dissenting.
I respectfully dissent. I disagree with the majority’s conclusions that (1) the Ohio courts applied law contrary to Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), (2) our review is not constrained by AEDPA and that de novo review is appropriate, and (3) Vasquez received constitutionally ineffective assistance of counsel. In my view, the majority and the district court have engaged in an impermissible reconstruction of Vasquez’s trial, failed to avoid the warping effects of hindsight, and declined to afford a strong presumption that Attorney Butler’s conduct fell within the wide range of reasonable professional assistance. Accordingly, I would reverse the judgment of the district court and deny the petition for a writ of habeas corpus.
I.
As the majority concedes, the district court erred by not according proper deference to the state court’s factual findings and conclusions of law under the Antiter-rorism and Effective Death Penalty Act of 1996 (AEDPA). 28 U.S.C. § 2244. The majority concludes, however, that de novo review is nonetheless appropriate because the Ohio courts applied law that was contrary to Strickland v. Washington, 466 U.S. at 668, 104 S.Ct. 2052; see Fulcher v. Motley, 444 F.3d 791, 799 (6th Cir.2006) (when a state court applies law that is “contrary to” clearly established federal law, a federal court is unconstrained by § 2254(d)(1) and de novo review is appropriate).
At the outset, I take issue with footnote 1 of the majority’s opinion. In it, the majority elects to “focus” its Strickland analysis on the Ohio trial court’s decision instead of the opinion rendered by the Ohio Court of Appeals. My colleagues base their decision on what they characterize as “an apparently unresolved question about the state decision a federal court should defer to under AEDPA[,]” namely, the majority queries whether an appellate court decision, governed by an abuse-of-discretion standard, constitutes “an adjudication] on the merits” under AEDPA. 28 U.S.C. § 2254(d) (a federal court shall not grant the writ “with respect to any claim that was adjudicated on the merits in State court proceedings unless....”).
However, it is well-established that a federal court reviewing a habeas petition should examine the decision of the last state court “to [render] a reasoned opinion on the issue[.]” Payne v. Bell, 418 F.3d 644, 660 (6th Cir.2005), cert. denied, 548 U.S. 908, 126 S.Ct. 2931, 165 L.Ed.2d 958 (2006); see Davie v. Mitchell, 547 F.3d 297, 315 (6th Cir.2008); Joseph v. Coyle, 469 F.3d 441, 450 (6th Cir.2006). Nonetheless, the majority contends that:
[w]here an appellate court reviews under an “abuse of discretion” standard, it is not adjudicating the merits of the claim so much as the merits of the decision under review. The question is only whether it fell into what is presumably a broad band of discretion. AEDPA, however, is meant to secure deference to the application of federal law by a state court in its “adjudication] on the merits” of a petitioner’s claim. 28 U.S.C. § 2254(d). Accordingly, AEDPA perhaps does not have anything to say about a state court’s decision that another state court did not abuse its discretion — it is concerned only with whether the actual adjudication of the merits of petitioner’s claim was contrary to, or an unreasonable application of, federal law.
In “focus[ing]” its analysis almost exclusively on the Ohio trial court opinion, the *123majority relies on Benge v. Johnson, 474 F.3d 236 (6th Cir.2007) as authority. In Benge, the Ohio Supreme Court analyzed the effect of an unobjected-to jury instruction in the context of plain-error analysis, not under the governing, and less burdensome, Strickland standard. We held that “[bjecause Benge could have met his burden under Strickland[] despite not being able to demonstrate plain error, [the Ohio Supreme Court’s] analysis did not constitute an ‘adjudication on the merits’ of Benge’s ineffective-assistance-of-counsel claim.” Benge, 474 F.3d at 246. Indeed, the Ohio Supreme Court’s analysis was extremely limited and consisted of only two sentences, holding that “even if the jury instruction is deemed improper, such an error will not mandate reversal unless it constitutes plain error. In other words, we must determine whether ‘but for the error, the outcome of the trial clearly would have been otherwise.’” State v. Benge, 75 Ohio St.3d 136, 661 N.E.2d 1019, 1025 (1996). The Ohio Supreme Court neither mentioned Strickland nor the Sixth Amendment when disposing of petitioner’s ineffective assistance of counsel claim.
In Danner v. Motley, 448 F.3d 372 (6th Cir.2006), we reached a similar conclusion. In our review of petitioner’s asserted Strickland error, there was “no indication in the trial court’s comments that it examined Danner’s Sixth Amendment claim. Nor [was] there evidence in the Kentucky Supreme Court’s opinion that it considered the Sixth Amendment at all in ruling on Danner’s claim.” Danner, 448 F.3d at 376. We noted in Danner that “[a]ny consideration of the Sixth Amendment contained within the state case law upon which the state courts relied is too attenuated to consider the Sixth Amendment claim to have been ‘adjudicated on the merits.’” Id.
