We address once again an area of the law of evidence with which the bench and bar appear to have continuing difficulty: the so-called similar-acts rule.1

Defendant was convicted by a jury on May 6, 1977, of first-degree murder and sentenced to life imprisonment. The Court of Appeals affirmed the conviction in a published opinion,2 prompting the defendant to file a request for review.3 In response, by order dated November 2, 1979, this Court appointed counsel to prepare and file an application for leave to appeal.

The single issue meriting attention is whether the trial court erred in admitting evidence of a separate and uncharged crime. After careful consideration we have concluded that the admission of the evidence was prejudicially unfair to the defendant and, pursuant to GCR 1963, 853.2(4), in lieu of granting leave to appeal, we reverse the decision of the Court of Appeals.

On November 1, 1976, a neighbor found Donald Mitchell’s body in the den of Mitchell’s condominium in the City of Novi. A 49-year-old bachelor, Mitchell had been strangled with his own bathrobe tie. He had been seen alive by his employer on October 29, 1976. Several items were missing from *305the condominium including a barometer, Fisher stereo speakers, a checkbook, and a credit card. Mitchell’s 1976 yellow Lincoln Continental automobile was also gone. The house was not in disarray, there was no sign of a break-in, and the neighbor who discovered the body had to use a key to enter the premises.

On October 30, 1976, the defendant, Golochowicz, appeared in Detroit at the home of Paul O’Clare and his brother Dennis. Defendant was driving Mitchell’s yellow Lincoln Continental and was also in possession of a pair of Fisher stereo speakers, a barometer and a checkbook, all identified at trial as belonging to the victim. The defendant later sold the year-old Lincoln Continental for $175. He also sold the vehicle’s spare tire, the stereo tape deck from the vehicle, the Fisher speakers and the barometer. He wrote three personal checks on the victim’s account and used the victim’s credit card to purchase a clock-radio.

The O’Clares both testified that on October 30, 1976, the day of his visit to them, the defendant offered to sell a Quasar television set and a Fisher stereo set to the O’Clares if they would accompany him to obtain the items. Paul O’Clare testified that the defendant claimed the items were in a condominium in the "Pontiac-Novi” area. The O’Clares declined the invitation.

At the time Mitchell’s body was discovered, a Fisher stereo set and Quasar television set were in the condominium. There was expert opinion testimony at the trial that fingerprints found in Mitchell’s home were the defendant’s.

On the basis of the so-called similar-acts statute, MCL 768.27; MSA 28.1050, the Michigan Rules of Evidence not having been adopted at the time of the trial, the prosecutor was able to introduce, *306over defense objection, the following testimony of Dennis O’Clare: On about November 4, 1976, the defendant contacted O’Clare and inquired whether he was interested in purchasing a Sony television set. O’Clare said he was interested. That prompted numerous telephone calls from defendant through the course of the night, encouraging O’Clare to obtain transportation so that the two of them could go to a certain location to examine the television set. Finally, at about 5 a.m., O’Clare took his roommate’s car and drove to the defendant’s home, and the two proceeded to a house on Muirland Street in Detroit where the defendant claimed the television set was located. After parking the car in the garage at defendant’s direction, O’Clare followed defendant to the back door of the house. Defendant opened the door without a key and told O’Clare to sit down in the living room, not to touch anything, and not "to go roaming” about. Defendant then went upstairs.

O’Clare testified that he went into the kitchen, turned on a light and, looking into the basement, saw a pool of blood at the bottom of the stairs. He went downstairs and discovered a man’s body lying face down with an electrical cord "coming from either side of his neck”. O’Clare further testified:

"Uhmmm, I went back upstairs. I yelled for Jerry and he couldn’t hear me. He was up on the second floor. So, like, I met him on the staircase of the second floor, just off the second floor landing, and he was carrying some — he had some guns he had taken out of a gun rack. I told him to set them on the floor. I told him, 'Come here. I saw a man lying on the floor.’ I said, 'I see someone on the floor. Somebody’s been killed on the floor.’ He just looked at me. I said, 'Come on. Just leave the stuff. Come here. I want to show you.’ We went back down the staircase and went down. We got half*307way down the stairs and I leaned over to show him. You could see the man’s legs halfway down the stairway — that’s when I first noticed him — but you couldn’t see any further than the man’s legs. He leaned over and said, 'Oh, let’s get out of here.’ ”4

The pair did not leave the home immediately. At defendant’s insistence, he and O’Clare removed the television set, some guns and calculators, lamps, and a number of other items from the home and put them in their car before finally leaving.

O’Clare testified further that the next day, while he and the defendant were driving to a pawn shop to sell some of the stolen goods, the defendant admitted that he had killed the man on Muirland Street and described how he had done so by beating the man over the head and then strangling him with an electrical cord from a coffee pot. Then, almost immediately, according to O’Clare, defendant denied that he had done it.

I

The question presented is whether the trial court committed reversible error in allowing into evidence Dennis O’Clare’s testimony concerning the Muirland Street homicide and related events, including defendant’s admission. The prosecutor offered this testimony pursuant to MCL 768.27; MSA 28.1050, which provides:

"In any criminal case where the defendant’s motive, intent, the absence of, mistake or accident on his part, or the defendant’s scheme, plan or system in doing an act, is material, any like acts or other acts of the defendant which may tend to show his motive, intent, the absence of, mistake or accident on his part, or the defendant’s scheme, plan or system in doing the act, in *308question, may be proved, whether they are contemporaneous with or prior or subsequent thereto; notwithstanding that such proof may show or tend to show the commission of another or prior or subsequent crime by the defendant.”

