrimination in violation of Batson . The OCCA could have reasonably determined on the record that Ms. Jamerson and the two identified non-minority jurors were not similarly situated. Therefore, the prosecutor's differential treatment of them would not be "evidence tending to prove purposeful discrimination to be considered at Batson 's third step." Miller-El II , 545 U.S. at 241, 125 S.Ct. 2317 (emphasis added).

More specifically, as noted, the prosecution indicated that it peremptorily dismissed Ms. Jamerson because she expressed hesitancy about viewing graphic photos. Mr. Grant argues that "[n]on-minority jurors Abrams and Dinwiddie expressed the same concerns about the gruesome photographs [as Ms. Jamerson], but were not stricken." Aplt.'s Opening Br. at 114. In our view, the OCCA would not have been unreasonable in reaching a contrary conclusion.

The OCCA could have reasonably concluded that Ms. Jamerson expressed more concern about viewing graphic photographs than Jurors Abrams and Dinwiddie. Upon questioning regarding her ability to look at graphic photographs, Ms. Jamerson said "I think it would be difficult. Even if it's not real, it's on TV, I have problems looking at it. ... [I]t's [too hard] sometimes to stop your emotions." R., Vol. IV, Trial Tr. I, at 226. She later stated, "I think I would have a hard time. Even on *953TV when it is not real I can't watch it. I don't want to see it." Id. at 227. Finally, the prosecutor asked, "[T]he key is do you have the stomach to do it, if you are called upon to serve the community?" and Ms. Jamerson responded, "If I was called on." Id. at 228. Ms. Jamerson thus expressed clear reluctance three times before she agreed that she could look at the graphic photographs and, notably, indicated that even when she knows the graphic images are not real (i.e., on TV), she has trouble watching them.

By contrast, Juror Abrams could reasonably be viewed as expressing a lesser degree of concern. Juror Abrams agreed that she could look at graphic photographs the first time that she was asked about the matter in voir dire. Specifically, the prosecutor told her that "you may see some things that are pretty graphic," and asked Ms. Abrams "how [she would] feel about that." Id. at 248. Juror Abrams responded, "I think it would be upsetting, but I think I can handle it." Id. The prosecutor said, "12 people have to listen to the facts of this case, have to look at the photographs and have to make a determination. Are you okay with that if you are called upon?" Id. Juror Abrams responded, "Yes." Id.

Juror Dinwiddie likewise could reasonably be viewed as expressing less concern about the gruesome photographs than Ms. Jamerson. When Juror Dinwiddie was seated, the prosecutor asked "do you understand that there may be some very graphic testimony, there may be some very graphic photographs that as a juror you have to look at." Id. , Trial Tr. III, at 31. And Juror Dinwiddie responded, "Yes, I understand." Id. The prosecutor added, "And is that something that you can do?" Id. And Juror Dinwiddie responded "Well, just like everyone else, it's not something anyone wants to do, but if that is what I'm being asked to do I believe I have the courage and the strength to do it." Id. (emphasis added).

In our view, this comparative juror evidence clearly indicates that the OCCA would not have been unreasonable in finding that Ms. Jamerson was not similarly situated to the non-minority jurors who expressed some concern about gruesome photographs-viz. , Ms. Jamerson expressed more concern about viewing graphic photographs than Jurors Abrams and Dinwiddie. Consequently, under Batson , the prosecutor's exclusion of Ms. Jamerson and retention of the other two would not be "evidence tending to prove purposeful discrimination to be considered at Batson 's third step." Miller-El II , 545 U.S. at 241, 125 S.Ct. 2317 (emphasis added).

Thus, the record strongly indicates that the OCCA was not unreasonable in finding that there was no showing of purposeful discrimination in violation of Batson -that is, for finding that "[t]he prosecutor's explanation for striking [Ms. Jamerson] was sufficiently race-neutral."26 Grant , 205 P.3d at 15. And there is nothing in the record that clearly and convincingly rebuts the presumption of correctness that attaches to the OCCA's factual findings regarding this matter. More generally, we can discern no ground on which to hold that the OCCA's Batson determination as to Ms. Jamerson was contrary to or an unreasonable application of Batson . Therefore, even reaching the merits of Mr. Grant's Batson *954claim, we would conclude that he cannot prevail.