Our decisions in Benge and Danner were primarily concerned with the state court’s cursory treatment of the petitioner’s Strickland claims and are therefore harmonious with the rule stated in Payne and Davie — we must review “the last state court [decision] to [render] a reasoned opinion on the issue[.]” Payne, 418 F.3d at 660 (emphasis added). Unlike the state court decisions in Benge and Danner, the state appellate court opinion in the present case engaged in reasoned and comprehensive analysis of Vasquez’s claims under Strickland. This case is readily distinguishable from Benge and Danner because those decisions completely overlooked Strickland. The concerns addressed in the Benge and Danner decisions are simply not implicated here.
Thus, I respectfully disagree with the majority’s approach and its characterization of Benge and Danner. I would adhere to the rule established by Payne and Bell and review the Ohio Court of Appeals decision because it was “the last state court to [render] a reasoned opinion on the issue.” Payne, 418 F.3d at 660 (6th Cir.2005).
II.
Next, the majority concludes that the Ohio courts applied law contrary to Strickland v. Washington by allegedly elevating the burden of persuasion necessary to sustain Vasquez’s ineffective assistance of counsel claim. The majority thereafter launches into de novo review, reconstructing the state court trial record, analyzing “cold” post-conviction hearing testimony, and eventually holding that Vasquez is entitled to habeas relief.
In Williams v. Taylor, 529 U.S. 362, 386, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000), the Supreme Court emphasized our limited role under AEDPA in adjudicating claims of state prisoners:
*124[I]t seems clear that Congress intended federal judges to attend with the utmost care to state-court decisions, including all of the reasons supporting their decisions, before concluding that those proceedings were infected by constitutional error sufficiently serious to warrant the issuance of the writ.... [Fjederal habe-as courts must make as the starting point of their analysis the state courts’ determinations of fact, including that aspect of a “mixed question” that rests on a finding of fact. AEDPA plainly sought to ensure a level of “deference to the determinations of state courts,” provided those determinations did not conflict with federal law or apply federal law in an unreasonable way. H.R. Conf. Rep. No. 104-518, p. Ill (1996) [U.S.Code Cong. & Admin.News 1996 p. 944]. Congress wished to curb delays, to prevent “retrials” on federal habeas, and to give effect to state convictions to the extent possible under law. When federal courts are able to fulfill these goals within the bounds of the law, AEDPA instructs them to do so.
The Strickland, prejudice standard requires the petitioner to “show that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Strickland, 466 U.S. at 694, 104 S.Ct. 2052. Here, the majority holds that the Ohio trial court’s decision was contrary to Strickland because its opinion states that prejudice occurs when “the result of petitioner’s trial or legal proceeding would have been different had defense counsel provided proper representation.” (Emphasis added.)1 According to the majority, the trial court’s omission of the words “reasonable probability” in its recitation of the prejudice standard resulted in its application of a different, more demanding measure of proof. Also, the majority notes that the Ohio Court of Appeals correctly articulated the Strickland prejudice standard, but later omitted the words “reasonable probability” in its discussion. My colleagues claim that the appellate court’s accurate recitation of the law neither cured the trial court’s alleged error nor evidenced the appellate court’s application of the proper burden of proof. I disagree.
In Williams, the Supreme Court defined AEDPA’s “contrary to” clause: “The word ‘contrary’ is commonly understood to mean ‘diametrically different,’ ‘opposite in character or nature,’ or ‘mutually opposed.’ ” 529 U.S. at 405, 120 S.Ct. 1495 (quoting Webster’s Third New International Dictionary 495 (1976)). “The text of § 2254(d)(1) therefore suggests that the state court’s decision must be substantially different from the relevant precedent of this Court.” Id.; Ivory v. Jackson, 509 F.3d 284, 291 (6th Cir.2007) (“state-court decision is considered ‘contrary to ... clearly established federal law' if it is ‘diametrically different, opposite in character or nature, or mutually opposed.’ ”) (citation omitted).
In interpreting § 2254(d), the Court employed the following example of an application of law that is “contrary to” clearly established federal law:
If a state court were to reject a prisoner’s claim of ineffective assistance of counsel on the grounds that the prisoner had not established by a preponderance of the evidence that the result of his criminal proceeding would have been different, that decision would be *125“diametrically different,” “opposite in character or nature,” and “mutually opposed” to our clearly established precedent because we held in Strickland that the prisoner need only demonstrate a reasonable probability that ... the result of the proceeding would have been different.”
Williams, 529 U.S. at 405-06, 120 S.Ct. 1495 (citing Strickland, 466 U.S. at 694, 104 S.Ct. 2052) (emphasis added).
