TYMKOVICH, Circuit Judge.
Mark T. Boyle is a state prisoner serving a sentence of 424 months after being convicted of multiple counts of aggravated criminal sodomy and sexual battery. Boyle petitioned for habeas corpus relief pursuant to 28 U.S.C. § 2254, and the district court denied the petition in all respects. We granted a certificate of ap-pealability (COA) on Boyle’s claims alleging ineffective assistance of trial and appellate counsel.
After carefully examining the record before us, we conclude Boyle is not entitled to relief on either claim. Specifically, we hold (1) Boyle is not entitled to an eviden-tiary hearing on his claim of ineffective assistance of trial counsel; (2) Boyle’s trial counsel was not constitutionally ineffective for failing to interview or call to the stand
certain witnesses; and (3) Boyle’s appellate counsel was not ineffective for failing to challenge the sufficiency of the evidence on direct appeal.
Exercising jurisdiction pursuant to 28 U.S.C. § 1291, we therefore AFFIRM.
I. Background
A. Factual Background
Boyle’s state court convictions stemmed from non-consensual sexual acts with two female victims: L.B. and C.G. On July 15, 1997, Boyle and L.B. met for the first time at a restaurant in Wichita, Kansas. They each consumed several drinks at the bar of the restaurant and began a consensual sexual relationship later that night. The relationship continued for several weeks.
During this time Boyle met C.G. On July 26, Boyle and C.G. had drinks at a restaurant and then a topless bar. After-wards, C.G. and Boyle drove to C.G.’s home. The two went up to C.G.’s bedroom, where Boyle began having anal sex with her until she told him to stop and he did. The next morning, C.G. and Boyle engaged in additional sexual acts. Although C.G. did not want the sexual contact, she felt she could not refuse. A few days later, C.G. told her doctor, and then the police, that she had been assaulted.
Within a week of his encounter with C.G., Boyle took L.B. to a Wichita restaurant called Café Chicago. L.B.’s memory about the night’s events was hazy. Although she recalled being with Boyle in a hotel room that night, she could not remember whether they had sex. She did recall engaging in consensual sex the following morning. Later that day, when L.B. began feeling pain and swelling in her rectal area, she became convinced Boyle had engaged in anal sex with her without her consent the previous night.
Based on these encounters, the district attorney charged Boyle with multiple counts of aggravated criminal sodomy, rape, and sexual assault. Key evidence in both cases was the testimony of three expert nurse witnesses. The nurses opined that non-consensual sexual activity likely occurred based on the following evidence: (1) certain bruises on both women, (2) “mounting injuries” on both women, (3) healing lacerations on C.G., and (4) a tear below the anal opening on L.B. Boyle was convicted of three counts of aggravated criminal sodomy and one count of sexual battery with regard to C.G. and L.B.
He was sentenced to 424 months in prison.
B. Procedural History
Boyle appealed his sentence to the Kansas Court of Appeals. He focused his appeal on (1) the nurses’ qualifications to serve as expert witnesses, and (2) the jury instructions on two of the counts. The court affirmed Boyle’s sentence and the Kansas Supreme Court denied review.
Proceeding pro se, Boyle then sought collateral review of his sentence under Kansas Statutes Annotated § 60-1507. Boyle raised multiple issues. Relevant to this appeal, he presented, for the first time, claims of ineffective assistance of trial and appellate counsel. The state trial court appointed counsel to represent Boyle in the collateral proceedings and held a hearing on the claims. The trial court determined Boyle was not entitled to an evidentiary hearing on any issue because he failed to supply sufficient evidence to show his entitlement to relief. The Kansas Court of Appeals affirmed and prevented Boyle from raising some of the
issues presented in the motion, although it did not explain any specific deficiency with respect to the ineffective assistance of counsel claim. The Kansas Supreme Court denied review.
Boyle next filed a federal habeas corpus petition pursuant to 28 U.S.C. § 2254. Among other claims, Boyle again argued he had received ineffective assistance of trial and appellate counsel. Boyle sought an evidentiary hearing to develop the record on the issue of whether his trial counsel was ineffective for failing to interview expert medical witnesses and certain other witnesses Boyle thought helpful to his case.
