I concur. The majority opinion fully justifies admitting evidence of the Michigan crime. I write separately to state an additional reason the evidence was admissible. It corroborated the complaining witness’s testimony in this case.
Sex crimes, more than others, tend to be committed in private. Usually, the only witnesses are the complaining witness and the accused. Often the victim testifies the defendant raped her, and the defendant testifies she is lying. Often, there is no physical evidence, especially when, as here, the defendant admits the intercourse but claims consent. In that situation, who should be believed? This can present difficult credibility questions for the jury, especially if there is no evidence corroborating either version. It is therefore particularly important that when corroborating evidence does exist, the jury be allowed to consider it. In making the momentous decision of guilt or innocence of rape, the jury needs all the evidence available; it should not be handicapped by withholding probative evidence supporting one party or the other.
Evidence that defendant committed rape under somewhat similar circumstances in Michigan a short time after the disputed events of this case strongly supports the inference that the victim’s testimony is the truthful one, not defendant’s. It might be coincidence that the complaining witness in this case and the victim in Michigan both claimed defendant raped them, and that both accounts contain common details. But that is unlikely. If a person claims the defendant committed rape, and the defendant denies it, the complaining witness might be lying. If, however, two people claim rape, and if their stories are sufficiently similar, the chance that both are lying, or that one is truthful and the other invented a false story that just happens to be similar, is greatly diminished. The jury can reasonably, and quite properly, infer that it is more likely both are truthful.
The more similar, and the more independent, the two accounts are, the greater the strength of the corroboration. The corroboration is most compelling when two persons who did not know each other, and could not have heard each other’s story, independently tell of a distinctive, unusual method *429of operation. If, for example, each victim independently stated that the defendant taunted them after the crime, gave them a quarter, and dared them to use it to call the police, the chance that each invented the same story would be almost nonexistent. The jury should be allowed to consider such highly relevant evidence.
When there are multiple similar accounts of rape, each is more likely to be true. Each mutually corroborates the other. This is a matter of logic, not prejudice. Evidence of the Michigan crime here does not merely show defendant is a bad person, or that he has a propensity to commit crimes; rather, it helps the jury decide who is telling the truth and to determine the facts. It tends in reason to show that defendant did in fact commit rape in this case.
The doctrine of corroboration is discussed at length in People v. Thomas (1978) 20 Cal.3d 457, 468-470 [143 Cal.Rptr. 215, 573 P.2d 433] (Thomas). There, we cited Court of Appeal cases that “suggest that evidence of prior sex offenses involving victims other than the prosecuting witness is generally admissible for the purpose of corroborating such witnesses without regard to the remoteness of the prior offenses or the lack of close similarity to the charged offense.” (Id. at p. 468.) We found that the rule stated in those cases “is too broad and must, to that extent, be disapproved.” (Ibid., italics added.) But then, as stated in People v. Pendleton (1979) 25 Cal.3d 371, 378 [158 Cal.Rptr. 343, 599 P.2d 649], we “reaffirmed” the doctrine when correctly applied.
We stated in Thomas that People v. Covert (1967) 249 Cal.App.2d 81 [57 Cal.Rptr. 220] “properly recognized that by reason of the unique circumstances of privacy and seclusion surrounding the commission of most sex offenses the determination of witness credibility plays a central role. Covert thereby found justification for the admission of evidence of defendant’s other sex offenses as corroborative of the prosecuting witness’ version of the event. We confirm, to an extent at least, the propriety of the Covert thesis, namely, that evidence of similar, nonremote offenses involving similar victims, declared admissible in [People v. Kelley (1967) 66 Cal.2d 232 [57 Cal.Rptr. 363, 424 P.2d 947] and People v. Cramer (1967) 67 Cal.2d 126 (60 Cal.Rptr. 230, 429 P.2d 582)] to show a common design or plan, may also assist in corroborating the prosecuting witness’ version of events. Yet, we sense an inherent danger. Were the theory to be held applicable in all sex offense cases, without regard either to remoteness or similarity, the ‘corroboration’ exception would absorb the general rule of exclusion in its entirety, permitting introduction of all prior sex offenses for purposes of corroborating the prosecuting witness.” (Thomas, supra, 20 Cal.3d at pp. 468-469, all but the final italics added.)
*430Because of these concerns, Thomas stated that “the remoteness and similarity restrictions expressed in Kelley and Cramer have particular application.” (Thomas, supra, 20 Cal.3d at p. 469.) We disapproved other cases that applied the corroboration doctrine, but only to the extent those cases conflict with the “principles” stated in Thomas. (Id. at pp. 469-470.) By citing People v. Kelley (1967) 66 Cal.2d 232 [57 Cal.Rptr. 363, 424 P.2d 947] and People v. Cramer (1967) 67 Cal.2d 126 [60 Cal.Rptr. 230, 429 P.2d 582], the Thomas court indicated that the degree of similarity needed for purposes of corroboration is like that for evidence showing a common design or plan. (See People v. Ewoldt, ante, 380 at pp. 395-396 [27 Cal.Rptr.2d 646, 867 P.2d 757].)