In the present case, to accept the majority’s position, a federal judge must conclude that the omission of the words “reasonable probability”2 from the state courts’ description of Strickland’s prejudice standard resulted in an application of a “diametrically different” burden of persuasion. Such a conclusion is not supported by the facts and the law.
In Woodford v. Visciotti, 537 U.S. 19, 23-24, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002) (per curiam), the Supreme Court held that a state court’s “occasional shorthand reference” to the Strickland prejudice standard “may perhaps be imprecise,” but it is not a repudiation of the standard that results in a decision that is contrary to clearly established federal law. In Vis-ciotti, the Court reversed the Ninth Circuit’s decision that had held that a state court’s use of the word “probable” was a standard of prejudice contrary to Strickland:
The Court of Appeals made no effort to reconcile the state court’s use of the term “probable” with its use, elsewhere, of Stricklands’ term “reasonably probable,” nor did it even acknowledge, much less discuss, the California Supreme Court’s proper framing of the question as whether the evidence “undermines confidence” in the outcome of the sen-fencing proceeding. This readiness to attribute error is inconsistent with the presumption that state courts know and follow the law. See, e.g., Parker v. Dugger, 498 U.S. 308, 314-316, 111 S.Ct. 731, 112 L.Ed.2d 812 (1991); Walton v. Arizona, 497 U.S. 639, 653, 110 S.Ct. 3047, 111 L.Ed.2d 511 (1990), overruled on other grounds, Ring v. Arizona, 536 U.S. 584, 122 S.Ct. 2428, 153 L.Ed.2d 556 (2002); LaVallee v. Delle Rose, 410 U.S. 690, 694-695, 93 S.Ct. 1203, 35 L.Ed.2d 637 (1973) (per curiam). It is also incompatible with § 2254(d)’s “highly deferential standard for evaluating state-court rulings,” Lindh v. Murphy, 521 U.S. 320, 333, n. 7, 117 S.Ct. 2059, 138 L.Ed.2d 481 (1997), which demands that state-court decisions be given the benefit of the doubt.
Visciotti, 537 U.S. at 24, 123 S.Ct. 357 (emphasis added). The Visciotti opinion also suggests that a state court’s accurate citation of Strickland evidences its application of the proper burden of proof. Visciotti, 537 U.S. at 22, 123 S.Ct. 357. Here, both state court decisions either correctly cite to Strickland or articulate the proper burden of persuasion standard.
Moreover, the Supreme Court has reversed our circuit on grounds substantially similar to the position espoused by the majority. In Jackson v. Holland, 80 Fed.Appx. 392 (6th Cir.2003) (unpublished), rev’d 542 U.S. 649, 124 S.Ct. 2736, 159 L.Ed.2d 683 (2004) (per curiam), we had held that a state court’s use of language at odds with Strickland evidenced its application of an elevated burden of proof:
Although the [state] court initially quoted the “reasonable probability” language of Strickland, it immediately followed *126this statement with language that “[i]n a post-conviction proceeding, the defendant has the burden of proving his allegations by a preponderance of the evidence.” The [state] court later stated “it is asking too much that we draw the inference that the jury would not have believed Hughes at all had Melissa Gooch testified.” Furthermore, the state court concluded that Jackson “failed to carry his burden of proving that the outcome of the trial would probably have been different but for those errors.6
* * *
Even keeping mindful of the deferential gaze that we must cast upon the Tennessee Court of Criminal Appeals’s decision, we must conclude that the state court merely paid lip service to Strickland’s reasonable probability standard before imposing an inappropriately stringent burden upon [petitioner]. Other than the initial articulation of the Strickland standard, there is absolutely no evidence that the state court analyzed Petitioner’s claim under the proper “reasonable probability” standard. Indeed, all evidence within the opinion is to the contrary.
Jackson, 80 Fed.Appx. at 407, n. 6, 409 (emphasis in original) (internal quotation and citation omitted).
The Supreme Court reversed, Holland v. Jackson, 542 U.S. 649, 654-55, 124 S.Ct. 2736, 159 L.Ed.2d 683 (2004) (per curiam), holding that the state court’s initial recitation of the correct Strickland standard sufficiently established that it understood and applied the proper burden of proof, notwithstanding its “shorthand” references to it throughout the remainder of its opinion:
The Sixth Circuit also erred in holding that the state court acted contrary to federal law by requiring proof of prejudice by a preponderance of the evidence rather than by a reasonable probability. The state court began by reciting the correct Strickland standard:
“ ‘[T]he defendant must show that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.’ ” App. to Pet. for Cert. 95 (quoting Strickland, 466 U.S. at 694, 104 S.Ct. 2052).
The Sixth Circuit nevertheless concluded that the state court had actually applied a preponderance standard, based on three subsequent passages from its opinion.