At the outset it should be recognized that this statute and its successor, MRE 404(b), stand as exceptions to the general rule of inadmissibility of evidence of a defendant’s other crimes. The policy consideration underlying genéral exclusion of similar bad-acts evidence for substantive purposes is the desire to avoid the danger of conviction based upon a defendant’s history of other misconduct rather than upon the evidence of his conduct in the case in issue.5

Given the potential for prejudice which inheres in the admission against a defendant of evidence of similar uncharged bad acts, the statutory and decisional exceptions to the general rule of exclusion of such evidence have been confined to a few narrowly defined circumstances and are required to meet a number of evidentiary safeguards to warrant proper admission into evidence. Those *309safeguards, accurately summarized by the Court of Appeals in People v Wilkins, 82 Mich App 260; 266 NW2d 781 (1978), and adequately supported by citation to precedent from this Court, required at the time of trial that before evidence of the defendant’s other misconduct may be admitted: (1) there must be substantial evidence that the defendant actually perpetrated the bad act sought to be introduced;6 (2) there must be some special quality or circumstance of the bad act tending to prove the defendant’s identity or the motive, intent, absence of mistake or accident, scheme, plan or system in doing the act and, in light of the slightly different language of MRE 404(b) we add, opportunity, preparation and knowledge; (3) one or more of these factors must be material to the determination of the defendant’s guilt of the charged offense; and (4) the probative value of the evidence sought to be introduced must not be substantially outweighed by the danger of unfair prejudice.

A

Applying the first of these safeguards to the present case, we find that there was substantial evidence that defendant killed the man found at the Muirland address. Testimony concerning the defendant’s familiarity with the premises, the fact that he was aware that he could enter the Muir-land home without a key, his instructions to witness Dennis O’Clare not to touch anything and not to roam about the house, his reaction upon being confronted with the fact that a dead man was lying in the basement, the subsequent removal of *310the deceased’s possessions, and finally the defendant’s admission, all combine to create a strong inference that defendant had previously been on the premises and had killed the man found in the basement.

B

The second and third of the safeguards described above present a more difficult problem.

The "special circumstance” of the second requirement refers to the relationship between the charged and uncharged offenses which supplies the link between them and assures thereby that evidence of the separate offense is probative of some fact other than the defendant’s bad character. Here, that fact is the identity of Mitchell’s killer, and the link is said to be the alleged uncommon similarity of the facts and circumstances surrounding the manner in which the two victims were killed, including the facts relating to the defendant’s access to the two residences and his theft and disposition of the victims’ property.

Where, as in this case, the only conceivable justification for admission of such similar-acts evidence is to prove the identity of the perpetrator, the link is forged with sufficient strength to justify admission of evidence of the separate offense only where the circumstances and manner in which the two crimes were committed are "[s]o nearly identical in method as to earmark [the charged offense] as the handiwork of the accused. Here much more is demanded than the mere repeated commission of crimes of the same class, such as repeated burglaries or thefts. The [commonality of circumstances] must be so unusual and distinctive as to *311be like a signature.” McCormick, Evidence (2d ed), § 190, p 449.7

It is because of the combined value of those two factors, the unique and uncommonly distinctive style employed by the defendant in committing the "substantially proved” uncharged similar oifense, and the same distinctive modus operandi employed in the charged oifense, that the jury is permitted to infer, if it believes the evidence, that both crimes were the handiwork of the same person, the defendant.8

It will not suffice that the "like act” be simply another crime of the same general category or even of the same specific character. It will not do simply to show, for example, that the defendant committed another murder. That information is likely to be used by an ordinarily reasonable juror for the very purpose for which evidence of bad character is required to be excluded, to show that *312the accused is a bad person who has murdered before and to invite the inference that he probably did so in this case. It is the uniqueness and the distinctiveness with which both crimes were committed, combined with the proof that the defendant committed the "like act”, that is the key. On that point this Court recently stated:

"It is the distinguishing characteristics which constitute the acts as similar within the meaning of MCL 768.27 and MRE 40403), not the fact that all constitute the same crime or are violative of the same statute. The distinguishing, peculiar or special characteristics which are common to the acts and thus personalize them are said to be the defendant’s 'signature’ which identifies him as the perpetrator * * *.” People v Major, 407 Mich 394, 398-399; 285 NW2d 660 (1979).

Consequently, if the trial court determines that there is substantial evidence that the defendant in fact committed the other or uncharged crime, it must then turn to the task of determining whether the manners or systems employed by the perpetrator of the uncharged crime and the crime in question were sufficiently "like” or "similar” and involved such distinctive, unique, peculiar or special characteristics as to justify an ordinarily reasonable juror to infer that both were the handiwork of the same person. If the trial court concludes the evidence is of that character, it may be admitted. If not, it is excluded. This is the essence of the second requirement of Wilkins, supra. We turn now to determine whether that requirement was met in this case.

Here a word is in order concerning the requirement that one or more of the distinctly different purposes named in the statute and evidence rule for the admission of similar-acts evidence must be *313material to assist, it is hoped, the bench an«i bar in understanding the analytical process which should inform the decision to admit or reject such evidence.