The district court denied the petition, but this court granted a COA on two issues: (1) whether Boyle’s trial counsel was ineffective for failing to interview or investigate witnesses for the defense, and (2) whether Boyle’s appellate counsel provided ineffective assistance.
II. Ineffective Assistance of Trial Counsel
Before arguing the merits of his case, Boyle asks this court to grant him an evidentiary hearing on his claim of ineffective assistance of trial counsel. We address that issue first.
A. Evidentiary Hearing
Because Boyle’s habeas petition is governed by the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), he can obtain an evidentiary hearing in federal court by (1) showing he was diligent in developing the factual basis for his claim in state court, 28 U.S.C. § 2254(e)(2) (2000);
Williams v. Taylor,
529 U.S. 420, 429-31, 120 S.Ct. 1479, 146 L.Ed.2d 435 (2000), and (2) asserting a factual basis that, if true, would entitle him to habeas relief.
Schriro v. Landrigan,
550 U.S. 465, 127 S.Ct. 1933, 1940, 167 L.Ed.2d 836 (2007);
Mayes v. Gibson,
210 F.3d 1284, 1287 (10th Cir.2000). This means he must show that, taking his allegations as true, he would prevail on his claim.
We address each of these requirements in turn.
1. Diligence
Section 2254(e)(2) prohibits a federal court from conducting an evidentiary hearing on a claim if the petitioner failed to develop the factual basis for it in state court. The Supreme Court has held “a failure to develop the factual basis of a claim is not established unless there is lack of diligence, or some greater fault, attributable to the prisoner or the prisoner’s counsel.”
Williams,
529 U.S. at 432, 120 S.Ct. 1479;
see also id.
at 431, 120 S.Ct. 1479 (noting that the word “ ‘fail’ connotes some omission, fault, or negligence on the part of the person who has failed to do something”). “Diligence will require in the usual case that the prisoner, at a minimum, seek an evidentiary hearing in state court in the manner prescribed by state law.”
Id.
at 437, 120 S.Ct. 1479.
Seeking an evidentiary hearing “in the manner prescribed by state law” does not mean the prisoner must actually obtain a hearing.
See, e.g., Barkell v. Crouse,
468 F.3d 684, 695-96 (10th Cir.2006) (holding that defendant’s request for an evidentiary hearing in state court — although denied— was in a manner prescribed by state law and therefore satisfied the diligence requirement);
Cannon v. Mullin,
383 F.3d 1152, 1177 (10th Cir.2004) (same). Petitioner need only show he “complied with what reasonably appeared to be the established state-law requirements ... even if his reasonable interpretation of state law turned out to be wrong.”
Barkell,
468 F.3d at 694.
The state courts are, of course, the final arbiters of when and how a state prisoner can obtain an evidentiary hearing
in their courts. But whether a habeas petitioner has shown “a lack of diligence” in failing to obtain an evidentiary hearing is a question of federal law decided by the federal habeas courts.
Id.
at 693-95;
see also Cannon,
383 F.3d at 1177 (noting the question of diligence is a question to be determined by the federal district court). To determine whether a petitioner has shown the requisite diligence under federal law, we look to the state law controlling at the time petitioner sought an evidentiary hearing. The petitioner can prove his diligence by showing he “could reasonably have believed” his request for an evidentiary hearing in state court met the requirements for such a hearing under then-existing state law.
See Barkell,
468 F.3d at 695-96. This is a legal question the federal habeas court must decide based on an objective standard; petitioner’s subjective thoughts are irrelevant.
Applying this framework, Boyle may escape the bar of § 2254(e)(2) only if he developed the basis of his ineffectiveness claim in the Kansas state courts, meaning he was diligent in pursuing his claim there. On the advice of counsel, Boyle first raised his claim during collateral review under Kan. Stat. Ann. § 60-1507. The Kansas trial court denied him an evidentiary hearing, and the Kansas Court of Appeals affirmed. Our task is to determine whether Boyle “could reasonably have believed” his request for an evidentiary hearing met the then-existing Kansas law requirements.