First was the statement that “[i]n a post-conviction proceeding, the defendant has the burden of proving his allegations by a preponderance of the evidence.” App. to Pet. for Cert. 95. In context, however, this statement is reasonably read as addressing the general burden of proof in postconviction proceedings with regard to factual contentions — for example, those relating to whether defense counsel’s performance was deficient. Although it is possible to read it as referring also to the question whether the deficiency was prejudicial, thereby supplanting Strickland, such a reading would needlessly create internal inconsistency in the opinion.
*127Second was the statement that “it is asking too much that we draw the inference that the jury would not have believed Hughes at all had Melissa Gooch testified.” App. to Pet. for Cert. 96. Although the Court of Appeals evidently thought that this passage intimated a preponderance standard, it is difficult to see why. The quoted language does not imply any particular standard of probability.
Last was the statement that respondent had “failed to carry his burden of proving that the outcome of the trial would probably have been different but for those errors.” Id., at 98. We have held that such use of the unadorned word “probably” is permissible shorthand when the complete Strickland standard is elsewhere recited. See Woodford v. Visciotti, 537 U.S. 19, 23-24, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002) (per curiam).
As we explained in Visciotti, § 2254(d) requires that “state-court decisions be given the benefit of the doubt.” Id. at 24, 123 S.Ct. 357. “[Rjeadiness to attribute error is inconsistent with the presumption that state courts know and follow the law.” Ibid. The Sixth Circuit ignored those prescriptions.
The judgment of the Sixth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion.
Id.; see also Urban v. Ohio Adult Parole Auth., 116 Fed.Appx. 617, 627 (6th Cir.2004) (unpublished) (discussing Visciotti and Jackson and rejecting petitioner’s argument that a state court’s shorthand description of Strickland, specifically, its statement that the “outcome probably would have been different,” resulted in its application of an elevated burden of proof under Strickland).
In the present case, it is undisputed that the Ohio Court of Appeals began its opinion by articulating the correct Strickland burden of persuasion standard. In the words of Judge Karpinski, writing for the Court of Appeals of Ohio, Eighth District:
A defendant must demonstrate that trial counsel’s performance fell below the objective standard of reasonable competence under the circumstances and that there exists a reasonable probability that, but for such deficiency, the outcome of the trial would have been different. State v. Bradley (1989), 42 Ohio St.3d 136, 538 N.E.2d 373.
State v. Vasquez, No. 82156, 2004 WL 35766, at *3 (Ohio Ct.App. Jan. 8, 2004). The court of appeal’s citation to State v. Bradley is significant because in it, the Ohio Supreme Court accurately followed and applied the “reasonable probability” prejudice standard of Strickland. Bradley, 538 N.E.2d at 380.
The concluding paragraphs of Judge Karpinski’s reasoned opinion on this ineffective assistance of counsel claim are consistent with the law of Strickland, although the words “reasonable probability” are not reiterated:
During his post-conviction hearing, defendant presented numerous witnesses who did not testify at his trial. Their absence, however, does not demonstrate his counsel was ineffective and that the outcome of his trial would have been different had they been witnesses during his trial.
On the contrary, the record shows that none of defendant’s witnesses offered any testimony to support defendant’s theory of innocence. None of their testimony challenges the victim’s account of the events leading to defendant’s convictions. We agree with the trial court’s conclusion that “not one of the witnesses *128who testified — neither the defendant, his family or others — offered any testimony that could have changed the outcome of defendant’s trial.” Trial Court Findings of Fact and Conclusions of Law, p. 25.
Vasquez, 2004 WL 35766, at *6.
Nonetheless, my colleagues conclude that, based on shorthand language used in the opinion, the Ohio Court of Appeals applied law that was contrary to Strickland v. Washington. I disagree.
In reaching this conclusion, the majority makes no concerted effort to distinguish the present case from Visciotti or Jackson. This is untenable given the substantially similar imprecise or shorthand language employed by the state court in Jackson, which arguably misdescribed Strickland’s burden of proof. Jackson, 80 Fed.Appx. at 407.
Also, the majority reasons that because the state trial court opinion consistently omits the words “reasonable probability,” such an omission creates a stronger inference of a heightened standard of proof than an opinion that correctly recites the standard once and thereafter misdescribes it.3 However, the state trial court opinion accurately cites both Strickland and State v. Bradley; see Visciotti, 537 U.S. at 22, 123 S.Ct. 357 (suggesting that a state court decision’s accurate citation to Strickland indicates that it knows the law and the applicable burden of proof).
Under Visciotti and Jackson, the state court opinions employed permissible shorthand language describing Strickland’s prejudice standard, which did not result in their application of a “diametrically different” burden of proof under Williams. See Visciotti, 537 U.S. at 24, 123 S.Ct. 357; Jackson, 542 U.S. at 654, 124 S.Ct. 2736; Williams, 529 U.S. at 405-06, 120 S.Ct. 1495; Urban, 116 Fed.Appx. at 627. The majority’s “readiness to attribute error is inconsistent with the presumption that state courts know and follow the law. It is also incompatible with § 2254(d)’s ‘highly deferential standard for evaluating statecourt rulings’ ... which demands that state-court decisions be given the benefit of the doubt.” Visciotti, 537 U.S. at 24, 123 S.Ct. 357 (emphasis added) (citation omitted).
