An Oklahoma jury convicted and sentenced Gilbert Ray Postelle to death in connection with the brutal killings of four people. On Memorial Day 2005, Postelle and two other assailants attacked Donnie Swindle at his home, murdering him along with three acquaintances. The raid apparently sprang from the Postelle family's grudge against Mr. Swindle alone; the *1207three other victims had no connection to the feud.
After an unsuccessful appeal and collateral action in state court, Postelle now pursues federal habeas corpus relief. He alleges the state prosecution violated several of his constitutional rights, including his Sixth Amendment right to counsel and his Eighth Amendment right against cruel and unusual punishment. Postelle raises three issues: (1) whether he received constitutionally adequate trial counsel; (2) whether he received constitutionally adequate appellate counsel; and (3) whether the unconstitutional presentation of victim-impact evidence at trial prejudiced his defense. He also asks to expand the scope of our review to include several new issues for which he has yet to receive a Certificate of Appealability.
For the reasons given below, we affirm denial of the writ and decline to extend the scope of our review.
I. Background
We base our description of Postelle's crimes on the Oklahoma Court of Criminal Appeals's (OCCA) account in Postelle v. State (Postelle I ), 267 P.3d 114 (Okla. Crim. App. 2011), as well as the jury's findings and other uncontested facts.
The background for these crimes begins with Earl Bradford "Brad" Postelle being thrown from his motorcycle in a single-vehicle accident. See id. at 124 & n.7 ; Tr. 1030-33. Brad suffered grave injuries, both physical and mental, as a result of the crash. See Postelle I , 267 P.3d at 124. Without apparent basis, he and his two sons-David and Gilbert Postelle-would eventually blame the accident on an acquaintance named Donnie Swindle. See id. at 124-25 ; Tr. 2239. And on Memorial Day 2005, that blame erupted into violence.
The day began with the Postelles hosting several friends at their home in Midwest City, Oklahoma. See Postelle I , 267 P.3d at 123-24 ; Tr. 1635, 2087. The house often served as a place to use methamphetamine, and this gathering was no different. Postelle I , 267 P.3d at 124. On this day, however, Gilbert and David Postelle resolved that "those responsible" for their father's injuries "were 'going to pay.' " Id.
That afternoon, the Postelles and three friends left the house, ostensibly to go target shooting. Id. at 124-25. After dropping off two of the passengers, however, their van did not follow its usual course to the riverbank. Id. at 125 ; see Tr. 2039, 2065. Instead, it rolled on toward the home of Donnie Swindle.
As they drove onto Swindle's property, he and a guest named Terry Smith approached the van. Postelle I , 267 P.3d at 125 ; see Tr. 2072. Gilbert Postelle promptly slid open the van door and shot Smith in the face with a military-style rifle. Postelle I , 267 P.3d at 124-25 & n.9. Gilbert and Brad then shot at Swindle, dropping him to the ground. Id. at 125. Next, David Postelle took Brad's gun and shot the bewildered Swindle in the head. Id. at 126. Gilbert then "turned and ran through [Swindle's] trailer, looking for others and firing his gun." Id. at 126. He came out through the back door and "chased down" a third victim, James Alderson. Id. Gilbert "shot [Alderson] as [he] tried to seek cover under a boat." Id. Gilbert then gunned down one final victim-Amy Wright-with three shots from behind. See id. at 123 & n.1, 126. The perpetrators then got back in the van and drove away. Id. at 126.
Oklahoma law enforcement eventually identified, arrested, and charged Gilbert Postelle with four counts of first-degree murder and one count of conspiracy to commit a violent felony. See id. at 123. In light of evidence depicting the events above, a jury convicted Postelle of all five crimes. See id. Then, despite mitigating evidence of "organic brain damage and *1208mental illness," "drug abuse from an early age," and a "chaotic and abusive upbringing," Postelle v. State (Postelle II ), No. PCD-2009-94, slip op. at 14 (Okla. Crim. App. filed Feb. 14, 2012), the jury sentenced Postelle to death, see Postelle I , 267 P.3d at 123.
Postelle challenged his conviction and sentence in the Oklahoma courts. On direct appeal, he argued-among other claims-that the State's use of victim-impact statements during the trial's sentencing phase violated his Eighth Amendment rights. See Brief ex rel. Gilbert Ray Postelle, Appellant at 78-80, Postelle I , 267 P.3d 114 (D-2008-934). The OCCA rejected the challenge on plain error review. See Postelle I , 267 P.3d at 142-43. Postelle then applied for post-conviction collateral relief. This time-again, among other claims-he contended his trial and appellate counsel had rendered constitutionally inadequate assistance. See Original Appl. Post Conviction Relief Death Penalty Case at 5-10, 47-49, Postelle II , slip op., (PCD-2009-94) [hereinafter PCR Appl.]. The OCCA rejected these arguments as well, again affirming the trial court. See Postelle II , slip op. at 9-17, 18-20.
Finally, Postelle sought protection from the federal courts. In September 2013, he filed this action in the Western District of Oklahoma for the writ of habeas corpus. See R., Vol. I at 10. Postelle based his petition, in relevant part, on the alleged constitutional violations just mentioned. See id. at 24-51. The district court denied relief. See id. at 584-85. Postelle now appeals that denial.
II. Analysis
Postelle asks us to overturn the district court with respect to three issues. First , he claims his trial counsel rendered ineffective assistance by not using the "Flynn Effect" as part of the mitigation strategy to help argue against a death sentence. See Aplt. Br. at 2. Second , he claims his appellate counsel rendered ineffective assistance by not challenging trial counsel's failure to use Flynn Effect evidence for death penalty-eligibility and mitigation purposes. See id. Finally , Postelle claims certain victim-impact evidence erroneously introduced in the sentencing phase was not harmless, but in fact prejudiced his defense. See id.
In appeals from orders denying a writ of habeas corpus, we review the district court's legal analysis de novo and its factual findings for clear error. Smith v. Duckworth (Smith II ), 824 F.3d 1233, 1241-42 (10th Cir. 2016). To qualify for the writ, however, a state prisoner must carry a heavy burden. Indeed, Congress has directed federal courts to give their state counterparts deference in all but the narrowest circumstances. See 28 U.S.C. § 2254(d). As relevant here, under the Antiterrorism and Effective Death Penalty Act (AEDPA), a state court must contradict or unreasonably apply "clearly established Federal law, as determined by the Supreme Court of the United States" as a prerequisite to federal habeas relief. Id. Mindful of that threshold inquiry, we turn to Postelle's claims.1
A. Ineffective Assistance of Counsel
The Sixth Amendment guarantees every accused "the right ... to have Assistance of Counsel for his defence."
*1209U.S. Const. amend. VI. The Supreme Court has interpreted this right to guarantee every criminal defendant a minimum quality of advocacy from a professional attorney. In Strickland v. Washington , 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), the Court held a criminal defendant could establish a violation of his right to counsel upon two related but distinct showings. First, he "must show that counsel's performance was deficient." Id. at 687, 104 S.Ct. 2052. In this context, only commission of "errors so serious that counsel was not functioning as the 'counsel' guaranteed ... by the Sixth Amendment" constitutes "deficient performance." Id. "Second, the defendant must show that the deficient performance prejudiced the defense." Id. This inquiry also looks to counsel's errors, this time to determine whether they were "so serious as to deprive the defendant of a fair trial" with a "reliable" result. Id.
Postelle challenges the adequacy of both his trial counsel and appellate counsel. See Aplt. Br. at 2; see generally Evitts v. Lucey , 469 U.S. 387, 105 S.Ct. 830, 83 L.Ed.2d 821 (1985) (establishing the right to effective appellate counsel). As far as this appeal is concerned, however, both claims derive from a single alleged error: trial counsel's failure to incorporate evidence of the Flynn Effect into Postelle's defense.
1. The Flynn Effect
We start with a short explanation of the Flynn Effect-an aspect of IQ testing upon which Postelle's petition heavily relies.
