STATE of Oregon, Respondent on Review, v. Douglas Leroy PITT, Petitioner on Review.
A jury convicted defendant of two counts of first-degree unlawful sexual penetration, ORS 163.411, and two counts of first-degree sexual abuse, ORS 163.427. Defendant appealed, assigning error to the trial court's order denying his motion in limine1 to exclude evidence of prior uncharged sexual misconduct involving the victim, A, and another individual, R. The Court of Appeals affirmed defendant's convictions, concluding that the trial court did not err in declining to exclude the challenged evidence because it was relevant to bolster the victim's credibility in identifying defendant as her abuser. State v. Pitt, 236 Or.App. 657, 237 P.3d 890 (2010). We granted review to decide whether the Court of Appeals properly affirmed the trial court's decision to deny defendant's motion in limine. For the reasons set forth below, we reverse the decision of the Court of Appeals and the circuit court.
Defendant was charged in Clatsop County Circuit Court with sexually abusing A, the four-year-old daughter of defendant's girlfriend, Snider. In the summer of 2001, defendant moved into a Lane County residence with Snider and A. Snider's sister, Finney, and Finney's four-year-old daughter, R, also lived at that residence. In September, defendant, Snider, and A moved to Clatsop County. Shortly after the move, Snider noticed a change in A's behavior. A began having nightmares, wetting the bed, and resisting being left alone. In December, A pointed to her genital area and told Snider that defendant “touches me there.” A referred to defendant as “Doug.” Snider then took A to the police station, where officers arranged to have A evaluated at the Lighthouse, a child abuse assessment center.
At the Lighthouse, Dr. Stefanelli conducted an interview and physical examination of A. During the examination, A disclosed that defendant had touched her genital area more than once. Stefanelli also found physical evidence consistent with sexual abuse. Based on a June 1999 medical examination of A that showed no physical signs of sexual abuse, Stefanelli diagnosed A as having been sexually abused between June 1999 and December 2001.
The Lighthouse staff subsequently referred A to Dr. Berdine, a clinical psychologist, for further evaluation. When Berdine met with A a few days later, A stated that defendant had touched her genital area when she lived in Clatsop County. A also disclosed that defendant had touched her genital area when she lived in Lane County, and that she had seen defendant touch R's genitals there too.
A and R were later interviewed at the Lane County Child Advocacy Center by the center's director, Broderick. Both interviews were conducted separately and were videotaped. During Broderick's interview with A, A stated that defendant had touched her genital area and that she had observed defendant touch R in a similar manner. When R was interviewed, she too disclosed that defendant had touched her genital area, and that she had observed defendant touch A as well.
Defendant was charged by indictment with two counts of first-degree unlawful sexual penetration and two counts of first-degree sexual abuse, based on conduct involving A that occurred in Clatsop County. Defendant's case was initially tried in 2002 and defendant was convicted of all charges. Defendant appealed his convictions, and the Court of Appeals reversed and remanded the matter for retrial. See State v. Pitt, 209 Or.App. 270, 147 P.3d 940 (2006), adh'd to on recons, 212 Or.App. 523, 159 P.3d 329 (2007) (reversal based on a violation of defendant's constitutional right to confront witnesses against him). This case concerns the proceedings on retrial.
On remand and before trial, defendant moved to exclude all evidence of prior uncharged misconduct involving A and R that took place in Lane County. At a hearing on defendant's motion, defendant contended that that evidence was inadmissible character evidence under OEC 404(3).2 Specifically, defendant argued that the evidence was irrelevant to prove “intent” under OEC 404(3) because his “defense is and has always been that this didn't happen, that he didn't do it, if it did happen, it wasn't him. And so the question of intent is not really at issue in this case.” Defendant added, “We're not arguing that this was an incidental touching, that it was an accidental touching, that it was a touching for some other purpose.” The state, on the other hand, asserted that proof of multiple incidents of sexual abuse “would bear upon the absence of mistake or accident, in addition to going to defendant's intent.”