E. Cumulative Error

Mr. Grant's final claim is that, even if we decline to grant habeas relief on any one of his aforementioned individual claims, we should nonetheless reverse the district court's denial of his petition under the cumulative-error doctrine.27 "The cumulative effect of two or more individually harmless errors has the potential to prejudice a defendant to the same extent as a single reversible error. The purpose of a cumulative-error analysis is to address that possibility." United States v. Rivera , 900 F.2d 1462, 1469 (10th Cir. 1990) (en banc). "A cumulative-error analysis aggregates all errors found to be harmless and analyzes whether their cumulative effect on the outcome of the trial is such that collectively they can no longer be determined to be harmless." Victor Hooks II , 689 F.3d at 1194 (quoting Cargle v. Mullin , 317 F.3d 1196, 1206 (10th Cir. 2003) ). "[S]uch errors will suffice to permit relief under [the] cumulative error doctrine only when the constitutional errors committed in the state court trial so fatally infected the trial that they violated the trial's fundamental fairness." Littlejohn II , 875 F.3d at 567 (second alteration in original) (emphasis added) (quoting Littlejohn I , 704 F.3d at 868 ); see John Grant , 727 F.3d at 1025 ("Only if the errors 'so fatally infected the trial that they violated the trial's fundamental fairness' is reversal appropriate." (quoting Matthews , 577 F.3d at 1195 n.10 )). "We cumulate error only upon a showing of at least two actual errors." Hanson , 797 F.3d at 852 (citing Fero v. Kerby , 39 F.3d 1462, 1475 (10th Cir. 1994) ).

"It is not lost on us, however, that 'as easy as the standard may be to state in principle, it admits of few easy answers in application.' " Littlejohn II , 875 F.3d at 567 (quoting John Grant , 727 F.3d at 1025 ). We elaborated on the point in John Grant :

Where and how, then, should a court draw the line between what's ordinary (and ordinarily harmless) and what's rare (and fundamentally unfair)? Especially when the errors we are called on to accumulate may be very different in kind (incommensurate) and involve separate aspects of the case (guilt versus penalty)? Our precedent doesn't say except to suggest that wherever the cumulative error line may fall, it is not crossed often.

727 F.3d at 1025.

The OCCA rejected Mr. Grant's cumulative-error arguments on direct appeal and on post-conviction review, finding no errors to cumulate. However, in resolving under the prejudice prong of Strickland Mr. Grant's ineffective-assistance claims based on counsel's alleged failure (1) to monitor his competency, and (2) to investigate and present mitigating evidence of *955organic brain damage, we have effectively assumed that counsel's performance was constitutionally deficient. See Cargle , 317 F.3d at 1207 ("These substantive prejudice components [e.g., under the Strickland rubric] essentially duplicate the function of harmless-error review."). And thus we are obliged to assess these two assumed errors in a cumulative-error analysis. See, e.g. , Littlejohn II , 875 F.3d at 568 ("Although we have held supra that his ineffective-assistance claim fails to evince the level of prejudice required under Strickland , we include the assumed error resulting from Mr. Rowan's alleged ineffective assistance in the cumulative-error analysis."); Cargle , 317 F.3d at 1207 ("[P]articular types of error considered here are governed in the first instance by substantive standards which already incorporate an assessment of prejudice with respect to the trial process as a whole: Strickland errors require us to assess whether there is a reasonable probability that counsel's deficient performance affected the trial outcome .... [S]uch claims should be included in the cumulative-error calculus if they have been individually denied for insufficient prejudice." (footnotes omitted)).

