We withdraw our prior opinion and hold that a defendant meets the prejudice prong of his ineffective assistance of counsel claim by demonstrating that he would have opted for a jury if his attorney had correctly advised him that he was ineligible for probation from the trial court. He does not have to show that the likely outcome of the jury trial he waived would have been more favorable than the court trial he had. We base our holding on a line of cases stretching from Hill v. Lockhart , 474 U.S. 52 , 59, 106 S.Ct. 366 , 88 L.Ed.2d 203 (1985), to Lee v. United States , --- U.S. ----, 137 S.Ct. 1958 , 198 L.Ed.2d 476 (2017), and our opinion in State v. Recer , 815 S.W.2d 730 , 731-32 (Tex. Crim. App. 1991). We disavow our contrary holding in Riley v. State , 378 S.W.3d 453 , 458 (Tex. Crim. App. 2012), to the extent that it conflicts with this opinion. We express no opinion about whether Appellant has demonstrated a reasonable probability that he would have opted for a jury if his attorney had correctly advised him about his probation eligibility but remand to the court of appeals to address that issue.

Background

Appellant was charged with aggravated sexual assault of a child and indecency with a child alleged to have occurred in July 2001. TEX. PENAL CODE ANN. §§ 21.11, 22.021 (West 1999). If convicted of either of those crimes he was not eligible for probation from a judge. TEX. CODE CRIM. PROC. ANN. ART. 42.12 § 3g(a)(1) (West 2001). But his attorney advised him otherwise. Appellant waived a jury and pleaded not guilty. After a bench trial, the judge found him guilty of both charges and sentenced him to 22 years in prison for aggravated sexual assault and ten years for indecency.

Appellant sought a new trial on grounds of ineffective assistance of counsel for the bad advice about his probation eligibility. The trial court denied the motion, and the court of appeals affirmed the ruling. Miller v. State , 2015 WL 3456783 (Tex. App.-Dallas, June 1, 2015) (mem. op., not designated for publication). We granted Appellant's petition for discretionary review which asked whether the lower court erred in finding that the deficient performance was not prejudicial under *499 Strickland v. Washington , 466 U.S. 668 , 104 S.Ct. 2052 , 80 L.Ed.2d 674 (1984).

Ineffective Assistance of Counsel

To prevail on a claim of ineffective assistance of counsel, a defendant must demonstrate two things: deficient performance and prejudice. Hill , 474 U.S. at 59 , 106 S.Ct. 366 ; Strickland , 466 U.S. at 694 , 104 S.Ct. 2052 . The question here is how to measure prejudice when the attorney's deficient performance-bad advice about probation eligibility-could not have affected the outcome of the defendant's trial but could only have affected the defendant's decision to waive a jury. Although the Hill opinion characterized itself as an application of Strickland , its prejudice focus differed enough from Strickland 's to present us with a choice between two standards: Strickland 's different outcome standard and Hill 's waived proceeding standard. For the reasons given below, we conclude that the choice between them depends on the possible result of the alleged deficient performance. In this case, since the attorney's wrong advice about probation could only have affected the defendant's choice to waive a jury, the correct measure of prejudice is Hill 's waived proceeding standard.

Strickland held that the measure of prejudice for IAC claims is whether the deficient performance might have affected the outcome of the proceeding that the defendant had. Strickland , 466 U.S. at 694 , 104 S.Ct. 2052 . The "ultimate focus" of the Strickland prejudice standard is "the fundamental fairness of the proceeding whose result is being challenged." Id. at 696 , 104 S.Ct. 2052 . This requires the reviewing court to examine "the totality of the evidence before the judge or jury" and ask whether "the result of the particular proceeding is unreliable because of a breakdown in the adversarial process that our system counts on to produce just results." Id. at 695-96 , 104 S.Ct. 2052 . If the deficient performance might have affected a guilty verdict, "the question is whether there is a reasonable probability that, absent the errors, the factfinder would have had a reasonable doubt respecting guilt." Id. at 695 , 104 S.Ct. 2052 . If the deficient performance might have affected a punishment verdict, the prejudice issue is whether there is a reasonable probability that, absent the errors, the sentencer would have assessed a more lenient punishment. See id. (in a death penalty case, "the question is whether there is a reasonable probability that, absent the errors, the sentencer ... would have concluded that the balance of aggravating and mitigating circumstances did not warrant death.").

On the other hand, if the deficient performance might have caused the waiver of a proceeding, the defendant's burden is to demonstrate a reasonable probability that the deficient performance caused the defendant to waive a judicial proceeding that he was otherwise entitled to have. Lee , 137 S.Ct. at 1965 ; Roe v. Flores-Ortega , 528 U.S. 470 , 484, 120 S.Ct. 1029 , 145 L.Ed.2d 985 (2000) ; Hill , 474 U.S. at 59 , 106 S.Ct. 366 . The focus then is on the defendant's decision making. Lee , 137 S.Ct. at 1966 . In Lee , where the attorney wrongly advised the defendant that he would not be deported if he pleaded guilty, 137 S.Ct. at 1962 , the error "affected Lee's understanding of the consequences" of his plea, so the prejudice issue was "whether there was an adequate showing that the defendant, properly advised, would have opted to go to trial." Id. at 1965 . In Flores-Ortega , where the defendant waived his right to appeal, the prejudice issue was whether he would have appealed but for his attorney's bad advice. 528 U.S. at 484 , 120 S.Ct. 1029 . In Hill , where the defendant claimed his attorney gave him *500 bad advice about parole, 474 U.S. at 56 , 106 S.Ct. 366 , the prejudice issue was whether the defendant still would have pleaded guilty if his attorney had correctly advised him about parole. Id. at 59 , 106 S.Ct. 366 .

