Grant Wilson appeals a modification made to his criminal sentence. He contends the prosecutor misstated to the sentencing court the facts underlying his conviction and the facts of a court case cited by Wilson as legal authority against the modification. A divided Court of Appeals panel could not agree on the errors or the appropriate standard of review to assess any resulting prejudice.

*844 State v. Wilson , No. 114567, 2016 WL 7324427 (Kan. App. 2016) (unpublished opinion). We granted review to consider those questions and now remand the case to the district court for a new hearing on the State's motion to correct an illegal sentence.

We hold prosecutorial error may occur during a sentencing proceeding before a judge. We also hold the analytical framework from State v. Sherman , 305 Kan. 88 , 378 P.3d 1060 (2016), applies in both the guilt and penalty phases of any trial-whether before a jury or judge. And based on the Sherman test, we hold there was reversible error at Wilson's sentencing hearing.

FACTUAL AND PROCEDURAL BACKGROUND

In 2007, Grant Wilson pled guilty to aggravated indecent solicitation of a child. He later failed to meet his probation terms and was eventually ordered to serve his underlying prison sentence of 32 months. In 2015, the State moved to correct an illegal sentence, arguing the district court erred by not imposing lifetime postrelease supervision as part of Wilson's original sentence. At the hearing on that motion, Wilson claimed lifetime supervision was grossly disproportionate to his offense, amounting to cruel or unusual punishment prohibited by Section 9 of the Kansas Constitution Bill of Rights. The court rejected Wilson's argument based on State v. Freeman , 223 Kan. 362 , 367, 574 P.2d 950 (1978), and granted the State's motion.

On appeal, Wilson raised two issues: (1) whether the prosecutor deprived him of a fair sentencing hearing by misstating facts in his case and those in an unpublished Court of Appeals decision cited as authority for Wilson's Freeman claim; and (2) whether the district court erred by rejecting his disproportionality argument under Freeman . The first claim intersects with the second to the extent Wilson argues the prosecutor's comments denied him a fair hearing on the following Freeman factor:

"The nature of the offense and the character of the offender should be examined with particular regard to the degree of danger present to society; relevant to this inquiry are the facts of the crime, the violent or nonviolent nature of the offense, the extent of culpability for the injury resulting, and the penological purposes of the prescribed punishment." Freeman , 223 Kan. at 367 , 574 P.2d 950 .

Wilson contends this factor supported his disproportionality argument because (1) the 13-year-old victim snuck out of her house and joined some high school students for a party where the crime occurred; (2) Wilson had just turned 18 years old and lacked the ability to make a "good decision" because of his age; (3) Wilson acknowledged he made "a terrible mistake" in having sex with the underage victim; (4) Wilson worked at his father's sprinkler installation business after getting out of prison and had a child he was trying to parent; and (5) there was little danger of Wilson reoffending.

The State set the stage for this prosecutorial error claim by replying to Wilson's arguments during the following colloquy:

"THE COURT: ... so we now get to Mr. Wilson's particular crime and I know I've had other cases in which there was violence involved.

"There does not appear that [violence] happened in this case but Mr. Wilson was certainly of the appropriate age to be held responsible and the fact that he didn't, you know, he put himself in a situation. I think there was drinking involved, is that right, [Prosecutor]?

"[PROSECUTOR]: Judge, actually that's one reason, if I can address the Court just briefly about those facts, with Mr. Wilson, he raped and sodomized a 13 year old girl. He digitally raped her and he sodomized her . Those are extreme crimes of violence. Your Honor, the facts of this case, the State charged only the aggravated indecent solicitation and agreed to a border box finding to place him on Community Corrections. He was placed on Community Corrections, violated the terms of his Community Corrections and was sentenced by the Court to go to Labette. He went to Labette, failed to complete Labette and his prison sentence was executed so ... he's clearly shown he's not necessarily amenable to do, it's a *845 crime of violence and that distinguishes itself completely from [ State v. ] Proctor [No. 104,697, 2013 WL 6726286 (Kan. App. 2013) (unpublished opinion),] which involved a police officer impersonating himself to be a child and the defendant showing up at the house to meet the child .

