dissenting,
TACHA, J., O’BRIEN, J., and TYMKOVICH, J., joining; and BRISCOE, J., joining Part II.B.This case requires us to interpret the words of a federal statute. That statute says writs of habeas corpus “shall not be granted with respect to any claim that was adjudicated on the merits in State court proceedings” unless the state court’s decision is contrary to a Supreme Court precedent, or unless it rests on an unreasonable application of the Court’s cases or an unreasonable reading of the facts before it. 28 U.S.C. § 2254(d). This language seemingly brooks no exception. Yet, rather than applying AEDPA’s deferential standard to the claims before us, the court today finds itself applying de novo review. How can this be?
The majority tells us this is because a state court does not adjudicate a claim on the merits unless its opinion adequately reflects that the court “considered]” all the evidence we later deem to be “material.” Maj. Op. at 1291. Under this view, the fact that the state court says it is adjudicating a claim on the merits is beside the point. If the state court, in the course of purporting to adjudicate the merits, employs evidentiary rules to exclude evidence we think “material,” it somehow cannot adjudicate the claim at all. The state court’s opinion might look like, talk like, walk like an adjudication on the merits, but it’s not the real thing.
I have substantive and procedural difficulties with this view. Substantively, it is inconsistent with AEDPA’s plain terms and structure, contradicts the decisions of several other circuits, and effectively frustrates AEDPA’s central purpose by getting us back in the business of grading state court procedures rather than focusing on the reasonableness of the results they reach. That isn’t to say state procedures don’t matter: when a state court renders a decision on the merits of a federal claim without considering all material evidence, it is surely more likely that its decision will be an unreasonable application of federal law, and thus reversible under § 2254(d). But the fact that a petitioner is more likely to satisfy AEDPA’s standard of review isn’t a reason to ignore that standard of review, or to disregard a state court’s decision by suggesting its adjudication of the merits is really no such thing.
Procedurally, there’s no need for any of this. Every member of this court agrees that, if the evidentiary standard embodied in the OCCA’s Rule 3.11 is no more onerous than the federal evidentiary standard for deciding ineffective assistance claims, then the OCCA’s denial of a Rule 3.11 motion amounts to an “adjudication on the merits” of a federal ineffective assistance claim. The majority and Judge Tymkovich just disagree on the state law question how best to read Rule 3.11. As the majority reads the Rule, it imposes an evidentia*1316ry standard higher than federal law; as Judge Tymkovich reads it, the Rule parallels federal law. We could and should have just asked the OCCA what its Rule means. After all, the OCCA, not this court, is the authoritative expositor of its own rules. By the simple expedient of certifying a question of state law to the OCCA, we likely could have produced a unanimous or near unanimous result in this case and avoided both a controversial interpretation of AEDPA and a circuit split on an important question of federal law.
I
At least in Mr. Wilson’s case, it is clear that the OCCA did adjudicate the merits of his claim. As the majority formulates his claim, Mr. Wilson alleged he received constitutionally ineffective assistance of counsel “with respect to [the presentation of] mental health mitigation [evidence] at the sentencing phase.” Maj. Op. at 1288. The OCCA spoke to this claim unambiguously:
¶ 108 Wilson first asserts that his attorney failed to fully investigate his mental health background or effectively assist Dr. Reynolds in preparation for his second stage testimony. Wilson has filed, contemporaneously with this issue, an application for an evidentiary hearing regarding ineffective assistance of counsel in an attempt to supplement the record with material not found in the record.
¶ 109 A review of the trial record shows trial counsel did put forth a mental health expert to rebut the State’s continuing threat contention and to mitigate punishment. At trial, Dr. Reynolds testified that he examined Wilson on three separate occasions. He also met with Wilson’s mother and was provided with Wilson’s medical records, school records and statements from people who knew Wilson. Reynolds testified that Wilson had a severe personality disturbance. Reynolds explained that Wilson had some unusual, bizarre types of thinking that would suggest that he is not in touch with reality at times. Reynolds [sic] testimony indicated that Wilson committed this crime as an intelligent but immature person, and that, because of his family support and his intelligence, he had the capability of being rehabilitated. The mere fact more evidence could have been presented is not, in itself, sufficient to show counsel was deficient. Douglas, 951 P.2d at 680. Reynold’s [sic] testimony was credible and well developed. We find Appellant has failed to carry his burden to show either deñcient performance by counsel, or prejudice from the omission of this specific evidence. [FN8]
FN8. Accordingly, we further find that Wilson’s application for an evidentiary hearing on this claim should be denied.