The likelihood of a better outcome from a waived or forfeited proceeding is not the correct prejudice standard because "we cannot accord any 'presumption of reliability' to judicial proceedings that never took place." Flores-Ortega , 528 U.S. at 483 , 120 S.Ct. 1029 (citing Smith v. Robbins , 528 U.S. 259 , 286, 120 S.Ct. 746 , 145 L.Ed.2d 756 (2000) ). In those circumstances the different outcome question may be relevant to the extent that it sheds light on whether the deficient performance really did affect the defendant's decision making, but it is not the measure of prejudice. See Lee , 137 S.Ct. at 1966 (defendant's decision to plead guilty "may not turn solely on the likelihood of conviction after trial"); Flores-Ortega , 528 U.S. at 486, 120 S.Ct. 1029 (possible merits of the forfeited appeal "may give weight to the contention that the defendant would have appealed," but they are not required to "satisfy the prejudice requirement where there are other substantial reasons to believe that he would have appealed"); Hill , 474 U.S. at 59 , 106 S.Ct. 366 (different outcome question relevant only to the extent it impacted decision to plead guilty). A prediction about the likely outcome of a trial is unnecessary and inappropriate in cases where the deficient performance is not "pertinent to a trial outcome, but is instead alleged to have affected a defendant's understanding of the consequences of his guilty plea." Lee , 137 S.Ct. at 1967, fn. 3 .

The dissenting opinions argue for application of Strickland 's different outcome standard because Appellant had a court trial. They maintain that Hill 's forfeited proceeding standard only applies to guilty pleas. But Strickland was a guilty plea, and Hill applied in Lafler v. Cooper , 566 U.S. 156 , 164, 132 S.Ct. 1376 , 182 L.Ed.2d 398 (2012), where the defendant had a jury trial. "[I]t is insufficient simply to point to the guarantee of a fair trial as a backstop that inoculates any errors in the pretrial process." Id. at 170 , 132 S.Ct. 1376 . Thus, choosing between Strickland 's different outcome standard and Hill 's forfeited proceeding standard does not depend on the simple guilty plea vs. trial formula.

Even if the different outcome standard did apply here, the dissenting opinions misread Strickland in two ways.

First, the dissent would ask whether a proceeding the defendant did not have would have been more favorable without the deficient performance; but Strickland asked whether the proceeding the defendant did have would have been more favorable without the deficient performance. 466 U.S. at 694 , 104 S.Ct. 2052 . Consistent with that approach, Weaver v. Massachusetts , --- U.S. ----, 137 S.Ct. 1899 , 1911, 198 L.Ed.2d 420 (2017), an IAC claim based on the attorney's failure to object to a closed courtroom, required a reasonable probability of a different outcome in the trial the defendant had or a showing that the trial he had was fundamentally unfair; Weaver did not demand proof that a better outcome would have resulted from a hypothetical trial that was not had.

A majority of the Supreme Court has explicitly refused to evaluate prejudice based on the likely outcomes of proceedings not had. See Lee , 137 S.Ct. at 1965 (when an attorney's deficient performance caused the defendant to accept a guilty plea instead of pursue a trial, "we do not ask whether, had he gone to trial, the result of that trial 'would have been different' "); Flores-Ortega , 528 U.S. at 483, 120 S.Ct. 1029 (the likelihood of a better outcome from a forfeited proceeding is not the *501 correct prejudice standard because such a proceeding enjoys no presumption of reliability). Such speculation is like asking "what might have occurred in an alternate universe." See United States v. Gonzalez-Lopez , 548 U.S. 140 , 150, 126 S.Ct. 2557 , 165 L.Ed.2d 409 (2006) (holding that proof of likelihood of a better outcome if a different attorney had represented the defendant was not required in case of denial of counsel of choice).

Second, the dissent misapplies Strickland by requiring a reasonable likelihood of a specifically better outcome, i.e., probation. Strickland required only a generically better outcome. 466 U.S. at 694 , 104 S.Ct. 2052 . Since "any amount of [additional] jail time has Sixth Amendment significance[,]" Cooper , 566 U.S. at 165 , 132 S.Ct. 1376 , quoting Glover v. United States , 531 U.S. 198 , 203, 121 S.Ct. 696 , 148 L.Ed.2d 604 (2001), requiring a likelihood of probation instead of merely a better outcome would impose an unconstitutional burden on the prejudice prong of ineffective assistance claims.

Recer vs. Riley

Although there is no Supreme Court authority for assessing prejudice by asking what might have happened in a trial that was never had, our own precedent has been inconsistent on this point. The competing views are presented by Recer and Riley .

Recer , 815 S.W.2d 730 , focused on the defendant's decision making in assessing prejudice from an attorney's bad advice about probation eligibility. To prevail on her claim of ineffective assistance Recer had to demonstrate that she was eligible for probation from the jury, her attorney's advice to elect the court for punishment was not based on a valid strategy, her decision to elect the judge for punishment was based on her attorney's bad advice, and she would have elected the jury for punishment but for her attorney's error. Id. at 731-32 . She did not have to prove a likelihood of a different outcome from a hypothetical jury. Id.

We later purported to follow Recer in Riley , 378 S.W.3d 453 , but we tacked on an additional requirement: proof that "the results of the proceeding [not had] would have been different had [the defendant's] attorney correctly informed him of the law." 378 S.W.3d at 458 (citing Recer , 815 S.W.2d at 731-32 ). We imposed that additional requirement without any supporting authority or rationale. 1

Stare decisis would not require adherence to Riley because proving a better outcome from a proceeding never had is so speculative as to be unworkable. See Paulson v. State , 28 S.W.3d 570 , 571-72 (Tex. Crim. App. 2000) (goals of stare decisis not advanced by adhering to unworkable precedent). Additionally, Riley 's own violation of stare decisis by silently overruling Recer undermines its precedential value. See Paulson, 28 S.W.3d at 574 (Keller, J., concurring) (in deciding whether to overrule precedent "we should take into account whether the decision under consideration was itself a violation of the principles of stare decisis "). But regardless of its unworkability and its failure to adhere to precedent, we cannot follow Riley because we are bound by the rulings of the United States Supreme Court on this topic.

*502 The highest authority on federal constitutional law is the United States Supreme Court. Ex parte Evans , 537 S.W.3d 109 , 111 (Tex. Crim. App. 2017), citing 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). Its pronouncements about federal constitutional law are binding on this Court. Evans , 537 S.W.3d at 111 , citing 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). Claims of ineffective assistance of counsel are grounded in the federal constitution. Evans , 537 S.W.3d at 111 , citing U.S. CONST. amend. VI. Therefore, we are bound by the Supreme Court's opinions about ineffective assistance of counsel.