In a letter opinion denying defendant's motion, the trial court applied the analysis set out in State v. Johns, 301 Or. 535, 725 P.2d 312 (1986).3 Under that analysis, the court found that: (1) the charges required the state to prove that defendant had acted with intent; (2) the Lane County incidents would have required intent as well; (3) A was the same victim and R was in the same class of victims; (4) defendant faced similar charges in Lane County for his conduct there; (5) the physical elements of the conduct were the same or very similar; and (6) the evidence was probative of defendant's knowledge but could confuse the jury. Examining those factors in whole, the trial court determined that the evidence was relevant “in determining one of the necessary elements of each count, the mental state.” The trial court further reasoned that, because defendant had indicated during the hearing on his motion that he may raise the defense that someone else committed the crime—specifically that A's statements stemmed from an alleged incident of sexual abuse of A by her father David in 1999 and that David, who had ongoing contact with A, could have been responsible for the physical evidence present—“then [the] identity of who committed the crime is at issue.” Therefore, the trial court concluded that the evidence could be admitted at trial, and it denied defendant's motion.
In accordance with that ruling, the state offered evidence at trial relating to defendant's charged conduct as well as his prior uncharged conduct. In particular, A testified that defendant had touched her genital area in Clatsop County (charged conduct) and that defendant had touched her and R in Lane County (uncharged conduct). R also testified that she thought, but was not certain, that defendant had touched her genital area in Lane County (uncharged conduct). Stefanelli further testified about the physical evidence that she had discovered in her examination of A, as well as her diagnosis that A had been sexually abused.
On cross-examination of the witnesses, defendant elicited testimony that young children are highly suggestible; that A and R lived in a chaotic environment during the time of the alleged abuse; that both defendant and R's father were named Doug and therefore the reference to “Doug” was potentially ambiguous; and that A's father David was accused of sexually abusing A in 1999 and had had ongoing contact with A since that time. Stefanelli also stated on cross-examination that she could not identify a particular individual as A's abuser based on the physical evidence present. Defendant did not otherwise raise an OEC 404(3) objection to the introduction of uncharged misconduct evidence during the trial on remand.
After closing arguments, the trial court instructed the jury to consider the evidence of defendant's prior uncharged conduct “for two limited purposes”:
“first, as to whether defendant acted with knowledge as to the alleged criminal conduct in this case, or second, as to the identity of the person who committed the allegations in this case, i.e., whether the defendant or someone else committed the alleged criminal conduct. Specifically you are not to draw the inference that the evidence of the other conduct makes defendant guilty of the charges in this case.”
The jury found defendant guilty of all charges.
Defendant appealed and assigned error to the trial court's order denying his motion in limine. Defendant reprised his pretrial arguments that the uncharged misconduct evidence was inadmissible.4
The Court of Appeals affirmed the trial court's decision to deny defendant's motion in limine, but it did so on narrower grounds than those relied on by the trial court. The Court of Appeals reasoned that the uncharged misconduct evidence was relevant and admissible under OEC 404(3) to prove the identity of the perpetrator in light of defendant's position that, if the sexual abuse had occurred, another person was responsible. The court acknowledged that the evidence did not comport with the “identity” exception to OEC 404(3) as discussed in Johns. In Johns, this court had explained that “evidence of other crimes offered to prove identity is strictly limited to crimes committed ‘by the use of a novel means or in a particular manner’ so as to earmark the acts as the handiwork of the accused.” 301 Or. at 551, 725 P.2d 312. Notwithstanding that stated limitation, the Court of Appeals explained in its opinion that, “[b]y arguing that [A] was abused by someone else, defendant contested the accuracy of the victim's identification of him as the perpetrator of the charged acts.” Pitt, 236 Or.App. at 663, 237 P.3d 890. The court further stated:
“The credibility of the victim's identification of defendant * * * was plainly a contested issue.
“The uncharged misconduct evidence was relevant to bolster the victim's credibility on that issue. A jury could conclude that [A] would be less likely to misidentify her abuser if she had been abused by the same person on other occasions and had seen the same person abuse [R] as well. Likewise, evidence that R had reported being abused by defendant bolstered the credibility of [A's] identification of defendant.”
Id. at 663–64, 237 P.3d 890 (internal citations omitted). The court concluded that “[t]he relevance of that evidence does not rely on an impermissible inference about defendant's character,” and it affirmed the trial court's ruling denying defendant's motion in limine. Id. at 664, 237 P.3d 890.