The Flynn Effect is an observed phenomenon believed to impact the accuracy of IQ testing. See generally John Matthew Fabian et al., Life, Death, and IQ: It's Much More Than Just a Score: Understanding and Utilizing Forensic Psychological and Neuropsychological Evaluations in Atkins Intellectual Disability/Mental Retardation Cases , 59 Clev. St. L. Rev. 399, 414-16 (2011). As the well-known scoring system makes clear, IQ testing does not aim to pinpoint the test subject's absolute intelligence. Instead, it attempts to measure his intelligence relative to the rest of the population. See Nancy Haydt et al., Advantages of DSM-5 in the Diagnosis of Intellectual Disability: Reduced Reliance on IQ Ceilings in Atkins (Death Penalty) Cases , 82 UMKC L. Rev. 359, 364 (2014). Accordingly, before administering a new IQ test to any one person, the creator must first "norm" it by scoring the performance of a sample group. Fabian et al., supra , at 414. Like zeroing a scale, this norming process identifies how someone of average intelligence should perform on the new test. See id. The test maker then keys that average performance to an IQ of 100 and constructs a "normal" bell curve of performance around that point. See id.; see also Haydt et al., supra , at 364 (explaining points on the IQ scale as corresponding to deviation from the mean on a normal curve). Assuming the sample group accurately represented the general population, the test should now be capable of identifying any single taker's relative intelligence. See Fabian et al., supra , at 414.
But in 1984, Dr. James Flynn published a study documenting an increase in average performance on IQ tests over time. See James R. Flynn, Tethering the Elephant: Capital Cases, IQ, and the Flynn Effect , 12 Psychol. Pub. Pol'y & L. 170, 172 (2006). Specifically, Flynn's findings indicated an upward creep of average IQ scores by about 0.33 points every year. See John H. Blume et al., *1210Of Atkins and Men: Deviations from Clinical Definitions of Mental Retardation in Death Penalty Cases , 18 Cornell J.L. & Pub. Pol'y 689, 700 (2009) ; Fabian et al., supra , at 414 (identifying a rate of 0.31 points per year); Richard J. Bonnie & Katherine Gustafson, The Challenge of Implementing Atkins v. Virginia: How Legislatures and Courts Can Promote Accurate Assessments and Adjudications of Mental Retardation in Death Penalty Cases , 41 U. Rich. L. Rev. 811, 838 (2007) (identifying a rate of 0.31 points per year).
Academic literature has since dubbed this phenomenon "the Flynn Effect," and it carries relatively straightforward implications for the accuracy of IQ testing: The performance of the sample group used to norm an IQ test is obviously static-frozen in time. But the average performance of all other test takers gradually improves with each passing year. Thus, just as a photo taken at dawn will not depict the brightness of noon, a sample group used to norm an IQ test in 1995 will not reflect average intelligence in 2005.2 On the contrary, because of the upward creep in average scores, we should expect a person of average intelligence in 2005 to score a 103 on an IQ test normed ten years earlier, rather than the usual 100. See Blume et al., supra , at 701. Conversely, if a person scores a 73 on an IQ test normed ten years before its administration, we may adjust his score downward to 70 to reflect his intelligence relative to today's general population. See id.3
2. Ramifications for Capital Punishment
Two lines of death penalty jurisprudence connect the Flynn Effect to this case.
The Supreme Court has read the Eighth Amendment's prohibition on "cruel and unusual punishments," U.S. Const. amend. VIII, to forbid " '[e]xcessive' sanctions." Atkins v. Virginia , 536 U.S. 304, 311, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002). The federal courts determine whether a "punishment is excessive" according to currently prevalent standards. Id. And in Atkins v. Virginia , 536 U.S. 304, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002) (quoting U.S. Const. amend. VIII ), the Supreme Court observed "a national consensus ... against" putting intellectually disabled persons to death. Id. at 316, 122 S.Ct. 2242.4 The Court therefore held states could not execute such persons, as that punishment would be "excessive" in the eyes of the Eighth Amendment. See id. at 314-17, 122 S.Ct. 2242.
In addition, states cannot prevent a court or jury from hearing relevant, mitigating evidence during a capital sentencing determination. In *1211Lockett v. Ohio , 438 U.S. 586, 98 S.Ct. 2954, 57 L.Ed.2d 973 (1978), a plurality of Justices took the position that-pursuant to the Eighth Amendment-states must permit capital juries to "consider[ ]" all proffered mitigating evidence respecting the "defendant's character[,] ... record[,] ... [or] the circumstance of the offense" "in all but the rarest kind of capital case." Id. at 604, 98 S.Ct. 2954 (Opinion of Burger, C.J.). Several years later, in Eddings v. Oklahoma , 455 U.S. 104, 102 S.Ct. 869, 71 L.Ed.2d 1 (1982), a majority of the Court adopted, expanded, and applied that rule. See id. at 112-15, 102 S.Ct. 869. Eddings clarified "that the sentencer in capital cases must be permitted to consider any relevant mitigating factor ." Id. at 112, 102 S.Ct. 869 (emphasis added). "Relevance" here takes the same meaning as in any other evidentiary context-that is, relevant evidence has some "tendency to make ... any fact ... of consequence ... more ... or less probable than it would be" otherwise. Tennard v. Dretke , 542 U.S. 274, 284, 124 S.Ct. 2562, 159 L.Ed.2d 384 (2004) (quoting McKoy v. North Carolina , 494 U.S. 433, 440, 110 S.Ct. 1227, 108 L.Ed.2d 369 (1990) ).
Flynn Effect evidence could potentially play an important role within each of these two jurisprudential veins. First, accounting for the Flynn Effect might impact whom states may execute consistent with the Eighth Amendment. This is because IQ is one of the metrics commonly used to identify intellectual disability. See, e.g., Atkins , 536 U.S. at 308 n.3, 309 n.5, 317 & n.22, 122 S.Ct. 2242. Second, when the death penalty is, in fact, available as a sentence, evidence of the Flynn Effect may bear on the sentencer's choice to issue it in lieu of a lesser punishment.
3. Postelle's Claims
Postelle adopts both of these potential uses in his claims of ineffective assistance.
First , Postelle argues his trial counsel should have used Flynn Effect evidence to help exempt him from the death penalty under Atkins . This is because, under Oklahoma law, his entitlement to an intellectual disability determination depended entirely on whether the court adjusted IQ scores to account for the Flynn Effect.
The Supreme Court has left to the states the task of "determining which offenders ... in fact" fall within Atkins 's ambit. Id. at 317, 122 S.Ct. 2242. To implement this directive, the State of Oklahoma has created a process whereby capital defendants may be adjudicated intellectually disabled either by the court prior to trial, see Okla. Stat. tit. 21, § 701.10b(E), or by the jury prior to determination of a sentence, see id. at § 701.10b(E)-(F). But a criminal defendant that scores a 76 or higher on any valid IQ test may not receive an Atkins determination under Oklahoma law. See id. at § 701.10b(C).
Postelle completed two separate IQ tests in 2006 and 2007 in anticipation of trial. Postelle II , slip op. at 12; Tr. 2870. He scored a 79 on the first and a 76 on the second. Postelle II , slip op. at 12. If adjusted for the Flynn Effect, Postelle contends, his two IQ scores would both have fallen to roughly 73. Aplt. Br. at 27-28.5 He therefore argues his trial counsel should have used Flynn Effect evidence to get him an eligibility determination, and claims his appellate counsel rendered ineffective assistance by not raising trial counsel's omission on direct appeal. See Aplt. Br. at 2, 46.
Second , Postelle claims his trial counsel was ineffective for not using Flynn Effect *1212evidence in his sentencing determination, again faulting appellate counsel for not raising this error on appeal. See id. at 2. This argument presents a more direct attack on the failure to utilize the Flynn Effect. So the logic goes, a capital defendant may use any relevant evidence to convince a jury not to return a death sentence. Thus, Postelle claims counsel rendered deficient and prejudicial performance by not mentioning the Flynn Effect in support of a lesser sentence. See Aplt. Br. at 2, 18-24.
For the reasons stated below, however, neither argument justifies habeas relief.
4. The Eligibility Argument
The OCCA's handling of Postelle's eligibility-based argument certainly warrants AEDPA deference.
The OCCA clearly rejected Postelle's eligibility-based argument in his application for post-conviction relief. See Postelle II , slip op. at 11-13, 18-20. It explained that any attempt to exempt Postelle from the death penalty by virtue of intellectual disability would have been fruitless. See id. at 13. We take this to mean counsel was wise to strategically omit the evidence, and-by extension-such omission could not have prejudiced Postelle's defense. See id. at 13. Postelle therefore could not fault appellate counsel for failing to raise a meritless claim of trial-counsel ineffectiveness on appeal. Id. at 19.
