Hugh Nicely appeals the district court’s denial of his petition for a writ of habeas corpus. Because the state court’s adjudication of Nicely’s ineffective-assistanee-of-counsel claim was not unreasonable, we affirm.

I.

In 1994, a Tennessee jury convicted Hugh Nicely on charges that he sexually abused his minor stepdaughter from the time she was six until she was ten. The evidence against him primarily consisted of the victim’s testimony detailing the abuse; testimony from the victim’s daycare teachers that the victim informed them that Nicely had abused her; photographs of the victim’s vagina; and testimony from the nurse who examined the victim and took the photographs, including her opinion that the photographs showed an absence of the victim’s hymen at the six o’clock position, something the nurse said was “consistent with” vaginal penetration. Nicely denied all wrongdoing and testified in his defense.

He was tried twice on these charges; the first trial ended in a mistrial when the jury was unable to reach a unanimous verdict. One primary difference between the two trials was the introduction and explanation of the photographs taken by the nurse offered in the second trial. Nicely was sentenced to fifty-three years in prison. Some of his convictions were reversed on appeal for reasons irrelevant here, but his sentence remained the same.

With the help of new counsel, Nicely petitioned for post-conviction relief in state court. He alleged, among other things, that his trial counsel was ineffective for failing to find and present testimony from a medical expert who could counter the nurse’s opinion that the photographs showed the absence of hymenal tissue. The court heard testimony from Nicely, his trial counsel, the nurse who examined the victim and testified at trial, and Dr. Matthew Seibel, among others. Dr. Seibel testified in part that, while he noted some “irregularities” on the victim’s hymen as depicted in the photographs, he found no place where the hymen was absent, including at the six o’clock position. This testimony was contrary to the nurse’s on the same point. Dr. Seibel further opined that he could not determine, based upon the photographs, whether the victim had been sexually penetrated.

Based upon Dr. Seibel’s testimony and the absence of any tactical reason offered by trial counsel for not seeking an expert’s opinion on the photos, the state trial court granted Nicely’s petition for relief on the basis of ineffective assistance of counsel. The Tennessee Court of Criminal Appeals reversed the grant of relief, and the Tennessee Supreme Court declined further review.

This federal habeas litigation followed. The district court accepted a magistrate *400 judge’s recommendation and denied relief on all claims. It certified one issue for appeal: “whether the petitioner’s trial counsel’s failure to obtain additional expert testimony ... was ineffective assistance of counsel.” Nicely timely appealed.

II.

We review the district court’s habeas decision de novo. Ayers v. Hudson, 623 F.3d 301, 307 (6th Cir. 2010). The Antiterrorism and Effective Death Penalty Act (“AEDPA”) applies because Nicely’s petition was filed after the statute’s effective date and because his claim was adjudicated on the merits by a state court. See 28 U.S.C. § 2254(d); Harrington v. Richter, - U.S. -, 131 S.Ct. 770, 784, 178 L.Ed.2d 624 (2011). Under AEDPA, relief is warranted only where a state court’s merits determination was “contrary to, or involved an unreasonable application of, clearly established Federal law,” or was “based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” 28 U.S.C. § 2254(d). To be unreasonable, “the state court’s decision must have been more than incorrect or erroneous[;] ... [it] must have been ‘objectively unreasonable.’ ” Wiggins v. Smith, 539 U.S. 510, 520-21, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003) (citations omitted). The threshold for “unreasonableness” is “substantially higher” than it is for incorrectness, Schriro v. Landrigan, 550 U.S. 465, 473, 127 S.Ct. 1933, 167 L.Ed.2d 836 (2007), satisfied only when a state-court decision “was so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement,” Harrington, 131 S.Ct. at 786-87.

III.

The only issue certified for our review is whether Nicely’s trial counsel was ineffective for not investigating and later presenting expert testimony.

A.

To succeed on a claim of ineffective assistance of counsel, a defendant must show that his lawyer’s performance was deficient and that the deficiency prejudiced his defense. Strickland v. Washington, 466 U.S. 668, 687, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). A court need not address both components of the inquiry if a defendant makes an insufficient showing on one. Id. at 697, 104 S.Ct. 2052. Both prongs are mixed questions of law and fact, id. at 698, 104 S.Ct. 2052, and are therefore reviewed under AEDPA’s “unreasonable application” prong when that statute applies, as it does here, Barnes v. Elo, 339 F.3d 496, 501 (6th Cir. 2003).