At the outset, we recognize that the form of the challenged ruling—here, an order denying a motion in limine—constrains our analysis of this case. A motion in limine is a motion made by a party, commonly prior to trial, seeking an order, as relevant here, to exclude certain evidence from the trial. See Black's Law Dictionary 1038 (8th ed 2004) (stating that definition). It is not uncommon for a party to file a motion in limine to keep damaging evidence, particularly evidence of a criminal defendant's prior bad acts, from the eyes and ears of the jury. As this court previously has explained, a motion in limine is preferred because, “[i]f counsel must wait to make an objection and receive a ruling in front of the jury, the client could be prejudiced even though the ruling was in the client's favor.” State v. Foster, 296 Or. 174, 183, 674 P.2d 587 (1983). Stated differently, “[a]n objection to evidence, with a motion to tell the jury to disregard it, is a poor alternative. The old cliché, ‘you can't unring a bell,’ still applies.” Id. at 182, 674 P.2d 587.
Yet relying on a motion in limine to limit evidence before trial presents its own obstacles. A challenge during trial to an offer of evidence provides greater context for the trial judge to assess issues concerning the admissibility of the proffered evidence in light of what has occurred in the course of the trial. In contrast, a trial judge ruling on a motion in limine must rely on the parties' representations and arguments about what they expect the evidence, including the challenged evidence, will demonstrate during trial. Those representations and arguments can be hypothetical and abstract. Moreover, as trial progresses, new circumstances may arise that directly or indirectly alter the admissibility or evidentiary value of certain pieces of evidence. A trial judge has discretion either to deny or to postpone ruling on a pretrial motion until more information is available. See id. at 183, 674 P.2d 587 (so stating). Unless a trial judge rules otherwise, a trial judge's decision to deny a pretrial motion to exclude evidence does not preclude a party from later objecting to the admission of the evidence during trial, nor does it prevent the trial judge from later revisiting the issue before or during trial. Accordingly, even if a trial judge has denied a pretrial motion to exclude evidence, the moving party (and other parties) are well advised to consider making the same or other objections, if warranted, when a party offers the evidence during trial. An objection during trial allows a judge to reevaluate the issue of admissibility in light of what has occurred at trial, including whether the anticipated evidence or the parties' arguments have changed since the court denied the motion in limine. An objection at trial to the admission of certain evidence also may help refine the evidentiary issues for appellate review.
In this case, defendant challenges the trial court's pretrial order denying his motion in limine and does not assign error to the admission of the prior bad act evidence during trial proceedings. As noted, defendant made no objections at trial to the introduction of the evidence of defendant's prior bad acts. Defendant's motion in limine preserved the issue that he now raises for appellate review. This court previously has recognized that a defendant properly may assign as error the denial of a pretrial motion in limine to exclude arguably irrelevant or prejudicial evidence. State v. Madison, 290 Or. 573, 575–76, 624 P.2d 599 (1981). In Foster, this court determined that, notwithstanding a lack of later relitigation during trial of the same issue, an objection to evidence in a motion in limine preserves the issue of admissibility, noting that the moving party made “a sufficient offer of proof of what would happen in the trial to permit the court to rule intelligently on the proprietary of the offered evidence,” and “the judge gave a final ruling.” 296 Or. at 183–84, 674 P.2d 587; see also State v. Cole, 323 Or. 30, 35, 912 P.2d 907 (1996) (concluding that the lack of later relitigation of an issue decided prior to trial, “even where a statute permits relitigation on a discretionary basis, does not render any claim of error associated with the ruling unpreserved.”).
Finally, we note that defendant's reliance on the trial court's pretrial ruling for his assignment of error does affect the scope of the record that an appellate court may consider. If a party rests his or her argument on appeal on a trial court's pretrial order declining to exclude certain evidence, we ordinarily will evaluate that argument in light of the record made before the trial court when it issued the order, not the trial record as it may have developed at some later point. We do not rule out the possibility that that conclusion may change due to unique circumstances in another case, such as an agreement of the parties or other developments during trial. But in the usual case, we will evaluate a claim of pretrial error on the basis of the same record that the trial court relied on in making the challenged ruling. See State v. Perry, 347 Or. 110, 116–17, 218 P.3d 95 (2009) (noting the focused inquiry on appeal when the defendant did not object to discrete pieces of evidence during trial or assign error to any ruling occurring after the jury was empanelled, but instead rested his argument on the trial court's ruling following a pretrial OEC 104 hearing).