The Oklahoma legislature established its statutory framework for implementing Atkins in 2006. See Okla. Stat. tit. 21, § 701.10b (effective July 1, 2006). Four years later, in Smith v. State , 245 P.3d 1233 (Okla. Crim. App. 2010), the OCCA deemed Flynn Effect evidence-"whatever its validity"-irrelevant to the statute's IQ cutoff. Id. at 1237 n.6. The defendant in Smith then sought federal habeas relief, arguing the Oklahoma court's decision contradicted Atkins . See Smith II , 824 F.3d at 1242. The district court denied the petition, and we affirmed. See id. at 1238. In so doing, we observed " Atkins does not mandate an adjustment for the Flynn Effect ... and 'no decision of the Supreme Court squarely addresses the issue.' " Id. at 1246 (quoting Hooks v. Workman (Victor Hooks II ), 689 F.3d 1148, 1170 (10th Cir. 2012) (brackets and ellipses omitted) ). Thus, Smith had no right to habeas relief because Oklahoma's treatment of the Flynn Effect did not contradict or unreasonably apply Supreme Court precedent. Id.
Though the Supreme Court's more recent decision in Hall v. Florida , 572 U.S. 701, 134 S.Ct. 1986, 188 L.Ed.2d 1007 (2014), did not bear on our analysis in Smith , see Lockyer v. Andrade , 538 U.S. 63, 71-72, 123 S.Ct. 1166, 155 L.Ed.2d 144 (2003), we nevertheless explained that Hall , like Atkins , "says nothing about application of the Flynn Effect to IQ scores in evaluating a defendant's intellectual disability." Smith II , 824 F.3d at 1246. Hall deals only with the standard error measurement-a feature of IQ testing already accounted for in Oklahoma's statute. See id. at 1245-46.
In light of our decision in Smith , Postelle's eligibility-based argument cannot further his claim of ineffective appellate counsel. Regardless of whether Postelle's counsel could have predicted it, Smith's experience clearly shows any attempt to pursue an Atkins exemption through Flynn Effect evidence would have failed. Indeed, the exact same argument failed in Smith , and Postelle gives us no reason to believe that his trial and appeal would have turned out any differently. His claim therefore falls far short of the requirements necessary to show prejudice under Strickland . See, e.g. , Grant v. Royal , 886 F.3d 874, 905 (10th Cir. 2018).
Accordingly, Postelle's claim of ineffective appellate counsel cannot draw support *1213from his eligibility-based argument. Far from contradicting or unreasonably applying Supreme Court precedent, the OCCA rendered sound analysis to reach a permissible result.
5. The Mitigation Argument
Postelle's mitigation-based argument presents a more complex analysis. In the end, however, it too fails to persuade us.
a. The State Court Adjudication
To begin, the state-court adjudication of the mitigation-based Flynn Effect argument differs markedly from that of the eligibility-based argument.
Postelle's only mention of the Flynn Effect as mitigation evidence in his state post-conviction briefing appears at the tail end of his eligibility-based argument. There his application states-without elaboration-that "even if counsel had been unsuccessful in obtaining a pre-trial finding that Mr. Postelle is [intellectually disabled], counsel could have still presented the evidence as mitigation during the second stage of his trial." PCR Appl. at 10.
Under its most reasonable interpretation, the OCCA opinion did not comment on this throw-away assertion. Cf. Johnson v. Williams , 568 U.S. 289, 133 S.Ct. 1088, 1095, 185 L.Ed.2d 105 (2013) ("[A] state court may not regard a fleeting reference to a provision of the Federal Constitution or federal precedent as sufficient to raise a separate federal claim."). Instead, it focused solely on the eligibility argument in rejecting Postelle's Flynn Effect-based ineffective assistance of counsel theory. See Postelle II , slip op. at 11-13 (expressly addressing eligibility without explicitly mentioning mitigation); id. at 13-17 (discussing the adequacy of Postelle's mitigation defense without mentioning the Flynn Effect); id. at 18-20 (incorporating the trial-counsel analysis into Postelle's appellate counsel claim). And even if we look at the decision most broadly, the only colorable partial reference to the mitigation-based argument is the OCCA's sweeping rejection of any ineffective assistance of appellate counsel claim premised on the mitigation defense. In relevant part, this followed from the OCCA's conclusion that trial counsel had not, in fact, rendered ineffective assistance in the mitigation phase. See id. at 19-20.
We will not, however, read the OCCA opinion to contradict the Lockett line of cases as Postelle argues we should. Under Postelle's reading, the OCCA held that Smith v. State compelled exclusion of Flynn Effect evidence from capital sentencing proceedings. Aplt. Br. at 14-15, 18, 25, 30, 35, 39. This would indeed contradict the Lockett line of cases, as evidence of the Flynn Effect clearly meets the low bar of relevance to the sentencing determination in light of Atkins . But this reading of the OCCA opinion is untenable.6 In excluding Flynn Effect evidence from the eligibility calculus, Smith v. State in no way addressed its use as mitigating evidence. See 245 P.3d at 1237. In addition, and more importantly, we see nothing in the OCCA opinion in this case making that leap. See Postelle II , slip op. at 11-13. Moreover, the OCCA ably applied the Lockett line of cases on Postelle's direct appeal, see Postelle I , 267 P.3d at 140-41, giving us little if any reason to believe it would ignore those cases on Postelle's state collateral review.
Accordingly, whether we read the OCCA to reject the mitigation-based argument silently or implicitly to sweep it into a broader analysis, our task would be the same. Postelle has neither asserted the *1214OCCA ignored his mitigation-based argument nor shown a contradiction of Lockett and its progeny. We thus ask only whether the OCCA reasonably applied Strickland in denying a claim to relief under Postelle's mitigation-based Flynn Effect theory. See Williams v. Trammell , 782 F.3d 1184, 1201-02 (10th Cir. 2015).7 We conclude that it did.
b. Deficient Performance
The OCCA reasonably concluded Postelle's trial counsel did not perform deficiently by omitting Flynn Effect evidence from the mitigation case.
As we have already mentioned, "our review of counsel's performance under the first prong of Strickland is a 'highly deferential' one." Byrd v. Workman , 645 F.3d 1159, 1168 (10th Cir. 2011) (quoting Hooks v. Workman (Danny Hooks ), 606 F.3d 715, 723 (10th Cir. 2010) ). "[C]ounsel is strongly presumed to have rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment." Grant , 886 F.3d at 903 (alteration in original) (quoting Victor Hooks II , 689 F.3d at 1187 ). Indeed, a showing of deficient performance requires proof that counsel's conduct was "not merely wrong," but "outside the wide range of professionally competent assistance." Id. (quoting Danny Hooks , 606 F.3d at 723 ). With regard to a charge of inadequate investigation in particular, we ask whether counsel's conduct was reasonable in light of all the circumstances. See *1215Newmiller v. Raemisch , 877 F.3d 1178, 1196 (10th Cir. 2017). And it is particularly relevant to this case that we make "[e]very effort ... to evaluate the conduct from counsel's perspective at the time." Grant , 886 F.3d at 903 (quoting Littlejohn v. Trammell (Littlejohn I ), 704 F.3d 817, 859 (10th Cir. 2013) ).
We recognize the Flynn Effect was not wholly foreign to criminal defense advocates at the time of Postelle's trial. So far as we can tell, the first discussion of the phenomenon by an American court appeared roughly five years earlier in a footnote to a federal district court opinion in Virginia. See Walton v. Johnson , 269 F.Supp.2d 692, 699 n.5 (W.D. Va. 2003), vacated , 440 F.3d 160 (4th Cir. 2006) (en banc). It was thereafter mentioned, though not much explained or discussed, in a dissent to an opinion of the Eleventh Circuit. See In re Hicks , 375 F.3d 1237, 1242-43 (11th Cir. 2004) (Birch, J., dissenting). In 2004, four years before Postelle's trial, the California Court of Appeals became the first tribunal to require adjustment of an IQ score to account for the Flynn Effect. See People v. Superior Court, 124 Cal.App.4th 806, 21 Cal.Rptr.3d 542, 568 (2004), vacated , 26 Cal.Rptr.3d 568, 109 P.3d 68 (Cal. 2005). Several other courts did the same within the next few years. See Walker v. True , 399 F.3d 315, 322-23 (4th Cir. 2005) (observing a place for Flynn Effect evidence under Virginia law); Wiley v. Epps , 668 F.Supp.2d 848, 894 (N.D. Miss. 2009). And the legal academy weighed in as well. See Bonnie & Gustafson, supra , at 837-38; Dora W. Klein, Categorical Exclusions from Capital Punishment: How Many Wrongs Make A Right? , 72 Brook. L. Rev. 1211, 1231-32 n.89 (2007).