MCL 768.27; MSA 28.1050, now supplanted by MRE 404(b), declares that

"like acts or other acts of the defendant which may tend to show his motive, intent, the absence of, mistake or accident on his part, or the defendant’s scheme, plan or system in doing the act, in question, may be proved [in any criminal case in which one or more of those factors, are material].”

The now applicable Michigan Rule of Evidence, 404(b), is somewhat different and adds to the foregoing factors four more: "opportunity”, "preparation”, "knowledge”, and "identity”.

The comparable Federal Rule of Evidence, 404(b), is still different. It lists most of the factors stated in our statute and evidence rule except that it refers only to "plan” instead of "scheme, plan or system in doing an act”.

Of the factors listed in our statute, rule, and decisions, it should be understood that not all are on the same plane.

As noted in Stone, The Rule of Exclusion of Similar Fact Evidence: America, 51 Harv L Rev 988, 1026, fn 190 (1938):

"Intent, absence of mistake, and identity are facts in issue — facta probanda. Motive, plan, or scheme are facta probantia, and may tend to show any facta probanda.”

Thus it becomes clearer that while evidence of similar acts might be offered to prove intent, absence of mistake or accident, or identity directly, *314evidence of plan, scheme or system in doing an act might be directly in issue or might be provable as the threshold evidence from which could be inferred the identity of the perpetrator as the ultimate point to be proved.

No conceptual template can be constructed by any appellate court by which trial courts and counsel can mechanically test which, if any, of the purposes described above are material within the meaning of our statute or rule. It is for the trial judge, upon close inquiry of counsel, in the absence of the jury, to satisfy himself, and, in the process, create an appellate record, that one or more of the named factors are material.

When requesting admission for substantive purposes of evidence of other misconduct by the defendant, the prosecutor’s first duty is to identify, with specificity, the purpose for which such evidence is admissible by identifying which of the purposes named in MRE 404(b) justifies admission of such evidence as "material”. While in some instances the evidence might be admissible for more than a single purpose, ordinarily that is not the case. Usually, only one among the several purposes described in the rule is material to the case and justifies admission of such evidence. Consequently, if the prosecutor does not announce the purpose for which the evidence is offered, the trial judge should require the prosecutor to identify the specific basis in the rule justifying its admission: to show motive, opportunity, knowledge, preparation, intent, the absence of mistake or accident on the part of the defendant, or to prove the defendant’s scheme, plan or system in doing an act, or because scheme is itself directly material, or because it tends to prove the defendant’s identity as the perpetrator of the crime in question.

*315Experience in the trial courtroom and review of trial records on appeal suggests rather incontrovertibly that, when asked by the trial judge to specify the grounds for admission of similar-acts evidence, prosecutors often loose a "shotgun” fusillade of reasons which typically include most, if not all, of the purposes named in the statute.9 Such a response hints, of course, if it does not demonstrate, that the prosecutor has an inadequate understanding of the correct application of the rule and is unclear as to precisely why the evidence is or is not admissible.

Similarly, trial judges, when admitting evidence of other crimes, should avoid doing so with the vague justification that "I’ll let it in for what it is worth”, or "I’ll allow it to show plan, scheme or system” when that is the basis suggested by the prosecutor, without requiring a showing by the prosecutor as to how such evidence is relevant to show plan, scheme or system, or how plan, scheme or system is material to the case, or, most importantly, whether the evidence, if indeed relevant and material, is not more unfairly prejudicial than probative of the proposition for which it is offered.

As we have said before, evidence of other misconduct is not admissible in this state to negate mistake or accident, to prove motive, to show intent, to demonstrate the defendant’s plan, *316scheme or system, or to prove his identity, unless one or more of those factors are genuinely in issue —not "in issue” in the sense that criminal intent, identity, motive, lack of accident or some criminal plan are nearly always in issue to some greater or lesser degree in every case, but in issue or "material” in the sense that they are genuinely controverted matters. A genuine controversy exists concerning such matters when the defendant, either by counsel’s opening statement, a motion in limine, the nature of cross-examination by the defense, or evidence offered by the defense, has made one or more of them an issue actually disputed in the case.

In the case at bar, a motion for mistrial was made by defense counsel because the prosecutor, in his opening statement, detailed the evidence he intended to introduce concerning the defendant’s alleged guilt of the Muirland Street killing. When the mistrial motion was denied, defense counsel objected to the admission of evidence of the killing and the activities relating to the theft and sale of that victim’s property.

The prosecutor insisted, however, that the evidence was admissible under the "similar-acts statute”. In the course of his argument in support of the mistrial motion, defense counsel stated:

"I would like to know before we proceed further what the materiality and what particular fact [the prosecutor] is trying to prove”.

The prosecutor, in answer, named virtually all of the purposes which appear in the statute as well as the separate purpose of proving identification:

’’[The Prosecutor]: All right. Under 766.27 [sic], the *317people are allowed to use, from want of a better term, similar acts. Under that statute—

"The Court: It’s the same as 728.1050 [sic]?

’[The Prosecutor]: I have the MCLA cite. It states you can use that to show motive and intent.

"We have an open murder charge and the court is well aware of the definition of murder, arid it is the intention of killing another human being with malice aforethought. The Detroit homicide tends to be relevant as to justiñcation or accident, and also the defendant’s plan, scheme or system in doing an act.