"THE COURT: All right. I saw back here when I was doing a bond condition that he had a rape and aggravated sodomy [at] that point in time back in 26 February, 2007. I'm not satisfied that, I am satisfied rather that the facts and circumstances of this crime do not cry out that this would be an unconstitutional sanction so I'm going to grant the State's motion. I find that the sentence was illegal; that the Court is required to sentence Mr. Wilson to lifetime post release so I pronounce lifetime post release." (Emphases added.)

On appeal, the State suggested the digital rape and sodomy comments were based on police reports, but those reports were not in the appellate record. The State acknowledged reciting the wrong facts about the Proctor decision.

The panel majority's holdings and the dissent's arguments

The panel majority first considered whether prosecutorial misconduct can occur in the context of a hearing on a motion to correct an illegal sentence. The majority concluded it could, so appellate review was proper. It cited other Court of Appeals decisions that had considered similar claims arising during hearings before a judge. Wilson , 2016 WL 7324427 , at *4 (discussing State v. Serrano-Garcia , No. 103651, 2011 WL 4357804 , at *3-4 [Kan. App. 2011] [unpublished opinion], State v. Roland , No. 101879, 2010 WL 1078454 , at *1-3 [Kan. App. 2010] [unpublished opinion], State v. Clelland , No. 93001, 2005 WL 1805250 , at *3-5 [Kan. App. 2005] [unpublished opinion] ).

The majority then considered whether to use the modified two-step analysis adopted in Sherman but ultimately decided to apply the "old" pre- Sherman test. It did so because Sherman was decided after the panel's oral arguments, which meant the parties had not had an opportunity to brief or argue its possible impact. Even so, the majority observed "application of the new framework would not make a difference in the outcome." 2016 WL 7324427 , at *4.

Moving to the merits, the majority held the prosecutor's comments that Wilson digitally raped and sodomized a child were unsupported by the record. It observed: "The factual basis for the plea, accepted by the court, consisted solely of Wilson's admission that he 'had sex with a minor under the age of 14 years old.' " 2016 WL 7324427 , at *5. No other evidence was presented.

As to the prosecutor's misstatement about the facts in Proctor , which the State conceded, the majority noted: " Proctor involved no impersonation and no police officer. Instead, Proctor was a 19-year-old male who took advantage of a 12-year-old boy." Wilson , 2016 WL 7324427 , at *6 ; see also Proctor , 2013 WL 6726286 , at *2 (Proctor knew the victim and lived with him for several months while committing the crimes). The majority held: "The prosecutor thus invited the district court to rely on inaccurate facts in Wilson's case and to compare them to inaccurate facts in Proctor's case." 2016 WL 7324427 , at *6.

It then concluded these misstatements were: (1) gross and flagrant misconduct because they were planned and violated well-established rules; (2) motived by ill will since they were "volunteered in response to the court's narrow question to the prosecutor whether drinking was involved-a question that went unanswered"; and (3) may have affected the district court's decision to impose lifetime postrelease supervision. 2016 WL 7324427 , at *7-8 ("In the context of this case, we determine the likely effect not on the verdict reached by the jurors but on the decision made by the district court."). The majority held those improper comments were prejudicial and denied Wilson a fair hearing. 2016 WL 7324427 , at *9.

The dissent viewed the circumstances differently. Judge Michael Buser argued the prosecutor's assertions about digital rape and sodomy were supported by: (1) the standard arrest report indicating Wilson was arrested for rape and aggravated sodomy of a child *846 under 14 years old; (2) an entry from the district court's docket sheet memorializing a hearing held on the same day the arrest report was filed, which states " 'Probable cause found' "; and (3) the appearance bond signed by the judge, which noted Wilson was charged with "RAPE < 14, AGG. SODOMY." 2016 WL 7324427 , at *13.