Hill , Flores-Ortega and Lee hold that the correct measure of prejudice for an attorney's deficient performance that might have caused a defendant to waive a judicial proceeding is whether there is a reasonable likelihood that the defendant would have opted for the proceeding if his attorney had performed adequately. In the context of this case, Appellant would have to demonstrate a reasonable likelihood that he would have opted for a jury if his attorney had correctly advised him about his probation eligibility from the trial court. He would not have to demonstrate a reasonable likelihood that the jury trial he waived would have yielded a more favorable result than the court trial he had.

Conclusion

We overrule Riley to the extent that it conflicts with the Hill , Flores-Ortega and Lee line of cases, and we reaffirm our Recer opinion. We remand this case to the court of appeals for analysis of prejudice consistent with this opinion. 2

Keel, J., delivered the opinion of the Court in which Hervey, Richardson, Yeary, Newell and Walker, JJ., joined. Newell, J., filed a concurring opinion in which Walker, J., joined. Keller, P.J., filed a dissenting opinion. Alcala, J., filed a dissenting opinion in which Keller, P.J., and Keasler, J., joined.

Newell, J., filed a concurring opinion in which Walker J., joined.

Appellant complains that he waived his right to a jury based upon bad advice from his attorney regarding his eligibility for probation. This resulted in his ineligibility to receive probation at punishment. There is no disagreement that this amounted to deficient performance. The disagreement is about what an Appellant must do to establish that he was prejudiced by his attorney's performance. I agree with the Court that Appellant need not prove that he would have actually received probation. He need not prove that receiving probation was even a realistic possibility. He must prove that his attorney's bad advice resulted in making a decision to waive rights or procedures he would not have otherwise waived had he been given correct advice. That's why I join the Court's opinion. I *503 write separately to explain why the dissents convinced me to do so.

At the outset, I disagree with framing the Court's opinion as a great expansion of the Sixth Amendment right to effective assistance of counsel rather than a faithful application of Supreme Court precedent. The Court does not cut a new standard out of whole cloth to evaluate Appellant's claims. Rather, the Court decides that the standard set out by the Supreme Court in Hill v. Lockhart -focusing upon how the deficient representation affected the defendant's decision-making-is best suited for addressing the type of ineffective assistance complaint lodged in this case. 1 But that standard has always been a part of the traditional Strickland v. Washington standard. 2

Rather, drawing a distinction regarding the standard of prejudice based upon whether there was "an entire trial proceeding" would greatly diminish the right to effective assistance of counsel and undermine the purpose behind both Strickland and Hill . When the Supreme Court decided Hill v. Lockhart it applied the Strickland standard to a completely different type of proceeding than the one in Strickland . 3 The focus was protecting a defendant's Sixth Amendment right to the effective assistance of counsel not the reliability of a particular type of proceeding. 4 Later, the Court made clear in Roe v. Flores-Ortega , that prejudice results from the denial of the "entire judicial proceeding" to which a defendant is entitled. 5 Yet requiring proof that a defendant would have actually received probation had he received proper advice places the emphasis upon the proceeding rather than the effectiveness of the representation. 6 The Supreme Court rejected this approach in Lafler v. Cooper . 7

There, the defendant chose to reject a plea bargain offer based upon the erroneous advice of counsel and proceed to trial. 8 The federal district court granted relief and the Sixth Circuit Court of Appeals affirmed, which lead to the State petitioning the Supreme Court to hold that relief was inappropriate. 9 The Court summarized the State's arguments as simply claiming that the defendant received a fair trial so he could not complain that counsel's advice was deficient.

*504 In the end, petitioner's three arguments amount to one general contention: A fair trial wipes clean any deficient performance by defense counsel during plea bargaining. 10

The Court rejected this argument, noting that it is insufficient to simply point to the guarantee of a fair trial as a backstop that inoculates any errors in the pre-trial process. 11 Yet, holding that Appellant was not prejudiced in this case because he received an entire trial proceeding adopts the same argument rejected by the Supreme Court.

It is true that the Supreme Court wrote in Lafler that a defendant must show "that the conviction or sentence, or both, under the offer's terms would have been less severe than under the judgment and sentence that in fact were imposed." 12 But that observation must be considered in the context of that case. The Supreme Court relied upon a record that showed Cooper had received a sentence that was 3 ½ times greater than he would have received under the plea. 13 That holding does not translate to a requirement that Appellant prove he actually would have received probation. Here, a lack of a probated sentence does not tell us anything because the trial court was not statutorily authorized to award such a sentence. In this case, Appellant has proven that he would have had a better outcome, at least as to punishment, had he not waived his right to a jury because it would have at least provided him with an opportunity he absolutely did not have under the proceeding he received.

This understanding is reinforced by the Supreme Court's recent decision in Lee v. United States . 14 There, the defendant was charged with possession of ecstasy with intent to distribute, and, as the Court described it, the defendant's "prospects of acquittal at trial were grim." 15 However, his primary concern was not how much time he might serve if found guilty. Rather, he was concerned that a conviction would result in deportation. His attorney incorrectly advised him that he would not be deported if he took the plea deal. He filed a motion to vacate his conviction and sentence arguing that his attorney provided constitutionally ineffective assistance. The Government argued that the defendant had not shown prejudice because he had "no viable defense at trial, he would almost certainly have lost and found himself still subject to deportation, with a lengthier prison sentence to boot." 16 According to the Government, the defendant's only hope at trial was that "something unexpected and unpredictable might occur that would lead to an acquittal." 17

The Supreme Court rejected the Government's argument because deportation was the determinative issue in the defendant's decision to accept the plea deal. 18 According to the Court, the defendant faced a choice between the certainty of being deported by accepting a plea and the "almost" certainty of being deported by risking a trial. That "almost" made all the *505 difference. 19

Appellant was faced with an analogous choice in this case. Though the issue was not deportation, Appellant faced a choice between the certainty that the judge could not award him probation and the almost certainty that a jury would not award him probation. As in Lee , that "almost" is enough to establish prejudice. As Professors Dix and Schmolesky have observed of our original opinion in this case, "In light of Lee , the plurality in Miller is apparently wrong." 20 I agree. And it is better to fix the error now on rehearing. That is why I join the Court.

Keller, P.J., filed a dissenting opinion.

