dissenting:
This appeal involves a brutal and shocking crime. Nevertheless, we are charged with the duty of assuring that the defendant receives a fair trial. I dissent because I conclude that the introduction of the testimony of Evelena Hohl was so unreliable and so prejudicial as to have deprived Colley of a fair trial.
The majority’s brief discussion of the issue of evidence of another alleged rape simply affirms the holdings of all the courts below, that Hohl’s testimony was relevant to prove Colley’s intent and identity. This court has held, however, that evidence of other crimes or wrongful acts is not looked upon with favor, and must be carefully scrutinized to determine its probative value. United States v. Back, 588 F.2d 1283, 1287 (9th Cir.1979). I believe that the admission of Hohl’s testimony to prove Colley’s intent and identity cannot withstand careful scrutiny.
Colley’s defense was based on the contention that he did not attack Ms. Jensen. He does not argue and did not testify that he attacked her but without the intent to commit rape. While intent is technically placed in issue by a plea of not guilty, the government apparently realized that Hohl’s highly inflammatory testimony was too marginally relevant to introduce as part of its direct case. The testimony, however, served no permissible purpose as rebuttal. Hohl’s testimony was presented after Colley’s testimony. In the absence of any defense testimony indicating an attack without the requisite intent, there was nothing to rebut on that issue. In any event, any probative value underlying the admission of Hohl’s testimony for the purpose of showing Colley’s intent was far outweighed by the danger of unfair prejudice.
Likewise, the admission of Hohl’s testimony to prove Colley’s identity cannot withstand that careful scrutiny and the balancing of probative value with unfair prejudice which our prior decisions have required. There was no critical need for Hohl’s testimony for purposes of identifying Colley as Jensen’s assailant because there was no serious question of Colley’s identification by Jensen. Colley and Jensen were casual acquaintances. Colley brought Jensen to the hospital. Jensen readily identified Colley as her attacker. This is not a case where the government is attempting to prove the identity of a masked or otherwise unidentified assailant by offering proof of the defendant’s past bad acts, committed in an identical manner, as probative of the defendant’s identity. The proper question in this case does not concern Jensen’s identification of Colley, but whether Colley committed the assaults upon Jensen for which he was charged.
The majority finds Hohl’s testimony probative of identity because it suggests Colley possessed a unique modus operandi. The evidence, however, does not show any unique modis operandi. The crime for which Colley was charged and tried, and the alleged attack upon Hohl do not constitute “signature” crimes. Colley was a casual acquaintance of Jensen’s; according to *992her testimony, he had asked her to go driving on the night of the attack. There is no evidence that Colley knew Hohl; she, testified that she had asked him for a ride prior to her alleged attack. According to her testimony, Hohl was raped; Jensen was not, although nothing intervened to prevent her from being raped had her assailant chosen to do so. Jensen was knifed repeatedly; Hohl was not. The location of the alleged Hohl attack, though close, was not the same as that of the Jensen attack. The closeness in time between the two attacks is suspect, given that Hohl did not report her alleged attack until after reading about Jensen’s attack in the paper, eight days later. The similarities between the two events were limited to the grabbing of hair, choking the victim, and expression of remorse after the attack. These cannot be said to be signature characteristics “sufficiently distinctive” from numerous other sexual assaults committed by persons other than the defendant.