"Here we have the defendant going to another home in Detroit and removing property from the home in Detroit with a Mr. O’Clare, and we return to our crime, the instant crime before the court. We do not have an eyewitness, but a homicide by ligature strangulation by the same kind found in Detroit, and the defendant is found with various items of personal property of the deceased. I state that goes to the plan or system of committing this crime. ” (Emphasis added.)

and,

"Once again we can use this pattern in Detroit to establish identification of the Novi homicide because once again the Novi, we have no eyewitness. We have circumstantial identification and we can use the plan or system as used in the Detroit homicide to establish identiñcation of the perpetrator of the Novi homicide.” (Emphasis added.)

Despite the prosecutor’s protestations that the evidence was admissible "to show motive and intent” and that the "Detroit homicide tends to be relevant as to justification or accident”, there was no basis whatever for admission of the evidence to prove motive or intent or to show justification or accident. There was no issue raised on the evidence in the case concerning the defendant’s intent to justify admission of the similar-crime evi*318dence on that ground. He did not, for example, claim accident, self-defense, insanity, or some other state of mind which suggests an innocent killing.

As the prosecutor appears ultimately to have recognized, the evidence of the Muirland Street killing could be admissible, if at all, only as tending to prove the identity of the killer of the deceased, Mitchell.

C

No eyewitnesses were able to place the defendant at the scene of the homicide in Novi at the time the deceased was killed. There was no direct evidence that the defendant was the killer. The identity of the killer was a critical element of the prosecution’s case. The only evidence pertaining to the defendant as the guilty party was entirely circumstantial: fingerprints of the defendant taken from the deceased’s residence, evidence of recent possession by the accused of the deceased’s property, sale of the property, use of the deceased’s credit cards, and the cashing by the defendant of three of the deceased’s checks. While, cumulatively, that was damaging circumstantial evidence pointing rather convincingly toward the accused as the killer and sufficient perhaps to sustain the conviction, it was such as to leave the identity of the killer the most doubtful element of the case and, consequently, a fact very much in issue. It was a point genuinely controverted, particularly in view of the defense counsel’s extensive and probing cross-examination of the fingerprint interpretation expert witnesses, whose testimony sought to place the accused at the scene of the killing, although whether before, during, or after the killing was not clear.

*319Clearly then, because the identity of Mitchell’s killer was material in the sense that it was a matter genuinely controverted, the issue of the applicability of the similar-acts statute was introduced into the case. Thus, the third safeguard was satisfied.

D

However, the fact that the killer’s identity was genuinely in issue and therefore material was not, of itself, a ticket for admission of evidence of the Muirland Street homicide and the circumstances surrounding it as proof of identity.

It was required, in addition, as we have indicated, that the court find from the evidence that the circumstances of the commission of the crime in question and that to which it was being compared as "like” or "similar”, both bore such unique, uncommon, and distinctive characteristics as to suggest the handiwork or signature of a single actor, the defendant.

In the case before us, we think distinctiveness of that quality was not shown.

Certainly there are features of commonality in the crimes, and with effort one can construct a list of similarities both as to what occurred and what did not occur.10

*320In both of the homicides, the victims were found dead with household items left wrapped around their necks, suggesting strangulation. In both cases, the defendant was seen in possession of the decedent’s personal property shortly after the homicide. In both cases, the defendant offered for sale a television set located at the premises of the deceased which would have to be retrieved with the help of another person. In both cases, other possessions of the deceased were offered for sale or stored at the same location. These are indeed similarities in the circumstances of the two homicides.

There are differences, however.

The defendant was on trial for murder. While the overall circumstances surrounding the commission of both crimes are relevant as bearing upon a common system in their commission, particularly the theft and subsequent sale of the deceased’s personalty, we look first to determine the degree of similarity and the uniquely distinguishing characteristics of the manner of commission of the crime charged — the homicide.

The evidence shows that the deceased, Mitchell, was bloodlessly garroted with the belt from his bathrobe. The cause of death, according to the medical examiner, was "asphyxia due to ligature strangulation”.

The deceased whose body was found at the Muirland Street address was discovered lying in "quite a bit” of blood. He had been severely beaten. According to the witness, O’Clare, the defendant stated:

*321'' 'Yeah, I hit him, you know. I hit him. I kept hitting him and, finally, he fell down and he wouldn’t lay there. I kept beating him and I had to run upstairs — I grabbed an electrical cord from a coffee pot and knelt on the back of his head or stood.’ He was holding him from behind the head and strangled him from behind, he said.”

While a factfinder would be warranted in concluding that both of the victims were strangled, the defendant’s statement to O’Clare does not describe the bloodless strangulation which the medical examiner described as the cause of Mitchell’s death.