Wilson v. State, 983 P.2d 448, 471-72 & n. 8 (Okla.Crim.App.1998) (emphasis added).
What else could be required for a court to adjudicate the merits of a claim? To be sure, in its footnote 8, the OCCA denied Mr. Wilson’s request for an evidentiary hearing to pursue collaterally developed evidence. One might read this as excluding collateral evidence from the OCCA’s consideration by dint of a state evidentiary rule (as the majority does), or as passing on the collateral evidence and finding it insufficient (as Judge Briscoe does). But either way, the OCCA went on to hold (at the least) that the trial record showed counsel’s performance in connection with the presentation of mitigating mental health evidence was neither deficient nor prejudicial, the two essential elements of any Sixth Amendment ineffective assistance claim under Strickland. The result the OCCA reached on either Strickland element might well be held unreasonable under § 2254(d) for any number of reasons, perhaps including in light of evidence *1317it failed to consider. But none of this diminishes the fact that, wrongly or rightly, the OCCA’s opinion adjudicated the merits of Mr. Wilson’s claim.
When federal courts deny Strickland claims, we do exactly the same thing that the majority contends the OCCA did here: apply rules of procedure and evidence, admit what’s allowed, exclude what’s not, and then say, as the OCCA did, “We find Appellant has failed to carry his burden to show either deficient performance by counsel or prejudice.” The only apparent difference is that the OCCA is not a federal court, and therefore does not use our wise and sensible procedural and evidentiary rules. Instead, as the majority would have it, the OCCA applies unwise and parochial state rules. As a result, a state court’s attempts to adjudicate the claims committed to its jurisdiction can unknowingly fail. Though the OCCA thinks it’s adjudicating the merits of a Sixth Amendment claim— and says what we say in an attempt to accomplish the feat — the majority treats the OCCA’s holding as no more than an ineffectual incantation.
Today’s result is not only inconsistent with the plain language of the OCCA’s decision, it is also inconsistent with the plain language of § 2254. In § 2254(e), Congress expressly anticipated that federal courts sometimes will have material new evidence or facts before them that the state court didn’t have the chance to consider. Yet, Congress did not provide for de novo review. Even in these circumstances, the plain language of the statute indicates that § 2254(d)’s deferential standard continues to apply. As Judge Cabranes has observed for the Second Circuit, this statutory feature forecloses the notion that material new evidence uncovered in federal habeas proceedings “has any impact whatsoever on the standard of review governing habeas claims.” Wilson v. Mazzuca, 570 F.3d 490, 500 (2d Cir.2009).
The majority’s contrary holding today cements a circuit split. The Second, Fifth, and Seventh Circuits have each held, as Judge Tymkovich and I would, that AED-PA’s mandates in subsection (d) regarding our standard of review “do not lose their force because an intervening evidentiary hearing is held in federal court” even if material new evidence emerges. Id. at 500; accord Pecoraro v. Walls, 286 F.3d 439, 443 (7th Cir.2002); Valdez v. Cockrell, 274 F.3d 941, 952 (5th Cir.2001).10 To be sure, additional “evidence obtained in such a [federal court] hearing is quite likely to bear on the reasonableness of the state courts’ adjudication,” but that fact alone supplies no reason “why it should alter the standard of federal review.” Pecoraro, 286 F.3d at 443 (emphasis in original). Instead, where (as here) the petitioner’s claim is governed by subsection (d)(1), “[a]ny new evidence uncovered in the federal proceeding is relevant only insofar as it assists the habeas court in determining whether the state court reached an unreasonable application of law.” Mazzuca, 570 F.3d at 500. In other words, federal courts “are directed to apply the same AEDPA standard that would otherwise be in force, now in light of the new information that has been obtained through” federal court proceedings. Id. at 501-02. If that new evidence renders the state court’s result deficient under § 2254(d) (as of course it may well), relief can then be granted.11
*1318By defining “on the merits” to mean, effectively, after giving adequate consideration to material evidence, the majority returns us to a bygone era. Before AEDPA, § 2254 allowed us to review a state court’s findings de novo whenever we found the state procedures inadequate. See 28 U.S.C. § 2254(d)(2) & (6) (1994) (repealed 1996) (providing for de novo review when “the factfinding procedure employed by the State court was not adequate to afford a full and fair hearing,” or when “the applicant did not receive a full, fair, and adequate hearing in the State court proceeding”). With AEDPA, Congress expressly abolished this approach. It told us to ask whether the substantive decisions made by state courts are reasonable, rather than critique the processes by which they reach those decisions. Today’s majority opinion does much to restore the anden regime: its rule that a state court cannot really adjudicate a claim on the merits unless it adjudicates it well — by considering the “right” evidence — -is immediately recognizable as a species of the full and fair hearing requirement Congress rejected. Cf. Valdez, 274 F.3d at 950 (rejecting a similar view to the majority’s because “[tjhis mandatory and all-encompassing language [of AEDPA] ... leaves no room for judicial imposition of a full and fair hearing prerequisite” formerly permitted by § 2254). In this way, the majority’s definition is at odds with the central purpose of the statute it is interpreting.