Judge Buser insisted the first comment was not prosecutorial error because "[i]n sentencing matters ..., district court judges rely on the presentence investigation (PSI) report. In this case, the PSI contained facts and information which also supported [the challenged factual assertion]." 2016 WL 7324427 , at *15. And, the dissent continued, while none of the listed pieces of the record explicitly stated Wilson "digitally" raped and sodomized the victim, this information was encompassed by the statute because "Kansas law defines sexual intercourse as 'any penetration of the female sex organ by a finger , the male sex organ or any object,' " referring to K.S.A. 21-3501(1). 2016 WL 7324427 , at *15. As to the second comment about Proctor , Judge Buser agreed the prosecutor incorrectly stated the case's facts but disagreed this constituted misconduct because the district court judge was presumed to know the law. 2016 WL 7324427 , at *15. Finally, the dissent insisted the pre- Sherman test was inapplicable to a prosecutorial error during a motion hearing before a district judge because it equated an alleged error tainting a jury with impact on a trial judge. 2016 WL 7324427 , at *11.

The State petitioned this court for review. Wilson did not ask us to review his Freeman challenge, which the panel avoided. Accordingly, we address only the prosecutorial error claim. See Supreme Court Rule 8.03(a)(4)(C) (2018 Kan. S. Ct. R. 54) ("The court will not consider issues not presented or fairly included in the petition."); Castleberry v. DeBrot , 308 Kan. 791 , 794-95, 424 P.3d 495 (2018). The State does not challenge the majority's holding relating to the Proctor misrepresentations, so that resolves in Wilson's favor on review. See State v. McBride , 307 Kan. 60 , 62, 405 P.3d 1196 (2017).

Jurisdiction is proper. K.S.A. 20-3018(b) (petitions for review of Court of Appeals decision); K.S.A. 60-2101(b) (Supreme Court has jurisdiction to review Court of Appeals decisions upon petition for review).

PRESERVATION

At the outset we must consider whether Wilson's prosecutorial error challenge is preserved for our review. On appeal, the State pointed out Wilson did not object to the alleged misstatements during the hearing, so it argued the issue was not properly preserved. The panel rejected this contention because a contemporary objection is not required to claim prosecutorial error during closing argument before a jury, so it reasoned that principle should extend to Wilson's sentencing proceeding before the judge. Wilson , 2016 WL 7324427 , at *4 ; see also State v. Miller , 293 Kan. 535 , 550, 264 P.3d 461 (2011) (while a contemporary objection is required for review of an evidentiary prosecutorial error claim, it is unnecessary to consider prosecutorial error during closing argument).

In its petition for review, the State identified one issue for this court to take up: Whether the prosecutor committed reversible error at the hearing on the State's motion to correct illegal sentence? This advances only a merits based challenge to the prosecutorial error question. As a result, we hold the State waived review of the panel majority's conclusion on preservation. See Supreme Court Rule 8.03(a)(4)(C) (Supreme Court will not consider issues not presented or fairly stated in the petition for review). We express no opinion whether a contemporaneous objection or other posthearing remedial motion is required to appeal a prosecutorial error claim arising from a nonjury setting.

ANALYSIS

To determine what test applies when an appellate court reviews an alleged prosecutorial error during a sentencing hearing, it is important to understand the protected right at stake.

Prosecutorial error jurisprudence recognizes a prosecutor's conduct can implicate a criminal defendant's due process *847 rights to a fair trial under the Fourteenth Amendment to the United States Constitution. Smith v. Phillips , 455 U.S. 209 , 219, 102 S.Ct. 940 , 71 L.Ed.2d 78 (1982) ("[T]he touchstone of due process analysis in cases of alleged prosecutorial misconduct is the fairness of the trial."); Sherman , 305 Kan. at 97 , 378 P.3d 1060 ("The jurisprudence surrounding policing prosecutors during trial has historically taken as its lodestar-in almost all cases-the due process requirements of the Fourteenth Amendment."). And our prosecutorial error test makes clear that criminal defendants have a constitutional right to a fair trial. As we explained in Sherman :

"To determine whether prosecutorial error has occurred, the appellate court must decide whether the prosecutorial acts complained of fall outside the wide latitude afforded prosecutors to conduct the State's case and attempt to obtain a conviction in a manner that does not offend the defendant's constitutional right to a fair trial . If error is found, the appellate court must next determine whether the error prejudiced the defendant's due process rights to a fair trial ." (Emphases added.) Sherman , 305 Kan. at 109 , 378 P.3d 1060 .