Boyle filed his state court petition in 2000. Under the controlling Kansas precedent at that time, to obtain an evidentiary hearing a petitioner must set forth enough background facts supporting his claim to “demonstrate that petitioner is entitled to relief.”
To be entitled to an evidentiary hearing on a post-conviction motion under K.S.A. 60-1507 the movant is required to allege a factual basis in the motion to support his claim for relief.... While corroboration of petitioner’s statements or allegations is no longer required, a petition must set forth a factual background, names of witnesses or other sources of evidence to demonstrate that petitioner is entitled to relief.
Sullivan v. State,
222 Kan. 222, 564 P.2d 455, 457 (1977). (citations omitted).
The diligence question in this case is close, but for purposes of this appeal we can assume he met the Kansas standard at the time he filed collateral proceedings. Even assuming compliance on the diligence front, we still conclude he is not entitled to habeas relief.
2. Entitlement to Hearing
Having assumed Boyle was diligent in developing the factual basis for his claim of ineffective assistance in state court, we next consider whether Boyle is entitled to relief on that claim in federal court. Boyle is entitled to an evidentiary hearing so long as his factual allegations, “if true, would entitle [him] to federal habeas relief.”
Schriro,
127 S.Ct. at 1940;
Mayes,
210 F.3d at 1287. Consistent with this standard, “an evidentiary hearing is unnecessary if the claim can be resolved on the record.”
Anderson v. Att’y Gen. of Kan.,
425 F.3d 853, 859 (10th Cir.2005);
see also Schriro,
127 S.Ct. at 1940 (“[I]f the record refutes the applicant’s factual allegations or otherwise precludes habeas relief, a dis
trict court is not required to hold an evi-dentiary hearing.”).
Kansas asks us to review Boyle’s entitlement to relief under the deferential standards set forth in 28 U.S.C. § 2254(d). Although its decisional language is unclear, the Kansas Court of Appeals seems to have rejected Boyle’s ineffectiveness claim on procedural, rather than substantive, grounds. The court held, “Boyle raises several issues in his 60-1507 motion that could have been raised on direct appeal .... All of the issues he attempts to raise have been or could have been dealt with on his direct appeal.”
Boyle v. State,
No. 86,730, 2002 WL 31274355 (Kan.Ct.App. Oct.4, 2002) (per curiam) (unpublished).
Whether or not we think the court’s decision involves the full-blown invocation of a procedural bar, it does not constitute an adjudication on the merits under our case law.
See generally Harris v. Poppell,
411 F.3d 1189, 1195-96 (10th Cir.2005) (citing cases).
See also Wilson,
536 F.3d 1064, 1082 (“When a state court has not examined [petitioner’s] non-record evidence, it has reached no conclusion ‘on the merits.’”);
Valdez v. Cockrell,
274 F.3d 941, 946-47 (5th Cir.2001) (“An adjudication on the merits occurs when the state court resolves the case on substantive grounds, rather than procedural grounds.” (internal quotation marks omitted)).
Because Boyle’s claim was not adjudicated on the merits in state court, § 2254(d)’s deferential standards of review do not apply to our merits determination.
See
28 U.S.C. § 2254(d) (setting forth a deferential standard of review for “any claim that was adjudicated on the merits in State court proceedings”). Without an adjudication on the merits, there is no state court decision to which we could defer.
See Barkell,
468 F.3d at 697-98. Therefore, Boyle does not have to prove the state court’s adjudication of his claim was in some way unreasonable,
see
28 U.S.C. § 2254(d)(1) and (d)(2), but rather that he is entitled to relief under our independent review of his ineffectiveness claim,
Barkell,
468 F.3d at 698.
Under the now familiar two-part standard for judging ineffective assistance, Boyle is entitled to relief if he can show by a preponderance of the evidence the following: (1) trial counsel was deficient, i.e., “counsel’s representation fell below an objective standard of reasonableness”; and (2) the deficiency prejudiced his defense, i.e., “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.”