In fact, the majority’s interpretation of § 2254(d) winds up treating state courts less like instruments of sovereign governments and more like federal agencies whose decisions Congress has instructed us to review by asking whether they were “within the scope of its lawful authority, [and also whether] the process by which it reaches that result [was] logical and rational” with an “explication [that] is not inadequate.” Allentown Mack Sales & Service, Inc. v. NLRB, 522 U.S. 359, 364, 374, 118 S.Ct. 818, 139 L.Ed.2d 797 (1998) (citation omitted). The problem is that AEDPA is not the APA. In AEDPA, Congress sought to give meaning to principles of federalism and comity and afford the judgments of state court respect and deference without regard to the length of their explications, directing us instead to extend “deference to the state court’s result, even if its reasoning is not expressly stated.” Aycox v. Lytle, 196 F.3d 1174, 1177 (10th Cir.1999). As the Supreme Court has emphasized, Congress instructed that a state court’s decision “does not even require awareness of [the Supreme Court’s] cases” to warrant deference under AEDPA, “so long as neither the reasoning nor the result of the state-court decision contradicts them.” Early v. Packer, 537 U.S. 3, 8, 123 S.Ct. 362, 154 L.Ed.2d 263 *1319(2002) (emphasis in original). Under the majority’s interpretation, however, it is impossible to defer to a state court’s result without first examining its reasoning — precisely what AEDPA and Early say the law is not.
In splitting with other circuits and holding that an “adjudication on the merits” under § 2254(d) requires a state court to give sufficient consideration to evidence we think material, the majority’s interpretation does not rely on AEDPA’s plain language, structure, or purpose. Instead, it defines the words “adjudicated on the merits” by reference to Black’s Law Dictionary. Maj. Op. at 1291. But as the Supreme Court has recognized, “[t]he term ‘on the merits’ has multiple usages.” Gonzalez v. Crosby, 545 U.S. 524, 532 n. 4, 125 S.Ct. 2641, 162 L.Ed.2d 480 (2005); see also Semtek Int'l Inc. v. Lockheed Martin Corp., 531 U.S. 497, 501-03, 121 S.Ct. 1021, 149 L.Ed.2d 32 (2001) (describing ambiguity of the term). Thus, the task before us is to ascertain its meaning in this statutory context. One handpicked citation to a single authority using an ambiguous term in a friendly way, and without any reference to the statutory regime before us, does not persuade.12
One may well complain that the OCCA’s adjudicative procedures are unjust, and argue that ineffective assistance claims cannot be fairly decided without reference to evidence outside the trial court record. See Maj. Op. at 1294. But even assuming this is true, Congress supplied a solution for the problem in AEDPA’s text. In subsection (b), Congress indicated that a petitioner need not exhaust any state procedure that is “ineffective to protect the rights of the applicant.” 28 U.S.C. § 2254(b)(1)(B); see also Kimmelman v. Morrison, 477 U.S. 365, 106 S.Ct. 2574, 91 L.Ed.2d 305 (1986). It is an open and interesting question whether the OCCA’s Rule 3.11, by limiting extra-record evidence in the consideration of Sixth Amendment claims, meets this threshold. The trouble in this particular case is that Messrs. Wilson and Wackerly have not invoked the protections of subsection (b) or advanced any challenge to Rule 3.11 on that basis, so it can do them no good. In the majority’s view, this seems to be reason enough to create another remedy, re*1320dressing § 2254(b)’s concern within § 2254(d) by refashioning what it means to adjudicate a claim. Doing so, however, generates a statutory anomaly. Under the majority’s view, not only is seeking a ruling from a state court using inadequate evidentiary procedures unnecessary (under subsection (b)), it is now metaphysically impossible (under subsection (d)). If we do not like the state court’s evidentiary procedures, then we will say not that the state adjudicated the petitioner’s claim unfairly, but that the state failed to adjudicate the petitioner’s claim at all. So it is that, under the reading of the statute the court adopts today, § 2254(b)(1)(B) seems to have little meaningful role left to play, another clue suggesting a problem with the interpretive path the majority has taken.13
II
The course we chart today is not only erroneous, it is unnecessary — and unnecessary in two distinct ways. First, the majority itself identifies a narrower basis supporting its result that does not depend on a controversial reading of AEDPA’s text; yet, without explanation, the majority then declines to rest on it. Second, simply certifying our dispute over what Oklahoma’s Rule 3.11 means to the institution able to interpret it authoritatively — the OCCA— could have entirely obviated the need for us to interpret § 2254(d)’s terms; yet, the majority eschews that course, too.