But hindsight makes this material deceptively easy to find. Indeed, prior to September 2008, only a small proportion of cases and secondary literature citing Atkins mentioned the Flynn Effect.8 In fact, a review of our own opinions, the opinions of the federal district courts in our circuit, as well as the courts of the states that comprise our circuit yields only a single mention of the Flynn Effect before Postelle's trial. That passing reference occurred in a footnote of Myers v. State , 130 P.3d 262, 268 n.11 (Okla. Crim. App. 2005), a decision upholding the jury's finding that the defendant was not intellectually disabled despite scoring in the 60s on several separate IQ tests. See id. at 267-68 & n.10. Thus, the notion that Postelle's counsel necessarily rendered substandard advocacy by not using the Flynn Effect as mitigating evidence in Postelle's 2008 trial cannot be sustained.
More importantly, though, Postelle gives us no reason to believe counsel ignored or failed to properly solicit expert advice on this subject. The Flynn Effect is not a legal concept. It is a phenomenon that might affect how IQ tests are administered, scored, and evaluated. We should thus expect that if the psychiatric community widely recognized Dr. Flynn's research prior to Postelle's trial, the defense's mental health expert, Dr. Ruwe, would have alerted counsel to its potential value. This is precisely the reason lawyers seek out expert assistance in the first *1216place. See American Bar Association, Guidelines for the Appointment and Performance of Defense Counsel in Death Penalty Cases , 31 Hofstra L. Rev. 913, 956 (2003) [hereinafter ABA Guidelines]; cf. Wilson v. Sirmons , 536 F.3d 1064, 1089 (10th Cir. 2008) ("[I]n many situations, the expert will know better than counsel what evidence is pertinent to mental health diagnoses and will be more equipped to determine what avenues of investigation are likely to result in fruitful information."), reinstated sub nom., Wilson v. Workman , 577 F.3d 1284 (10th Cir. 2009) (en banc); id. at 1133 (Tymkovich, J., concurring in part and dissenting in part) ("When investigating a defendant's mental health, counsel by necessity often relies on expert assistance."). Indeed, just as "[c]ounsel's own observations ..., while necessary, can hardly be expected ... to detect ... conditions" like intellectual disability, ABA Guidelines, supra , at 956 (footnote omitted), we should likewise not expect the lawyer to know more than the clinical neuropsychologist about the fine details of scoring IQ, see id. at 1002 ("[T]he provision of high quality legal representation in capital cases requires a team approach that combines the different skills, experience, and perspectives of several disciplines ." (emphasis added) ). And in this case, despite having been evaluated by two mental health professionals, Postelle points us to no evidence in this voluminous record indicating either expert alerted counsel to the existence of the Flynn Effect.
Admittedly, the academic literature gives some indication that the Flynn Effect may have been "outside the ken of many mental health clinicians" at the time of Postelle's trial. Bonnie & Gustafson, supra , at 856. But of course, that fact further weakens the case for deficient performance; the less well-known it was in the mental health community, the less likely a competent attorney would identify it through due diligence. And indeed, given Dr. Ruwe's extensive experience doing mental-health evaluations for the purpose of litigation, see Tr. 2846, we think it all the more reasonable for Postelle's counsel to have thought Dr. Ruwe would offer the most promising mental health-based mitigation arguments.
This is not to say an attorney can abdicate all responsibility for handling scientific or technical evidence. On the contrary, counsel's "managerial role" requires "continue[d] exercise [of] supervisory authority over" expert witnesses and advisors to "ensur[e] that [they] examine[ ] [necessary] sources of information." Wilson , 536 F.3d at 1089 (majority opinion). But having provided Dr. Ruwe with the information necessary to test Postelle's intelligence, we think counsel's reliance on expert advice in the administration and scoring of Postelle's testing was at least within the bounds of professional competence. Cf. id. at 1089-90 (stating counsel can reasonably rely on expert opinion "once either the expert or counsel has consulted all readily available sources" of mitigating evidence, id. at 1089 ).9
Neither is the defense team's supposed failure to recognize the potential importance of the Flynn Effect altogether surprising. Had either counsel or expert consulted the Diagnostic and Statistical Manual of Mental Disorders (DSM) current at the time of Postelle's trial, they *1217would have found no mention of the Flynn Effect at all. See Am. Psychiatric Ass'n, Diagnostic and Statistical Manual of Mental Disorders 41-49 (4th ed., text rev. 2000). Not until 2013-five years after Postelle's trial-did the DSM reference the Flynn Effect, and even then only vaguely as a "[f]actor[ ] that may affect test scores" on account of "overly high scores due to out-of-date test norms." Am. Psychiatric Ass'n, Diagnostic and Statistical Manual of Mental Disorders 37 (5th ed. 2013).
And all of this simply assumes counsel never discovered the Flynn Effect. But see Aplt. Br. at 6 ("Defense counsel was unaware of, or disinterested in the Flynn Effect ...." (emphasis added) ). But in fact, the record does not foreclose the possibility that Postelle's defense team knew about the Flynn Effect and made a strategic choice to omit it from the mitigation case. As the OCCA may have recognized, the difference of a few IQ points was not some magical key to success. And the possible marginal benefit of raising the issue carried with it the additional risk of provoking a "battle of the experts" which could have detracted from the relatively strong evidence of mental impairment Postelle did put on. Cf. Aplt. Br. at 40-41 (acknowledging that Flynn Effect evidence might provoke debate even if valid). We discuss this tradeoff more fully below with regard to prejudice. But it suffices to say we disagree that any choice to omit Flynn Effect evidence from the mitigation case would necessarily amount to an egregious practice error.
In sum, Postelle's petition appears to fall well short of showing the plainly incompetent and unprofessional conduct necessary to support a charge of deficient trial-counsel performance. Thus, the OCCA applied Strickland reasonably to determine Postelle's counsel had not performed deficiently, and we defer to its judgment.
c. Prejudice
Postelle has also failed to make a strong showing that the omission of Flynn Effect evidence prejudiced his mitigation defense.
Strickland 's prejudice prong requires Postelle to "demonstrate a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different." Grant , 886 F.3d at 905 (quoting Littlejohn v. Royal (Littlejohn II ), 875 F.3d 548, 552 (10th Cir. 2017) ). This means the errors must "undermine [our] confidence in the outcome" of Postelle's sentencing. Newmiller , 877 F.3d at 1197 (quoting Strickland , 466 U.S. at 694, 104 S.Ct. 2052 ). Of course, in a case such as this, a single juror's choice to impose a sentence less than death meets that standard. Littlejohn II , 875 F.3d at 553. Even still, "[t]he likelihood of a different result must be substantial, not just conceivable." Newmiller , 877 F.3d at 1197 (quoting Harrington v. Richter , 562 U.S. 86, 112, 131 S.Ct. 770, 178 L.Ed.2d 624 (2011) ).
In considering whether an inadequate investigation prejudiced a habeas petitioner, we "reweigh the evidence" on both sides, this time accounting for the petitioner's proposed additions. Littlejohn II , 875 F.3d at 553 (quoting Victor Hooks II , 689 F.3d at 1202 ). This exercise also requires us to account for how the state would have responded to the omitted evidence. E.g., id.
Postelle's mitigation case had several clear focal points.
The first of these was Postelle's difficult upbringing. The jury heard how Postelle's mother had abused and starved him as a young child. Court's Ex. 12 at 2654 (introduced at Tr. 2695). It heard how she "made [him] feel rejected and unloved." Video recording: Patsy Postelle Mitigation *1218Testimony (Defendant's Ex. 2, introduced at Tr. 2844). It heard how, when she eventually gave him up to his grandparents, he was "malnutrishioned, filthy, and [had] sores on [his] bod[y]." Id. The jury heard how thereafter Postelle's mother "refused to take telephone calls from [him or] have anything further to do with [him]." Id. The jury was also told that Brad Postelle's new girlfriend obstructed the father-son relationship. See id. ; Tr. 2712-13, 2747. Finally, the jury learned that, despite all of this, Postelle selflessly cared for his bedridden grandfather, see Tr. 2749, 2814, and later, his handicapped father, see Tr. 2785-86.
