¹ All references to the Wisconsin Statutes are to the 2019-20 version unless otherwise noted.

² The State initially recommended three years’ initial confinement and three years’ extended supervision on Count 1, consecutive to any sentences Nederhoff was then serving. However, after the circuit court raised a question about the legality of that recommendation, the State conceded that the extended supervision portion of the sentence could not exceed two years, and it therefore altered its recommendation to “four and two.” five-year sentences on each of the three counts in this case, consisting of three years’ initial confinement and two years’ extended supervision. The court ordered that those sentences would be concurrent to each other but consecutive to any other sentences that Nederhoff was serving. The court further determined that Nederhoff was not entitled to any sentence credit because his sentences in this case were consecutive to his revocation sentence, and “the applicable … credit would have been applied against … Nederhoff’s underlying offenses” in that prior case. ¶11 Nederhoff filed a postconviction motion for plea withdrawal, arguing that his pleas were not knowing, intelligent, and voluntary because: (1) Stanek had inaccurately informed him that the plea agreement required the State to follow the sentence recommendation in the PSI and to recommend concurrent sentences; and (2) Nederhoff inaccurately believed that because he had refused to sign the signature bond in this case, he would be entitled to sentence credit from the date of his arrest until the date of his sentencing. In addition, Nederhoff asserted that Stanek was ineffective by failing to ensure that Nederhoff understood the correct terms of the plea agreement before he entered his pleas. Nederhoff also contended that his first attorney in this case, Aaron Olson, was ineffective by advising him that he would be entitled to sentence credit if he refused to sign the signature bond. ¶12 At the postconviction hearing, Olson testified that he had no independent recollection of representing Nederhoff, but that court records indicated he had appeared with Nederhoff for a bond hearing. He did not remember any conversation with Nederhoff about whether Nederhoff should sign a signature bond. ¶13 Stanek conceded during the postconviction hearing that he had misinformed Nederhoff about the terms of the plea agreement by erroneously telling him that the State had agreed to follow the sentence recommendation in the PSI and to recommend concurrent sentences. Stanek explained that the plea negotiations were conducted over email, and he had conveyed the wrong offer to Nederhoff because he had overlooked an email containing an updated plea offer from the State. He testified that he did not object when the State made its recommendation during the sentencing hearing because he was confused and thought he might have been wrong about the terms of the plea agreement. ¶14 On cross-examination, Stanek testified that the overall defense strategy was to “[m]inimize [Nederhoff’s prison] exposure,” given that the original charge against him was a Class D felony with a four-year repeater enhancer. When asked whether Nederhoff had “[gone] along with the overall strategy of, let’s … negotiate, minimize the damage and the impact as much as possible,” Stanek responded, “I think so.” ¶15 With respect to sentence credit, Stanek testified on direct examination that he could not remember exactly what he told Nederhoff, but it was “possible” he stated Nederhoff would receive “substantial credit.” On cross-examination, Stanek testified that he “may have” told Nederhoff that he would receive credit from the start of this case to the point when his extended supervision was revoked. On redirect examination, Stanek testified that he “may have” had a conversation with Nederhoff at some point about the specific amount of sentence credit that he believed Nederhoff would receive, but he could not remember when that discussion might have occurred. ¶16 Nederhoff testified that Stanek told him the State’s plea offer was to amend the original charge to three Class I felonies, to order a PSI, to follow the sentence recommendation in the PSI, and to recommend that the sentences run concurrently to each other and to the other sentences he was serving. He testified that he accepted the plea deal, as communicated to him by Stanek, because he “thought [he’d] get concurrent time, and [he] thought that sounded like a good deal.” He further testified that when he signed the plea questionnaire, he “wanted to make sure that … [the sentences] were all going to be ran concurrent,” and he was assured that they would be. Nederhoff testified that he would not have accepted the State’s plea offer had he known that “it was going to be just three Class I Felonies, with no other deal.” ¶17 Nederhoff also testified that before he entered his pleas, both Olson and Stanek told him that he would receive sentence credit in this case for the time that he spent in custody prior to sentencing—which ultimately amounted to 417 days, or about fourteen months. He testified that he would not have taken the plea deal if he had known it was possible that he would not receive any sentence credit. Nederhoff conceded that, during the plea hearing, he became aware that “if it was