A
In the first place, after exerting so much effort to explain why § 2254(d) deference doesn’t apply, the majority tells us it makes no difference. Even if the OCCA’s decisions were adjudications on the merits and so entitled to AEDPA deference, the majority says, it would still reverse them under § 2254(d)(1). Maj. Op. at 1294. In other words, we learn that the entire dispute about what it takes to adjudicate a claim on the merits under § 2254(d) is unnecessary to the majority’s resolution of these appeals. Our usual course in such circumstances is to assume without deciding that AEDPA deference is due and then *1321reverse because, even affording such deference, the state court’s decision could not stand. The majority, however, declines to follow this more restrained path and never tells us why it prefers a broader holding that, in its own view, isn’t needed.
I confess that I find even the majority’s narrower course unpersuasive on its own terms. The Supreme Court has cautioned that § 2254(d)(l)’s “ ‘contrary to’ and ‘unreasonable application’ clauses [have] independent meaning.” Terry Williams v. Taylor, 529 U.S. 362, 405, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000). Yet, the majority’s opinion isn’t clear about which clause the OCCA has violated. It first tells us that the OCCA’s decisions should be reversed as “contrary to ... clearly established Federal law.” Maj. Op. at 1294 (quoting § 2254(d)(1)). In order for this to be true, the state court must have either applied a rule that contradicts a Supreme Court decision, or else confronted a set of facts materially identical to a relevant Supreme Court precedent and reached a different result. Terry Williams, 529 U.S. at 405-06, 120 S.Ct. 1495. Still, the majority’s opinion does not tell us what Supreme Court holding the OCCA misstated or what result runs contrary to a decision of the Supreme Court. Instead, in the very next sentence, the majority switches to the “unreasonable application” theory, asserting that “[t]o decide a legal claim without regard to the evidence is surely unreasonable.” Maj. Op. at 1294 (emphasis added). But this also misstates the applicable statutory test: we do not sit to decide whether state courts behave unreasonably; the question we may ask is only whether their applications of federal law, as embodied in Supreme Court precedents, are unreasonable. See Carey v. Musladin, 549 U.S. 70, 127 S.Ct. 649, 166 L.Ed.2d 482 (2006). And the majority identifies no Supreme Court holding the OCCA has contradicted or unreasonably applied. It does supply a quote from Terry Williams, but the quoted passage is from Part III of Justice O’Connor’s separate concurrence, which did not speak for the Court. Maj. Op. at 1294. Then the majority cites a list of our own precedents supporting its conclusion that we do not owe deference to adjudications under Rule 3.11. Id. at 1294. There is, of course, a similar list of Tenth Circuit cases supporting just the opposite rule, and this intra-circuit split of panel authority is precisely why we granted en banc rehearing in the first place. See Maj. Op. at 1287. So even if we could somehow leave aside the difficulty that we are not the Supreme Court, our inconsistent panel precedents on this point have certainly not clearly established the state of the law.
Having said all this, the majority does not share my concerns with its alternative holding. Given its apparent confidence in its alternative and narrower holding that reversal is required even if AEDPA applies, there is no need for the court to pursue its primary and broader holding that AEDPA does not apply. Restraint would normally caution against a broader course when the narrower will do. Why issue a sweeping and controversial legal holding about AEDPA’s meaning when a fact-bound holding about the statute’s application is so readily at hand?