Postelle's mitigation case also focused on the role methamphetamine played in his life from an early age. This included evidence of family members cooking and using meth during Postelle's childhood. See Tr. 2698, 2753, 2767, 2770, 2793-94. The jury also heard about Postelle's initiation into meth use himself at the age of twelve or thirteen, including openly ingesting the drug in front of his father and other family members. See Tr. 2698, 2711, 2728, 2752. In fact, the evidence even indicated that Brad Postelle supplied meth to little Gil. Tr. 2765.
Finally, and most relevant to this appeal, the mitigation case concluded with expert testimony regarding Postelle's mental function. See Tr. 2844-96. Dr. Ruwe had run roughly thirty different tests on Postelle. Tr. 2848. He testified that Postelle had "pretty significant [neurocognitive] impairments," along with "pretty severe psychological difficulties." Tr. 2849. These included "pretty pronounced problems with remembering information" and "difficulty with reasoning, especially verbal reasoning." Tr. 2850. Dr. Ruwe told the jury that Postelle gave off "clear[ ] indications of paranoia[ and] disorganized thinking." Tr. 2851. And, according to Dr. Ruwe's assessment, the parts of Postelle's brain responsible for "impulse control, making good decisions, [and] well reason[ed] judgments" had not developed fully or normally by the time of the murders. Tr. 2856; see also Tr. 2857 ("Generally, what that tells us is that the drugs, along with the normal developmental process, makes it more difficult for [Gilbert Postelle] to make good decisions."). Postelle even displayed "symptoms consistent with a post traumatic stress disorder." Tr. 2863.
Moreover, Dr. Ruwe explained, Postelle had not always been this way. Based on tests administered before Postelle dropped out of school, Dr. Ruwe concluded the young Postelle "was performing pretty consistently in the average, unlike current testing." Tr. 2852. "[H]e probably would have had a learning disability or would have qualified for special services," but he had markedly better brain function. Tr. 2852. And though it was somewhat speculative, Dr. Ruwe attributed most of Postelle's mental difficulties "to the longstanding chronic use of drugs and methamphetamine, which are known to have a pretty pronounced impact on cognitive functions, especially memory." Tr. 2853-54.
This testimony culminated in an apparent attempt to cast Postelle as a victim of circumstance, rather than a lost cause. He knew right from wrong. See Tr. 2865. He was capable of choosing between the two. See Tr. 2866. He was young and would continue to mature. See Tr. 2868. Perhaps the structured environment of prison was the best way to reform him. See Tr. 2866.
Though Postelle makes much of Dr. Ruwe's concession that he did not reach the level for "a diagnosis of [intellectual disability]," Tr. 2861; see Aplt. Br. at 5, 6, 7, 32, 38, 44, 49, this comment was not overly prejudicial in context or even necessarily wrong in light of Flynn Effect evidence. Indeed, Dr. Ruwe testified Postelle was not intellectually disabled but instead *1219"in the borderline range" of intellectual functioning. Tr. 2861. His main point was that Postelle fell "pretty close to" an intellectual disability diagnosis, which would "start somewhere around [an IQ of] 70." Tr. 2861. Of course, that would have also been true of Postelle's Flynn Effect-adjusted scores. In fact, Dr. Ruwe placed Postelle "in the 5th percentile range" of relative intelligence, meaning "95 out of 100 ... [people probably] function[ ] at a higher level." Tr. 2862. With Flynn Effect-adjusted scores, Postelle would still have slotted into a relatively similar band. See Haydt et al., supra , at 364. Furthermore, Dr. Ruwe clarified that people are "able to function pretty well until they get down into that borderline range [ ]." Id. (emphasis added). There, said Dr. Ruwe, people like Postelle "start developing more pronounced problems with ... some of the things that we typically take for granted," including "working in the competitive employment force" and "independent living." Id.
Postelle also stresses the "uniquely mitigating" nature of intellectual-disability evidence as support for incorporating the Flynn Effect. See Aplt. Br. at 12-13, 45. But this is not a case where counsel simply ignored such evidence or used it against the defendant. Cf. Smith v. Mullin , 379 F.3d 919, 939-44 (10th Cir. 2004) (finding ineffective assistance where counsel completely failed to incorporate a substantial body of mental health evidence into the mitigation defense). As we have just explained, Postelle's counsel stressed his poor mental health and severe intellectual difficulties in the final stanza of the mitigation case. Our proper focus is thus the marginal benefit of Flynn Effect evidence in light of the other evidence presented, as well as its marginal potential cost to the overall mitigation strategy.10
We have no reason to believe an attempt to use Flynn Effect evidence would have gone unchallenged by the prosecution either. As we mentioned above, incorporation of this evidence might have caused the mitigation phase to devolve into a confusing and tangential "battle of experts" on the validity and practical significance of the Flynn Effect. Not only would this have risked distracting the jury from more salient issues, it might also have alienated jurors sensitive to a defense that appeared to be focusing on minor side issues. Indeed, even absent a central focus on IQ scores, the state still brought Postelle's much-higher nonverbal IQ to light. See Tr. 2875-77. And though the Flynn Effect could have opened the door to an argument that Postelle's IQ might actually fall below 70 after accounting for standard error, see Aplt. Br. at 25, that point cuts both ways. Indeed, the state itself introduced the margin-of-error concept to attack Postelle's IQ-based argument. See Tr. 2871.
In sum, the OCCA reasonably concluded the omission of Flynn Effect evidence did not prejudice Postelle's defense, and that application of Strickland warrants deference on both the trial-counsel claim and the appellate-counsel claim.
*1220* * *
In light of the above, the district court did not err in rejecting Postelle's mitigation-based ineffective assistance claims. The OCCA's rejection of these claims on the merits warrants deference under federal habeas law.
B. Victim Impact Evidence
Postelle also claims the erroneous introduction of victim-impact evidence during the penalty phase of his trial prejudiced his defense. Again, we disagree.
In Booth v. Maryland , 482 U.S. 496, 107 S.Ct. 2529, 96 L.Ed.2d 440 (1987), the Supreme Court interpreted the Eighth Amendment to prohibit capital juries from considering evidence of a crime's impact on the victim and his family as part of its sentencing decision. See id. at 501-02, 107 S.Ct. 2529. Such evidence, the Court explained, "may be wholly unrelated" to the defendant's "blameworthiness" because he "often will not know the victim, and therefore will have no knowledge about the existence or characteristics of the victim's family." Id. at 504, 107 S.Ct. 2529. Moreover, the Court reasoned that capital defendants "rarely select their victims based on whether the murder will have an effect on anyone" else. Id. Victim-impact evidence might thus contribute to death sentences premised upon "factors about which the defendant was unaware, and that were irrelevant to the decision to kill." Id. at 505, 107 S.Ct. 2529. In sum, victim-impact evidence had no place in the jury's sentencing task: "determining whether the death penalty is appropriate in light of the background and record of the accused and the particular circumstances of the crime ." Id. at 507, 107 S.Ct. 2529 (emphasis added).
But the Court revisited Booth a few years later in Payne v. Tennessee , 501 U.S. 808, 111 S.Ct. 2597, 115 L.Ed.2d 720 (1991). Payne recognized "the assessment of harm" resulting from a crime "has understandably been an important concern of the criminal law" in determining both guilt and punishment. Id. at 819, 111 S.Ct. 2597. The Court thus held states could present "evidence of the specific harm caused by the defendant" to the jury at sentencing, id. at 825, 111 S.Ct. 2597, including "evidence about the victim and about the impact of the murder on the victim's family," id. at 827, 111 S.Ct. 2597.