To establish prejudice from counsel’s deficiency, a defendant must show a “reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Strickland, 466 U.S. at 694, 104 S.Ct. 2052. A “reasonable probability” is a probability “sufficient to undermine confidence in the outcome.” Id. “The likelihood of a different result must be substantial, not just conceivable.” Harrington, 131 S.Ct. at 792. A verdict or conclusion strongly supported by the record is less likely to have been affected by errors than one with weak record support. See Strickland, 466 U.S. at 696, 104 S.Ct. 2052. When AED-PA applies, the ultimate question for a federal court is not whether its confidence in the outcome is sufficiently undermined by counsel’s errors, but rather whether the state court’s prejudice determination was reasonable. See Lockyer v. Andrade, 538 U.S. 63, 75-76, 123 S.Ct. 1166, 155 L.Ed.2d 144 (2003); Williams v. Taylor, *401 529 U.S. 362, 411, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000).

B.

Nicely’s claim concerns his attorney’s failure to investigate and later call an expert to counter testimony from the state’s expert explaining photographs of the victim’s hymen taken during a physical examination. Julie Rosof, the nurse practitioner who physically examined the victim shortly after she reported the abuse, offered the jury her opinion that the photographs showed a complete absence of hymenal tissue at the six o’clock position. Rosof also testified that there has never been a female that did not have hymenal tissue at that position when born, and that medical professionals routinely rely on this fact in making medical diagnoses. The nurse further testified that the absence of hymenal tissue was “consistent with” penetration of the victim’s vagina, though she could not say she deemed the medical finding “definite proof’ of penetration. Notably, she disagreed that the physical findings were “also consistent with there not having been penetration.” (Emphasis added.) Rosof confirmed that the only physical finding she relied upon when testifying at Nicely’s trial that her exam was consistent with penetration was the absence of hymenal tissue at the six o’clock position.

Rosof s opinion regarding the absence of hymenal tissue at the six o’clock position, it turns out, is not beyond dispute — there is an alternative medical interpretation of the photos. After his conviction, Nicely found and solicited from Dr. Matthew Sei-bel, a medical doctor board certified in pediatrics and specializing in child abuse and neglect, a medical opinion that the photographs show a fully intact hymen, including at the six o’clock position.

During post-conviction proceedings in state court, Dr. Seibel testified that, although he could not see hymenal tissue at the six o’clock position in some of the photographs, he was unable to tell from those photos whether the tissue was actually absent or had simply flattened down due to gravity and thus only appeared to be absent. However, based upon photographs of the victim taken while in a different position (in the “knee-chest” position, as opposed to the “frog leg” position), Dr. Seibel was able clearly to observe the full presence of hymenal tissue at the six o’clock position. Dr. Seibel did acknowledge “some irregularities,” including a small crease at the six o’clock position, but explained that “a lot of time we’ll see little seams or creases in that area,” and that such irregularities “are, in fact, normal findings in a lot of children. We see them quite often.” (He later stated that hyme-nal irregularities and absences are entirely different matters, and he would never document an irregularity as an absence or vice versa.) Explaining why one might see hymenal tissue in some positions but not in others, Dr. Seibel said the following:

[With] a complete transection, [the hymen] will be absent with the child supine, it will be absent with the child in [the] knee-chest [position]. If it’s absent it’s absent. Sometimes we will see it where it’s laying down or flattened in the supine position but then drops down almost like a curtain in knee-chest; and it’s there; you just couldn’t see it in the supine position.