The question before us is how to determine prejudice when counsel tells a defendant that he could get probation in a bench trial when, in fact, he could get probation only in a jury trial. The usual test for prejudice, articulated in Strickland v. Washington , is whether there is a reasonable probability that the result of the proceeding would have been different. 1 Applying this usual test would require asking whether the defendant would have actually obtained probation in a jury trial. However, the Court holds that a different prejudice standard applies, requiring a showing merely that the defendant would have opted for a jury trial. In support of this holding, the Court cites cases involving plea agreements and the failure to file an appeal. 2 Each of those cases involved the deprivation of an entire judicial proceeding: the deprivation of a trial or the deprivation of an appeal. 3

This case is different. Appellant pled "not guilty." He had a trial. It was a trial before a judge instead of a trial before a jury-arguably, the wrong kind of trial. But that means the cases cited by the Court are not directly on point. Instead, those cases are, at most, analogous to the present case, and so the question is whether we should treat a trial in which the defendant is deprived of a jury like the entire deprivation of a trial. In answering that question, an instructive case to examine is Weaver v. Massachusetts. 4

Weaver involved a claim that the defendant received the wrong kind of trial: a trial that was not open to the public. 5 The deprivation of a public trial is structural *506 error. 6 The Supreme Court addressed whether the fact that the error was structural changed the way that a prejudice analysis in an ineffective-assistance claim was to be conducted. 7 The Court explained that "structural error" is the kind of error that, if preserved, results in reversal without a harm analysis. 8 The Court concluded, however, that the structural nature of an error did not necessarily dictate a deviation from the usual Strickland prejudice analysis. 9 The Court assumed, but did not decide, that the structural nature of an error might affect the Strickland prejudice analysis if one of the rationales for labeling a particular error as structural was that it rendered a trial fundamentally unfair and if that rationale were shown to be applicable in the defendant's case. 10 The Court ultimately concluded that the defendant had not made such a showing, and because the defendant had not met the traditional test for prejudice-showing a reasonable probability of a different outcome-the defendant was not entitled to relief. 11

Whether the underlying jury-trial violation in this case is a structural error is questionable, at best. The Supreme Court has recognized only "a very limited class of errors" as structural. 12 The constitutional right to a jury trial does not appear in the lists of structural error recited by the Supreme Court. 13 When a violation of the constitutional right to a jury trial occurs because an element of the offense has not been submitted to the jury, a harm analysis applies, 14 and, significantly, that harm analysis may appropriately focus on whether the jury would have found the element if it had been submitted. 15

*507 Moreover, it is not clear that the constitutional right to a jury trial is implicated in this case. Appellant's claim is that he should have gotten a jury trial so that he would have a chance at getting probation. In noncapital cases, the federal constitutional right to a jury trial does not extend to the discretionary determination of what punishment to assess. 16 The right to a jury trial on the issue of probation is based solely in statute, 17 and a statutory error cannot be structural. 18

And Weaver illustrates that, even if the deprivation of a jury trial that was alleged to have occurred here were structural, the usual prejudice analysis under Strickland might still be appropriate. And even if a deviation from the usual prejudice analysis were warranted, it might be warranted only for punishment-what the defendant says he wanted the jury for.

I believe that the Court's holding today is an unnecessary expansion of the Supreme Court holdings on which it relies. I respectfully dissent, and I also join Judge Alcala's dissent.

Alcala, J., filed a dissenting opinion in which Keller, P.J., and Keasler, J., joined.

Should this Court greatly expand upon Supreme Court precedent describing the law for claims of ineffective assistance of counsel so as to permit a defendant to obtain a new trial when an exhaustive review of the evidence introduced in the guilt and punishment phases of his trial shows that the outcome of his case was not prejudiced by his attorney's complained-of bad advice? I respectfully disagree with this Court's majority opinion that appears to answer this question "yes." Rather than expanding a defendant's federal constitutional right to effective counsel in a manner that goes beyond the parameters set forth by the Supreme Court, I would apply that Court's precedent in Strickland v. Washington , 466 U.S. 668 , 104 S.Ct. 2052 , 80 L.Ed.2d 674 (1984). Applying that precedent to this case, I would affirm the court of appeals' judgment against Arthur Franklin Miller, Jr., appellant, upholding his convictions for aggravated sexual assault of a child and indecency with a child, for which he received sentences of twenty-two years' imprisonment and ten years' imprisonment, respectively. The court of appeals' judgment affirmed the trial court's ruling denying appellant's motion for new trial, in which he had presented a claim of ineffective assistance of trial counsel. Appellant's claim contended that his trial attorney misadvised him that the trial court could consider a probated sentence if it convicted him at a court trial. The State concedes that trial counsel performed deficiently in that respect and that appellant has established the first of the two required Strickland prongs for establishing an ineffective-assistance claim. The sole question in this appeal, therefore, is whether appellant has proven the second Strickland prong requiring him to show that he was prejudiced by counsel's bad *508 advice. This Court's majority opinion agrees with appellant that, under these circumstances, courts should limit the prejudice analysis solely to a consideration of the defendant's decision-making process-that is, courts should consider only whether the defendant would have waived his right to a jury trial in favor of a court trial if he had been properly advised about his ineligibility for a probated sentence after a court trial. I, however, conclude that when, as here, a defendant's waiver of a jury trial resulted in a court trial rather than in a waiver of the entire trial proceedings, then appellate courts must consider whether it is reasonably probable that counsel's erroneous advice affected the outcome of the proceedings as a whole. Here, therefore, the relevant question before the court of appeals was whether the trial court properly rejected appellant's ineffective-assistance claim because, regardless of trial counsel's bad advice about appellant's probation ineligibility after a court trial, the probable outcome after a jury trial would have been the near equivalent of the actual outcome of the court trial that was had. Because appellant has not demonstrated prejudice under the applicable law set forth by the Supreme Court, I would affirm the court of appeals' judgment. 1

I. Background

Because it involves an entire guilt and punishment trial on the merits and is at the rehearing stage in this Court, this case's background is lengthy. I examine its background by discussing the trial on the merits, the motion for new trial, the court of appeals' opinion, and this Court's plurality opinion on original submission of the petition for discretionary review, including the grounds for rehearing that decision.