Because the OCCA did not conduct a cumulative-error analysis (much less one involving these precise errors), we must perform our own de novo, employing the well-established standard found in Brecht v. Abrahamson , 507 U.S. 619, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993). See, e.g. , Littlejohn II , 875 F.3d at 568 (conducting a de novo cumulative-error analysis under Brecht , where "the cumulative-error claim advanced here differs from the claim that the OCCA confronted"); Cargle , 317 F.3d at 1224 ("[T]he OCCA did not recognize and address the collective errors we have before us here. Thus, we address the issue de novo under the Brecht standard ...."). In doing so, we inquire whether the identified harmless errors, in the aggregate, "had [a] substantial and injurious effect or influence in determining the jury's verdict." Brecht , 507 U.S. at 637, 113 S.Ct. 1710 (quoting Kotteakos v. United States , 328 U.S. 750, 776, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946) ); accord Littlejohn II , 875 F.3d at 568 ; Cargle , 317 F.3d at 1224. "An error may be deemed to have a substantial and injurious effect under Brecht 's rubric when a 'conscientious judge [is left] in grave doubt about the likely effect of an error on the jury's verdict.' " Littlejohn II , 875 F.3d at 568 (alteration in original) (quoting O'Neal v. McAninch , 513 U.S. 432, 435, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995) ); accord Welch , 639 F.3d at 992.

We have no such grave doubt here. We turn first to counsel's assumed error in failing to investigate and present mitigating evidence of organic brain damage. Based especially on our analysis in Part II.B.4.c, supra , "we do not believe the Strickland prejudice question [as to this individual error] is a close one." Littlejohn II , 875 F.3d at 570. In other words, the harmlessness of this individual error is not a close call. Specifically, the particular nature of the organic-brain-damage evidence that Mr. Grant identified (notably, coming from Dr. Gelbort) was qualitatively weak in its likely mitigating force. Moreover, this evidence-absent any indication that the symptoms associated with such brain damage could be treated-may well have caused (if presented) some damage to the substantial mitigation case trial counsel did present to the jury, which notably focused in large part on Mr. Grant's mental-health problems. However, akin to our reasoning in Littlejohn II -which took into account the fact that generally "[e]vidence of organic mental deficits ranks among the most powerful types of mitigation evidence available," Littlejohn I , 704 F.3d at 864 -we are willing to assume arguendo that *956"some modest prejudice" resulted from the assumed unreasonable failure of Mr. Grant's counsel to investigate and present organic-brain-damage evidence to the sentencing jury, Littlejohn II , 875 F.3d at 570.

As for the other assumed error-i.e., involving the failure to monitor Mr. Grant's purported decline into incompetency-we can hardly do more. After all, as we explicated in Part II.B.3.c, supra , we must accept the OCCA's related-but distinct and independent-determination that Mr. Grant was in fact substantively competent. Therefore, any prejudice stemming from the unreasonable failure of Mr. Grant's counsel to monitor his descent into incompetency cannot rise above the modest level, when Mr. Grant was in fact competent .

In Littlejohn II , in facing a similar showing of purported cumulative error, we reasoned:

From a purely additive or sum-of-the parts perspective, the three dashes of modest prejudice that we have assumed here ... hardly constitute, in the aggregate, a recipe for the kind of prejudice that would render Mr. Littlejohn's resentencing proceeding fundamentally unfair or cause us to have grave doubts about whether the errors affected the jurors' verdict, especially when viewed in the context of the State's substantial case in aggravation.

875 F.3d at 570. This logic applies with even greater force here, where we are collectively assessing the modest prejudicial effect of only two errors, not three; and the State was able to present here a substantial aggravation case-which was sufficient to win the jury's verdict regarding no less than four aggravating factors, instead of only two in Littlejohn , see Littlejohn I , 704 F.3d at 824 -and notably where the State would very likely have underscored in its case, in response to organic-brain-damage evidence, arguments supporting the continuing-threat aggravator, which we have considered in settings like this one, "perhaps [the] most important aggravating circumstance." Littlejohn II , 875 F.3d at 564 (alteration in original) (quoting John Grant , 727 F.3d at 1017 ). In short, from "a purely additive or sum-of-the parts perspective," Littlejohn II , 875 F.3d at 570, we are hard-pressed to conclude that these errors-viewed collectively-"fatally infected" Mr. Grant's proceedings, Matthews , 577 F.3d at 1195 n.10 (quoting Young , 551 F.3d at 972 ), so as to cause fundamental unfairness.