Strickland v. Washington,
466 U.S. 668, 688, 694, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984);
accord Sperry v. McKune,
445 F.3d 1268, 1274-75 (10th Cir.2006);
Anderson,
425 F.3d at 859.
“There is a strong presumption that counsel’s performance falls within the wide range of professional assistance^] the defendant bears the burden of proving that counsel’s representation was unreasonable under prevailing professional norms and that the challenged action was not sound strategy.”
Williamson v. Ward,
110 F.3d 1508, 1514 (10th Cir.1997) (quoting
Kim-melman v. Morrison,
477 U.S. 365, 381, 106 S.Ct. 2574, 91 L.Ed.2d 305 (1986)). “[A] particular decision not to investigate must be directly assessed for reasonableness in all the circumstances, applying a heavy measure of deference to counsel’s judgments.”
Strickland,
466 U.S. at 691, 104 S.Ct. 2052.
Of particular note here is that claims of ineffective assistance of counsel are not designed to allow defendants to relitigate trial errors that should have been raised on direct appeal. State habeas petitioners, in particular, are not allowed to bring their state trial court errors into federal court under the guise of catch-all ineffective assistance of counsel claims. We cannot allow this practice for the simple reason that our task on habeas is limited to upholding federal law.
See
28 U.S.C. § 2254(a) (“The Supreme Court, a Justice thereof, a circuit judge, or a district court shall entertain an application for a writ of habeas corpus in behalf of a person in custody pursuant to the judgment of a State court
only
on the ground that he is in custody in violation of the Constitution or laws or treaties of the United States.” (emphasis added)).
Applying this standard here, we conclude that even if the facts Boyle alleges were true, he has not met the second prong of
Strickland’s
ineffectiveness test: He has failed to show that even if counsel were deficient in the ways he alleges, such deficiency would have prejudiced his defense. There is simply no 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;
see also Boyd v. Ward,
179 F.3d 904, 915 (10th Cir.1999) (concluding, after examining the “totality of the evidence,” there was “no reasonable probability that the jury would have reached a different verdict”).
Medical Experts
Boyle first complains of trial counsel’s failure to interview and call to the stand any expert witnesses. Even assuming some deficiency on the part of counsel, we agree with the district court that Boyle has not shown the requisite prejudice. Given the weight of the evidence against him, and the highly speculative nature of the allegations made by Boyle to support ineffectiveness, there was “no reasonable probability that, had counsel not committed the errors [Boyle] now claims were committed, the outcome of the case would have been different.”
See Boyd,
179 F.3d at 915. The main issue at trial was whether C.G. and L.B. had consented to Boyle’s sexual contact. Both L.B. and C.G. testified they had not consented, and the nurses who examined the women- bolstered their testimony that the sexual contact was non-consensual. At trial, Boyle’s defense was that L.B. and C.G. had consented to the contact. In his habeas petition, he fails to show how calling expert medical witnesses would have bolstered that defense.
Boyle has failed to show, for example, that medical experts could have reached a conclusion regarding consent contrary to the conclusions reached by the nurses. He supplied no evidence or convincing argument that medical testimony could support his claim. And the speculative witness is often a two-edged sword. For as easily as one can speculate about favorable testimony, one can also speculate about unfavorable testimony.
See, e.g., United
States v. Snyder,
787 F.2d 1429, 1432 (10th Cir.1986) (rebutting defendant’s assertion additional testimony would have been helpful by concluding “it is at least as reasonable, and maybe more so, to speculate that the testimony of those witnesses would have damaged defendant’s case”).
Although the medical experts may have provided helpful testimony on direct examination, the admissions and qualifications elicited by prosecutors on cross examination may have been damaging.
See, e.g., Parker v. Scott,
394 F.3d 1302, 1322 (10th Cir.2005) (noting that a certain witness not called to the stand “may have been able to corroborate evidence detrimental to” defendant). This is why the decision of which witnesses to call is quintessentially a matter of strategy for the trial attorney.