The doctrine of corroboration is closely akin to another rationale for admitting other crimes evidence that we have recognized recently—the doctrine of chances. In People v. Robbins (1988) 45 Cal.3d 867, 879-880 [248 Cal.Rptr. 172, 755 P.2d 355], we quoted Wigmore’s explanation of the doctrine: “ ‘the doctrine of chances—the instinctive recognition of that logical process which eliminates the element of innocent intent by multiplying instances of the same result until it is perceived that this element cannot explain them all. Without formulating any accurate test, and without attempting by numerous instances to secure absolute certainty of inference, the mind applies this rough and instinctive process of reasoning, namely, that an unusual and abnormal element might perhaps be present in one instance, but that the oftener similar instances occur with similar results, the less likely is the abnormal element likely to be the true explanation of them.’ ”
This explanation applies here. When one person claims rape, the unusual and abnormal element of lying by the complaining witness may be present. But when two (or more) persons tell similar stories, the chances are reduced that both are lying or that one is telling the truth and the other is coincidentally telling a similar false story.
In People v. Tassell (1984) 36 Cal.3d 77, 84 [201 Cal.Rptr. 567, 679 P.2d 1], we dismissed the doctrine of corroboration by peremptorily stating that it was “laid to rest in People v. Thomas, supra.” We did not, however, lay the doctrine to rest in Thomas, supra, 20 Cal.3d 457; we merely explained it and limited its application to its proper realm. Justice Mosk argues that principles of stare decisis prevent us from overruling Tassell. (People v. Ewoldt, supra, ante at p. 408 (dis. opn. of Mosk, J.).) As the majority correctly notes (id. at p. 401, fn. 5), Tassell itself violated stare decisis by ignoring or tacitly overruling prior settled authority. (See, e.g., People v. Alcala (1984) 36 Cal.3d 604, 634, fn. 18 [205 Cal.Rptr. 775, 685 P.2d 1126] [Tassell “effectively overruled” Thomas and People v. Kelley, supra, 66 Cal.2d 232].) *431Tassell is the aberration, not the preceding line of authority that this court reaffirms today.
I thus agree with Thomas both in its affirmation of the doctrine of corroboration and in its expression of caution. Thus, evidence of other crimes that meets the similarity requirement for evidence of a common design or plan is also admissible under Evidence Code section 1101 to corroborate the complaining witness. As the majority opinion explains, however, such evidence, even if it meets the threshold requirements for admission, is subject to the trial court’s discretion to exclude evidence under Evidence Code section 352 if its probative value is substantially outweighed by its prejudicial effect. (Maj. opn., ante, at pp. 426-427.)
Trial courts should decide on a case-by-case basis whether to exclude as unduly prejudicial other crimes evidence that is otherwise admissible to corroborate the complaining witness. The court should weigh the probative value against the prejudicial effect. To do so, it should consider such factors as the degree of similarity, the remoteness of the other crime, the independence of the complaining witness from the victim of the other crime, whether the other crime is particularly inflammatory relative to the instant crime, and whether defendant has been convicted of the other crime.
Favoring admission is any factor increasing the probative value. The more similar and less remote the incidents, the more probative the evidence. Additionally, the more independent the victims, the less likely they influenced each other, and the greater the tendency of each to corroborate the other. For example, if two victims who never heard of each other each accuse the same person of raping them, and if each account contains some bizarre or otherwise unique element, the evidence would be extremely probative. Another factor supporting admission would be a conviction for the other crime. (Maj. opn., ante, at p. 427.)
Favoring exclusion is any factor reducing the probative value. The less similar and more remote the incidents, and the less independent the witnesses, the less probative the evidence. The court must also weigh the prejudice. Any evidence of other crimes contains some potential undue prejudice. (People v. Ewoldt, supra, ante, pp. 403-404.) When the other crime is particularly inflammatory relative to the instant crime, the prejudice may be greater. Additionally, when the other crime has not resulted in a conviction, that factor weighs against admission, although that would be but one of the factors to consider, and does not itself disqualify the evidence from admission.
Here, as the majority correctly finds, the court acted well within its discretion in admitting evidence of the Michigan crime. Defendant had been *432convicted of the crime, and its facts do not appear to be particularly inflammatory compared to those of this case. It is not remote, and the two victims, thousands of miles apart, were apparently completely unaware of each other when they reported the crimes.
As the majority explains, the two victims’ accounts also contain significant similarities. Not knowing what was to happen in Michigan, the victim in this case accused defendant of doing what he later did in Michigan, that is, stealing the victim’s automated teller machine card and forcing her to reveal her personal identification number during the incident in which he raped her. The doctrine of chances shows that it is less likely the victim here is lying than it would be without the Michigan evidence. Evidence of the Michigan crime enhances the credibility of the victim here.
For these reasons, in addition to the reasons stated in the majority opinion, evidence of the Michigan crime was properly admitted.