Admittedly, the above is cast in terms of the prosecutor's efforts to obtain a "conviction," but this does not confine prosecutorial error to a trial's guilt phase. One's fair trial right is equally protected in a penalty phase. See State v. Kleypas , 272 Kan. 894 , 1083-84, 40 P.3d 139 (2001) (discussing in a death penalty case review for claimed prosecutorial misconduct during the penalty phase and holding it "is similar to the standard applied in the guilt phase."), overruled on other grounds by Kansas v. Marsh , 548 U.S. 163 , 126 S.Ct. 2516 , 165 L.Ed.2d 429 (2006). So to the extent our caselaw typically describes prosecutorial error in the guilt-phase context when obtaining a "conviction" or a "verdict," we must remember our concern is comporting with the due process right to a fair trial. Those concerns remain for sentencings.

A review of out-of-state caselaw also shows many states have held prosecutors can commit error in nonjury settings. See, e.g., Liggett v. People , 135 P.3d 725 , 733-35 (Colo. 2006) (defendant had a bench trial and the court analyzed whether prosecutorial misconduct occurred during closing argument); State v. Mosley , 853 N.W.2d 789 , 801-03 (Minn. 2014) (defendant was convicted after a bench trial and the court addressed an issue of whether prosecutor committed misconduct by eliciting inadmissible character evidence); Com. v. Francis , 445 Pa. Super. 353 , 360-61, 665 A.2d 821 (1995) (recognizing "trial judges are presumed to ignore prejudicial material when acting as factfinder"; holding prosecutorial error affected the factfinder at the bench trial); State v. Kehdy , 120 Hawaii 418 , 2009 WL 1805908 , at *5-6 (Hawaii App. 2009) (unpublished opinion) (rejecting State's assertion the prosecutorial misconduct was harmless beyond a reasonable doubt simply because it was a bench trial); State v. Myers , 2004-Ohio-478 , 2004 WL 226105 , at *5 (Ohio App. 2004) (unpublished opinion) (considering alleged prosecutorial error occurring in a bench trial). We are aware of no case with a contrary holding, nor are we cited to any.

The Wilson dissent points out some state courts have a distinctive test addressing circumstances like those we have here. Wilson , 2016 WL 7324427 , at *18, citing: Deeds v. State , 2014 WY 124 , 335 P.3d 473 , 480 (2014) (defendant's burden to show a due process violation in the context of a sentencing hearing before the bench by showing " 'the trial court relied upon the [prosecutorial] statements in sentencing' " him); State v. Tierinni , 144 Conn.App. 232 , 240, 71 A.3d 675 (2013) (rejecting traditional test for prosecutorial impropriety in favor of two-pronged inquiry: " 'first, did the information at issue contain some minimal indicium of reliability; second, if it did not, did the trial court substantially rely on this improper information in fashioning its ultimate sentence?' "); State v. Rivera , No. 30836, 2013 WL 4512056 , at *4 (N.M. App. 2013) (unpublished opinion) ("When matters are tried before a judge, we presume the judge has disregarded improper comments or evidence unless the record clearly indicates otherwise," and "we do not treat [those] issue[s] as a standard prosecutorial misconduct claim."). But despite their distinct tests, they begin by recognizing prosecutorial error can occur in these settings.

*848 We hold prosecutorial missteps may implicate due process rights to a fair trial in this nonjury context.