We turn to defendant's evidentiary arguments on review. OEC 404(3) governs the admissibility of prior bad acts evidence, and it prohibits the admission of evidence of “other crimes, wrongs or acts” if the evidence is offered to “prove the character of a person in order to show that the person acted in conformity therewith.” The rule, as applied to criminal trials, recognizes the long-standing principle “that a defendant should not be convicted because he is an unsavory person, nor because of past misdeeds, but only because of his guilt of the particular crime charged.” Christopher B. Mueller & Laird C. Kirkpatrick, 1 Federal Evidence § 4:21, 691–92 (3d ed 2007). As applied to this matter, the rule prohibits the admission of evidence offered to prove that, because defendant may have sexually abused young girls before, he subsequently acted in conformity with that behavior.
Nonetheless, OEC 404(3) permits the introduction of evidence bearing on a person's character if the state offers it for a noncharacter purpose. OEC 404(3) provides a nonexclusive list of permissible purposes for which prior bad act evidence may be admitted at trial which, as pertinent to this case, includes proof of intent, absence of mistake or accident, and identity. See Johns, 301 Or. at 544, 725 P.2d 312 (observing that OEC 404(3) is a rule of inclusion). OEC 404(3) “allows Oregon judges to resort to any theory of logical relevance * * * that does not run afoul of the ‘propensity to commit crimes or other acts' prohibition.” Johns, 301 Or. at 548, 725 P.2d 312.
In examining evidence to be admitted under the categories or exceptions to OEC 404(3), a trial judge first must determine whether the prior act is relevant to the issues being tried. Johns, 301 Or. at 549, 725 P.2d 312; see also State v. Pinnell, 311 Or. 98, 109–10, 806 P.2d 110 (1991) (preliminary ruling concerning the admissibility of prior bad act evidence is a question of logical relevancy which the trial judge must decide under OEC 404(1)). As this court has explained, “[t]he admissibility of evidence of other crimes must not be based upon the relationship of the evidence to one of the listed categories, rather it must be based on its relevancy to a fact at issue in the trial * * * other than proving a propensity to commit certain acts.” Johns, 301 Or. at 549, 725 P.2d 312. The party offering the evidence has the burden of showing that the proffered evidence is relevant and probative of some noncharacter purpose. State v. Garrett, 350 Or. 1, 6, 248 P.3d 965 (2011); State v. Pratt, 309 Or. 205, 210, 785 P.2d 350 (1990). We will reverse the trial court if the trial court committed legal error in concluding, based upon the evidence before it on defendant's motion in limine, that the evidence was relevant.5 See State v. Titus, 328 Or. 475, 481, 982 P.2d 1133 (1999) (relevance determination reviewed for errors of law).
As noted, the trial court reasoned that defendant's uncharged Lane County misconduct was relevant to prove intent, absence of mistake or accident, and identity, and the Court of Appeals affirmed based on identity alone. Turning to the Court of Appeals opinion first, we agree with that court's conclusion that the evidence was not properly admitted under the traditional identity exception recognized by this court. That exception permits the admission of evidence proving that a defendant operated in a novel or distinctive manner that identifies that defendant as the perpetrator. See State v. Johnson, 313 Or. 189, 196, 832 P.2d 443 (1992) (prior bad acts offered to prove identity by modus operandi requires “a very high degree of similarity between the charged and uncharged crimes” as well as a distinctive methodology “so as to earmark the acts as the handiwork of the accused”); Johns, 301 Or. at 551, 725 P.2d 312 (“[T]o prove identity the prior acts must be a ‘signature’ crime.”). Mere repetition of crimes of a similar class, such as repeated assaults or robberies, without more, is not sufficient. Pinnell, 311 Or. at 109–10, 806 P.2d 110. Rather, there must be some unique and repeated feature that makes the act stand out. Id. The acts at issue here fail to meet that requirement.
However, as noted, the Court of Appeals went beyond the stated modus operandi exception to recognize that evidence can be admitted to prove identity if it is relevant to bolster a witness's identification. More specifically, the court found that A's ability to recognize and identify defendant as the party who had abused her on this occasion could be supported by evidence that, on a past occasion, (1) defendant had abused A; (2) A had seen defendant abuse R; and (3) defendant had abused R. The Court of Appeals erred in applying that rationale. In our view, bolstering A's identification of defendant with defendant's prior uncharged conduct against A and R constituted an impermissible propensity purpose under these circumstances, because the reasoning relies on an impermissible character inference about defendant. See Charles Alan Wright & Kenneth W. Graham, Jr., 22 Federal Practice and Procedure: Evidence § 5246, 515 (1978) (use of prior bad acts evidence to bolster witness's identification is not a legitimate use of other crimes evidence under FRE 404(b)).
