It’s not every day we overturn a state jury verdict for first-degree murder when the defendant admits he received a fair trial and no one questions that his conviction is supported by overwhelming evidence. It’s not every day we exacerbate a split of authority over the recognition of a new constitutional right, and do so despite warning signs from the Supreme Court against our course. And it’s not every day we refuse to rehear a panel decision that every single state within our jurisdiction has urged us to revisit. Today, we do all these things, and I respectfully dissent.1
*1257Before his state trial, Michael Williams’s lawyer persuaded him not to accept a plea offer that would have reduced the charge against him from first-degree murder to second-degree murder and given him a shot at a ten-year sentence. After rejecting the plea offer, Mr. Williams received a trial that he concedes was fair. At that trial, the government amassed overwhelming evidence of his guilt,2 the jury unanimously reached a guilty verdict, and the trial court ultimately sentenced Mr. Williams to life in prison without the possibility of parole. After all this, Mr. Williams complained that his trial lawyer improperly convinced him to reject the government’s pre-trial plea offer. The Oklahoma Court of Criminal Appeals (OCCA) agreed that Mr. Williams’s lawyer had behaved “improperly],” OCCA Op. at 20 (internal quotation marks omitted), and reduced his sentence from life without the possibility of parole to life with the possibility of parole, the lowest sentence available under Oklahoma law for someone convicted of first-degree murder. While Mr. Williams sought the full benefit of the forgone plea, the OCCA recognized that a ten-year sentence was only possible in a second-degree murder case under state law, and that it had no power to compel the State’s executive branch to revive that lesser charge.
Pursuing the matter further in federal court, Mr. Williams argued that the OCCA’s decision violated the Sixth Amendment of the United States Constitution by failing to afford him more relief. The panel majority adopted the same view. So it is that, under the panel’s holding, defendants in our circuit who are offered pre-trial plea agreements (and surely that’s most of them) can now take a shot at trial knowing that, if it doesn’t go well, they will still have a chance at reviving the forgone plea offer, as long as they can show counsel failed to advise them properly. And they may do all this even if the trial they elected proved to be fair, the evidence produced against them is incontestable, and restoring the lost plea will require a federal court to compel a state prosecutor to revive a new and different set of charges than those endorsed by a grand jury, selected by the prosecutor for proof at trial, and found by the jury. Plainly, this holding represents a significant new federal intrusion into state judicial functions and a revamping of the separation of powers, one that unsurprisingly conflicts with the decisions of a number of other courts, including the Utah Supreme Court.3 No doubt all these factors con*1258tributed to the unusual decision of every state within our circuit (and some outside it) to join an amicus brief urging us to rehear this case en banc.
The OCCA’s decision deserved more deference from us. In the provision of the Antiterrorism and Effective Death Penalty Act (AEDPA) that governs our review of cases like this one, Congress has instructed us to defer to state court decisions unless they are “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court.” 28 U.S.C. § 2254(d)(1). As it happens, no decision from the United States Supreme Court has ever held (or even hinted) that a lawyer’s bad advice to reject a plea offer gives rise to a violation of the Sixth Amendment, or any other provision of federal law. Neither does a conventional Strickland analysis compel such a novel result. And whether the OCCA chooses to engage in an elaborate Sixth Amendment exegesis is neither here nor there: we are not here to grade state court opinions but owe “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) (emphasis in original). Indeed, the Supreme Court has told us that a state court’s decision “does not even require awareness of’ Strickland or other Supreme Court 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 (2002) (emphasis in original). That is hardly the case here, and as a result, we were statutorily obliged to leave the OCCA’s decision undisturbed. See Knowles v. Mirzayance, — U.S. -, 129 S.Ct. 1411, 1420, 173 L.Ed.2d 251 (2009) (explaining the “doubly deferential” review owed to state court decisions under Strickland and § 2254(d)(1)).
Even if we could somehow overlook the absence of any clearly established federal law suggesting that the OCCA’s decision was improper and allow Mr. Williams, AEDPA notwithstanding, to argue for an extension of existing Sixth Amendment jurisprudence (that’s really what he seeks), his project remains flawed on its own terms. The flaw resides in a conflation of Strickland’s two analytically distinct elements. Even assuming that Mr. Williams’s lawyer gave him bad advice in the pre-trial plea bargaining process, that alone is not enough to establish a Sixth *1259Amendment violation. To make out a Sixth Amendment violation under Strickland, a lawyer’s bad advice (or deficient performance) must also prejudice the defendant by infringing some legal entitlement due him. That is, a lawyer’s bad advice must have some adverse impact on the defendant’s legal rights — say, by denying him his constitutionally guaranteed right to cross-examine a witness, or his constitutionally protected right to present exculpatory proof at trial. See Strickland v. Washington, 466 U.S. 668, 695, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984); Nix v. Whiteside, 475 U.S. 157, 186-87, 106 S.Ct. 988, 89 L.Ed.2d 123 (1986); Terry Williams v. Taylor, 529 U.S. 362, 393 n. 17, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000); Lockhart v. Fretwell, 506 U.S. 364, 372, 113 S.Ct. 838, 122 L.Ed.2d 180 (1993). Underscoring all this, the Supreme Court has told us over and again that the Sixth Amendment right to effective assistance of counsel is not itself the font of substantive legal rights, but an instrumental right designed to ensure that the criminal defendant’s lawyer sees to it that he receives all the protection to which other laws entitle him. See United States v. Cronic, 466 U.S. 648, 658, 104 S.Ct. 2039, 80 L.Ed.2d 657 (1984); United States v. Gonzalez-Lopez, 548 U.S. 140, 147, 126 S.Ct. 2557, 165 L.Ed.2d 409 (2006).