“Transection,” Dr. Seibel later explained, “is where the hymen has torn through to the vaginal wall and typically does not heal back in scar fashion; ... it just heals open, and [there will be] a complete absence of hymen in that area extending back to the vaginal wall.” Dr. Seibel stated unequivocally in regards to hymenal tissue at the six o’clock position: “[I]t’s absolutely there, sir. You can see it.” *402 And: “there’s not a complete transection. There is hymenal tissue all the way around, sir.” Finally: “The hymen’s there, sir. You can clearly see it in the knee-chest [position].” 1

On cross-examination, Dr. Seibel agreed that, had he examined the victim, he would have made a medical diagnosis of sex abuse, but would have done so based entirely on the verbal history the victim provided, not the physical findings, which he deemed equivocal. (“History,” Dr. Seibel explained, “is all the medical, social information, the history of present illness and any other pertinent historical verbal issues which help you to make a medical diagnosis.”). Though he agreed that the physical examination was “consistent with” penetrating injury, he stated that a normal exam, too, is “consistent with” a penetrating injury. In other words, the photographs of the victim’s hymen did not prove the existence of a penetrating injury; they were simply consistent with an injury, such that he could not rule out a penetrating injury, though even a perfectly normal hymen, Dr. Seibel explained, can be consistent with a history-based finding of penetration.

After considering this and other testimony, the state trial court concluded that counsel was deficient for failing to call a medical expert to refute Rosof, and that the deficiency prejudiced the defense. With respect to prejudice, it noted that, given the “two almost identical trials,” it could “more easily analyze the effects of the photos/slides on the second trial[.]” The first one did not involve photos; the second one did. The first ended in deadlock; the second in a conviction. In the trial court’s view, given the different outcomes, “the significance of the photographs/slides can not be discounted in their importance and the effects of the photos and interpretation of the photos cannot be minimized.” It found a reasonable probability that the result of the proceeding would have been different had the jury heard Dr. Seibel’s testimony.

The Tennessee Court of Criminal Appeals disagreed. With respect to prejudice, the court explained as follows:

[W]e disagree with the conclusions of the post-conviction court regarding the effect that Dr. Seibel’s testimony would have had at the trial. Certainly, his testimony regarding the intactness of the victim’s hymen differed from that of Nurse Rosof-Williams. However, whether her hymen was completely intact was not the ultimate issue. Dr. Seibel acknowledged that, even given his different interpretation of the photographs, he would not have disagreed that the victim had been sexually abused and that, in fact, the “irregularity” which he observed at the six o’clock position on the victim’s hymen was consistent with a penetrating injury. As to this, Dr. Sei-bel’s opinion would appear to have been helpful to the prosecution, not the defense. Additionally, we note that, although the petitioner presented expert proof at trial that the large size of his penis would have prevented his penetrating the victim, Dr. Seibel was not asked at the evidentiary hearing whether the irregularities he observed from the photographs of the victim’s vagina were consistent with penetration by a large penis. His response to this question could have undercut the petitioner’s claim of “impossibility.” Thus, we conclude that the petitioner failed to establish that trial counsel either was ineffective in not presenting this testimony at *403 trial or that the petitioner was prejudiced by the fact this did not occur.

C.

We do not address counsel’s pre-trial performance because we conclude that the Tennessee Court of Criminal Appeals’ prejudice determination was reasonable.

Although we observe material differences in the two opinions from Rosof and Dr. Seibel regarding injury to the victim’s hymen, reasonable judges could view them as largely consistent with one another. Specifically, one could reasonably read Dr. Seibel’s statements that the physical examination was “equivocal in nature” 2 and that he had “some questions about ... the area at 6:00” as largely consonant with Rosofs opinion regarding the absence of hymen at that same position. And Dr. Seibel’s opinion regarding the presence of “irregularities” in the victim’s hymen that are “consistent with” a finding of penetrating injury can reasonably be considered in accord with Rosofs identical opinion. This overall similarity in the testimony from both experts suggests that the absence of Dr. Seibel’s opinion regarding the photographs had little or no effect on the ultimate outcome, and may have even bolstered the state’s case. It was reasonable for the state court implicitly to conclude that the only opinion Dr. Seibel might offer that would undermine confidence in the verdict would be that the state of the victim’s hymen, as depicted in the photos, was inconsistent with any finding of vaginal penetration, an opinion Sei-bel was not able to provide.

In addition, there is other evidence in the record, though not expressly considered by the state court in its opinion, that supports its prejudice determination. See Holder v. Palmer, 588 F.3d 328, 341 (6th Cir. 2009) (noting that when AEDPA applies, federal courts consider the result or decision reached by the state court, not necessarily the reasoning used to get there); Neal v. Puckett, 286 F.3d 230, 246 (5th Cir. 2002) (en banc) (per curiam) (“[Ojur focus on the ‘unreasonable application’ test under Section 2254(d) should be on the ultimate legal conclusion that the state court reached and not on whether the state court considered and discussed every angle of the evidence.”).