While in both instances a television set and stereo equipment were stolen from the residence of the deceased, a very considerable amount of other personal property of various kinds was stolen from the two residences ranging from an automobile, a barometer, a credit card and blank checks from the Mitchell residence to lamps, rifle, shotguns, a desk calculator and slide projector from the Muir-land Street residence. The essential similarity between the two crimes then is in the theft and disposition of personal property from the homes of recently deceased bachelors, but there is nothing in the way the two crimes were committed, save the fact that both persons were strangled, although one had first been beaten and the other had not, which necessarily suggests the employment of a special, peculiar, or unique method so distinctive as to mark the killer of both victims as the same person. ,

It is, of course, a matter of judgment whether the fact that two unmarried male victims were both strangled, one bloodlessly and one after a beating, and their personal property of various kinds ranging from an automobile to a desk calcu*322lator stolen from their residences and later sold to friends of the defendant, is evidence of the existence of such distinctive characteristics common to both events as to suggest a personalized modus operandi. We think there was not the requisite "distinguishing, peculiar or special characteristics” of which we spoke in Major, supra. It is a close question, however, and close questions arising from the trial judge’s exercise of discretion on matters concerning the admission of evidence do not call for appellate reversal because the reviewing justices would have ruled differently. Reversal is warranted only if the resolution of the question by the trial court amounted to an abuse of discretion. The decision upon a close evidentiary question by definition ordinarily cannot be an abuse of discretion.11

*323Nevertheless, we conclude without hesitancy that admission of the evidence of the Muirland Street homicide was reversibly erroneous, not because we find an abuse of discretion in the determination that the two crimes were of the requisite similarity and bore the requisite distinctive "signature” of a single actor, but because we are persuaded that the evidence of the uncharged homicide was so unfairly prejudicial when weighed against its limited and tenuous probative worth that, in its admission, the defendant was denied a fair trial.

Because of the familiar rule that evidence of the accused’s uncharged misconduct is inadmissible for *324the unfair damage it is thought to do in most cases, the accused here was entitled to stand trial for the Mitchell killing alone, without introduction of evidence of a separate and uncharged homicide. He was presumptively entitled to have his guilt or innocence of the Mitchell killing determined by a jury evaluating evidence of that case alone, uninfluenced by the distracting, indeed horrifying, evidence of the defendant’s admission of a separate homicide unless, for compelling reasons, evidence of the other crime would be more helpful to assist the jury in determining the identity of the killer than it would be prejudicial to the defendant.

The identity of Mitchell’s killer was in doubt. The prosecutor was relying entirely upon circumstantial evidence that the defendant was guilty. The evidence had to be derived solely from inferences to be drawn from the accused’s unexplained recent possession of the deceased’s car and other items of personal property in circumstances strongly suggesting theft of them, plus the fingerprint expert’s opinion evidence of the defendant’s presence, at some time, inside the deceased’s condominium apartment. The result was that the identity of Mitchell’s killer was the weakest link in the prosecution’s case. For that reason, the prosecutor turned, and understandably so, to evidence of the Muirland Street killing to fortify, with further circumstantial evidence, his claim that the accused was Mitchell’s killer. We do not question the general proposition that similar-acts evidence may be properly admissible for such a purpose. But this case presents the classic example as to why trial courts should be

"stricter in applying [the] standards of relevancy when the ultimate purpose of the [evidence] is to prove identity or the doing by the accused of the criminal act *325charged than they are when the evidence is offered on the ultimate issue of knowledge, intent or other state of mind.” McCormick, Evidence (2d ed), § 190, p 452.

As a general rule, if the evidence of the accused’s identity as the perpetrator of the crime in question is strong or essentially uncontroverted, there is no need for evidence of other crimes to prove identity, which evidence is only circumstantial at best. On the other hand, if evidence of the identity of the criminal actor is weak or tenuous, revelation that he has committed an unrelated similar crime may, by reason of its tendency to distract the jury from the idéntification issue, tempt it to compromise or ignore that critical element of the case while focusing on the clearer proof of the defendant’s other misconduct. The dangerous result may well be, and indeed is likely to be, the jury’s conclusion that whatever the strength of the identification evidence in the case, the defendant is demonstrably a bad person and should be imprisoned anyway.

It is, at least in part, for those reasons that the trial court, when similar-acts evidence is offered to prove identity, should insist upon a showing of a high degree of similarity in the manner in which the crime in issue and the other crimes were committed. We regard the requirement to be a heavy burden upon the prosecution to show that the manner in which the crime charged and the other crimes were committed was marked with special characteristics so uncommon, peculiar and distinctive as to lead compellingly to the conclusion that all were the handiwork of the defendant because all bore his distinctive style or "touch”.

Even where evidence of that distinctiveness is produced, a trial court should be especially alert to the paramount consideration that ruling on the *326admission of such evidence is more a matter of a careful weighing process in the exercise of discretion than the mechanical application of a rule. The overriding policy is to protect the accused from unfair prejudice. That policy can be carried out only if the trial court is alertly sensitive to the need to balance the probative value of the proffered evidence against its prejudicial impact by weighing the likelihood that, as Professor McCormick put it, the jury will "be roused by the evidence to overmastering hostility”.12

In passing upon the admissibility of evidence of similar crimes for the purpose of proving identity especially, the manifest danger of unfairness suggests that the presumptive direction of the trial court’s exercise of discretion should be toward excluding the evidence. An ultimate decision in the opposite direction, to admit the evidence, should be taken only when thé trial court is convinced that the policy of the rule of exclusion will not be offended because, in view of the probative force of the evidence to prove identity, there is little or no danger that the jury, aided by a limiting instruction, will misconstrue the proper purpose of the evidence or, upon learning of the other crime, be stirred by "such passion * * * as to [be swept] beyond a rational consideration of [the defendant’s] guilt or innocence of the crime on trial”.13

In this case, the evidence of the manner of commission of the uncharged homicide, because of its manifest weakness to show the unique and distinctive hand of a single actor in both crimes, had limited probative force as proof of the defendant’s identity as Mitchell’s killer. The inevitable *327countereffect was to heighten the prejudicial impact of the evidence.