B
The majority’s principal holding is needless in a second and even more fundamental way. The question whether the OCCA’s passage in Wilson quoted above constitutes an “adjudication on the merits” must be resolved only if the OCCA’s denial of the Rule 3.11 motion in its footnote 8 is not itself an adjudication on the merits. Mr. Wackerly’s case turns even more heavily on Rule 3.11’s meaning because there the OCCA dealt with his present *1322claim only by denying his Rule 3.11 motion.14
It is accepted by everyone on this court that if, in the course of denying an evidentiary hearing under Rule 3.11, the OCCA reviews a petitioner’s proffered evidence under standards as favorable to the petitioner as Strickland’s, we owe that denial deference under AEDPA without respect to what else the OCCA does or does not say. Harris v. Poppell, 411 F.3d 1189, 1196 (10th Cir.2005) (“[I]f the OCCA rejected Mr. Harris’s claim under a standard that is equally or more favorable to him relative to the federal standard, the state court’s decision constitutes an adjudication of the federal claim despite citing no federal decisions.”); see also Romano v. Gibson, 239 F.3d 1156, 1164 (10th Cir.2001). My colleagues just disagree about whether Rule 3.11 imposes on petitioners a standard more or less onerous than Strickland because it requires petitioner to present “clear and convincing evidence [of] a strong possibility” of ineffective assistance. On the one hand, Judge McConnell says “clear and convincing evidence” requires more than federal law, which requires a petitioner to show a Strickland violation by a mere preponderance of the evidence. On the other hand, Judge Tymkovich argues that Rule 3.11 requires less of petitioners than Strickland because a petitioner need only show “a strong possibility” of ineffective assistance by clear and convincing evidence.
It seems to me that, given the absence of any Oklahoma decision on point, we can do no better than guess at which interpretation of Rule 3.11 (if either) is right. But, more importantly, there is no reason for us to guess. The OCCA has authority to decide definitively what the phrase “clear and convincing evidence” means in the context of its own rule. We should have just asked. If the OCCA had agreed with the majority’s interpretation of its Rule, a very large majority (if not all) of us would have agreed that no deference would be owed to the OCCA’s decision in Wackerly,
*1323and we would have had to proceed to confront the § 2254(d) question in Mr. Wilson’s case. But if the OCCA interpreted Rule 3.11 in the manner Judge Tymkovich does, as imposing a lesser burden on petitioners than Strickland, everyone on this court would have treated the denial of Rule 3.11 applications as implicit adjudications of ineffective assistance claims and applied AEDPA deference to the OCCA’s decisions in both Wilson and Wackerly— and done so without needing to address whether the OCCA’s additional language in Mr. Wilson’s case amounted to an adjudication of the merits of his claim under § 2254(d). Asking the OCCA to interpret its own rule rather than venture a guess ourselves would have had the benefit, as well, of “ ‘helping] build a cooperative federalism’ ” by “giv[ing] meaning and respect to the federal character of our judicial system, recognizing that the judicial policy of a state should be decided when possible by state, not federal, courts.” Pino v. United States, 507 F.3d 1233, 1236 (10th Cir.2007) (quoting Lehman Bros. v. Schein, 416 U.S. 386, 391, 94 S.Ct. 1741, 40 L.Ed.2d 215 (1974)).15
Instead of asking the OCCA, the majority ventures a guess about the meaning of state law. Yet, the OCCA can simply tell us tomorrow what Rule 3.11 means (or revise it, or repeal it), and we will then be obliged to follow its understanding, regardless of our court’s interpretation today. Future panels of this court likewise remain free to certify the question of Rule 3.11’s meaning to the OCCA. Given all this, the shelf life of the majority’s construction of state law could be short indeed. Precisely because of this fact, and in recognition of certification’s useful role in promoting a cooperative federalism, federal courts of appeals do not often (ever?) devote such substantial en banc efforts to the adjudication of state law questions. The majority’s refusal to certify is made more puzzling when doing so might well have allowed us to avoid issuing a divided opinion, offering a questionable interpretation of an important federal statute, and cementing a split among the federal circuits. Our first in*1324terpretive imperative should be to do no harm, and I do not see why we should be so anxious to charge into such a breach when it might be well and honorably avoided.
I respectfully dissent.