For these reasons, I respectfully disagree with the majority’s conclusion that the Ohio courts applied law contrary to Strickland v. Washington. The majority has made a fundamental error. Our review of Vasquez’s petition for a writ of habeas corpus should be constrained by AEDPA deference, not de novo.
III.
The final issue is whether the Ohio Court of Appeals unreasonably applied Strickland to the facts. 28 U.S.C. § 2254(d)(1). I concur with the majority opinion insofar as it holds that Vasquez cannot base his ineffective assistance of counsel argument on his failure to investigate Karra Vasquez and JoAnn Kitchen. Again, pursuant to Payne and Davie, my analysis focuses on the state appellate court opinion because it is the last state court to render a reasoned decision on this claim.
A state court decision unreasonably applies federal law “if the state court identifies the correct governing legal principle from the Supreme Court’s decisions but unreasonably applies that principle to the facts.” Irick v. Bell, 565 F.3d 315, 320 (6th Cir.2009) (internal quotation omitted). *129“[A] federal habeas court may not issue a writ under the unreasonable application clause ‘simply because that court concludes in its independent judgment that the relevant state-court decision applied clearly established federal law erroneously or incorrectly.’ ” Id. (quoting Bell v. Cone, 535 U.S. 685, 694, 122 S.Ct. 1843, 152 L.Ed.2d 914 (2002) (internal quotation and citation omitted)). “[T]he question under AEDPA is not whether a federal court believes the state court’s determination was incorrect[,] but whether that determination was unreasonable — a substantially higher threshold.” hick, 565 F.3d at 320 (quoting Owens v. Guida, 549 F.3d 399, 404 (6th Cir.2008) (internal quotation and citation omitted)). After assuming my proper role, I would hold that the Ohio courts did not unreasonably apply Strickland.
For Vasquez to succeed on a claim for ineffective assistance of counsel, he must satisfy the well-rehearsed two-prong Stñckland test:
First, the defendant must show that counsel’s performance was deficient. This requires showing that counsel made errors so serious that counsel was not functioning as the “counsel” guaranteed the defendant by the Sixth Amendment. Second, the defendant must show that the deficient performance prejudiced the defense. This requires showing that counsel’s errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable.
466 U.S. at 687, 104 S.Ct. 2052.
Moreover:
[j]udicial scrutiny of counsel’s performance must be highly deferential. It is all too tempting for a defendant to second-guess counsel’s assistance after conviction or adverse sentence, and it is all too easy for a court, examining counsel’s defense after it has proved unsuccessful, to conclude that a particular act or omission of counsel was unreasonable.... Because of the difficulties inherent in making the evaluation, a court must indulge a strong presumption that counsel’s conduct falls within the wide range of reasonable professional assistance; that is, the defendant must overcome the presumption that, under the circumstances, the challenged action might be considered sound trial strategy. There are countless ways to provide effective assistance in any given case. Even the best criminal defense attorneys would not defend a particular client in the same way.
Id. at 689, 104 S.Ct. 2052 (internal quotation and citation omitted; emphasis added).
This high level of deference means that “[w]e address not what is prudent or appropriate, but only what is constitutionally compelled.” United States v. Cronic, 466 U.S. 648, 665, n. 38, 104 S.Ct. 2039, 80 L.Ed.2d 657 (1984) (emphasis added). In addition, “ ‘[strategic choices made after thorough investigation of law and facts relevant to plausible options are virtually unchallengeable; and strategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional judgments support the limitations on investigation.’” Barnes v. Elo, 231 F.3d 1025, 1029 (6th Cir.2000) (quoting Strickland, 466 U.S. at 691, 104 S.Ct. 2052).
At the state court level, Vasquez’s ineffective assistance of counsel claim challenged Butler’s exclusion of certain family members as potential defense witnesses. Because I agree with the majority that Vasquez cannot base his ineffective assistance of counsel claim on the exclusion of Karra Vasquez and JoAnn Kitchen, my analysis focuses solely on Butler’s performance regarding Becky Shaffer. Thus, if Butler’s decision to exclude Becky Shaf*130fer from his potential defense witnesses was, as the Ohio courts concluded, based on reasoned trial strategy and judgment, the issue is resolved.