consecutive, [he] would not get … time credit. But if it was concurrent, [he] would.” However, he did not change his mind at that point and inform the court that he did not want to enter guilty pleas. ¶18 The circuit court made a factual finding that Stanek “really believed” that the State had agreed to recommend that Nederhoff’s sentences in the instant case be concurrent to each other and to any other sentences he was serving. The court further found that Stanek was “confused” about the terms of the State’s offer and had conveyed his erroneous understanding of the offer to Nederhoff. Nevertheless, the court also found that the record “clearly demonstrate[d]” that Nederhoff understood at the time he entered his pleas that the court was not required to comply with the parties’ plea agreement when imposing sentence. The court reasoned that Nederhoff’s knowledge of that fact “cure[d]” any misinformation about the plea agreement that he had received. ¶19 The circuit court therefore concluded that Nederhoff had failed to establish the prejudice prong of his ineffective assistance claim. The court did not specifically address Nederhoff’s claim that his pleas were not knowing, intelligent, and voluntary. Nederhoff now appeals. DISCUSSION ¶20 To withdraw a guilty plea after sentencing, a defendant must show by clear and convincing evidence that a refusal to allow plea withdrawal would result in manifest injustice. State v. Dillard, 2014 WI 123, ¶36, 358 Wis. 2d 543, 859 N.W.2d 44. A defendant may establish the existence of manifest injustice in several ways. Id., ¶37. As relevant here, a defendant may demonstrate manifest injustice by showing that his or her plea was not knowing, intelligent, and voluntary, or by showing that his or her trial attorney was constitutionally ineffective. Id., ¶¶37, 84. Both of those inquiries present questions of constitutional fact. Id., ¶¶38, 86. As such, under both inquiries, we will uphold the circuit court’s factual findings unless they are clearly erroneous, but we independently review the application of constitutional principles to the facts. See id. ¶21 A plea that was not entered knowingly, intelligently, and voluntarily violates fundamental due process, and a defendant may therefore withdraw the plea as a matter of right. Id., ¶37. “[A]ffirmative misinformation about the law provided by the prosecutor and defense counsel can support a holding that withdrawal of a plea of guilty or no contest must be permitted because the plea is uninformed and its voluntariness is compromised.” Id., ¶39. To determine whether misinformation warrants granting plea withdrawal in a particular case, a court must “review the totality of the circumstances, including the record of the postconviction hearing.” Id., ¶40. A defendant is entitled to plea withdrawal based on misinformation that he or she received if the defendant presents a “persuasive account” of why, absent the misinformation, he or she would not have entered a plea and would have instead gone to trial. Id., ¶52. ¶22 To establish ineffective assistance of trial counsel, a defendant must prove that his or her trial attorney performed deficiently, and that the deficient performance prejudiced the defense. Id., ¶85. To establish deficient performance, the defendant must show that trial counsel’s performance fell outside the wide range of professionally competent assistance. Id., ¶88. To demonstrate prejudice, the defendant must show a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different. Id., ¶95. In the plea withdrawal context, this requires the defendant to establish a reasonable probability that, but for counsel’s errors, the defendant would not have pled guilty and would have instead insisted on going to trial. State v. Bentley, 201 Wis. 2d 303, 312, 548 N.W.2d 50 (1996). ¶23 In this case, the State does not dispute that Stanek misinformed Nederhoff about the terms of the State’s plea offer, and it concedes that Stanek performed deficiently in that regard. The State also does not dispute Nederhoff’s assertion that Olson performed deficiently by advising him that he would be entitled to sentence credit for the time he spent in custody before his sentencing in this case. The State argues, however, that Nederhoff has not established prejudice because he has failed to show that, absent his trial attorneys’ deficient performance, he would not have pled guilty and would have instead gone to trial. For the same reason, the State contends Nederhoff has not established that his pleas were not knowing, intelligent, and voluntary due to the misinformation he received. Stated differently, the State asserts that under either analysis, Nederhoff has failed to show that the misinformation “made a difference” with respect to his decision to plead guilty. ¶24 The State acknowledges that the circuit court did not perform the analysis described above when deciding Nederhoff’s plea withdrawal motion. Instead, the court determined that Nederhoff had failed to establish prejudice because the court had informed Nederhoff during the plea colloquy that it was not bound by the parties’ plea agreement, and that knowledge cured any misinformation Nederhoff had received about the plea agreement’s terms. The State concedes that the court’s prejudice analysis was “incorrect” because a circuit court’s power to reject a