To be sure, cumulative-error analysis is not confined solely to this perspective. As we recognized in Cargle , harmless individual errors may possess "an inherent synergistic effect," 317 F.3d at 1221, such that the errors may be "collectively more potent than the sum of their parts," Littlejohn II , 875 F.3d at 571 ; see Hanson , 797 F.3d at 853 (discussing Cargle 's application of this principle). But Mr. Grant makes no argument that the two errors that we have assumed here possess any "particularized synergies." Aplt.'s Reply Br. at 47; see Aplt.'s Opening Br. at 120-21 (discussing two other "set[s] of violations with synergistic effect"). Therefore, this synergy principle of Cargle does not avail Mr. Grant.

In sum, we cannot conclude that, viewed collectively, the two ineffective-assistance errors that we have assumed here-relating to counsel's failure (1) to monitor Mr. Grant's competency, and (2) to investigate and present mitigating evidence of organic brain damage-had a substantial and injurious effect on the jury's consideration of Mr. Grant's case and, more specifically, we are not in grave doubt about the likely effect of these errors (in the aggregate) on the jury's verdict either in the guilt or penalty phase. Accordingly, we affirm the *957district court's denial of this final aspect of Mr. Grant's petition.

IV. MOTION TO EXPAND COA

Mr. Grant has filed a motion to expand the COA to add a sixth claim that the "jury was prevented from considering mitigation evidence in violation of the Sixth, Eighth, and Fourteenth Amendments by mechanistic application of state evidentiary rules." Aplt.'s Mot. for COA at 3. Specifically, Mr. Grant's claim addresses two groups of expert reports: (1) eight reports authored by Dr. Curtis Grundy ("Grundy Reports"), who testified for Mr. Grant during the penalty stage of his trial, and (2) ten psychological reports on Mr. Grant authored by other doctors that Dr. Grundy relied on in forming his opinions ("other expert reports"). The sentencing court excluded both sets of reports and the OCCA found the exclusion to be within the sentencing court's discretion.

Specifically, the OCCA found that (1) the Grundy Reports were cumulative of Dr. Grundy's testimony, and (2) the other expert reports were properly excluded absent the live testimony of their authors. The OCCA offered a detailed explanation as to the latter ruling:

By excluding the [other expert reports], the trial court avoided placing undue emphasis on writings by authors who were never asked to come to court and testify about them. Many of these reports contain information and terminology which might be confusing to someone outside the world of psychology and psychiatry. Liberal admission of such documents could turn trials into paper wars, and undermine the fundamental preference for live testimony subject to cross-examination. Further, admitting a stack of evaluations by non-testifying experts runs a serious risk of confusing the jury.

Grant , 205 P.3d at 20 (footnotes and citation omitted). Moreover, the OCCA indicated that, even if Mr. Grant were able to demonstrate that the trial court abused its discretion by excluding all of the reports at issue, Mr. Grant would not be entitled to relief because he failed to demonstrate that he was "prejudiced by the trial court's decision." Id. In particular, even though the jurors were prevented from reviewing the actual reports, Dr. Grundy provided "a comprehensive summary of [their contents]" through his live testimony. Id. In sum, the OCCA upheld the trial court's exclusion of the other expert reports.

In his § 2254 petition, Mr. Grant argued to the district court that the OCCA's affirmance of the sentencing court's exclusion of both sets of reports was contrary to clearly established federal law. The district court rejected this argument. Mr. Grant now seeks a COA on this additional issue. We deny relief.

A. Legal Standard

"[A] prisoner who was denied habeas relief in the district court must first seek and obtain a COA" before he may secure a merits review on appeal. Miller-El I , 537 U.S. at 335-36, 123 S.Ct. 1029 ; see 28 U.S.C. § 2253(c)(1). "Under AEDPA, a COA may not issue unless 'the applicant has made a substantial showing of the denial of a constitutional right.' " Slack v. McDaniel , 529 U.S. 473, 483, 120 S.Ct. 1595, 146 L.Ed.2d 542 (2000) (quoting 28 U.S.C. § 2253(c) ).