The Ohio Court of Appeals concluded that “the record shows that none of defendant’s witnesses offered any testimony to support defendant’s theory of innocence. None of their testimony challenges the victim’s account of the events leading to defendant’s convictions. We agree with the trial court....” In reaching its conclusion, the appellate court relied on the following findings of fact set forth in the state trial court decision:
Butler testified that after he was assigned to the case about six weeks after arraignment, he reviewed the Court’s file, the clerk’s office docket, met with his client, ascertained that his client would not sign a speedy trial waiver pursuant to the Court’s concern about time, and proceeded immediately to conduct research and prepare for trial. Butler filed the appropriate discovery motions [and] met with his client at subsequent scheduled pre-trials.
* * *
Butler explained that he didn’t subpoena the family members because they were uncooperative and he had to keep in mind the other allegations of sexual molestation, specifically from his sister’s daughter. He [] had to be concerned that family members would be aware of the rape allegation by [Vasquez’s] four-year-old niece in California.
The defendant testified that while in County Jail, even he could not reach his wife, his sister-in-law Becky [Shaffer] or his mother-in-law, JoAnn Kitchen: none of them responded [and][h]is wife stopped visiting him.
* * *
[] “Becky” Shaffer testified [that] she and her husband had gone to bed and did not see the [victim or her mother and father] the night [the victim] stated that Robert Vasquez raped her.
[Becky Shaffer] also suspected [Vasquez] of sexually abusing her son.
Pursuant to 28 U.S.C. § 2254(e)(1), the district court is required to apply a presumption of correctness to state court findings of fact unless the petitioner offers clear and convincing evidence rebutting this presumption. See Warren v. Smith, 161 F.3d 358, 360-61 (6th Cir.1998). In its review of Vasquez’s habeas petition, the district court abandoned this presumption and did not circumscribe its analysis to comport with AEDPA deference.
In fact, the district court’s opinion challenged only one finding of fact upon which the state appellate court relied, specifically, the fact that “[Becky Shaffer] suspected or was concerned that Vasquez had sexually abused [her son.]” The district court found that “Becky Shaffer [actually] testified that she took her son Aaron to the hospital on the night of August 4, 2000, because Steve Loomis and ‘the other police officer’ [Zbikowski] told her to do so, [] not because she suspected or was concerned that Vasquez had sexually abused [her son.]” Id. The district court further noted that:
[a] closer analysis of Becky’s testimony [ ] makes this finding a significant stretch, if not an outright mischaracteri-zation. This is especially so when considering the convoluted cross-examination questioning involving multi-part questions that elicited a “yes” and Becky’s repeated testimony that she only took her son in the ambulance because Steve Loomis and another police officer directed her to do so.
*131A review of the post-conviction hearing transcript, however, does not rebut the state appellate court’s reliance upon this factual finding by clear and convincing evidence. The post-conviction transcript contains the following colloquy between the trial court, Vasquez’s counsel, and Becky Shaffer (during Shaffer’s cross-examination):
The Court: Ma’am, did you tell the hospital personnel the reason you were there [at the hospital] was because [A.L.] had made the allegation—
Becky Shaffer: Yes.
The Court: — of rape and you wanted to see whether or not your son had been abused?
Becky Shaffer: Yes.
Vasquez’s Counsel: I’m sorry, What was your answer?
The Court: She said yes.
Vasquez’s Counsel: Thank you very much.
Vasquez’s Counsel: You said this is an emotional situation, correct?
Becky Shaffer: Yes.
Vasquez’s Counsel: So obviously this is a very serious situation?
Becky Shaffer: Very.
Vasquez’s Counsel: And you suspected that the same thing might have happened to your son, that’s how serious it was.
Becky Shaffer: I didn’t know. Yes.
Moreover, the remaining factual findings relied upon by the state appellate court remained uncontroverted by the district court.4 Specifically, there is ample evidence supporting the appellate court’s decision to affirm the state trial court’s ruling that Vasquez’s family members, including Becky Shaffer, were uncooperative and unsympathetic to his defense. Butler’s decision to forgo contacting Becky Shaffer is not so “incredible” amid swirling family-based allegations of spousal abuse and separate family-based allegations of sexual molestation. In addition, Becky Shaffer was slated to testify for the prosecution. Moreover, during the post-conviction hearing, Becky Shaffer admitted that she suspected that Vasquez had sexually abused her son, a damning accusation if revealed during his rape trial, and arguably evidencing Butler’s reasonable fears about Vasquez’s family’s proclivity for damaging and harmful testimony.
The majority claims that Butler knew that “Becky Shaffer was ideally situated to provide information on the only avenue of defense available to Vasquez” and that she *132was likely to have “relevant information.” The majority also asserts that Butler could have contacted Becky Shaffer independently. These statements are speculative. Furthermore, according to Vasquez’s own testimony, credited by the state trial court, he had been “abandoned by his family” during the six weeks before his trial and had made several attempts to contact Becky Shaffer, but to no avail. Vasquez’s concern regarding his family’s abandonment appropriately influenced Butler’s reasoned and strategic decision not to pursue certain defense witnesses, including Becky Shaffer.