Even still, as the Court has recently pointed out, " Payne 'specifically acknowledged its holding did not affect' Booth 's prohibition on [characterizations of and] opinions about the crime." Bosse v. Oklahoma , --- U.S. ----, 137 S.Ct. 1, 2, 196 L.Ed.2d 1 (2016) (per curiam) (emphasis added) (quoting Ledbetter v. State , 933 P.2d 880, 890 (Okla. Crim. App. 1997) ); see also id. (" ' Booth ... held that the admission of a victim's family members' characterizations and opinions about the crime, the defendant, and the appropriate sentence violates the Eighth Amendment,' but no such evidence was presented in Payne , so the Court had no occasion to reconsider that aspect of the decision." (quoting Payne , 501 U.S. at 830 n.2, 111 S.Ct. 2597 ) ). Thus, in conducting capital sentencing proceedings, state courts must still take care to exclude evidence that goes beyond a victim's subjective suffering and strays into description of the defendant's conduct.
The Oklahoma courts ignored that prohibition in this case. At the prosecution's request, several family members of the victims read prepared statements to the jury. See Tr. 2653-55, 2657-60. As relevant to this appeal, James Alderson's brother described Alderson's gruesome injuries. "On advice of the funeral director," he said, "we decided not to allow [our mother] to view Jimmy's body." Tr. 2654. Due to "[t]he disfigurement caused by *1221head and facial wounds," Alderson's mother therefore "never had a chance for "a final goodbye." Id. Amy Wright's mother then testified to "know[ing] that [Amy] was chased from her home and shot in the back ... without apparent reason." Tr. 2657. As the court below determined, this testimony overstepped the fine line between Booth and Payne , effecting a constitutional violation. R., Vol. I at 626. And because the OCCA did not enforce Booth 's restrictions on appeal, see id. ; see also Postelle I , 267 P.3d at 142-43 (relevant discussion), we do not grant it deference on this issue. See 28 U.S.C. § 2254(d).
But this does not entitle Postelle to automatic relief. Instead, much like in the ineffective assistance context, he must show the error of admitting impermissible victim-impact evidence prejudiced his defense. See Welch v. Workman , 639 F.3d 980, 1002 (10th Cir. 2011) ; see also Brecht v. Abrahamson , 507 U.S. 619, 637-38, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993) (establishing the standard). This analysis requires us to consider all of the evidence from both stages of the trial. See Welch , 639 F.3d at 1004 ; see also Lockett v. Trammel , 711 F.3d 1218, 1239 (10th Cir. 2013) ("In evaluating whether the unconstitutional portions of the ... statement had a substantial and injurious effect on the jury, we must consider it in the context of all of the aggravating and mitigating evidence."). In so doing, we ask whether the jury still would have found the aggravating circumstances outweighed the mitigating factors without the testimony in question. See Lockett , 711 F.3d at 1239-40 ; DeRosa v. Workman , 679 F.3d 1196, 1240 (10th Cir. 2012) ; Welch , 639 F.3d at 1004.
We have already described the mitigation case Postelle presented. It included testimony regarding abuse, neglect, mental illness, intellectual difficulties, and the corrosive influence of drugs.
But the jury thought that evidence could not outweigh the aggravating circumstances. In particular, the jury found the same two aggravating circumstances present in all four murders: First, it found Postelle "knowingly created a great risk of death to more than one person" in each case. Dir. App. R., Vol. VIII at 1550-53. This is, of course, an obvious conclusion to draw in a case involving multiple homicides. Second, the jury found the murders "especially heinous, atrocious, or cruel." See id. According to the instructions, this meant the jury found beyond a reasonable doubt that (1) "either torture ... or serious physical abuse of the victim[s]" preceded the murders, and (2) "the murder[s] [were] ... extremely wicked or shockingly evil ... outrageously wicked or vile ... [or] pitiless, designed to inflict a high degree of pain, or utter indifference to or enjoyment of the suffering of others." Id. at 1521. In this context, " 'torture' means the infliction of either great physical anguish or extreme mental cruelty." Id. Moreover, a finding of "serious physical abuse" or "great physical anguish" must include a finding "that the victim experienced conscious physical suffering prior to ... death." Id.
The victim-impact statements could not have been decisive in a single juror's balancing. The testimony in question centers on two main points: (1) that James Alderson had disfiguring head and facial wounds, Tr. 2654, and (2) that Amy Wright ran for her life only to be randomly killed, Tr. 2657. Both of these statements were substantially redundant and relatively mild when compared to other evidence.
The state had already introduced substantial evidence of Mr. Alderson's wounds. It had presented detailed testimony from a forensic pathologist in the office of the state medical examiner regarding Mr. Alderson's autopsy. See Tr. 1450-56.
*1222That witness recounted Mr. Alderson's "compound open fracture" on the top of his head, Tr. 1454, as well as how a bullet had hit his jaw bone, see Tr. 1455. Clear diagrams of Mr. Alderson's face and body had aided this description. See Ex. 161, 163. A second witness, a forensic consultant, had also described Mr. Alderson's injuries to the jury. Tr. 1549-52. And all of this testimony had followed the state's introduction of a close-up photo of Mr. Alderson's body from the crime scene, showing the open wound on the top of his head. See State's Ex. 54 (introduced at Tr. 1402). Mr. Alderson's brother's statement thus contributed little to the jury's understanding of Mr. Alderson's disfigurement.
So too, the guilt-phase evidence had already painted a more vivid image of Amy Wright's final moments. When police arrived at the scene, they found bullet casings in the trailer, see Tr. 1498-1500; see also State's Ex. 100, 101, 105, 108, 110, and the back door flung wide open, see Tr. 1249. They discovered Ms. Wright's body face down on the grass, obviously shot to death. See Tr. 1243-45. She was in close proximity to a solid metal fence that hemmed in the property, see Tr. 1388, and she had gravel and grass under her fingernails as though she had been clawing at the ground or other objects, see Tr. 1406. Despite being outside in a junk-filled scrapyard, see Tr. 1562, Ms. Wright was not wearing her shoes, see Tr. 1552. According to the autopsy, she had sustained three gunshot wounds, all from behind. See Tr. 1457. Pulling together this evidence, the forensic consultant offered his interpretation of the events: The four victims had been at ease within the trailer home when they experienced some sort of "blitz attack." Tr. 1561. With regard to Ms. Wright in particular, the evidence was consistent with her attempting to flee only to be shot in the midst of escape. See Tr. 1569-70. Surely, then, the jury had already concluded Ms. Wright ran for her life before being gunned down for no other reason than that she was in the wrong place at the wrong time.
Our review of the record thus compels the conclusion that the victim-impact evidence erroneously admitted did not affect Postelle's sentence. In fact, even if we assumed for the sake of argument that the jury should have also heard Flynn Effect evidence as part of the mitigation case, see Aplt. Br. at 52-53 (arguing for a cumulative-error analysis), we do not think the impermissible victim-impact statements could have influenced even a single juror. These statements simply told the jurors what they each already knew.
We thus affirm denial of Postelle's habeas petition on this ground.
C. Expansion of the Certificate of Appealability
As a final matter, we reject Postelle's request to expand the scope of his appeal.
Postelle has asked us to expand the scope of his appeal to cover three additional issues. First , he wishes to challenge the district court's ruling that Oklahoma did not contradict or unreasonably apply Atkins in sentencing him to death. See Mot. Broaden Certificate of Appealability at 15-17 [hereinafter COA Mot.]. Second , he seeks permission to appeal the district court's determination that Oklahoma did not contradict or unreasonably apply the Lockett line of cases in refusing to admit David Postelle's lesser sentence as mitigating evidence. Id. at 3-8. Finally , he requests an appeal of the district court's procedural ruling denying him a stay and abatement or leave to amend his habeas petition. See id. at 8-15.
When a federal district court denies a state prisoner's petition for habeas corpus, he has no absolute right to appeal. See *1223Slack v. McDaniel , 529 U.S. 473, 480-81, 120 S.Ct. 1595, 146 L.Ed.2d 542 (2000). We may, however, grant the petitioner permission to challenge the district court's resolution of discrete, specified issues through a Certificate of Appealability. See 28 U.S.C. § 2253(c). But a petitioner must make "a substantial showing of the denial of a constitutional right" to justify our doing so. Id. at § 2253(c)(2). This means he "must demonstrate that reasonable jurists would find the district court's assessment of the [relevant] constitutional claim[ ] debatable or wrong." Slack , 529 U.S. at 484, 120 S.Ct. 1595 ; see Miller-El v. Cockrell , 537 U.S. 322, 327, 123 S.Ct. 1029, 154 L.Ed.2d 931 (2003). In considering whether a petitioner has made such a showing, we incorporate Congress's mandate of deference to state court decisions. Dockins v. Hines , 374 F.3d 935, 938 (10th Cir. 2004).