On hearing the testimony describing the defendant’s guilt of a separate and uncharged murder, the jurors can only have been deafened to the court’s instruction that the evidence might be considered only as another link in the chain of circumstantial evidence of the identity of Mitchell’s killer and not as a suggestion that the defendant was an admitted murderer and therefore probably guilty of the crime charged. As Justice Cardozo once said of comparably prejudicial evidence in another context, "[t]he reverberating clang of those accusatory words would drown all weaker sounds.” Shepard v United States, 290 US 96, 104; 54 S Ct 22; 78 L Ed 196 (1933).

For these reasons, we conclude that the evidence of the Muirland Street homicide was more unfairly prejudicial than probative of the identity of the deceased’s killer.

Consequently, we reverse the decision of the Court of Appeals, and remand for a new trial.

(concurring in part and dissenting in part). This case concerns whether it was error requiring reversal for the trial court to allow a witness to testify at trial about defendant’s perpetration of a different but similarly committed homicide under the similar-acts statute, MCL 768.27; MSA 28.1050. The purpose of the similar-acts testimony was to show the common identity of the perpetrator of the two murders by showing a similar distinctive "scheme, plan, or system in doing the [two] act[s]”. The importance of the similar-acts testimony was that the prosecutor’s *328other evidence tying defendant to the charged crime was wholly circumstantial, whereas there was an admission and other testimony tying defendant to the signature crime. The principal defense was a failure by the state to prove the charge beyond a reasonable doubt.

I

At the outset, let it be understood that this opinion does not disagree in any way with Justice Ryan’s helpful and masterful exposition of the underlying law. We accept that fully.

In particular, we accept Justice Ryan’s following four-pronged primary admission requirement:

"(1) there must be substantial evidence that the defendant actually perpetrated the bad act sought to be introduced; (2) there must be some special quality or circumstance of the bad act tending to prove the defendant’s identity or the motive, intent, absence of mistake or accident, scheme, plan or system in doing the act and, in light of the slightly different language of MRE 404(b) we add, opportunity, preparation and knowledge; (3) one or more of these factors must be material to the determination of the defendant’s guilt of the charged offense; and (4) the probative value of the evidence sought to be introduced must not be substantially outweighed by the danger of unfair prejudice.” (Footnote omitted.)1

*329The majority recognizes the satisfaction of prongs (1) and (3), so we will not consider them further. While we agree that the factual satisfaz tion of prong (2) is a close question, we believe, unlike the majority, that the positive case is stronger. As a consequence, in balancing the probative value of the evidence with the danger of unfair prejudice in prong (4), we more readily reach the conclusion that the probative value of the evidence is not substantially outweighed by the danger of unfair prejudice.

II. Distinctive Characteristics

The majority states that the necessary test before admission of such evidence is:

"that the court find from the evidence that the circumstances of the commission of the crime in question and that to which it was being compared as 'like’ or 'similar’, both bore such unique, uncommon, and distinctive characteristics as to suggest the handiwork or signature of a single actor, the defendant.”

In People v Kelly, 386 Mich 330; 192 NW2d 494 *330(1971), the defendant was convicted of rape and armed robbery. He burst into a Detroit hotel room at approximately 7 a.m. on August 26, 1967. He forced the victim at gun point to place a pillowcase over her head. He then raped her twice. During the attack, he wore no clothes. After the attack, he fled with the complainant’s tape recorder and $74 in cash.

At trial, a prosecution witness testified that the defendant had also appeared in a Detroit hotel room at 7:30 a.m. on March 16, 1968. The victim and her fiance were sharing the hotel room which they locked before going to sleep. After entering the room, the defendant forced the boyfriend into the bathroom at gun point. He locked the bathroom door, and he forced the victim to put a pillowcase over her head. He then raped her three times. Once again, he wore no clothes during the rapes. He also took a watch, a check and approximately $75 in cash.

In Kelly, we held that the testimony was properly admitted because the later act of the defendant tended to show a scheme, plan, or system in performing these similar acts under the similar-acts statute. In comparing the factual record of Kelly with the case at bar, we are supported in finding that the murders met the distinctive-characteristics prong. In both cases, evidence indicated that a systematic plan or scheme was employed by the defendants in brutally attacking the victims at different locations. In Kelly, the principal distinctive characteristic or signature of the defendant was that the women were forced to place pillowcases over their heads before being raped. In Golochowicz, it was the fact that both bachelor decedents were apparently strangled with household items. Moreover* in both cases, personal property *331of the victims was taken by the defendants, and, on each occasion that property was removed, it was not exactly the same.

Our conclusion that the accomplishments of the murders were etched with the signature of the defendant profits from a further comparison of the plans or schemes in the two cases. The defendant in Kelly entered two different Detroit hotel rooms without force and raped two different women several times. In addition, the defendant wore no clothing during the attacks. Similarly, the murders in Golochowicz were accomplished in accordance with a distinctive systematic plan or scheme. The defendant entered two different homes without force and strangled the bachelor decedents. In addition, the murders took place within a few days of each other, and the property stolen from the homes was either sold to the same individual or stored in the same location. In short, the holding and analysis employed in Kelly reinforces our holding that the murders in the case at bar met the distinctive-characteristics prong.