To make this showing, a petitioner must demonstrate that "reasonable jurists could debate whether (or, for that matter, agree that) the petition should have been resolved in a different manner or that the issues presented were 'adequate to deserve encouragement to proceed further.' " Id. at 475, 120 S.Ct. 1595 (quoting *958Barefoot v. Estelle , 463 U.S. 880, 893 n.4, 103 S.Ct. 3383, 77 L.Ed.2d 1090 (1983), superseded on other grounds by statute , 28 U.S.C. § 2253(c)(2) ). Furthermore, "AEDPA's deferential treatment of state court decisions must be incorporated into our consideration of a habeas petitioner's request for COA." Dockins v. Hines , 374 F.3d 935, 938 (10th Cir. 2004) ; accord Davis v. McCollum , 798 F.3d 1317, 1320 (10th Cir. 2015) ; Dodd v. Trammell , 753 F.3d 971, 999 (10th Cir. 2013). Therefore, we must determine whether reasonable jurists could debate whether Mr. Grant was entitled to habeas relief on this additional claim (or believe that it was worthy of encouragement to proceed further).

With this legal framework in mind, we address Mr. Grant's arguments to expand the COA. We ultimately conclude that Mr. Grant has not made a sufficient showing to warrant issuance of a COA.

B. Excluded Evidence

1. Grundy Reports

Mr. Grant argues that the OCCA's exclusion of the Grundy Reports was contrary to Skipper v. South Carolina , 476 U.S. 1, 106 S.Ct. 1669, 90 L.Ed.2d 1 (1986), a death-penalty case. But Mr. Grant's argument misjudges Skipper 's reach. In Skipper , the Supreme Court held that the trial court infringed on the defendant's well-established federal "right to place before the sentencer relevant evidence in mitigation of punishment" when the court excluded certain evidence that it deemed cumulative. 476 U.S. at 4, 8, 106 S.Ct. 1669. There, the defendant sought to admit at his sentencing hearing the testimony of two jailers and a regular visitor at the jail. See id. at 3, 106 S.Ct. 1669. The proposed witnesses would have testified regarding the defendant's good behavior in jail while awaiting trial, thus allegedly mitigating the prosecution's argument that, if sentenced to life in prison, the defendant posed a threat to the safety of other inmates. See id. at 4, 8, 106 S.Ct. 1669. The State argued that the testimony was properly excluded because it was merely cumulative of other testimony of the defendant and his former wife that his behavior in jail had been satisfactory. See id. at 8, 106 S.Ct. 1669.

The Court concluded that the additional evidence was not cumulative and that its exclusion constituted reversible error. See id. at 8, 106 S.Ct. 1669. The Court explained:

The evidence petitioner was allowed to present on the issue of his conduct in jail was the sort of evidence that a jury naturally would tend to discount as self-serving. The testimony of more disinterested witnesses-and, in particular, of jailers who would have had no particular reason to be favorably predisposed toward one of their charges-would quite naturally be given much greater weight by the jury.

Id. (emphasis added). In light of the prosecutor's contention that the defendant posed a continuing threat even if incarcerated, the Court concluded that "it appears reasonably likely that the exclusion of evidence bearing upon petitioner's behavior in jail (and hence, upon his likely future behavior in prison) may have affected the jury's decision to impose the death sentence." Id. Therefore, the Court declared the exclusion of the evidence "reversible error." Id.

Here, Mr. Grant argues that the exclusion of the Grundy Reports as cumulative was contrary to Skipper . He acknowledges, however, that Dr. Grundy's testimony-which the jury heard-and the Grundy Reports both reflect Dr. Grundy's medical opinions of Mr. Grant's mental illness. But he argues that the reports are not cumulative of Dr. Grundy's testimony *959because they contain different examples of Mr. Grant's behavior, and therefore their exclusion is contrary to Skipper .