Even now, on the State’s motion for rehearing en banc, Mr. Williams still has not identified any legal entitlement due him that was denied him or infringed. This is so for the simple and disarming reason that a defendant has no legal entitlement to accept and enforce a pre-trial plea offer. In fact, the only potential source of law anyone has suggested that might entitle a defendant to accept a pretrial plea bargain is the Due Process Clause of the United States Constitution. Yet, because the Bill of Rights “speaks in explicit terms to many aspects of criminal procedure,” the Supreme Court has set a high bar for us to pass before we may conscript the Due Process Clause into service on behalf of new and unenumerated constitutional rights. Medina v. California, 505 U.S. 437, 443, 112 S.Ct. 2572, 120 L.Ed.2d 353 (1992) (“In the field of criminal law, we have defined the category of infractions that violate ‘fundamental fairness’ very narrowly based on the recognition that, beyond the specific guarantees enumerated in the Bill of Rights, the Due Process Clause has limited operation.” (internal quotation marks omitted)); see also id. at 445, 112 S.Ct. 2572 (“[Pjreventing and dealing with crime is much more the business of the States than it is of the Federal Government.... [W]e should not lightly construe the Constitution so as to intrude upon the administration of justice by the individual States.” (internal quotation marks omitted)). Under Medina, state criminal procedures and convictions will not be held to run afoul of the Due Process Clause unless they “offend[ ] some principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental.” Id. at 445, 112 S.Ct. 2572 (internal quotation marks omitted).
No principle of justice rooted in our constitutional order entitles a defendant to receive, accept, or enforce a plea offer. Very much to the contrary, the Supreme Court has explained (repeatedly) that a plea offer is a matter of executive grace— not constitutional right or even contract— and so affords the defendant no enforceable rights unless and until the plea is embodied in the judgment of a court. See Weatherford v. Bursey, 429 U.S. 545, 97 S.Ct. 837, 51 L.Ed.2d 30 (1977); Mabry v. Johnson, 467 U.S. 504, 104 S.Ct. 2543, 81 L.Ed.2d 437 (1984), abrogated on other *1260grounds by Puckett v. United States, — U.S. -, 129 S.Ct. 1423, 173 L.Ed.2d 266 (2009); Whiskey Cases, 99 U.S. 594, 25 L.Ed. 399 (1878); see also United States v. Norris, 486 F.3d 1045, 1048, 1053 (8th Cir.2007) (en banc) (plurality).4
At the end of the day, there’s no authority anywhere suggesting that Mr. Williams suffered a Sixth Amendment violation. It follows that the OCCA, in deciding to remit Mr. Williams’s sentence from life without the possibility of parole to life with the possibility of parole, was more generous to him than any federal law required. But no good deed goes unpunished.5 Mr. Williams’s arguments, adopted by the pan el, mistake the deference due state court decisions under AEDPA, conflate two distinct portions of the Strickland test, and disregard fundamental principles of Due Process jurisprudence — leading us, in the end, to intrude improperly into state criminal proceedings and charging decisions belonging to the executive, not judicial, department. For all these reasons, I respectfully submit, we should have reheard this case.
. While voting for rehearing en banc, I have not sought panel rehearing. The matter before us is one of “exceptional public importance,” meriting en banc review, 10th Cir. R. 35.1(A), and I surely believe the panel got it wrong. But, I cannot say, as I must to request panel rehearing, that my panel colleagues "overlooked or misconstrued” any argument or evidence presented to them. See 10th Cir. R. 40.1(A). In addition, because we *1257have, sadly, lost a panel colleague to the academy, a vote among the remaining two panel members would likely result in a tie.
. In 1997, someone entered the home of Larry and Dolores Durrett with a gun. The gunman shot Mr. Durrett three times in his sleep, and twice more when the victim tried to pursue him. Mr. Durrett later died of his wounds. During a routine traffic stop the next day, police discovered Mr. Williams and his girlfriend, Debra Smith, with packed suitcases and a rifle matching the shell casings left at the Durretts’s home. At trial, several witnesses reported hearing Mr. Williams threaten to kill Mr. Durrett over a botched drug deal. Evidence also revealed that Mr. Williams's friend and eventual co-defendant, Stacy Pearce, drove Mr. Williams to the Durretts’s home the day of the murder and watched Mr. Williams exit the car with a gun in hand. Mr. Pearce testified that when Mr. Williams returned to the car he confessed to shooting Mr. Durrett. Ms. Smith also testified that Mr. Williams confessed to her that he killed Mr. Durrett.
