filed a dissenting opinion.

I agree with the concurrence that Hill v. Lockhart is not controlling.1 The issue in this case is—or should be—whether the rule announced in Moussazadeh III applies retroactively.2 The concurrence considers it “more appropriate to accord retroactive status to a new rule that once was the rule than to a new rule that is truly new, in the sense that it has never been the rule before.”3 This point is well-taken, but there is another consideration that I believe weighs in favor of this. Court’s adherence to the Teague v. Lane4 retroac-tivity standard: the State’s reliance interest in the finality of its plea bargains. In light of our holding in Moussazadeh II, from at least 2002 to 2012 the State was on notice that any time parole eligibility was not an “element of the plea agreement,” it could safely leave the plea record silent as to this point.5 It did so with the understanding that this Court would typically interpret this silence .as an indicator of the non-elemental status of parole eligibility.6 The State should not now suffer the consequences of this potentially deliberate silence when it had good reason to rely upon Moussazadeh II—or at least good reason to think that any clear “break[ ]” from this law (such as occurred in Moussazadeh III) would mean that Teague weighs against retroactivity.7

The plea in this case took place in 2009—after Moussazadeh II but before Moussazadeh III. Because I would hold Moussazadeh III non-retroactive on collateral review, Moussazadeh II requires Evans to show that parole eligibility rose to the level of an “affirmative element of the plea agreement.”8 I do not believe Evans has made this showing. His claim should be denied. Because the Court grants relief, .1 respectfully dissent. ...

OPINION

delivered the opinion of the Court

Applicant was charged with causing serious bodily injury to a child under Texas Penal Code section 22.04(a)(1). After the State abandoned the deadly weapon allegation, he pled guilty with a 50-year cap, and the trial court sentenced him to 50 years in prison. Applicant now claims that his plea was involuntary because his attorney mi-sadvised him about the effect of a deadly weapon finding on his parole eligibility. He says that if his attorney had correctly advised him, he would have insisted on going to trial. The habeas court found those claims to be true and recommended that we grant relief. The habeas court’s findings are supported by the record.1 The only question we face is whether the law as it existed when Applicant’s conviction became final entitles him to relief. We conclude that it does.

Applicant’s claim for relief is grounded in the federal constitution. U.S. Const. amend. VI (rights to jury and counsel). The ultimate authority on federal constitutional law is the U.S. Supreme Court. U.S. Const. art. VI, cl. 2; Marbury v. Madison, 5 U.S. 137, 177-78, 1 Cranch 137, 2 L.Ed. 60 (1803); Hernandez v. State, 988 S.W.2d 770, 772 (Tex. Crim. App. 1999); State v. Guzman, 959 S.W.2d 631, 633 (Tex. Crim. App. 1998). The Supreme Court’s pronouncements about federal constitutional law are binding on this Court. Ex parte Ramey, 382 S.W.3d 396, 397 (Tex. Crim. App. 2012); Coronado v. State, 351 S.W.3d 315, 317 (Tex. Crim. App. 2011); Coble v. State, 330 S.W.3d 253, 270 (Tex. Crim. App. 2010). Thus, the validity of Applicant’s claim must be judged in accordance with applicable U.S. Supreme Court precedent.

A defendant is entitled to effective assistance of counsel in the guilty plea context. Hill v. Lockhart, 474 U.S. 52, 59, 106 S.Ct. 366, 88 L.Ed.2d 203 (1985). To prevail on a claim of ineffective assistance of counsel due to bad advice about parole eligibility, a defendant “must show that there is a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” Id. at 58-59, 106 S.Ct. 366. Accord Ex parte Moussazadeh, 361 S.W.3d 684, 691 (Tex. Crim. App. 2012) (“Moussazadeh III”).2

Because Applicant’s claim meets the Hill formula, and' Hill predated the finality of his conviction, we grant relief. The judgment is vacated, and. Applicant is remanded to the custody of the Bell County Sheriff to answer the charges set out in the indictment.

filed a concurring opinion in which Hervey, J., joined.

The Court says that this case is controlled by Hill v. Lockhart,1 but I read Hill differently. In Hill, the Supreme Court explicitly declined to decide whether erroneous advice about parole eligibility could ever be deemed constitutionally ineffective assistance:

In the present case the claimed error of counsel is erroneous advice as to eligibility for parole under the sentence agreed to in the plea bargain. We find it unnecessary to determine whether there may be circumstances under which erroneous advice by counsel as to parole eligibility may be deemed constitutionally ineffective assistance of counsel, because in the present case we conclude that petitioner’s allegations are insufficient to satisfy the Strickland v. Washington requirement of “prejudice.” Petitioner did not allege in his habeas petition that, had counsel correctly informed him about his parole eligibility date, he would have pleaded not guilty and insisted on going to trial.2

The only thing Hill decided was that, even if erroneous advice about parole eligibility could ever be deemed ineffective assistance, the advice in the case before the Court failed to satisfy the Strickland standard.3 But the Supreme Court did not decide that erroneous advice about parole eligibility could ever be constitutionally ineffective assistance, and in a line of cases culminating in Moussazadeh II,4 we held that such erroneous advice could be deemed constitutionally ineffective assistance only when parole eligibility was an element of the plea bargain.5 Because there is no binding Supreme Court precedent on whether erroneous advice about parole eligibility can ever be deemed ineffective assistance, the newness of the rule in Moussazadeh III6 must be judged by this Court’s own precedent.

Nevertheless, I concur in the result because I believe that, despite our general adherence to Teague,7 several factors weigh in favor of retroactivity. First, the rule in Moussazadeh III was once the old rule, and we have come full circle.8 It seems more appropriate to accord retroactive status to a new rule that once was the rule than to a new rule that is truly new, in the sense that it has never been the rule before. In addition, the type of claim before us is one that is generally raised on collateral review, and the new rule here was announced on collateral review. It is unnecessary to decide whether any of these factors alone would be sufficient to accord retroactive effect to a new rule.9 The combination of these factors is,'in my judgment, sufficient to accord retroactive effect here.

With these comments, I concur in the Court’s judgment.