We conclude, however, that no reasonable jurists could debate that the OCCA's decision to exclude the Grundy Reports was not contrary to or an unreasonable application of Skipper . The OCCA could have reasonably concluded that the facts of Skipper are distinguishable. More specifically, it could have reasonably concluded that no circumstances in this case-including with respect to the contents of the Grundy reports themselves-would have naturally led Mr. Grant's jury to accord "much greater weight" to the Grundy Reports than the live testimony from Dr. Grundy. See Skipper , 476 U.S. at 8, 106 S.Ct. 1669. Accordingly, we conclude that no reasonable jurist could debate the correctness of the decision to deny Mr. Grant a COA regarding the exclusion of the Grundy Reports or deem the matter one that was worthy of encouragement to proceed further.

2. Other Expert Reports

With regard to the other expert reports, Mr. Grant asserts that they contain "critical mitigating evidence ... that Dr. Grundy's testimony did not, and could not fully address." Aplt.'s Mot. for COA at 12. Accordingly, Mr. Grant argues that the exclusion of the other expert reports for lack of authentication violates Lockett and its progeny. As we noted in Part II.C.1, supra , the Lockett Court held that the Constitution "require[s] that the sentencer, in all but the rarest kind of capital case, not be precluded from considering, as a mitigating factor, any aspect of a defendant's character or record and any circumstances of the offense that the defendant proffers as a basis for a sentence less than death." 438 U.S. at 604, 98 S.Ct. 2954.

However, we conclude that no reasonable jurists could debate that the OCCA's decision to exclude the other expert reports was not contrary to or an unreasonable application of Lockett . More specifically, the OCCA could have reasonably determined that the evil that Lockett addressed was not present here. The other expert reports concern Mr. Grant's mental illness and Mr. Grant was never prevented from presenting evidence of his mental illness as a mitigating factor. The record is replete with evidence of Mr. Grant's mental illness, including Dr. Grundy's testimony that Mr. Grant had schizophrenia. Rather, by excluding the other expert reports, the court prevented Mr. Grant from submitting evidence of additional examples and proof of Mr. Grant's mental illness. The OCCA could have reasonably determined that such an exclusion is not a concern of Lockett - viz. , the OCCA could have reasonably concluded that Lockett does not stand for the proposition that every scrap or scintilla of evidence bearing on a defendant's mitigation issue-here, mental illness-must be admitted without consideration of the rules of evidence. In sum, no reasonable jurists could debate the conclusion that Mr. Grant has not made a substantial showing that the OCCA's decision is contrary to or an unreasonable application of Lockett .

Mr. Grant also relies on Crane v. Kentucky , 476 U.S. 683, 690, 106 S.Ct. 2142, 90 L.Ed.2d 636 (1986), but to no better effect. In Crane , the Court struck down a state's "blanket exclusion of ... testimony about the circumstances of [a defendant's] confession" as unconstitutional. 476 U.S. at 690, 106 S.Ct. 2142. But, the Crane Court carefully limited its holding, noting that the Court has "never questioned the power of States to exclude evidence through the application of evidentiary rules that themselves serve the interests of fairness and reliability-even if the defendant would *960prefer to see that evidence admitted." Id. Once again, no reasonable jurists could debate that the OCCA's decision to uphold the exclusion of the other expert reports was not contrary to or an unreasonable application of Crain . Unlike Crain , the OCCA did not endorse a wholesale exclusion of evidence relating to one mitigating factor. Instead, the OCCA upheld the sentencing court's discretionary application of evidentiary rules that favor live testimony (which permits adversarial testing) and guard against juror confusion; like those alluded to in Crain , such rules "serve the interests of fairness and reliability." Id. In our view, no reasonable jurists could debate the conclusion that Mr. Grant has not made a substantial showing that the OCCA's decision to uphold the exclusion of the other expert reports was contrary to or an unreasonable application of Crain .

* * *

In sum, we conclude that no reasonable jurist could debate the correctness of the district court's decision to deny Mr. Grant a COA regarding the exclusion of the Grundy Reports and the other expert reports, or deem the matter one that was worthy of encouragement to proceed further. Accordingly, we deny Mr. Grant's motion to expand the COA.

V. CONCLUSION

In sum, for the reasons explicated above, we AFFIRM the district court's decision denying Mr. Grant's § 2254 petition for a writ of habeas corpus. We also DENY Mr. Grant's motion to expand the COA.