. Last year, the Supreme Court granted certiorari to resolve the dispute between the circuits on this score, Bell v. Kelly, - U.S. -, 128 S.Ct. 2108, 171 L.Ed.2d 228 (2008) (mem.), but later dismissed the writ as improvidently granted, - U.S. -, 129 S.Ct. 393, 172 L.Ed.2d 353 (2008) (mem.).
. The majority seeks to distinguish Mazzuca on the basis of its procedural posture. The *1318majority emphasizes that Mazzuca "does not consider whether or not the state court made an adjudication on the merits, but instead asks how new evidence brought at habeas should affect the federal court's evaluation of the state court’s decision given that there was an adjudication on the merits.” Maj. Op. at 1295, n.2. This distinction fails for the simple reason that the federal evidentiary hearing in Mazzuca occurred precisely because the state court record lacked relevant material evidence. See Wilson v. Mazzuca, 2005 WL 27543, at *1 (2d Cir.2005) (unpub.) ("[W]ith the record before us, we are unable to assess with confidence whether strategic considerations accounted for certain acts or omissions by defense counsel .... We therefore remand to the District Court to afford Wilson’s trial counsel the opportunity to explain his actions.” (quotation marks omitted)). This finding carries the necessary implication that the state court in Mazzuca failed to consider that material non-record evidence — exactly what the majority says happened here. Yet, the Second Circuit accepted the state court's decision as an adjudication on the merits and applied AEDPA deference — exactly what the majority does not do here.
. The majority takes a similar approach to its definition of the word "claim,” relying on carefully chosen authorities from other contexts rather than examining the statutory context in which the word is found. See Maj. Op. at 1291. For example, the majority lifts one definition of "claim” from Black's Law Dictionary (which contains six entries for the word), cites authority that itself describes the term as ambiguous, see United States v. Ripa, 323 F.3d 73, 82-83 & n. 10 (2d Cir.2003), and relies on an analogy to Fed.R.Civ.P. 12(b)(6). With respect to Rule 12(b)(6), moreover, the majority suggests that a plaintiff whose complaint contains insufficient factual averments has no claim at all because his suit will be dismissed "for failure to state a claim.” Maj. Op. at 1291. But the phrase "failure to state a claim,” like "AEDPA deference,” is just shorthand for the text of the actual rule, which allows dismissal “for failure to state a claim upon which relief can be granted." Fed. R.Civ.P. 12(b)(6) (emphasis added). In a case dismissed under Rule 12(b)(6), the plaintiff surely advanced a claim, but relief could not be granted for that claim because the facts he averred in support of it were not sufficient to warrant relief under law. The majority occasionally even goes so far as to suggest that a "claim” consists not only of facts supporting a legal theory, but also all the evidence that in turn proves those facts. See Maj. Op. at 1291 (“The merits question presented is whether, in light of non-record evidence, trial counsel's performance satisfied constitutional standards.”) (emphasis added). As Judge Tymkovich well explains, this false equivalence between a claim and all the evidence supporting it also undergirds the majority’s flawed analogy to the requirement that habeas petitioners exhaust their state remedies. See Tymkovich Dissent at 1322-23. The point is that, when interpreting an ambiguous term, context matters.
. The court does not mention one of the chief arguments for de novo review advanced by the parties in this case, one based on Rompilla v. Beard, 545 U.S. 374, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005), and Wiggins v. Smith, 539 U.S. 510, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003). In those cases, a state court rejected a petitioner's claim of ineffective assistance of counsel by finding that he had not satisfied the deficient performance element of Strickland, without reaching Strickland’s remaining prejudice element. In each case, the Supreme Court held that the state court's deficient performance decision was an unreasonable application of Strickland under § 2254(d)(1). Rompilla, 545 U.S. at 389, 125 S.Ct. 2456; Wiggins, 539 U.S. at 534, 123 S.Ct. 2527. Because the state courts in Rompilla and Wiggins had not issued any decision on the question of prejudice, the Supreme Court proceeded to review that question de novo. The petitioners suggest these cases allow us to apply de novo review to issues not addressed by the state courts. But the majority abandons this argument for good reason: in each case, issuance of the writ was consistent with § 2254(d)(1) because the state court's result on the deficient performance question was an unreasonable application of Strickland. Nothing in Rompilla and Wiggins suggests we may avoid our obligation to find something unreasonably wrong with a state court's decision under (d)(1) or (d)(2) before issuing a writ of habeas corpus.