A. Trial on the Merits

Appellant is the grandfather of A.M. and the step-grandfather of K.F. The State charged appellant with indecency with a child and aggravated sexual assault of A.M. and sexual offenses against K.F. See TEX. PENAL CODE §§ 21.11, 22.021. Appellant pleaded not guilty to the charges, and the case was called for trial.

Before the trial began, the trial court admonished appellant about his rights and determined that appellant was freely and voluntarily waiving his right to a jury trial and opting to proceed with a bench trial. In making that determination, the trial court discussed with appellant the fact that he had a right to a jury trial, that a jury could consider a probated sentence in his case, and that no one had promised him anything in exchange for his waiver of a *509 jury. Although he did specifically advise appellant that a jury could consider a probated sentence, the trial judge did not inform appellant that a consequence of a court trial rather than a jury trial was that the judge could not consider that type of sentence under the applicable law. The judge specifically confirmed with appellant that "no one has forced [him] or threatened [him] or promised [him] anything or done anything" to induce him to waive a jury trial.

At the court trial, each party gave opening statements and closing arguments, and eight witnesses testified, four for each side. The State called the investigating peace officer and the forensic interviewer, each of whom discussed certain evidence that either corroborated or was consistent with the allegations of abuse. The State's remaining witnesses were the two complainants. A.M., who was twenty-two years old by the time of trial, testified that, when she was around ten years old, appellant engaged in inappropriate behavior that "eventually led to sexual abuse." Appellant would make "awkward comments that [she] was uncomfortable with," try to undo her bathing suit top, and try to watch her change clothes and shower. A.M. testified that appellant would take her for a ride on his two-seat jet ski and refuse to take her back until she kissed him on the lips. She testified that appellant would awaken her at night and ask her to sit on his lap, show her pornography, and rub and penetrate her vagina with his fingers. A.M. testified that this occurred on more than one occasion. A.M. explained that she was afraid to report the abuse because appellant led her to believe that he would stop financially supporting her father if she told anyone and that "nobody would believe [her], and everybody would hate [her]." K.F. testified that, when she was approximately four years old, appellant forced her to touch his penis when she sat on his lap to play a computer game.

Appellant's defense witnesses included the defense investigator, who presented evidence that contradicted the complainants' testimony about the timeline of abuse. Appellant's three remaining witnesses were appellant's and the complainants' relatives. They disagreed with certain details regarding the complainants' accounts of the abuse. Furthermore, appellant's daughter additionally testified that she believed the allegations were instigated by K.F.'s mother and that K.F. had essentially indicated that her allegation against appellant was not true. After closing arguments, the trial court found appellant guilty of the charges for the conduct against A.M. and not guilty of the charges pertaining to K.F.

At the punishment phase of trial, the parties proceeded to make arguments as if appellant could be considered for community supervision by the trial court. A.M. was the sole witness called by the State to provide substantive testimony. A.M. testified that she felt that, as a result of the abuse, her "entire life[ ] has been violated," that she suffered from depression and trust issues from the abuse, and that she "lost [her] family" by coming forward. When asked whether she believed that appellant deserved probation, she responded that he did not because "he took things away from me that can never be replaced."

The parties' opposing closing arguments discussed whether appellant was a suitable candidate for community supervision. On the one hand, the State argued that community supervision was inappropriate due to the severe impact the abuse had on A.M. On the other hand, trial counsel argued that a probated sentence in addition to counseling and therapy was appropriate because appellant had no prior convictions for felony offenses, and the absence of any *510 claims of abuse by appellant in recent years suggested that he had changed. Additionally, trial counsel noted that, because appellant was seventy-eight years old and in "poor health," any prison sentence would functionally be a life sentence. Counsel also observed that appellant would be required to register as a sex offender. Without commenting about any rationale for declining to impose the community supervision that had been requested by trial counsel, the trial court sentenced appellant to twenty-two years' imprisonment for the aggravated sexual assault and ten years' imprisonment for the indecency with a child by contact.

B. Motion for New Trial

After receiving new counsel for purposes of appeal, appellant filed a written motion for new trial alleging ineffective assistance of counsel on the basis that trial counsel had erroneously promised him that he would receive probation if he was found guilty by the trial court. This advice was erroneous because, under the former law that applied to appellant's offenses that were committed in 2001, only a jury could recommend a probated sentence if he was found guilty of the offenses. See TEX. CODE CRIM. PROC. ANN. Art. 42.12 §§ 3(g), 4, 5 (West 2006). 2 Appellant argued that he would have elected a jury trial rather than a bench trial had trial counsel properly advised him that the trial court could not award a probated sentence. In support of the motion, appellant included affidavits from himself and two of his sons. Trial counsel did not submit an affidavit in response to the allegations.

The trial court then held an evidentiary hearing at which four witnesses testified. Appellant and several of his family members testified that trial counsel had repeatedly urged him to waive a jury trial and proceed with a bench trial because, if the trial judge found him guilty, the judge would give him probation due to his elderly age, lack of criminal history, and the weakness of the State's case. The trial court denied appellant's motion for new trial. Although he did not make written findings of fact and conclusions of law, the trial judge did make oral findings and conclusions on the record. Following the trial court's ruling, appellant timely appealed. 3

*511 C. Intermediate Appeal and Petition for Discretionary Review

On appeal, the court of appeals upheld the trial court's ruling denying appellant's motion for new trial. Miller v. State , No. 05-14-01065-CR, 2015 WL 3456783 , at *1 (Tex. App.-Dallas June 1, 2015) (mem. op., not designated for publication). In its arguments to that court, the State conceded that trial counsel's performance objectively fell below reasonable professional norms and thus constituted deficient performance, satisfying the first prong under Strickland . Id. at *2. The court of appeals' analysis, therefore, turned solely on whether counsel's deficient performance had prejudiced appellant under the second prong of Strickland . Id . at *3. The court of appeals' description of the issue was "whether there was a 'reasonable probability' that the result of the proceeding would have been different if appellant's attorney had given him correct advice; that is, whether a jury would have sentenced appellant to probation." Id. (citing Riley v. State , 378 S.W.3d 453 , 458 (Tex. Crim. App. 2012) ). The court of appeals determined that appellant had failed to show that he was prejudiced by counsel's erroneous advice because (1) the trial judge could have disbelieved appellant's assertion that he would not have waived a jury if he had known that the judge was unable to assess probated sentences, and (2) appellant had failed to show a reasonable probability that a jury would have awarded him the probated sentences. Id. at *4-5.