First, when considering the strength of the state’s case and the effect the photos had on the jury when viewed alongside Rosofs testimony, it is reasonable to attribute little significance to the fact that Nicely’s first trial ended in deadlock while the second ended in a conviction; evidence apart from Rosofs explanation of the photographs can reasonably explain the difference in outcomes. Trial counsel testified at the evidentiary hearing, for example, that the victim seemed “more upbeat, and sunny” at the first trial, while at the second, she was “more shy, or withdrawn,” making it harder to effectively cross-examine and impeach her credibility. Counsel also recalled that the prosecutor focused more in the first trial on penetration and more in the second on masturbation, thereby making physical evidence of penetration less relevant to the jury’s verdict.

Second, it is reasonable to find significant importance in the testimony from the victim’s daycare teachers, who testified that they learned of the abuse from the victim’s friend. This testimony substantially bolstered the victim’s story and could have informed the jury’s verdict.

Third, Nicely’s own medical expert, Dr. Steven Miniat, after viewing photographs *404 of the victim’s vagina while she was in the knee-chest position, testified at trial that “[t]he absence of the hymen there would dictate that one should have increased awareness and should probe farther” into whether there was sexual abuse. He further acknowledged that he would have contacted state authorities had he examined the victim and observed what the photographs depicted. Finally, he agreed with Rosof that the absence of hymen was consistent with penetration. One could reasonably conclude that Dr. Miniat’s testimony in this regard would cut strongly against any favorable testimony Dr. Seibel could have provided in Nicely’s defense.

In sum, because the decision of the Tennessee Court of Criminal Appeals was not unreasonable, habeas relief is not warranted.

IV.

Next, Nicely contends that the district court erred by not allowing him an evidentiary hearing. We are inclined not to consider this issue in light of the warden’s objection that it was not included in the certificate ■ of appealability (“COA”). See 28 U.S.C. § 2253(c); Perkins v. McQuiggin, 670 F.3d 665, 669 (6th Cir.) (“Our review of a petitioner’s § 2254 motion is limited to those issues specified in the certificate of appealability.”), cert. granted, — U.S.-, 133 S.Ct. 527, 184 L.Ed.2d 338 (2012); see also Gonzalez v. Thaler, — U.S.-, 132 S.Ct. 641, 649, 181 L.Ed.2d 619 (2012) (noting that “the failure to obtain a COA is jurisdictional, while a COA’s failure to indicate an issue is not[,]” thereby rendering defects in the certificate susceptible to forfeiture). Nonetheless, assuming, without deciding, that Nicely’s challenge is properly before us, we reject it. Nicely asked the district court for an evidentiary hearing so he could try to clarify Dr. Seibel’s earlier testimony, which Nicely claims was severely misconstrued by the Tennessee Court of Criminal Appeals in reversing the grant of relief. The district court denied that request, relying on the Supreme Court’s recent decision in Cullen v. Pinholster, — U.S.-, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011). That was correct. In Cullen, the Court explained that federal habeas review governed by § 2254(d)(1) is limited “to the record that was before the state court that adjudicated the claim on the merits.” Id. at 1398. The decision precludes Nicely from further developing his Strickland claim in federal court. Nicely attempts to work around Cullen by arguing that the need for Dr. Seibel’s further testimony arose for the first time on appeal, after the close of the evidence in the post-conviction proceeding. But he does not explain why that matters. Because his ineffective-assistance claim was decided on the merits, review under § 2254(d)(1) is limited to evidence the state court had before it. Cullen admits of no exception that would apply here. Cf. Ballinger v. Prelesnik, 709 F.3d 558, 561-63 (6th Cir. 2013) (concluding that an ineffective-assistance claim was adjudicated on the merits in state court and thus barred from further factual development in a federal evidentiary hearing, notwithstanding the state court’s decision disallowing the petitioner from developing the claim in an evidentiary hearing).

V.

For these reasons, we affirm the judgment of the district court.