In People v Oliphant, 399 Mich 472; 250 NW2d 443 (1976), the defendant was convicted of rape and gross indecency. At trial, the prosecutor called three witnesses who testified that the defendant had raped them in the same manner as he had raped the complainant. There were a number of circumstances to the rapes which were repeated in each incident. There were also a number of significant differences among the various attacks. Nevertheless, this Court held that the testimony about the signature acts was admissible under the similar-acts statute.

The factual similarities between the charged crime and the signature crimes in Oliphant show that all four rapes occurred during a five-month *332period, involving college-age women. The contacts with the women were made by the defendant in public places and were initially friendly. In each case, race and marijuana were topics of conversation.

Moreover, all the women entered defendant’s car voluntarily. They rode with the expectation that the defendant was to take them to a specific place. A deviation was made, however, from the intended route with an excuse by the defendant not likely to cause fear in the victims. They were then driven to an area unfamiliar to them where the forced intercourse took place.

Before the intercourse took place, all four women were told to submit to the acts or they would be harmed by a weapon, although a weapon was never seen. The women gave in to defendant’s crudely communicated demands because of fear for their persons if they did not. None of their clothing was torn or ripped in the attacks.

In all of the incidents, the victims had an opportunity to escape from the defendant. They did not flee because at the time the opportunity arose they did not feel the need to do so since the defendant was still friendly. After the change in defendant’s attitude towards them, they had no opportunity to escape.

There were also a number of important differences between the charged crime and the signature crimes. The three testifying witnesses entered defendant’s car immediately. The complainant, however, entered the car soon after the meeting. Moreover, the four women were not all approached in the same way; the defendant met them under different circumstances. In two of the attacks, there was also an accomplice with the defendant.

*333Additional differences were that two of the women were riding in a car conversing with a friendly man when, upon coming to an area unfamiliar to them, he became threatening and demanding. The other two accepted a ride with a friendly man and accompanied him on an errand to an unfamiliar apartment, whereupon he became threatening and demanding.

Only one victim attempted to physically resist the defendant, and she was struck by him with his fist. Moreover, two of the women were forced to dance without clothing before the defendant raped them. Three were given a ride home, but one victim escaped from defendant’s car while being driven home. Finally, two of the women went to the police with knowledge of defendant’s name, address, college identification and car license number after the defendant voluntarily provided them with the information.

When comparing the factual records of Oliphant and Golochowicz, it becomes immediately apparent that the overall cumulative impact of the similar-acts evidence in the two cases is important in finding a common signature. Like the similar acts in Oliphant, the murders in the instant case were not committed with as unusual a distinguishing characteristic as the pillowcase over the head in Kelly. Nevertheless, the Court in Oliphant found the necessary distinctive characteristic to admit the evidence of similar acts, notwithstanding the significant differences which existed between the charged crime and the signature crimes. Thus, the holding in Oliphant, where important differences existed in defendant’s scheme or plan in the various rapes, is strong precedent for our finding that the distinctive-characteristics prong was met in the case at bar where the differences between the two murders were less significant.

*334In conclusion, under the distinctive-characteristics test established in the opinion of the majority, we believe that the similar and unique details which intersect the patterns of the homicides indicate that the murders were perpetrated by the same killer. Our review of the details common to the killings show that both victims were bachelors who had been found dead in their homes with household items wrapped around their necks, indicating death by strangulation. Both dwellings had been entered without force; the doors and windows were found intact. Moreover, personal property of the victims was removed at different times, sold to the same individual, or stored in the same location. In fact, a witness testified that the defendant had been in possession of personal property owned by both the decedents shortly after their killings. In addition, in both cases the defendant had offered for sale to the same individual the television sets owned by the victims.

In short, we believe that the special details common to both murders are so distinctive and unique as to mark the killer of both victims as the same person.

III. More Probative Than Prejudicial

The majority bases its reversal on the conclusion that the "evidence of the uncharged homicide was so unfairly prejudicial when weighed against its limited and tenuous probative worth that, in its admission, the defendant was denied a fair trial”. We respectfully disagree.

Since the proof in the instant case rested solely on circumstantial evidence, it is without question that a similarly committed murder where the *335defendant can be pointed to as the perpetrator would be relevant and probative evidence in a trial for a different murder. See State v Johnson, 210 Kan 288, 293-294; 502 P2d 802, 808 (1972). It is uncontested in the instant case that the similar acts in question were relevant. See MRE 401.

The question becomes whether the probative value of the similar acts is unfairly and substantially outweighed by their prejudicial impact. Unfair prejudice within the context of the rules of evidence means an undue tendency of a factfinder to be influenced in rendering a decision on an improper basis such as an emotional one.2

What would be the natural impact of the evidence of the signature murder? There is no reason to believe that it would unduly excite the jury any more than the charged murder, because both involved murder by strangulation. It is true that the signature murder was bloody and the charged murder was not, but there is nothing to indicate that the signature murder was provokingly gruesome.

What the jury would naturally be struck by is the similar modi operandi, the numerous and strikingly similar common elements: that the murderer entered the victim’s home without force, that the victim was a bachelor living alone, that the murder was accomplished by strangulation by wrapping a household object around the victim’s neck, that personal property was removed piecemeal from the victim’s residence, that defendant remained in possession of some of the victim’s property, and that defendant offered some of the *336victim’s property for sale — in fact, offered property of both victims for sale to the same person.