Reasonable jurists could not debate the correctness of the additional district court rulings Postelle seeks to appeal.
1. The Atkins Claim
We need not hear appeal on Postelle's Atkins claim because our precedent forecloses relief.
We have already twice held that Oklahoma's rejection of the Flynn Effect as irrelevant to the Atkins analysis does not contradict or unreasonably apply Atkins . See Smith , 824 F.3d at 1244-46. Postelle recognizes these holdings, but argues that other Supreme Court precedents, specifically Lockett and its progeny, compel states to consider the mitigating value of Flynn Effect evidence in applying Atkins . COA Mot. at 17. But the Lockett line of cases only applies to relevant evidence, and only as it relates to the choice to impose the death penalty. Thus, because we have already determined that a state-consistent with Atkins -may deem Flynn Effect evidence irrelevant to death penalty eligibility , our precedents still preclude Postelle's argument. Reasonable jurists therefore could not debate this issue. See, e.g., United States v. Tafoya , 557 F.3d 1121, 1129 (10th Cir. 2009). Accordingly, we must deny Postelle permission to raise it on appeal.
2. David Postelle's Sentence as Mitigating Evidence
Neither can we permit Postelle to appeal the district court's ruling on the exclusion of David Postelle's sentence as mitigating evidence.
When Postelle raised this issue in state court, the OCCA correctly noted the constitutional requirement that "the proffered evidence ... relate to the defendant's personal circumstances, [that is], his character, record or circumstances of the offense." Postelle I , 267 P.3d at 141. To be sure, some courts have determined that evidence of a codefendant's sentence is relevant to capital sentencing. See id. at 140-41 (collecting cases). But the question we must answer is whether the OCCA unreasonably applied or contradicted Lockett and its progeny in rejecting Postelle's argument and taking the opposite stance. And the presence of a legitimate controversy regarding the relevance of a codefendant's sentence, see COA Mot. at 7, indicates the Lockett line of cases does not answer the question. Thus, even if the OCCA was ultimately wrong, reasonable jurists could not debate that its decision deserves deference under federal habeas law. And despite Postelle's argument to the contrary, see id. at 6, the severity of the sentence at issue cannot alter this analysis.
We therefore deny Postelle permission to appeal this issue as well.
3. The Actual Innocence Claim
Finally, Postelle wishes to appeal the district court's procedural ruling denying *1224him a stay and abatement or else leave to amend his habeas petition. This challenge to a matter of procedure requires a somewhat different analysis, but ultimately does not warrant further review.
We begin with some additional factual background. Postelle submitted his original habeas petition to the district court on September 3, 2013. See R., Vol. I at 10. Over a year and a half later, he moved for the district court to stay and abate habeas proceedings or otherwise permit him to amend his petition with an actual-innocence claim. See id. at 488. This was because David Postelle had contacted Postelle's attorneys and confessed to the murders. See id. at 491. David claimed he had directed Postelle not to discuss the events of the crime with anyone-even his own counsel-and that Postelle did not actually shoot anyone. See id. at 518-21 (David Postelle letter). On account of this new evidence, Postelle moved for the court to stay federal proceedings to allow him to exhaust claims of actual innocence (and possibly interference with counsel) in the Oklahoma courts. See id. at 491. In the alternative, he asked for leave to amend his petition "to add facts, argument, and authorities based upon the confession of his older brother." Id. He further asked for leave to newly assert "a constitutional claim of actual innocence." Id.11
The district court denied these requests. See Dkt. 71 at 2. First, it thought a stay and abatement "inappropriate in this situation" because Postelle sought "to exhaust claims to add to his existing petition" rather than "exhaust claims that are included in his petition." Id. at 3. Next, the district court rejected Postelle's request to amend his petition, reasoning that the new claims did not derive from the same operative facts as the old claims, and therefore the amendment could not relate back to the original habeas petition. See id. at 3-5. Accordingly, the amendment would fall outside the statute of limitations. Id. at 4. Equitable tolling was also unavailable because David's confession was not reliable, nor did it present "new" evidence in the relevant sense. Id. at 6-8.
The district court thus construed Postelle's motion not as a request for a stay or leave to amend, but as a second habeas petition. See id. at 9. Accordingly, it transferred the petition to this court for us to decide whether to authorize it as such. Id. ; see 28 U.S.C. § 2244(b)(3)(A). We reserved judgment on that question pending resolution of this appeal. See In re Postelle , No. 16-6237, unpublished order at 3 (10th Cir. Oct. 18, 2016).
Postelle now moves for permission to appeal the district court's rejection of his requests for a stay or leave to amend his petition. See Mot. at 8-15. To appeal a procedural ruling in a habeas action, "a litigant ... must demonstrate that" the "ruling ... is itself debatable among jurists of reason" Buck v. Davis , --- U.S. ----, 137 S.Ct. 759, 777, 197 L.Ed.2d 1 (2017). We deny this motion because reasonable jurists could not debate the district court's procedural ruling.
To begin, Postelle gives the court no reason to doubt the district court's holding that it could not stay and abate habeas proceedings for exhaustion of a claim not yet before the court. See Mot. at 11-15.
*1225And indeed, the Supreme Court's decision in Rhines v. Weber , 544 U.S. 269, 125 S.Ct. 1528, 161 L.Ed.2d 440 (2005), approving of the stay-and-abatement process for raised but unexhausted claims does not appear to apply to claims not yet raised at all. See id. at 275-79, 125 S.Ct. 1528 (discussing the procedure only with regard to "mixed" petitions containing both exhausted and unexhausted claims). We thus deem the district court's resolution of that issue undisputably within its discretion.
Neither does Postelle directly address the district court's treatment of the amendment issue. And rightfully so, as reasonable jurists could not debate it either.
A habeas petition "may be amended or supplemented as provided in the rules of procedure applicable to civil actions." 28 U.S.C. § 2242. Under the Federal Rules of Civil Procedure, a party generally has an absolute right to amend his pleading once within 21 days of its service. See Fed. R. Civ. P. 15(a)(1)(A). Once that window has passed-and in this case it certainly has-a party needs his opponent's consent or leave of court to file an amendment. See id. at 15(a)(2). And even though courts should grant this leave freely, see id. , federal habeas law strictly limits the circumstances under which an amendment can relate back to the original petition filing. As relevant here, it may do so "if and only if ... the proposed amendment does not seek to add a new claim or to insert a new theory into the case." United States v. Espinoza-Saenz , 235 F.3d 501, 505 (emphasis added) (quoting United States v. Thomas , 221 F.3d 430, 431 (3d Cir. 2000) ); see also Woodward v. Williams , 263 F.3d 1135, 1142 (10th Cir. 2001) ("Although this petition was brought under § 2254 rather than § 2255, we see no reason to treat the issue differently."). Postelle's amendment exclusively seeks to add new claims to the case under new theories of the facts. Accordingly, his actual-innocence claim could not possibly relate back to the original filing absent equitable tolling. Postelle thus attacks only the district court's treatment of the equitable tolling issue.
He cannot prevail on this theory. To be sure, a valid claim of actual innocence might render Postelle's amendment timely and bypass the relation-back problem altogether. See Gibson v. Klinger , 232 F.3d 799, 808 (10th Cir. 2000). But "[t]o be credible, such a claim requires petitioner to support his allegations of constitutional error with new reliable evidence-whether it be exculpatory scientific evidence, trustworthy eyewitness accounts, or critical physical evidence-that was not presented at trial." Schlup v. Delo , 513 U.S. 298, 324, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995). Moreover, "[t]o establish the requisite probability" that a miscarriage of justice will occur absent equitable tolling, "the petitioner must show that it is more likely than not that no reasonable juror would have convicted him in the light of the new evidence." Id. at 327, 115 S.Ct. 851.