sentencing recommendation made pursuant to a plea agreement does not cure misinformation provided to a defendant about his or her plea. See State v. Dawson, 2004 WI App 173, ¶15, 276 Wis. 2d 418, 688 N.W.2d 12. The State argues, however, that we may nevertheless affirm the court’s decision denying plea withdrawal using the correct legal analysis, as Nederhoff has not established that he would have rejected the State’s plea offer and gone to trial absent the misinformation he received. ¶25 In support of its argument in that regard, the State relies primarily on our supreme court’s decision in Dillard. In that case, the defendant was charged with armed robbery, as a persistent repeater, and false imprisonment, as a repeater. Dillard, 358 Wis. 2d 543, ¶16. The defendant ultimately agreed to plead no contest to the armed robbery charge, and in exchange the State agreed to dismiss the persistent repeater enhancer and the false imprisonment charge. Id., ¶¶16, 51. The persistent repeater enhancer would have exposed the defendant to a mandatory life sentence without the possibility of extended supervision. See id., ¶17. After the defendant was sentenced on the armed robbery charge, it came to light that he did not meet the criteria for sentencing as a persistent repeater. Id., ¶¶6, 19, 33. The defendant then moved to withdraw his plea, arguing that it was not knowing, intelligent, and voluntary because of the misinformation he had received about the applicability of the persistent repeater enhancer. Id., ¶34. ¶26 Our supreme court agreed that the defendant’s plea was not knowing, intelligent, and voluntary. Id., ¶69. The court noted that the defendant had initially intended to go to trial due to a known weakness in the State’s case. Id., ¶¶41-43. The court then cited the defendant’s testimony that, despite his initial intention to go to trial, when he considered the prospect of a life sentence without the possibility of extended supervision, he “couldn’t take that chance.” Id., ¶44. The defendant had also testified that the greatest benefit of the plea deal was eliminating the persistent repeater enhancer and that he entered his plea to “make the life without parole go away.” Id., ¶45. In addition, the court cited the defendant’s trial attorney’s testimony that the State’s offer to drop the persistent repeater enhancer was the most significant factor motivating her recommendation that the defendant accept the plea deal, as well as a letter counsel had sent the defendant that corroborated her testimony in that regard. Id., ¶48. ¶27 The State asserts that in this case, the record “does not support Nederhoff’s assertion that if he had been properly informed, he would have rejected the [plea] offer and would have gone to trial.” The State stresses that here, unlike in Dillard, there was no weakness in the State’s case that may have motivated Nederhoff to go to trial rather than accepting a plea deal. The State also asserts that there is “no corroborating support for [Nederhoff’s] assertion that if he had known the State would recommend a sentence of six years consecutive to his revocation sentence, he would have gone to trial.” The State further notes that the difference between the maximum penalty Nederhoff faced under the plea agreement and the maximum sentence he would have faced if convicted at trial was 6.5 years. Ultimately, the State asserts Nederhoff has not shown that “he would have rejected a plea offer that resulted in a five-year sentence and would have chosen to take a weak case to trial, risking a 29-year sentence.” ¶28 There are several problems with the State’s argument. First, the State asserts that, unlike in Dillard, there was no known weakness in the State’s case against Nederhoff. Dillard does not stand for the proposition, however, that absent an identified weakness in the State’s case, a defendant cannot prove that he or she would have rejected a plea offer and instead gone to trial. ¶29 In support of its argument regarding the lack of an identified weakness in the State’s case, the State also relies on the United States Supreme Court’s decision in Lee v. United States, 137 S. Ct. 1958 (2017). The Lee Court stated, “As a general matter, it makes sense that a defendant who has no realistic defense to a charge supported by sufficient evidence will be unable to carry his burden of showing prejudice from accepting a guilty plea.” Id. at 1966. The State also cites Lee for the proposition that a defendant with no viable defense “will rarely be able to show prejudice from accepting a guilty plea that offers him a better resolution than would be likely after trial.” Id. ¶30 The State fails to acknowledge, however, that the Lee Court declined the government’s request to adopt a per se rule “that a defendant with no viable defense cannot show prejudice from the denial of his right to trial.” Id. Instead, the Court noted that when the consequences of a conviction after trial and by plea are “similarly dire, even the smallest chance of success at trial may look attractive. For example, a defendant with no realistic defense to a charge carrying a 20-year sentence may nevertheless choose trial, if the prosecution’s plea offer is 18 years.” Id. at 1966-67. As such, the fact that Nederhoff has not identified a weakness in the State’s case against him is not dispositive of