It would appear natural that the jury would have its attention attracted to the several points of similarity between the two crimes rather than be excited or provoked by the nature of the signature crime which, after all, was little different from the one charged. As a consequence, the evidence would tend to effect its legitimate function of calling attention to a common signature rather than emphasizing that the defendant was a bad man.

In reaching a decision on whether to exclude evidence on grounds of unfair prejudice, the trial judge must give consideration to the probable effectiveness or lack of effectiveness of a limiting instruction.3 Such a determination lies solely within the province of the trial judge’s discretion, and in the instant case a forceful instruction was given.

We agree with the majority that People v Wilkins, 82 Mich App 260; 266 NW2d 781 (1978), establishes the proper balance of factors in determining whether similar-acts evidence is more probative than prejudicial. The Court of Appeals, in relying on People v Oliphant, 399 Mich 472, 490; 250 NW2d 443 (1976), and other Michigan cases, stated that a trial court in determining whether similar acts are too prejudicial, should consider many factors, inter alia, “the availability of less *337prejudicial sources of proof, the necessity of the evidence to prove an element of the prosecutor’s case, the defendant’s theory of the case, the tendency of the evidence to inflame the passions of the jury, and its potential for confusing issues in the case”. 82 Mich App 270-271.

Applying these factors to the case at bar, we find that the admission of the similar-acts testimony was not more prejudicial than probative. First, there were no other sources of proof less prejudicial because the prosecutor could only marshal circumstantial evidence to prove his case. Second, it is clear from the circumstantial nature of the other evidence that the similar-acts evidence was necessary in proving the prosecutor’s case.

Third, the defendant primarily attacked the credibility of testimony establishing the circumstantial links of the murder to the defendant. Defense counsel rested on the argument that the state had not proved beyond a reasonable doubt that the defendant had committed the murder with which he was charged. It is very doubtful that the similar-acts evidence would have provoked a purely emotional decision by the jury because (1) the striking aspect of the evidence was the similarity of the cumulative details of the defendant’s modus opérandi in committing the murders, and (2) the defendant’s theory of the case was that the state had not met its burden of proof which would have led the jury to focus on a number of elements rather than exclusively focusing on defendant’s character.

The similar-acts evidence, as does all relevant evidence, causes prejudice when it implicates a defendant. The impact of this evidence, however, *338could not have led the jury to decide the case on a purely emotional basis because the defendant had not raised non-participation as his only defense or placed his character at issue as his defense. See MRE 404(a). Thus, the jury’s attention was not focused solely on defendant’s character. This fact supports our conclusion that the jury could not have, in its deliberations, decided that the defendant was guilty only because of an emotional response without basing most of its reasons for the decision on reasonable and logical inferences drawn from all the evidence which a jury makes in reaching a decision on guilt or innocence. In short, we hold that because of defendant’s theory of the case the jury was not so inflamed by the concept that defendant was connected to a similar murder that the jury could not have followed the trial court’s limiting instruction on the use of the similar-acts testimony during its deliberations.4

*339Finally, the admission of the testimony linking the defendant to a different but similarly committed homicide did not confuse the issues in the case. The jury is assumed to understand why most evidence is being introduced. There was no error in light of the fact that a limiting instruction was given and in light of the simple theory of the defense. See generally 6 Wigmore, Evidence (Chadbourn Rev), § 1904, pp 747-748.

Moreover, the question whether the probative value of similar-acts evidence is substantially outweighed by unfair prejudice has been held to be discretionary with the trial judge. E.g., People v Duncan, 402 Mich 1, 14; 260 NW2d 58 (1977); People v Oliphant, supra, 399 Mich 494, fn 10. See also People v DerMartzex, 390 Mich 410; 213 NW2d 97 (1973). In People v Talley, 410 Mich 378, 386-387; 301 NW2d 809 (1981), we stated that our standard of review in testing for an abuse of discretion is a narrow one. Thus, we stated that the standard reads as follows in a criminal context:

" 'Where, as here, the exercise of discretion turns upon a factual determination made by the trier of the facts, an abuse of discretion involves far more than a difference in judicial opinion between the trial and appellate courts. The term discretion itself involves the idea of choice, of an exercise of the will, of a determination made between competing considerations. In order to have an "abuse” in reaching such determination, the result must be so palpably and grossly violative of fact and logic that it evidences not the exercise of will but perversity of will, not the exercise of judgment but defiance thereof, not the exercise of reason but rather of passion or bias.’ ”

*340Under the circumstances of the case, we find no abuse of discretion.

Conclusion

We agree that a four-pronged test should be used in resolving the admissibility of similar acts. We also believe that the similarity of unique circumstances found in the murders met the distinctive-characteristics prong under the test.

We take exception, however, to the majority’s conclusion that the prejudicial impact of the evidence outweighed the probative value. Under the factors discussed above, we hold that the probative value was established. Moreover, the balancing of such factors is an exercise of judicial discretion. Thus, in the case at bar, we find that the trial judge did not abuse his discretion in finding that the similar-acts evidence was more probative than prejudicial. In lieu of granting leave, we would affirm the trial court and the Court of Appeals.