So what test should apply? When the parties argued this appeal to the panel, Kansas courts referred to claims that a prosecutor's comments denied a defendant's due process rights to a fair trial as "prosecutorial misconduct." See State v. Barber , 302 Kan. 367 , Syl. ¶ 4, 353 P.3d 1108 (2015). The then-effective test was set out in State v. Tosh , 278 Kan. 83 , Syl. ¶¶ 1, 91 P.3d 1204 (2004). Under Tosh , an appellate court first decided whether the prosecutor's remark being complained about was outside the wide latitude allowed in discussing evidence. 278 Kan. at 85 , 91 P.3d 1204 . If so, the court made what was described as a "particularized harmlessness inquiry," assessing three factors: (1) whether the misconduct was gross and flagrant; (2) whether it showed ill will on the prosecutor's part; and (3) whether the evidence against the defendant was of such a direct and overwhelming nature that the misconduct likely had little weight in the jurors' minds. 278 Kan. at 93-96 , 91 P.3d 1204 . No factor was individually controlling; but before the third could override the first two, an appellate court had to be able to say the harmlessness tests of both K.S.A. 60-261 and Chapman v. California , 386 U.S. 18 , 87 S.Ct. 824 , 17 L.Ed.2d 705 (1967), were met. Tosh , 278 Kan. 83 , Syl. ¶ 2, 91 P.3d 1204 .

As mentioned, while Wilson's appeal was pending with the panel, we modified the two-step analytical framework for claims that a prosecutor's trial behavior requires reversal. See Sherman , 305 Kan. at 109 , 378 P.3d 1060 . Sherman renamed such claims " 'prosecutorial error,' " saving the pejorative " 'prosecutorial misconduct' " label for more egregious transgressions. 305 Kan. at 107, 114 , 378 P.3d 1060 ; see State v. Chandler , 307 Kan. 657 , 695, 414 P.3d 713 (2018).

Sherman did not disturb the preexisting standard for whether the prosecutorial action complained about was improper, i.e., the action was outside the wide latitude afforded prosecutors. 305 Kan. at 104 , 378 P.3d 1060 ("The well-developed body of caselaw defining the scope of a prosecutor's 'wide latitude' ... will continue to inform our review of future allegations of prosecutorial error."). Sherman modified Tosh 's second step only:

"If error is found, the appellate court must next determine whether the error prejudiced the defendant's due process rights to a fair trial. In evaluating prejudice, we simply adopt the traditional constitutional harmlessness inquiry demanded by Chapman . In other words, prosecutorial error is harmless if the State can demonstrate 'beyond a reasonable doubt that the error complained of will not or did not affect the outcome of the trial in light of the entire record, i.e. , where there is no reasonable possibility that the error contributed to the verdict.' We continue to acknowledge that the statutory harmlessness test also applies to prosecutorial error, but when 'analyzing both constitutional and nonconstitutional error, an appellate court need only address the higher standard of constitutional error.' [ Citations omitted.]" 305 Kan. at 109 , 378 P.3d 1060 .

In doing so, Sherman further noted:

"Multiple and varied individualized factors can and likely will affect the Chapman analysis in future cases. Every instance of prosecutorial error will be fact specific, and any appellate test for prejudice must likewise allow the parties the greatest possible leeway to argue the particulars of each individual case. Thus, appellate courts should resist the temptation to articulate categorical pigeonholed factors that purportedly impact whether the State has met its Chapman burden. Appellate courts must simply consider any and all alleged indicators of prejudice, as argued by the parties, and then determine whether the State has met its burden- i.e. , shown that there is no reasonable possibility that the error contributed to the verdict. The focus of the inquiry is on the impact of the error on the verdict. While the strength of the evidence against the defendant may secondarily impact this analysis one way or the other, it must not become the primary focus of the inquiry. As has often been repeated, prejudice can exist even 'in a strong case.' [ Citation omitted.]" 305 Kan. at 110-11 , 378 P.3d 1060 .

*849 Sherman provides the best measure to evaluate the prosecutorial error in the context of Wilson's sentencing hearing before a district court judge. To begin with, Sherman was decided more than two years ago and the parties have had the opportunity to address its application. See Gaudina v. State , 278 Kan. 103 , 106, 92 P.3d 574 (2004) (changes in the law generally apply to cases not yet final). And we discern no prejudice to either side by using Sherman . More importantly, the Tosh language for the second prong unnecessarily diverts attention from the pivotal fair trial inquiry. As Sherman points out, there is little need to consider whether an error was gross and flagrant or motivated by ill will because what is paramount for fair trial concerns is defining any impact from an error on the proceeding's outcome. Sherman , 305 Kan. at 93 , 378 P.3d 1060 . That is just as true in this nonjury setting.