. Mr. Williams very carefully argues that the panel's decision did not implicate a federal circuit split. Not only is this contention itself at least arguably incorrect, see United States v. Springs, 988 F.2d 746, 749 (7th Cir. 1993), it also conspicuously bypasses the fact that the panel's decision is inarguably in conflict with the decisions of various state courts. See Williams v. Jones, 571 F.3d 1086, 1108 (2009) (Gorsuch, J., dissenting) (collecting authority); see also Carmichael v. People, 206 *1258P.3d 800, 810-11 (Colo.2009) (Coats, J„ dissenting). No doubt, this split in authority contributed to the Supreme Court’s previous decision to grant certiorari to resolve the question. See Hoffman v. Arave, 455 F.3d 926, 942-43 (9th Cir.2006), cert. granted, - U.S. -, 128 S.Ct. 532, 532-33, 169 L.Ed.2d 371 (2007), vacated as moot, 552 U.S. 117, 128 S.Ct. 749, 169 L.Ed.2d 580 (2008). Today’s concurrence (but not Mr. Williams) seeks to downplay the panel’s conflict with certain of these decisions. That is difficult to do. The Utah Supreme Court explicitly held, in contravention to the panel’s holding, that "a fair trial for the defendant generally negates the possibility of prejudice.” State v. Greuber, 165 P.3d 1185, 1190 (Utah 2007). While the court provided an alternative holding and (unremarkably) reserved for future review questions not presented to it (none of which is presented in our case), its primary holding unambiguously conflicts with the panel’s. Similarly, the Missouri Court of Appeals plainly held that "effective assistance of counsel generally is not implicated unless counsel’s challenged conduct had some effect on the reliability of the trial process.” Bryan v. State, 134 S.W.3d 795, 802 (Mo.Ct.App. 2004). Consistent with this rule, in McClendon v. State, 247 S.W.3d 549, 555-56 (Mo.Ct. App.2007), the Missouri court merely permitted an evidentiary hearing when an attorney “affirmatively misinformed” the defendant about the maximum punishment possible after (his constitutionally guaranteed) trial, a circumstance that the court took pains to emphasize was not presented in Bryan, and which likewise is not presented here.
. Contrary to Mr. Williams’s suggestion, Hill v. Lockhart, 474 U.S. 52, 106 S.Ct. 366, 88 L.Ed.2d 203 (1985), does not provide this right. See Williams, 571 F.3d at 1098 (Gorsuch, J., dissenting). In Hill, counsel performed deficiently when he convinced the defendant to accept a plea offer. Id. at 59, 106 S.Ct. 366. As a result, the defendant was denied his constitutional entitlement to a trial. The same simply cannot be said here. This isn’t to suggest, however, that a district court may not take discretionary action (such as the exclusion of evidence) to prevent the government from gaming an unfair advantage from a defendant’s reliance on a plea agreement that hasn’t yet been reduced to a judgment of the court. See Norris, 486 F.3d at 1049.
. Mr. Williams's suggestion that my position is "at odds” with the OCCA’s decision gets things backward. My position is that, under AEDPA, we should have deferred to the OCCA’s decision. Of course, this is because I think the result the OCCA reached was more generous than any federal law required. But we often affirm for different reasons than those given by the court under review, and when we do so it is hardly reasonable to describe us as "at odds” with them. To the contrary, this is exactly what AEDPA requires of us: deference to state court decisions, whether perfect or not, so long as they are not contrary to clearly established federal law. Mr. Williams’s related argument that the prejudice question is beyond the scope of the COA is likewise misguided. As the entire panel recognized, we had to ascertain what violation occurred, if any, in order to determine whether the OCCA adequately remedied the violation. Indeed, the panel per curiam followed precisely this course in its analysis. See Williams, 571 F.3d at 1090-92 ("In fashioning the appropriate remedy for ineffective assistance of counsel, the remedy should be tailored to the injury suffered from the constitutional violation .... ” (internal quotation marks omitted)); id. at 1095 (Gorsuch, J., dissenting). In any event, the panel was of course free to expand the COA to reach the question as it did. See Young v. Sirmons, 551 F.3d 942, 974 n. 1 (10th Cir.2008); see also Valerio v. Crawford, 306 F.3d 742, 764 (9th Cir.2002) (en banc). And even if one assumed that the panel somehow wasn’t entitled to reach the prejudice question, that would only make the case for review stronger — not weaker. There is, after all, no doubt that we can review and undo ultra vires holdings. Mr. Williams’s proposal that we somehow couldn't review such a holding would perversely, and quite incorrectly, insulate from review decisions that are both erroneous and extra-jurisdictional. That’s never been the law.