On appellant's petition for discretionary review challenging the standard employed by the court of appeals to assess whether he was prejudiced by counsel's error, a plurality of this Court affirmed the judgment of the court of appeals. 4 In analyzing the proper standard for evaluating prejudice, we considered the articulations of the prejudice inquiry that have been advanced for two distinct sets of circumstances. We observed that, in the first circumstance in which there has been an entire trial or appellate proceeding, the proper inquiry is whether counsel's deficient performance during that proceeding undermined the confidence in the outcome of the proceeding. 5 In the second circumstance in which a defendant has entered a plea of guilty or no contest and thus no trial has occurred, we observed that the prejudice inquiry focuses on whether counsel's deficient performance during the plea proceedings resulted in a defendant's decision to completely forgo a trial or appellate proceeding. 6 A plurality of this Court determined that in appellant's case, because there had been an entire trial proceeding for which the confidence in its outcome could be evaluated, we would apply the ordinary Strickland prejudice standard that asks whether, but for counsel's error, there was a reasonable probability of a different outcome. We upheld *512 the court of appeals' affirmance of the trial court's denial of appellant's motion for new trial.

Shortly after this Court issued its original opinion in this case, the U.S. Supreme Court decided Lee v. United States , --- U.S. ----, 137 S.Ct. 1958 , 198 L.Ed.2d 476 (2017), and this Court granted appellant's motion for rehearing to determine if Lee undermined our reasoning. A majority of this Court concludes that it does. As I discuss more fully in the next section, I respectfully disagree with this Court's characterization of Lee as having any impact on the instant case, and I continue to believe that this case was correctly decided by the plurality opinion issued on original submission of appellant's petition for discretionary review.

II. The Proper Prejudice Standard for Ineffective Assistance Claims

As the discussion below shows, the Supreme Court has recognized that the prejudice prong for establishing ineffective assistance of counsel varies depending on whether a defendant opted for a trial proceeding or instead waived a trial and decided to enter a plea of guilty or no contest to a charge. In the first situation, the prejudice inquiry focuses on the probable outcome of the proceedings in the absence of counsel's error, whereas in the latter situation, the inquiry focuses on whether, but for counsel's error, the defendant would not have pleaded guilty and would have instead pursued a trial. Here, I explain that, because there was a complete trial proceeding, it is appropriate to examine how appellant's attorney's deficient performance affected the outcome of that proceeding, rather than more narrowly considering only the effect of counsel's error on appellant's decision to waive a jury.

A. Applicable Law for Establishing Prejudice Depends on Whether There was a Trial or a Waiver of an Entire Trial Proceeding

The law applicable to the prejudice prong can be divided into two spheres, one of which includes jury and court trial proceedings, and the other of which includes guilty and no-contest pleas that forgo trial proceedings. I discuss the law applicable to these two spheres before explaining why the prejudice standard in the instant case should fall within the first sphere.

1. Sphere One: Prejudice in a Court or Jury Trial

Under Strickland , in cases asserting ineffective assistance of counsel based on claims that counsel performed deficiently during either the guilt or punishment phases of trial, courts must apply the prejudice standard that focuses on whether there is a reasonable probability that the result of the proceedings would have been different but for the attorney's deficient performance. See Strickland , 466 U.S. at 694 , 104 S.Ct. 2052 . The Strickland two-prong standard to establish an ineffective-assistance claim requires a defendant to show that, under the totality of the evidence, (1) counsel's performance fell below an objective standard of reasonableness under prevailing professional norms, and (2) but for such deficient performance, a reasonable probability exists that the result of the proceeding would have been different. Id. at 688, 694-95 , 104 S.Ct. 2052 . The Court explained that a "reasonable probability is a probability sufficient to undermine confidence in the outcome." Id. at 694 , 104 S.Ct. 2052 .

With respect to the required prejudice showing in situations involving a trial, the Strickland test acknowledges the "strong presumption of reliability" that courts "normally apply ... to judicial proceedings[.]" Smith v. Robbins , 528 U.S. 259 , 286, 120 S.Ct. 746 , 145 L.Ed.2d 756 (2000)

*513 (citing Strickland , 466 U.S. at 696 , 104 S.Ct. 2052 ). To establish prejudice, a defendant must prove actual prejudice; theoretical prejudice is not enough. Roe v. Flores-Ortega , 528 U.S. 470 , 482, 120 S.Ct. 1029 , 145 L.Ed.2d 985 (2000). To prove prejudice, a defendant must show how " 'specific errors of counsel undermined the reliability of the finding of guilt' " or the punishment assessed. Id. (quoting United States v. Cronic , 466 U.S. 648 , 659 n. 26, 104 S.Ct. 2039 , 80 L.Ed.2d 657 (1984) ). Furthermore, actual prejudice is considered under an objective standard rather than being based on speculation about the possibility of jury nullification or some other extreme event. Strickland , 466 U.S. at 694-95, 104 S.Ct. 2052 . The Strickland Court observed that the "assessment of prejudice should proceed on the assumption that the decisionmaker is reasonably, conscientiously, and impartially applying the standards that govern the decision." Id. at 695 , 104 S.Ct. 2052 .

The Strickland prejudice standard applicable to ordinary jury and court trials, therefore, considers whether there is a reasonable probability that, but for counsel's mistake, the defendant would have received a more favorable outcome either at the guilt or punishment phase of trial. But what about those cases in which a defendant pleads guilty or no contest and there are no court or jury trial proceedings to consider? As I discuss next, the Supreme Court has held that, in those cases, the prejudice standard looks to the waiver of the trial rather than the outcome of the proceedings.