United States v. Miller,
643 F.2d 713, 714 (10th Cir.1981) (“Whether to call a particular witness is a tactical decision and, thus, a ‘matter of discretion’ for trial counsel.”);
Boyd,
179 F.3d at 915 (describing decisions regarding impeaching witnesses and introducing evidence as matters of “trial strategy and tactics”).
Boyle has also failed to show why counsel needed to obtain any additional information from expert medical witnesses. Counsel’s cross-examination of the prosecution’s expert nurse witnesses evidenced a sufficient understanding of the nature of the evidence against Boyle and its potential weaknesses.
See, e.g.,
R., Kan. D. Ct., No. 97-1539, Vol. 4 at 139-46 (Cross-examination of Nurse Rosenberg); Vol. 5 at 119-26 (Cross-examination of Nurse Flowers);
id.
at 144-55 (Cross-examination of Nurse Schunn). Defense counsel even elicited from the government’s expert nurse witnesses that it was possible the injuries on L.B. and C.G. could have stemmed from consensual vaginal sex, lack of lubrication, and even the re-aggravation of a previous injury.
See id.,
Vol. 4 at 145-46;
id.,
Vol. 5 at 123-24, 128.
Counsel’s cross-examination of the prosecution’s other witnesses also shows the lack of prejudice stemming from counsel’s chosen strategy. At one point, for example, defense counsel engaged in a discussion with the Wichita police chemist about two specialized laboratory procedures the chemist used on certain evidence.
See id.,
Vol. 5 at 99-102. The first test was called an acid phosphatase spot test, or AP Spot Test for short. And the second test involved the use of a Lumalight to spot certain substances by making them fluoresce. Trial counsel had sufficient knowledge and familiarity with these tests to convey to the jury, in the light most favorable to Boyle’s defense, how the tests worked and what the chemist discovered.
See id.
Given the nature of the evidence arrayed against him and his chosen defense, there is no reason to think consulting medical experts or putting them on the witness stand would have changed the outcome of Boyle’s trial. Therefore, even assuming counsel was deficient in the manner alleged by Boyle, such deficiency did not prejudice his defense.
Café Chicago Bartender
Boyle has also failed to show how the testimony of the Café Chicago bartender — even if favorable to him — would have changed the outcome of his case. The unwanted sexual contact Boyle was convicted of, with regard to L.B., occurred in a hotel room in which only Boyle and L.B. were located. Wfliile the two had drinks at the Café Chicago around 9:00 p.m., nothing unusual was alleged to have happened at the bar. Only once Boyle and L.B. were alone in the hotel room, hours later, did the non-consensual sexual contact occur. The relevance of the Café Chicago bartender’s testimony is therefore highly questionable.
See
Fed R. Evid. 401;
Medina v. Barnes,
71 F.3d 363, 367
(10th Cir.1995) (rejecting petitioner’s claim of ineffective assistance based on counsel’s failure to obtain witness testimony which would have been “at most cumulative, and of limited probative value”).
Once again, Boyle sets forth speculative assertions in attempting to prove his claim; these assertions are clearly insufficient to carry his burden of proving prejudice.
See Strickland,
466 U.S. at 693, 104 S.Ct. 2052 (“Attorney errors come in an infinite variety and are as likely to be utterly harmless in a particular case as they are to be prejudicial.”);
id.
at 694, 104 S.Ct. 2052 (“The 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.”). Simply put, Boyle’s allegations regarding the testimony of the Café Chicago bartender, even if true, do not rise to a level where they undermine confidence in the outcome of his trial.
Because Boyle would not be entitled to relief on his ineffectiveness claim even if his allegations were true, he is not entitled to an evidentiary hearing on the claim.
See Schriro,
127 S.Ct. at 1940 (“[I]f the record refutes the applicant’s factual allegations or otherwise precludes habeas relief, a district court is not required to hold an evidentiary hearing.”). Consequently, the district court did not abuse its discretion in refusing to grant Boyle such a hearing.