Separately, the majority suggests we dissenters are "concerned” that lawyers will sandbag us by withholding evidence in state court and then producing it in federal court to obtain de novo review. Maj. Op. at 1294. Our concern, however, is instead with arriving at the most likely reading of a law Congress passed, using conventional tools of statutory interpretation. As the above statutory analysis reflects, an assumption that lawyers will behave unethically does not inform my (or anyone’s) conclusions about what Congress meant by the phrase "adjudicated on the merits.”
. The discussion in the OCCA’s opinion in Wackerly makes this clear:
¶ 37 In addition to the ineffective assistance of counsel claims raised by Appellant in his Brief-in-Chief, he filed an Application for an Evidentiary Hearing asserting three additional claims which he contends are supported by evidence not in the record but which was available to defense counsel at the time of trial. In this application, Appellant requests an evidentiary hearing on: 1) the failure of defense counsel to impeach Michelle Wackerly; 2) the failure of defense counsel to produce evidence during the Franlcs hearing to establish that Agent Franchini misrepresented information regarding Michelle Wackerly's recollection of the location of the gun; and 3) counsel’s failure to investigate and present available evidence which would have warranted different verdicts and sentences in the first and second stages of trial.
¶ 38 Appellant requests this evidentiary hearing based upon Rule 3.11(B)(3)(b), Rules of the Oklahoma Court of Criminal Appeals, Title 22, Ch. 18, App. (1999). This rule allows an appellant to request an evidentiary hearing when it is alleged on appeal that trial counsel was ineffective for failing to “utilize available evidence or adequately investigate to identify evidence which could have been made available during the course of trial .... ” Once an application has been properly submitted along with supporting affidavits, this Court reviews the application to see if it contains “sufficient evidence to show this Court by clear and convincing evidence there is a strong possibility trial counsel was ineffective for failing to utilize or identify the complained-of evidence.” Rule 3.11 (B)(3)(b)(i).
¶ 39 Upon review of the application and supporting exhibits, we find Appellant has shown this Court that trial counsel could perhaps have accessed other information in preparing for trial. However, Appellant has not shown by clear and convincing evidence a strong possibility that defense counsel was ineffective for failing to utilize or identify the complained-of evidence. Accordingly, we decline to grant Appellant’s application for an evidentiary hearing.
Wackerly v. State, 12 P.3d 1, 13-14 (Okla.Crim.App.2000) (footnote omitted).
. The majority emphasizes that the State didn't move for (even though it expressed openness to) certification. Maj. Op. at 1299-1300. But, as the majority also acknowledges, we surely have the power to certify questions on our own motion, see, e.g., Elkins v. Moreno, 435 U.S. 647, 662, 98 S.Ct. 1338, 55 L.Ed.2d 614 (1978); Kansas Judicial Review v. Stout, 519 F.3d 1107, 1120 (10th Cir.2008), and, as it happens, sua sponte certification is the norm, see 17A Charles Alan Wright et al., Fed. Prac. & Proc. § 4248, at 509 (3d ed.2007) ("[ojrdinarily a court will order certification on its own motion"). The majority suggests certification isn't warranted because the question in this case "is not how we should interpret and apply the 'clear and convincing' standard (which we would never have occasion to do) but what past decisions of the OCCA meant when they employed that standard." Maj. Op. at 1299-1300. But no one has suggested that we are in the business of applying a state evidentiary rule to parties in a federal habeas proceeding. Instead, the unavoidable fact is that, for us to determine which federal standard of review pertains to Messrs. Wilson’s and Wackerly’s federal habeas petitions, we need to know the meaning of a state evidentiary rule. The applicability of AEDPA turns on the meaning of OCCA Rule 3.11. We routinely ask states to interpret their own rules and statutes when doing so will aid in the disposition of federal claims and lawsuits, and we should do the same here. To the extent that the majority seems concerned that the OCCA could somehow tell us what Rule 3.11 means only prospectively, and not what it meant at the time it adjudicated Wilson and Wackerly, this concern is also misplaced. We could have simply asked the OCCA what its Rule 3.11 meant at the time it issued Wilson and Wackerly, without regard to any later amendments or legal developments (if any happen to exist). Notably, this is exactly the course we took in the case the majority relies upon, Pino, 507 F.3d at 1236-38, asking the state court there the meaning of a state statute as of 2003, before certain later statutory amendments.