As the Supreme Court has stated:
[t]he reasonableness of counsel’s actions may be determined or substantially influenced by the defendant’s own statements or actions. Counsel’s actions are usually based, quite properly, on informed strategic choices made by the defendant and on information supplied by the defendant. In particular, what investigation decisions are reasonable depends critically on such information. For example, when the facts that support a certain potential line of defense are generally known to counsel because of what the defendant has said, the need for further investigation may be considerably diminished or eliminated altogether. And when a defendant has given counsel reason to believe that pursuing certain investigations would be fi-witless or even harmful, counsel’s failure to pursue those investigations may not later be challenged as unreasonable.
Strickland, 466 U.S. at 691, 104 S.Ct. 2052 (emphasis added).
The district court and majority rely heavily on Ramonez v. Berghuis, 490 F.3d 482, 484 (6th Cir.2007) in support of their conclusion that in making his decision regarding Becky Shaffer, Butler was not functioning as the counsel guaranteed by the Sixth Amendment. In Ramonez, however, the excluded defense witnesses actually witnessed the altercation that gave rise to the defendant’s conviction. Ramonez, 490 F.3d at 488 (concluding that attorney’s decision to exclude defense witnesses was constitutionally deficient because defendant insisted “months before trial” that the three witnesses at issue could testify to what they saw when they witnessed the altercation, and that each witness could testify that defendant did not force his way into the victim’s home). Here, however, Becky Shaffer did not witness the crime. In fact, there were no witnesses to the alleged crime other than A.L. and Vasquez. Becky Shaffer’s favorable testimony, if any, would be limited to impeachment, only. Thus, Ramonez is easily distinguishable from the present case.
The majority also holds that Butler was ineffective in failing to pursue potential impeachment evidence contained in an Emergency Medical Services report form (EMS run sheet). The hearsay evidence contained in the EMS report can be summarized as follows: When A.L. described the attack to the paramedics, she allegedly stated that “he put it in.” The EMS run sheet also reports that A.L. “took a shower later that day.”
The state appellate court did not rule on this issue, and the record does not reveal whether this claim was abandoned on appeal. Nonetheless, the state trial court decision (the last state court to render a reasoned decision on this matter) ruled that Butler’s failure to impeach A.L. with the EMS run sheet did not prejudice Vasquez’s defense because “it was consistent with her trial testimony.” In its finding of facts, the state trial court found as follows:
*133(B) Claim that Counsel Did Not Cross-examine Victim Regarding EMS “Run Sheet”:
At the hearing, Butler was presented with an unauthenticated copy of an Emergency Medical Services “run report” of the evening that [A.L.] was transported to the hospital. The defendant asked questions of Butler concerning the fact that an unidentified person wrote thereon: “pt. st. he put it in her.” Apparently defendant was attempting to show, as he claimed via his Investigator’s affidavit in his Supplement, that this notation “is in complete contradiction to [A.L.’s] trial testimony”. Tom Pavlish’s Affidavit at para. 4. Butler testified that his understanding of this case was that it was oral sex, more specifically, that [A.L.] accused the defendant of putting his tongue in her- — he had no information that it was anything else.
The majority contends, however, that this conclusion is not a factual finding entitled to deference under § 2254(e)(1) because “this [ ] reasoning is actually a holding that Vasquez has not carried his burden to show prejudice resulting from his attorney’s failure to identify the run sheet as useful impeachment evidence.” I disagree. The trial court’s finding was based in part on its determination that A.L.’s trial testimony was consistent with her EMS run sheet statements. At the very least, it is a mixed question that rests on a finding of fact, which is entitled to AEDPA deference. See Williams, 529 U.S. at 389, 120 S.Ct. 1495.
To establish prejudice, Vasquez “must show that there is a reasonable probability that, but for [Butler’s failure to discover this impeachment evidence], the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.” Strickland, 466 U.S. at 694, 104 S.Ct. 2052. In addition, AEDPA deference requires that we review the state trial court’s decision for “unreasonableness] — a substantially higher threshold.” Irick, 565 F.3d at 320 (quoting Owens, 549 F.3d at 404) (internal quotation and citation omitted).
I disagree with the majority’s holding that Butler’s failure to impeach A.L. with the information contained in the EMS run sheet creates a reasonable probability that but for his failure to do so, the result of his trial would have been different. See Strickland, 466 U.S. at 694, 104 S.Ct. 2052. A.L.’s statements, as reflected in the EMS run sheet, do not contradict her trial testimony. A.L. testified at trial that Vasquez “lick[ed][her] private spot ... [with] his tongue,” and answered “both” when asked on direct examination whether he had touched her on “the inside or outside or both.” Thus, her statement that “he put it in” is consistent with her trial testimony.5
*134As we have explained, “in order to establish prejudice, the [ ] evidence ... must differ in a substantial way — in strength and subject matter — from the evidence actually presented....” Hill v. Mitchell, 400 F.3d 308, 319 (6th Cir.2005). As the majority concedes, Butler repeatedly impeached A.L. and other state witnesses concerning A.L.’s trustworthiness and ability to tell the truth.