Taking the evidence of defendant's uncharged misconduct against A first, the Court of Appeals reasoned that evidence that defendant sexually abused A on a prior occasion tends to show that A's identification of defendant as her abuser on the occasion charged was accurate. Although that reasoning may be properly applied in some circumstances, in this case, defendant and A had continuous, ongoing contact with one another. Thus, given A's familiarity with defendant, the force of the “bolstering” evidence rests primarily on an inference that, on the occasion charged, defendant acted consistently with his character to sexually abuse A. That inference constitutes an impermissible propensity purpose that goes beyond proving that A could recognize and relay information about who had abused her. Evidence that A previously had witnessed defendant abuse R and that R had reported being abused, offered to bolster A's ability to identify defendant on the occasion charged, similarly relies on an impermissible inference that defendant acted consistently with his asserted character to abuse young girls.
Because we conclude that evidence of defendant's prior bad acts against A and R was not properly admitted to prove identity, we must examine whether the evidence was properly admitted to prove intent and absence of mistake or accident.6 On that issue, defendant argues—as he did in his argument in support of his pretrial motion—that his uncharged Lane County conduct was irrelevant for those purposes because his defense was that the charged acts never occurred and, if the acts did occur, he did not commit the acts. In defendant's view, had he admitted committing the underlying charged acts, but then claimed that he had acted, for instance, by mistake or inadvertence, then intent would be directly at issue during trial and the prior misconduct evidence would be admissible to establish intent. By not admitting the charged act, defendant claims that the first question for the jury in this case was whether he committed the sexual act charged. Until the record established that fact, defendant argues, the prior misconduct evidence offered to prove that he acted intentionally was irrelevant, except under the impermissible theory that the charged acts had occurred because defendant had a propensity to commit such acts. Stated differently, defendant asserts that the relevance of evidence of intent thus depends upon a prior finding of fact or a stipulation that establishes that defendant committed the charged act.
In State v. Leistiko, 352 Or. 172, 202 P.3d 857 (2012), this court held that other crimes evidence is relevant to prove that a defendant acted intentionally when the record establishes, in fact, that the defendant committed the charged act. See also State v. Gailey, 301 Or. 563, 568, 725 P.2d 328 (1986) (holding that evidence that the defendant possessed stolen property from a separate burglary was not admissible to prove that the defendant intended to burglarize the victim's home when the defendant had not admitted the act—i.e., that he was in the victim's home). More specifically, when this court decided Johns, the court based its reasoning on the “doctrine of chances,” as distilled from the work of John Henry Wigmore and Edward J. Imwrinkelried. Under that doctrine, a jury may infer that multiple acts of repeated conduct make it unlikely that the person who performed those acts acted inadvertently. Johns, 301 Or. at 552–55, 725 P.2d 312. As this court explained in Leistiko, application of the doctrine of chances assumes that the act has occurred “either because (as usually) it is conceded, or because the jury [is] instructed not to consider the evidence from this point of view until they find the act to have been done.” 352 Or. at 184, 282 P.3d 857 (quoting John Henry Wigmore, 2 Evidence § 302, 245 (Chadbourn rev.1979)). When the act is so established by the defendant's stipulation, or when the trial court instructs the jury that it first must find from the evidence that the alleged act occurred before it may consider the evidence of other crimes to determine intent, uncharged misconduct evidence that is sufficiently similar to the charged acts can be admitted to prove that the defendant acted with a culpable state of mind. Id. at 184–85, 282 P.3d 857. In the absence of a stipulation by the defendant, unless the state first introduces evidence sufficient to allow the jury to find that the charged act occurred, a court cannot properly admit the uncharged misconduct evidence as relevant to prove intent.7
The state argues that, even absent proof that the charged act occurred, defendant's not guilty plea alone was sufficient to make the evidence of uncharged misconduct relevant. The state contends that a plea of not guilty results in “a denial of every material allegation in the accusatory instrument.” ORS 135.370. Therefore, the state claims that it had the burden to prove each element of the charge and that the prosecution, not the defense, controls how it will prove its case. We disagree with the state's view. In making those assertions, the state relies on State v. Leland, 190 Or. 598, 630, 227 P.2d 785 (1951), aff'd sub nom Leland v. Oregon, 343 U.S. 790, 72 S.Ct. 1002, 96 L.Ed. 1302 (1952), in which this court stated that the prosecution has the right to prove its case “up to the hilt and to choose its own way of doing so * * *, subject only to the rules of evidence and the standard of fair play which should govern the prosecution of every criminal case.” (Emphasis added.) Here, as noted, admitting evidence of uncharged misconduct when the record did not contain evidence sufficient to permit the jury to find that the charged acts had occurred violates the evidentiary rule prohibiting the use of prior bad acts to prove propensity. A defendant's not guilty plea does not obviate that rule.