“An inference of identity from prior crimes can only arise when the elements of the prior offenses and the charged offense, singly or together, are sufficiently distinctive to warrant an inference that the person who committed the prior offense also committed the offense on trial.” United States v. Powell, 587 F.2d 443, 448 (9th Cir.1978). Thus, this circuit has allowed such evidence when the acts were strikingly similar. United States v. Bailleaux, 685 F.2d 1105,1109-10 (9th Cir.1982) (“bizarre” extortion attempt in both cases based on poisoning food in store; poison used, wording of notes, payment drop-off procedure, and contact via radio station similar in each case); United States v. Andrini, 685 F.2d 1094, 1096-97 (9th Cir.1982) (method of starting fire by filling plastic bleach bottle with gas, puncturing, stuffing in rag, and lighting rag similar). In contrast, however, we disallowed such evidence when the characteristics of the two acts were not “distinctive, but [were] similar to numerous other crimes committed by persons other than the defendant.” Powell, 587 F.2d at 448 (marijuana smuggling; in both cases large amounts of the drug were stored in the smugglers’ residence; insufficiently distinctive); United States v. Webb, 466 F.2d 1352, 1353 (9th Cir.1972) (robbery one: two men, ski masks, used pistol; robbery two: one man, different kind of mask, used pistol; insufficiently distinctive). For the reasons given above, I believe that under careful scrutiny, the evidence is marginal for the purpose of showing any unique modus operandi. Whatever minimally probative value Hohl’s testimony had to show intent and identity was far outweighed by the danger of unfair prejudice.
“Unfair prejudice” has been defined as an undue tendency to suggest decision on an improper basis, as for example, an emotional basis. Back, 588 F.2d at 1285 (quoting the Notes to the Advisory Committee on the proposed Fed.Rules of Evidence).
Unfair prejudice results from an aspect of the evidence other than its tendency to make the existence of a material fact more or less probable, e.g., that aspect of the evidence which makes conviction more likely because it provokes an emotional response in the jury or otherwise tends to affect adversely the jury’s attitude toward the defendant wholly apart from its judgment as to his guilt or innocence of the crime charged.
Bailleaux, 685 F.2d at 1111 (emphasis in original). Of course, the use of prior bad act evidence will always be prejudicial; indeed, all evidence which tends to establish the guilt of a defendant is, in one sense, prejudicial to the defendant, but that does not mean that such evidence must be excluded. Id. When the probative value of the evidence is substantially outweighed by the danger of unfair prejudice, however, such evidence, although relevant, should be excluded. Id. We have recognized that no matter how conscientious jurors are, they will find it difficult to compartmentalize their mental processes and consider only the probative aspects of evidence while disregarding the fact that a defendant has a bad character or a criminal record as shown by his prior bad acts. Id. at 1112. The jury in this case could not help but be *993emotionally and adversely affected by Hohl’s testimony that Colley raped her.
Such evidence becomes even more suspect and disfavored when its source is unreliable. In this case, Hohl did not come forward until after she read of Colley’s arrest for Jensen’s attack in the newspaper, eight days after Hohl’s alleged rape. She testified that after she had driven with Colley to the location where he allegedly raped her, she had been drinking whiskey out of an “Oly” beer bottle prior to the attack, and that the bottle was left at the scene. She led police to the site by finding the only beer bottle in the area, but that bottle was found to bear a “Miller” label. Hohl, an admitted abuser of alcohol and drugs, also testified to several suicide attempts and repeated treatments by a psychologist. The state produced no corroboration of her testimony. For these reasons, I believe that she must be viewed as an unreliable witness, and her testimony typical of that which any troubled and unstable person might give. There is no suitable response to such testimony other than outright denial. At the same time, the efforts to refute the damaging testimony detract from the defense of the charged crime.
It is also of significance that the original information charged Colley with the Hohl offense as well as the Jensen offense. As far as our record indicates, he was not tried on those charges. Yet the state brought the same facts in through the back door at the Jensen trial without the necessity of proving the Hohl offense beyond a reasonable doubt. Finally, I observe that without Hohl’s testimony, Colley’s first trial for the Jensen offense resulted in a hung jury.
Evidence of commission of a prior offense to prove a defendant’s identity and intent to commit a crime should be limited to cases where there is a real and immediate need to prove the defendant’s intent and identity, where the evidence clearly establishes intent and identity, and where there can be no serious question as to the reliability of the evidence that the defendant committed the prior bad act. Compliance with this standard will ensure that probative value truly outweighs unfair prejudice. I must conclude that Hohl's testimony was so unreliable and so prejudicial that the state court’s refusal to exclude her testimony was fundamentally unfair.