Wilson's appeal highlights why our move away from the harsher "prosecutorial misconduct" tag should be helpful. Many times in both the briefing and oral argument, the State's counsel, who was also the prosecutor at the sentencing hearing, defensively explained he was not motivated by ill will-even when the questioning was just about the reasonable possibility the error affected the judge's ruling. And while we appreciate prosecutors, as public servants, are appropriately sensitive to the "misconduct" label, these cases are more about the error's potential effect on the outcome than the prosecutor's professionalism or ethics. That focus should not get lost. Saving the "misconduct" brand for the most egregious circumstances is intended to avoid such distractions.

Applying the Sherman test, we hold the prosecutor was outside the wide latitude afforded when arguing the State's motion to correct an illegal sentence. We hold further the State fails to meet its burden to show there is no reasonable possibility this prosecutorial error contributed to the district court's decision, i.e., the proceeding's outcome.

To begin with, the prosecutor's statement that Wilson digitally raped and sodomized his victim was unsupported by evidence. The police reports referenced by the dissent are not in the appellate record. But even if they were, their contents were only allegations-not evidence. And contrary to the dissent's implication, there is no mention of digital rape or sodomy in the presentence investigation report in the appellate record.

Apparently, the prosecutor believed it was appropriate for the district court to consider claims unsupported by evidence in ruling on Wilson's Freeman challenge. If that were so, that belief was wrong. See State v. Atkisson , 308 Kan. 919 , Syl. ¶ 7, 425 P.3d 334 (2018) ("When a discretionary decision requires fact-based determinations, a district court abuses its discretion when the decision is based on factual determinations unsupported by the evidence."); State v. Gonzalez , 290 Kan. 747 , 757, 234 P.3d 1 (2010) (same). Similarly, an appellate court employs a bifurcated standard when reviewing a district court's decision whether a sentence is cruel or unusual under Section 9 of the Kansas Constitution Bill of Rights. And the first inquiry is whether there is sufficient support for the district court's factual findings, i.e., substantial competent evidence. State v. Mossman , 294 Kan. 901 , 906, 281 P.3d 153 (2012).

What the prosecutor did was ask the district court to base its decision on allegations unsupported by evidence. When a prosecutor argues facts outside the evidence, the first prong of the prosecutorial error test is met. Chandler , 307 Kan. at 678-79, 414 P.3d 713 .

As for the impact of that error, it appears reasonably possible the prosecutor's prompt pushed the court into considering unsupported allegations rather than evidence. As the panel majority concluded:

"[T]he colloquy between the court and the prosecutor shows that before the prosecutor's misstatements, the court considered Wilson's specific conduct to be nonviolent. But immediately after the prosecutor's misstatements, the court appears to have changed its mind. The effect of the prosecutor's words on the district court was thus *850 weighty and persuasive." Wilson , 2016 WL 7324427 , at *9.

The majority also correctly points out "the district court would be justified in relying on the prosecutor's statements of facts and law, particularly where, as here, the argument at the hearing was prepared by the State in support of its own motion." 2016 WL 7324427 , at *8.

The factual misstatement about the underlying crime is sufficient standing alone to reverse the district court's decision. For that reason, we need not delve further into the misstatements about Proctor . Plus, as the dissent appropriately observes, there was no reference by the district court to Proctor "in any way in evaluating the first Freeman factor or making [its] ultimate ruling." Wilson , 2016 WL 7324427 , at *18.

That said, we recognize lawyers routinely have different interpretations about a judicial decision's legal effect, as well as whether meaningful differences in facts distinguish one case from another. Because of that, circumstances resulting in a claim that a prosecutor stepped outside appropriate bounds as an advocate when discussing caselaw to a judge may be harder to sort out on appeal. We need not untangle that bird's nest in this decision.

We affirm the Court of Appeals judgment reversing the district court on the issue subject to review and remand the case to the district court to consider again the constitutional question, i.e., whether imposing lifetime postrelease supervision on Wilson would be grossly disproportionate to his offense under Freeman . If Wilson's argument is unsuccessful, the district court then can decide whether the State may prevail on its motion to correct an illegal sentence.