When the trial court ruled on defendant's motion in limine, it applied the doctrine of chances, as construed in Johns, and determined that the uncharged acts were sufficiently similar to the charged acts to be relevant to prove intent. However, at the time of the ruling, the trial had not occurred and, consequently, the record contained no evidence sufficient to support a finding that the charged acts had occurred. Defendant had not stipulated that he had touched A's genital area in Clatsop County. Before trial, and in the absence of a stipulation by defendant that he had touched A's genital area in Clatsop County, evidence of defendant's uncharged misconduct involving A and R in Lane County was only conditionally relevant. That is, the uncharged misconduct evidence might have become relevant if certain conditions had occurred concerning the trial court record. First, in the absence of defendant's stipulation, the state would have to introduce evidence at trial sufficient to permit the factfinder to find beyond a reasonable doubt that, in fact, defendant had touched A's genital area in Clatsop County, as charged.8 Second, the court would have to instruct the jury that it could not consider the evidence of defendant's uncharged misconduct for any purpose unless it first found as a fact that defendant had touched A's genital area in Clatsop County, as charged. If the record demonstrated that those conditions were satisfied, then the court could allow the state to offer and the jury to consider the evidence of defendant's uncharged conduct for the purpose of deciding whether defendant had committed the charged acts intentionally.
The trial court, however, did not decide that the evidence of defendant's uncharged misconduct was conditionally admissible in the manner that we have described above. It ruled instead, without conditions, that the uncharged misconduct evidence was relevant and admissible. In light of the record that existed when the trial court addressed the motion in limine, the court's ruling was erroneous.
The state also cites State v. McKay, 309 Or. 305, 787 P.2d 479 (1990), for the proposition that prior bad acts may be admitted to demonstrate a defendant's sexual predisposition for a particular victim. In McKay, this court indicated that a trial court could properly admit evidence of the defendant's past sexual contacts with a minor female victim under OEC 404(3) because that kind of uncharged misconduct evidence could “demonstrate the sexual predisposition this defendant had for this particular victim, that is, to show the sexual inclination of defendant towards the victim, not that he had a character trait or propensity to engage in sexual misconduct generally.” Id. at 308, 787 P.2d 479. Before the trial court, the state did not advance an argument that the uncharged misconduct evidence involving A was admissible under OEC 404(3) and McKay as evidence of defendant's sexual predisposition toward a specific victim, A, and the trial court did not base its decision to admit the evidence on that theory. The trial court did not instruct the jury to consider that evidence for the permissible, noncharacter purpose of deciding whether defendant was sexually predisposed toward A, and thus, that potential theory of admissibility is unavailable as an alternative basis for affirmance. Finally, even if McKay somehow applied to defendant's uncharged misconduct regarding A, that would not justify admitting evidence concerning defendant's uncharged misconduct with R. Defendant's asserted sexual misconduct with R does not tend to prove that he had a sexual disposition toward a different individual, A.
The final question for this court is whether the trial court's error was harmless. Defendant argues that it was not harmless because the state introduced and relied on the prior bad acts evidence throughout the trial. The state, for its part, asserts that any evidentiary error was harmless, but the state fails to explain that argument. We agree with defendant. The main issue in this case was whether the charged abuse occurred. The court's evidentiary ruling permitted the jury to consider the uncharged misconduct evidence before it decided whether defendant had committed the charged acts. That created a risk that the jury would use the uncharged misconduct evidence for an impermissible propensity purpose—i.e., to decide that, because defendant had committed the uncharged acts, his character was such that he again would act in the same manner and commit the charged acts. Thus, this is not an occasion where there was little likelihood that the error affected the verdict. See, e.g., State v. Naylor, 291 Or. 191, 196–98, 629 P.2d 1308 (1981) (applying Or Const, Art VII (Amended), § 3); State v. Davis, 336 Or. 19, 34–35, 77 P.3d 1111 (2003) (same).
The decision of the Court of Appeals is reversed. The judgment of the circuit court is reversed, and the case is remanded to the circuit court for further proceedings.