2. Sphere Two: Prejudice in Guilty or No Contest Pleas Based on a Waiver of Trial

The Supreme Court in Hill v. Lockhart decided that, when a defendant waives an evidentiary trial and resolves his case with a plea of guilty or no contest, the Strickland prejudice prong must examine whether 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." Hill v. Lockhart , 474 U.S. 52 , 59, 106 S.Ct. 366 , 88 L.Ed.2d 203 (1985). Hill's attorney had misinformed him about his eligibility for parole under the sentence that he had agreed to in a plea bargain. Id. at 60 , 106 S.Ct. 366 . Because there were no court or jury trial proceedings to examine to determine whether Hill had been prejudiced by his attorney's deficient performance, the Hill Court considered the only proceedings that had actually occurred in the case: the trial waiver that had resulted in the plea of guilty. See id . at 59, 106 S.Ct. 366 . 7 In subsequent decisions, the Supreme Court has indicated that there can be no presumption of reliability in the absence of any factual trial proceedings. See, e.g., Flores-Ortega , 528 U.S. at 483, 120 S.Ct. 1029 ("Put simply, we cannot accord any 'presumption of reliability' to judicial proceedings that never took place.") (internal citation omitted). Thus, under circumstances in which a defendant has resolved his case by plea, the Supreme Court has *514 recognized that the relevant inquiry is whether counsel's deficiency resulted in the defendant being denied the "entire judicial proceeding itself[.]" Id.

B. This Case Falls Within the First Sphere

Applying Supreme Court precedent to this case, I conclude that for two reasons appellant's complaint presents a claim of ineffectiveness that is subject to an ordinary Strickland prejudice analysis that focuses on the outcome of the proceedings. 8 First, as with the ordinary Strickland prejudice standard, there was an entire trial that occurred that can be examined for its regularity and for any prejudicial impact caused by deficient performance. Given that there was an evidentiary trial in the instant case, this Court has an entire record of the proceedings that occurred so that we may discern whether the reasonably probable outcome of the proceedings would have been different with a jury trial that did not occur as compared to the bench trial that did occur. Such an approach would be consistent with the Supreme Court's observation in Strickland that, in order to warrant relief from a conviction or sentence due to counsel's errors in a trial proceeding, those errors must have been "so serious as to deprive the defendant of a fair trial, a trial whose result is reliable." Strickland , 466 U.S. at 687, 104 S.Ct. 2052 .

The Supreme Court's decision in Lafler v. Cooper supports the view that an ordinary Strickland prejudice outcome analysis is the appropriate framework for resolving appellant's claim. 566 U.S. 156 , 164, 132 S.Ct. 1376 , 182 L.Ed.2d 398 (2012). In Cooper , the Supreme Court considered the appropriate prejudice standard for a situation where a defendant has had an entire trial proceeding but complains that he would not have gone to trial if his attorney had properly conveyed the State's plea offer to him, which he claims he would have accepted rather than proceed to trial. Id. at 160 , 132 S.Ct. 1376 . The Court explained,

[H]ere the ineffective advice led not to an offer's acceptance but to its rejection. Having to stand trial, not choosing to waive it, is the prejudice alleged. In these circumstances a defendant must show that but for the ineffective advice of counsel there is a reasonable probability that the plea offer would have been presented to the court ( i.e., that the defendant would have accepted the plea and the prosecution would not have withdrawn it in light of intervening circumstances), that the court would have accepted its terms, and that the conviction or sentence, or both, under the offer's terms would have been less severe than under the judgment and sentence that in fact were imposed.

Id. at 163-64 , 132 S.Ct. 1376 . The prejudice standard discussed in Cooper considered the totality of the record, including the explanation that a defendant must show that the plea offer that was not conveyed to him was less severe than the judgment and sentence that was imposed against him at a trial. Id. Prejudice in *515 Cooper did not focus solely on whether the defendant would have accepted a plea offer had it been conveyed to him. See id. Similar to Cooper , the appropriate prejudice standard here properly considers the totality of the record, including whether it is probable that the judgment and sentence that a reasonable jury would impose would be less severe than what was in fact imposed by the trial judge at the court trial.

Second, the cases on which this Court's majority opinion relies are procedurally distinguishable because they pertain to errors by counsel that deprived a defendant of an entire judicial proceeding, such as in cases where the defendants pleaded guilty or no contest, as compared to here where appellant pleaded not guilty and was found guilty at an evidentiary trial. In the instant case, the critical difference is that appellant was not completely denied a proceeding, as the defendant was in those cases on which this Court's majority opinion relies: Hill v. Lockhart , Roe v. Flores-Ortega , and Lee v. United States .

In Hill , there was no trial to examine for the reliability of the proceeding because all that had occurred in that case was a guilty plea. But, here, as in the punishment-phase court trial in Strickland , there are trial proceedings whose outcome may be examined for its reliability.

Roe v. Flores-Ortega is similarly distinguishable. In that case, counsel failed to file a notice of appeal and thereby waived the defendant's right to an appeal without his consent. 528 U.S. at 484, 120 S.Ct. 1029 . Like the decision to plead guilty or proceed to trial, the decision to pursue an appeal belongs solely to a defendant. Id. at 477 , 120 S.Ct. 1029 . Because, as in Hill , no presumption of reliability could be afforded to an appeal that never took place, proving prejudice required showing that, absent counsel's deficient performance, the appeal would have been made. Id. at 483-84 , 120 S.Ct. 1029 .

Lee v. United States is also inapposite. In Lee , the Supreme Court addressed whether under the Hill consideration of the prejudice prong for situations in which a defendant has entirely waived a trial, a court should consider a defendant's likelihood of success at trial had he insisted on going to trial. Lee , 137 S.Ct. at 1966. Lee was an immigrant who was charged with narcotics offenses that, if convicted, would make him subject to deportation. Lee complained that counsel misadvised him that his convictions would not subject him to deportation. Based on the erroneous advice of counsel, Lee pleaded guilty to avoid deportation. As a result of counsel's error, Lee was both convicted and deportable. Although Lee had no viable defense at trial, and thus would have almost certainly been found guilty at trial and still subject to deportation, the Supreme Court held that the likelihood of success at trial alone does not govern the prejudice analysis under these circumstances where a defendant waived his entire right to a trial. Because Lee may have rationally concluded that even a minuscule chance at acquittal and avoiding deportation was preferable to deportation plus a conviction, the decision to proceed to trial was still the proper focus of the prejudice inquiry where the entire trial was waived. This Court's majority finds that Lee undermines the rationale of our original opinion because it indicates that, even if appellant had virtually no chance of receiving probation from a jury, appellant's decision between a jury trial and a bench trial is the relevant inquiry for prejudice. However, the reasoning of Lee is inapplicable here. As noted, Lee simply addresses what factors should be considered when determining if a defendant has shown that he "would have insisted on going to trial"

*516 under the Hill prejudice standard. But Hill 's prejudice inquiry is crafted for situations where the defendant was deprived of an entire judicial proceeding. Here, appellant did insist on going to trial and was not deprived of an entire judicial proceeding. Thus, appellant's case is not one where counsel's error forfeited an entire judicial proceeding but rather an error which calls into question the reliability of the outcome. In this situation, Strickland 's ordinary prejudice inquiry is appropriate to determine if appellant received constitutionally deficient representation.