whether, had he been properly informed of the terms of the State’s plea offer, he would have rejected that offer and gone to trial. ¶31 Next, the State’s claim that there is “no corroborating support for [Nederhoff’s] assertion” that he would have gone to trial “if he had known the State would recommend a sentence of six years consecutive to his revocation sentence” does not properly frame the relevant inquiry. The operative question is whether Nederhoff would have rejected the State’s plea offer and gone to trial had he been correctly informed of the actual terms of the State’s offer. At the time Nederhoff made the decision to accept the plea deal, he did not know what sentence the State would ultimately recommend. Under the correct terms of the plea agreement, the State did not agree to recommend a six-year sentence; it was instead free to argue for whatever sentence it deemed appropriate. The fact that the State ultimately recommended a six-year sentence is therefore irrelevant to our analysis of whether Nederhoff’s pleas were knowing, intelligent, and voluntary, despite the misinformation he received. ¶32 A similar flaw exists in the State’s assertion that Nederhoff has not shown that “he would have rejected a plea offer that resulted in a five-year sentence and would have chosen to take a weak case to trial, risking a 29-year sentence.” Although the circuit court ultimately imposed concurrent five-year sentences in this case (albeit consecutive to Nederhoff’s revocation sentence), Nederhoff did not know what sentences the court would impose at the time he entered his pleas. Had Nederhoff been correctly informed of the terms of the State’s plea offer, he would have merely known that the State was free to argue at sentencing, that it was not required to follow the PSI recommendation, and that it was not required to recommend concurrent sentences. The question is whether, having been so informed, Nederhoff would have rejected the State’s plea offer and chosen to go to trial. ¶33 Finally, the State asserts that the difference between the maximum penalty Nederhoff faced at trial (29 years) and the maximum penalty he faced under the correct terms of the plea agreement (22.5 years) was 6.5 years.3 The State suggests that, given that difference, Nederhoff cannot show that he would have rejected the State’s offer and taken a weak case to trial.

³ Throughout this case, the State has consistently asserted—and Nederhoff has not disputed—that the maximum penalty Nederhoff faced on the original Class D felony charge, with the repeater enhancer, was twenty-nine years’ imprisonment. For instance, both the criminal complaint and the Information stated that the maximum penalty for the Class D felony charge was twenty-five years’ imprisonment, and that because Nederhoff was a repeater, that maximum penalty could be increased by up to four years under WIS. STAT. § 939.62(1)(b) because his prior conviction was for a felony. The State’s calculation of the maximum penalty for the original Class D felony charge, with the repeater enhancer, was erroneous. The State correctly asserted that the maximum penalty for a Class D felony was twenty-five years’ imprisonment. See WIS. STAT. § 939.50(3)(d). However, the State erred by asserting that the applicable paragraph of the repeater statute was WIS. STAT. § 939.62(1)(b), which provides: “A maximum term of imprisonment of more than one year but not more than 10 years may be increased … by not more than 4 years if the prior conviction was for a felony.” (Emphasis added.) Section 939.62(1)(b) is inapplicable because, as noted above, the maximum term of imprisonment for a Class D felony charge is twenty-five years, which is not between one and ten years. The applicable paragraph of the repeater statute is instead § 939.62(1)(c), which states: “A maximum term of imprisonment of more than 10 years may be increased … by not more than 6 years if the prior conviction was for a felony.” (Emphasis added.) Thus, as properly calculated, Nederhoff’s maximum sentence on the original Class D felony charge, with the repeater enhancer, was thirty-one years’ imprisonment, rather than twenty-nine. Accordingly, the difference between the maximum penalty Nederhoff faced at trial and the maximum penalty he faced under the correct terms of the plea agreement was 8.5 years, not 6.5 years.

⁴ We note one other issue that the circuit court may choose to consider on remand. During the postconviction hearing, both Stanek and Nederhoff testified that Stanek misinformed Nederhoff about the terms of the State’s plea offer. On appeal, the State does not argue that Nederhoff otherwise became aware of the correct terms of the plea offer before he entered his pleas. However, the plea questionnaire and waiver of rights form that Nederhoff signed described the plea agreement as follows: “Plead to three (3) counts of possession. PSI.” The court summarized the plea agreement in similar terms at the beginning of the plea hearing, and when the court subsequently asked Nederhoff whether its summary was consistent with his “understanding of what you thought was going to happen here today,” he responded, “Yes, Your Honor.” Given these facts, the court could choose to consider on remand whether Nederhoff actually misunderstood the terms of the State’s plea offer at the time he entered his pleas.