Most importantly, after according proper AEDPA deference, I do not agree that the Ohio courts unreasonably applied Strickland to the facts. Butler’s conduct regarding Becky Shaffer was “reasonable precisely to the extent that [his] reasonable professional judgment ] supported] th[at] limitation[ ] on [his] investigation^ Strickland, 466 U.S. at 691, 104 S.Ct. 2052, and the EMS run sheet does not contain impeachment evidence that is substantially different in strength and subject matter from the impeachment evidence presented at trial. Hill, 400 F.3d at 318-19.
“The question under AEDPA is not whether a federal court believes the state court’s determination was incorrect but whether that determination was unreasonable—a substantially higher threshold.” Schriro v. Landrigan, 550 U.S. 465, 473, 127 S.Ct. 1933, 167 L.Ed.2d 836 (2007). Here, the Ohio state court decisions were not unreasonable.
In conclusion, the following comments by the Ohio trial court judge to Vasquez regarding the performance of his attorney are illuminating:
Based upon the fact that I had to sit through this trial just like everybody else did, it’s probably one of the most depressing things I had to do in the last eight years, quite frankly. I’m not looking forward to saying that my last trial was probably my worst, but I think you[r] attorney did everything he could to try to convince the jury that you weren’t guilty as charged. I think Mr. Butler didn’t have a lot to work with.
IV.
For these reasons, I respectfully dissent. I would reverse the judgment of the district court and deny the petition for a writ of habeas corpus.
. Most of the district court's opinion focuses on its conclusion that the Ohio courts unreasonably applied Strickland. In contrast, the majority's opinion concludes that the Ohio courts violated the "contrary to” clause, and thereafter engages in de novo review of petitioner's ineffective assistance of counsel claim.
. As previously noted, the Ohio Court of Appeals recited the correct Strickland, standard, verbatim, at the beginning of its opinion.
In evaluating [petitioner's] contention that the lower court should have reopened the proof in the collateral proceeding to allow Melissa Gooch to testify, the Tennessee Court of Appeals also stated that Gooch's statement “does not lead to the conclusion that the jury would have decided the case differently had it heard Gooch testify.”
. Because the majority focuses its analysis on the state trial court opinion, it is necessary for me to address the trial court's omission of the words "reasonable probability” instead of reviewing only the appellate court’s decision.
. Attorney Butler was appointed to represent Vasquez approximately six weeks before his trial. On the same day that he was appointed, Butler met with Vasquez and asked him to execute a speedy trial waiver. Vasquez refused. Butler filed appropriate discovery requests, researched and prepared for trial, and met with Vasquez on several occasions. Butler decided to exclude Becky Shaffer because she was not present at the time the alleged rape occurred (she was asleep in a separate apartment, two floors above the basement level, which was the alleged crime scene), she was a close family member (Vasquez was married to her sister), and that Butler was concerned that Vasquez’s family members would testify or reveal details about the separate molestation allegations levied by Vasquez’s niece in California. When Butler eventually reached JoAnn Kitchen, she informed him that the family ”w[as] not going to cooperate [][] in any way” with his defense, and that she had recently learned that Vasquez had been abusing her daughter (Vasquez’s wife, Karra) for "an extended period of time.” During this same six-week period, Vasquez testified that he had been "abandoned by his family,” including the same family members the majority now contends could have aided in his defense. Vasquez made several attempts to contact Becky Shaffer, but she never responded.
. Vasquez also argues that Attorney Butler was constitutionally ineffective for failing to cross-examine A.L. regarding an ambiguous notation in the EMS run sheet, that she "took a shower later that day.” In its finding of facts, the state trial court found as follows:
Mr. Pavlish states in his affidavit that information in the EMS report also states "pt. st. she took a shower later that day” is inconsistent with Stephen Loomis's testimony that "immediately after the incident [A.L.] fell asleep in his truck during the drive home and that he carried her into his home as she slept.” Tom Pavlish Affidavit at para. 6. At the hearing, counsel asked Butler questions concerning this notation, to which Butler replied that the night [A.L.] was transported, she had reported an incident that took place two weeks prior.
The potential impeachment of A.L. regarding this collateral matter is only marginally relevant to the issue of whether the sexual molestation occurred. For this reason, it is not "reasonably probable” that the outcome of *134the trial would have been different had this collateral impeachment occurred. Most importantly, the more demanding standard of an unreasonable application of Strickland is clearly not met