Postelle faults the district court for conducting a "deep dive into the merits" of his actual-innocence claim to carry out this analysis. Id. at 11. According to Postelle, the district court "should have [only] asked ... whether [David's] affidavit raised a debatable question of ... Postelle's ineligibility for the death penalty." Id. But that simply is not the law. Even assuming it constituted "new reliable evidence," but cf. Hubbard v. Pinchak , 378 F.3d 333, 340 (3d Cir. 2004) ("A defendant's own late-proffered testimony is not 'new' because it was available at trial."), the question before the district court was whether any reasonable juror could have found Postelle guilty beyond a reasonable doubt in light of David's statement. On appeal, we would review its *1226decision for abuse of discretion. See Espinoza-Saenz , 235 F.3d at 503 ; cf. Carter v. Bigelow , 787 F.3d 1269, 1278 & n.6 (10th Cir. 2015) (observing the abuse-of-discretion standard applies to denials of motions to supplement as well as motions to amend).
Accordingly, the question before us now is whether reasonable jurists could debate the district court's exercise of discretion concluding reasonable jurists could still convict Postelle in the face of David's statements. For obvious reasons, this is not a close call. David Postelle's letter cannot wash away the mountain of other evidence presented at trial, including a directly contradictory account from another eyewitness. Reasonable jurors could easily disregard David's account. Accordingly, the court below clearly and undisputably rendered a decision within the bounds of its discretion on this issue.
* * *
For these reasons, we deny Postelle a Certificate of Appealability covering additional issues.
III. Conclusion
For the forgoing reasons, we AFFIRM the district court's denial of habeas relief and DENY Postelle's motion to expand the Certificate of Appealability.
The parties have also briefed tangential matters of preservation and procedure inherent to the federal habeas process. See Aplt. Br. at 35-38; Aple. Br. at 23-37. As the following makes clear, however, we have bypassed these issues and gone straight to the substance of Postelle's appeal. See, e.g., Nielander v. Bd. of Cty. Comm'rs , 582 F.3d 1155, 1166 (10th Cir. 2009) ; Revilla v. Gibson , 283 F.3d 1203, 1214 (10th Cir. 2002).
The dissent at times characterizes this phenomenon as producing "bias in the IQ tests" administered to Postelle. Dissent at 1226. To clarify, the Flynn Effect does not skew IQ test results based on the identity or personal characteristics of the test taker. Rather, the Flynn Effect tells us that an IQ test indicates the test taker's intelligence relative to the norming group, and that the average intelligence of the norming group will often lag behind the average intelligence of the general population at the time of the test's administration.
We neither endorse nor reject the Flynn Effect as a scientific matter. But cf. Aplt. Br. at 23, 40, 42 (arguing the Flynn Effect's validity). Our analysis does not depend on its validity. Instead, we assume for the sake of argument that the Flynn Effect is indeed a feature of intelligence testing that counsels in favor of the personalized IQ score revisions Postelle proposes.
The Atkins opinion used the then-accepted nomenclature "mentally retarded." E.g. Atkins v. Virginia , 536 U.S. 304, 316, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002). Our opinions have since adopted the now-preferred term "intellectually disabled." See Smith v. Duckworth (Smith II ), 824 F.3d 1233, 1242 (10th Cir. 2016).
One of Postelle's proposed adjustments also accounts for a supposed norming error specific to the IQ test in question. See Aplt. Br. at 22 & n.8, 26-27 & n.10. Again, we assume the validity and accuracy of Postelle's IQ adjustments for the sake of argument.
The dissent also rejects Postelle's reading of the OCCA opinion. The dissent proceeds to de novo review on a different theory, which we address more fully below. See infra p. 1214 n.7.
The dissent proceeds to de novo review because, in its view, "[b]y failing to address Postelle's [mitigation] argument ..., the OCCA failed to adjudicate Postelle's claim on its merits." Dissent at 1232. But when a state court "addresses some but not all of a [habeas petitioner's] claims," we presume that the court silently rejected remaining claims on the merits. Johnson v. Williams , 568 U.S. 289, 133 S.Ct. 1088, 1094, 185 L.Ed.2d 105 (2013). Of course, a petitioner can rebut that presumption by giving us "some reason to think some other explanation for the state court's decision is more likely." Harrington v. Richter , 562 U.S. 86, 99-100, 131 S.Ct. 770, 178 L.Ed.2d 624 (2011). To do so, however, the petitioner must point to some "indication" that the state court ignored the claim or rejected it on grounds of state procedure. Id. at 99, 131 S.Ct. 770.
Yet Postelle has never so much as attempted to argue that the OCCA ignored his mitigation-based claim. See Aplt. Br. at 14-18 (applying the OCCA's discussion of the eligibility-based argument to the mitigation-based argument); Reply Br. 8-13 & n.2 (same). And though the OCCA could have procedurally barred the claim under Oklahoma rules of post-conviction procedure, see Postelle II , slip op. at 11, neither Postelle nor the dissent gives us any reason to believe the OCCA did so, cf. James v. Ryan , 733 F.3d 911, 915 (9th Cir. 2013) (observing that the express application of a state procedural bar can rebut the Richter presumption). And we think that explanation quite unlikely given the OCCA considered the merits of Postelle's other ineffective assistance claims after noting the procedural bar. See Postelle II , slip op. at 11-17. Thus, if we did read the OCCA opinion to omit any discussion of the claim, the law would compel us to presume the claim's silent rejection on the merits. Postelle would then bear the burden to show "there was no reasonable basis for the state court to deny relief" under Strickland. Richter , 562 U.S. at 98, 131 S.Ct. 770.
Moreover, in concluding the OCCA ignored Postelle's mitigation-based claim, the dissent's position raises the question of whether Postelle fairly presented that claim to the OCCA in the first place. See, e.g., Grant v. Royal , 886 F.3d 874, 890-92 (10th Cir. 2018). And if the OCCA did indeed reject the claim on state procedural grounds, the dissent would have to hold those grounds inadequate before it could justify habeas relief. See Walker v. Martin , 562 U.S. 307, 315, 131 S.Ct. 1120, 179 L.Ed.2d 62 (2011) ; see also Aple. Br. at 30-33 (defending Oklahoma's procedural bar for ineffective assistance claims not raised on direct appeal). Though we have bypassed these issues given our interpretation of the OCCA opinion and resolution of the merits, the dissent simply leaves them unresolved. In sum, the OCCA's rejection of this claim warrants deference under AEDPA's standards.
The dissent has identified many appearances of the Flynn Effect in judicial opinions and academic literature prior to Postelle's 2008 trial. See Dissent at 1230 & n.1. But the question is not whether Postelle's counsel could have found references to the Flynn Effect after knowing to look for them. The question is whether the Flynn Effect featured so prominently in capital cases and literature prior to Postelle's trial that any failure to discover it would indicate severe professional incompetence. Indeed, the dissent's analysis succumbs to the hindsight bias our court has cautioned against. See, e.g., Grant , 886 F.3d at 903.
Our analysis does not state or imply that a capital defense attorney may "delegate development of the overall litigation strategy" to an expert witness. Dissent at 1226. Clearly that responsibility falls to counsel alone. But where, as here, counsel has decided to argue poor mental health and diminished cognitive function as mitigating factors justifying a lesser sentence, counsel may indeed presume that a qualified expert in the field of clinical neuropsychology would apply that expertise to the chosen strategy.
We disagree with the dissent's prejudice analysis on this point. As a scientific and practical matter, there just is not much difference between an IQ slightly below 75 and an IQ slightly above 75-certainly not so much as to fairly indicate Postelle definitively "lack[ed]," rather than possessed, "the intellectual capacity to bear full culpability for his crimes." Dissent at 1234. Indeed, the concept of standard error would itself rebut such an inference. See Hall v. Florida , 572 U.S. 701, 134 S.Ct. 1986, 2000, 188 L.Ed.2d 1007 (2014) ("An IQ score is an approximation , not a final and infallible assessment of intellectual functioning." (emphasis added) ). And if evidence of possible intellectual disability is so strongly mitigating, then evidence of functioning in the borderline range is at least somewhat mitigating-and plainly not aggravating.
For the sake of argument, we accept Postelle's contention that "a truly persuasive demonstration of actual innocence ... would render [his] execution ... unconstitutional, and warrant federal habeas relief if there were no state avenue open to process such a claim." COA Mot. at 15 (quoting Herrera v. Collins , 506 U.S. 390, 417, 113 S.Ct. 853, 122 L.Ed.2d 203 (1993) ). But cf. Doe v. Jones , 762 F.3d 1174, 1176 n.5 (10th Cir. 2014) (acknowledging the Supreme Court has never resolved whether such a claim exists).