See id.; Anderson,
425 F.3d at 858.
B. Ineffective Assistance of Trial Counsel
As we have just shown, even taking Boyle’s factual allegations as true, he cannot show he was prejudiced by any alleged deficiency in counsel’s performance. He therefore has not made out a claim of ineffective assistance of trial counsel.
See Strickland,
466 U.S. at 691, 104 S.Ct. 2052 (“An error by counsel, even if professionally unreasonable, does not warrant setting aside the judgment of a criminal proceeding if the error had no effect on the judgment.”). The district court correctly denied habeas relief to Boyle on this claim.
III. Ineffective Assistance of Appellate Counsel
Finally, Boyle argues his state appellate counsel provided ineffective assistance by failing to raise the issue of sufficiency of the evidence on direct appeal. Finding no possibly meritorious arguments supporting this position, Boyle’s federal habeas counsel submitted an
Anders-brief
on the issue.
See Anders v. California,
386 U.S. 738, 87 S.Ct. 1396, 18 L.Ed.2d 493 (1967). Although Boyle has submitted a supplemental pro se brief reiterating his position, we agree with counsel there are no legally non-frivolous arguments supporting Boyle’s claim.
Appellate counsel cannot be deficient for failing to raise a frivolous argument.
Smith v. Robbins,
528 U.S. 259, 278, 120 S.Ct. 746, 145 L.Ed.2d 756 (2000). The Sixth Amendment’s right to counsel “does not include the right to counsel for bringing a frivolous appeal.”
Smith,
528 U.S. at 278, 120 S.Ct. 746;
see. also Cargle v. Mullin,
317 F.3d 1196, 1202 (10th Cir.2003) (“[0]f course, if the issue is meritless, its omission will not constitute deficient performance.”). To prevail on a sufficiency-of-the-evidence claim, a defendant must meet the high burden of showing “upon the record evidence adduced at the trial no rational trier of fact could have found proof of guilt beyond a reasonable doubt.”
Jackson v. Virginia,
443 U.S. 307, 324, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979); see
also State v. Knetzer,
3 Kan.App.2d 673, 600 P.2d 160, 161 (1979) (“In challenging the
sufficiency of the evidence, the defendant faces a heavy burden.”).
Sufficient evidence was adduced at trial for a rational fact-finder to have found proof of Boyle’s guilt beyond a reasonable doubt. The main issue at Boyle’s trial was whether or not C.G. and L.B. had consented to sexual contact with Boyle. Both women testified they did not give consent, and expert nurse witnesses presented additional evidence to corroborate that testimony. Because the Kansas appellate courts would have viewed all the evidence in the light most favorable to the government, any sufficiency-of-the-evidence argument would have been entirely frivolous.
See State v. Parker,
282 Kan. 584, 147 P.3d 115, 124 (2006);
Knetzer,
600 P.2d at 161;
State v. Lewis,
27 Kan.App.2d 380, 5 P.3d 531, 536 (2000).
Boyle claims appellate counsel should have argued L.B. had not testified credibly at trial and he points to various facts that allegedly cast doubt on the credibility of her testimony. But as the Kansas appellate courts have made clear they do not weigh the credibility of witnesses on appeal, this argument would have been a sure loser in Boyle’s state appeal.
See, e.g., Knetzer,
600 P.2d at 161 (“It is well-recognized that it is not the function of the appellate courts to weigh conflicting evidence, to pass on the credibility of witnesses, or to redetermine questions of fact. The reviewing court is concerned only with evidence that supports the trial court’s findings, not with evidence that might have supported contrary findings.”);
Lewis, 5
P.3d at 536 (noting that in reviewing a sufficiency-of-the-evidence claim, the court will not “pass on the credibility of witnesses, or redetermine factual questions”).
Accordingly, we agree with the district court’s disposition of this claim.
IV. Conclusion
For the reasons set forth above, the district court’s denial of Boyle’s petition for a writ of habeas corpus is AFFIRMED.