In sum, the cases on which this Court's majority opinion relies are inapplicable because those cases involve waivers of the entire trial that "arguably led not to a judicial proceeding of disputed reliability, but rather to the forfeiture of a proceeding itself." Flores-Ortega , 528 U.S. at 483, 120 S.Ct. 1029 . Because here there was an evidentiary court trial at which appellant's claims were fully litigated before the trial judge, the prejudice determination must be made by considering the reliability of that judicial proceeding.

III. This Case Should Not Be Remanded to the Court of Appeals

This Court's majority opinion holds that the court of appeals erred by failing to properly consider whether appellant was prejudiced with respect to his decision to waive his right to a jury trial, and it remands the case to the court of appeals purportedly for it to consider that matter. I disagree with this Court's remand order for two reasons.

First, remand is unnecessary in this case because the court of appeals already answered the precise matter that this Court now asks it to consider on remand. In its opinion, the court of appeals stated, "The trial court was not required to accept appellant's claim that he would have acted differently had he received correct advice. And even if the trial court accepted this claim, appellant was also required to establish that correct advice would have changed the result of the proceeding." Miller , 2015 WL 3456783 , at *4. Given that the court of appeals has already determined that the trial court was within its discretion to deny the motion for new trial because the judge was not required to believe appellant's claim that he would have chosen a jury trial if he had been correctly advised, this Court's remand of that precise question to the court of appeals is a waste of judicial resources and an unnecessary delay to the resolution of this case.

Second, the trial judge's decision not to sentence appellant at the lower range of punishment strongly suggests that it is not reasonably probable that appellant would have been awarded a probated sentence if he had proceeded with the jury trial instead of waiving it. Appellant's sole complaint about his counsel's deficient performance is that the trial judge was prohibited from considering a probated sentence. But the jury most likely would not have awarded a probated sentence even though the law permitted them to consider it. When this Court has considered a defendant's ineffective assistance claim regarding the punishment phase of trial, the thrust of our analysis focuses on whether the attorney's error affected the sentence that was imposed in the case. See Ex parte Cash , 178 S.W.3d 816 , 818 (Tex. Crim. App. 2005) (citing Woodford v. Visciotti , 537 U.S. 19 , 22-23, 123 S.Ct. 357 , 154 L.Ed.2d 279 (2002) ) (noting that when it is alleged that counsel performed deficiently at the punishment phase of trial, defendant must prove that there is a reasonable probability that, but for counsel's errors, the sentencing jury would have *517 reached a more favorable penalty-phase verdict). In Ex parte Cash , under analogous circumstances in which the defendant claimed that his counsel's error had prevented the jury from considering probation, this Court observed that the jury's imposition of a forty-year sentence against Cash was a strong indicator that, even if it had been presented with the option of sentencing Cash to ten years or less and awarding him probation, the jury would not have exercised that option. Id. at 819.

As in Cash , it is not reasonably probable that a jury would have awarded probated sentences for appellant's convictions for aggravated sexual assault of a child and indecency with a child by contact based on the facts of the offense and the judge's imposition of significant prison terms for both offenses. 9 The facts in this case involved the molestation of appellant's grandchild on several occasions. The complainant testified in extensive detail about the sexual abuse, and her testimony regarding the abuse was supported by the testimony of police investigators. The testimony showed that appellant would go into the complainant's room at night and would expose her to pornography before touching and rubbing her vagina. The complainant was around nine or ten years old at the time of these incidents. The complainant indicated that she was afraid to report the abuse because appellant paid her father's bills and she feared that her family would become homeless if appellant went to jail. The complainant also expressed how traumatic the abuse had been and that its consequences continued years afterwards. The trial court assessed twenty-two years in prison for the aggravated sexual assault offense, which was more than four times the statutory minimum of five years. There is nothing to suggest that the jury would have assessed a sentence at less than half of what the trial court assessed, given the egregious nature of the offense and the facts of the case. The trial court assessed a ten-year sentence for the offense of indecency with a child by contact, and the law would have permitted the jury to have recommended that this prison term be probated. The trial court's sentence, however, was five times the statutory minimum of two years in prison, and half of the maximum sentence permitted under the law. It is clear, therefore, that the trial court did not consider the offense worthy of a minimum term or a lighter sentence. There is nothing to suggest that the jury would have decided that a more lenient probated sentence was appropriate under these circumstances. From an objective viewpoint, I conclude that a reasonable jury would not likely have awarded appellant a probated sentence for these two sexual abuse offenses such that confidence in the actual outcome is undermined. I, therefore, agree with the court of appeals' conclusion that appellant failed to show that trial counsel's deficient performance prejudiced him.

I note further that under this Court's majority opinion, this Court departs from the familiar standard used to gauge prejudice from deficient performance. Here, appellant has expressed no complaint that counsel's deficient performance prejudiced the guilt or innocence phase of the trial that was conducted by the trial court that heard from the two complaining witnesses and six other witnesses. Yet, because this Court's majority opinion declines to evaluate the reliability of the outcome of the *518 trial proceedings that did occur, appellant may receive a windfall of an entirely new trial. I conclude further that, even if this Court's majority opinion were correct that appellant's jury waiver was negatively impacted by counsel's mistaken advice that the trial court could consider a probated sentence, appellant complains, at most, about the punishment phase only. Any relief should be limited accordingly regardless of whether the prejudice standard looks to the waiver only or the outcome of the proceedings.

IV. Conclusion

Appellant has not shown that there was a reasonable probability of changing the outcome of the proceeding in that he probably would not have obtained a probated sentence even if his attorney had properly advised him that the trial court could not consider him for that type of sentence. Accordingly, appellant has failed to establish ineffective assistance of counsel. I, therefore, would affirm the judgment of the court of appeals affirming the two sexual abuse convictions and sentences assessed against appellant at a bench trial.