Williams v. State

Smith, Justice,

dissenting.

I respectfully dissent. I am convinced that appellant did not receive a fair trial, primarily because of the admission into evidence *810 of ten prior crimes for which he was never charged. The ten extrinsic offenses were highly prejudicial, did not fit any recognized exception to the general rule excluding such evidence, and were never clearly proven by the state. I would reverse and remand for a new trial.

1. The majority opinion affirms the trial court’s admission into evidence of 10 killings for which Williams was never charged, based on the “victims’ commonality of appearance and behavior,” Williams’ alleged “negative attitude” toward poor young blacks, and the majority’s conclusion that “the sheer number of victims with common characteristics each logically connected with Williams by hairs and fibers tends to show a pattern of killings.” I disagree. The trial court’s admission of the highly prejudicial other crimes evidence was erroneous under any fair reading of our prior decisions, and the majority’s approval of the other crimes evidence in this case distorts beyond recognition any exception to the general rule excluding evidence of other crimes committed by the accused.

“On a prosecution for a particular crime, evidence which in any manner shows or tends to show that the accused has committed another crime wholly distinct, independent, and separate from that for which he is on trial, even though it be a crime of the same sort, is irrelevant and inadmissible, unless there be shown some logical connection between the two from which it can be said that proof of the one tends to establish the other.” Bacon v. State, 209 Ga. 261 ( 71 SE2d 615 ) (1952). The reasons for excluding evidence of other crimes are well known. “Courts that follow the common-law tradition almost unanimously have come to disallow resort by the prosecution to any kind of evidence of a defendant’s evil character to establish a probability of his guilt... The state may not show defendant’s prior trouble with the law, specific criminal acts, or ill name among his neighbors, even though such facts might logically be persuasive that he is by propensity a probable perpetrator of the crime. The inquiry is not rejected because character is irrelevant; on the contrary, it is said to weigh too much with the jury and to so overpersuade them as to prejudge one with a bad general record and deny him a fair opportunity to defend against a particular charge. The overriding policy of excluding such evidence, despite its admitted probative value, is the practical experience that its disallowance tends to prevent confusion of issues, unfair surprise and undue prejudice.” Michelson v. United States, 335 U. S. 469, 475-76 (69 SC 213, 93 LE 168) (1948). An exception to the general rule excluding other crimes evidence arises where such evidence “is substantially relevant for some other purpose than to show a probability that [the defendant] committed the crime on trial because he is a man of criminal character,” Walraven v. State, 250 Ga. 401, 407 ( 297 SE2d 278 ) *811 (1982).

Purposes for which other crimes evidence may be offered include motive, intent, absence of mistake or accident (both are aspects of intent), plan or scheme, and identity. Id. at 408 . Our cases state clearly that for evidence of extrinsic offenses to be admissible for one or more of these purposes the state must establish that: (1) the defendant was the perpetrator of the extrinsic offense, and (2) there is a sufficient similarity or connection between the extrinsic offense and the offense charged, such that proof of the former tends to prove the latter. Walraven v. State, supra; Kilgore v. State, 251 Ga. 291, 296 ( 305 SE2d 82 ) (1983); Hamilton v. State, 239 Ga. 72, 73 ( 235 SE2d 515 ) (1977). Moreover, (3) “ [i]t has long been the rule in Georgia that evidence of an independent crime is never admissible unless the prejudice it creates is outweighed by its relevancy to the issues on trial.” Robinson v. State, 246 Ga. 469, 470 ( 271 SE2d 786 ) (1980). See Tuzman v. State, 145 Ga. App. 761, 763 ( 244 SE2d 882 ) (1978); Carroll v. State, 143 Ga. App. 796 ( 240 SE2d 197 ) (1977).

The second prong of this test is in reality a specific application of the general requirement that evidence be logically relevant to a disputed issue at trial. Walraven v. State, supra, at 407. Depending on the issue to which the extrinsic offense is addressed, the state may be required to prove a high degree of similarity between relevant characteristics of the extrinsic offenses and the charged crimes, or it may only have the burden of showing a logical connection between crimes which are essentially dissimilar. Bacon v. State, supra, at 263; State v. Johnson, 246 Ga. 654, 657 ( 272 SE2d 321 ) (1980) (Hill, J., dissenting); United States v. Beechum, 555 F2d 487 (5th Cir. 1977), vacated on other grounds, 582 F2d 898 (1978).

The majority opinion, while undertaking an exhaustive factual review of the details of other crimes evidence introduced against Williams, wholly fails to identify the purpose for which the other crimes evidence was introduced, or to apply to its fact statement the well-established three-part test for admissibility. Instead, the opinion hints vaguely that proof of some sort of “pattern” of mass murders of Atlanta black youth emerges from the evidence and that this pattern justified admission of the other crimes evidence. 1 Under *812 its shadowy “pattern” rubric, the majority opinion approves of the wholesale admission, in a highly publicized and emotionally charged trial, of ten uncharged homicides which, admittedly, bear varying degrees of similarity to the Payne and Cater killings for which Williams was on trial. The majority does so with very little explanation of its basis for decision, managing to do violence to our prior rule governing admission of other crimes evidence, and to confuse both bench and bar at the same time.

(a) In this case, the only arguable purposes for introducing evidence of the ten uncharged offenses were to show Williams’ identity as the killer by proving “plan or scheme” and “modus operandi.”

To be admissible under the “plan or scheme” exception, the extrinsic offense and the crime charged must be connected in the sense that they have “such a concurrence of common features that the various acts are naturally to be explained as caused by a general plan of which they are the individual manifestations.” United States v. Goodwin, 492 F2d 1141, 1153 (5th Cir. 1974) (quoting Wigmore). In other words, the crimes must be mutally dependent, so that one naturally flows from and can be explained by another. See Natson v. State, 242 Ga. 618, 620 ( 250 SE2d 420 ) (1978); United States v. O’Connor, 580 F2d 38, 41 (2d Cir. 1978). In this context evidence of other crimes must be offered to show the accused’s identity as the perpetrator of the charged crimes by showing that he acted in accordance with a larger goal, not merely that he acted with a propensity to commit crimes generally. See Wright & Graham, Federal Practice and Procedure § 5244, pp. 500-501 (murder of three fellow heirs to estate by defendant, each by a different method, admissible to show “plan or scheme” in trial for murder of third co-heir).

In this case there was obviously no such connection between the two charged crimes and the ten extrinsic offenses sufficient to justify their admission to prove a larger plan or goal of which the murders of Payne and Cater were a part. Even assuming that Williams was proven to be the perpetrator of the ten extrinsic offenses (an assumption which I amunwillling to make, see pp. 816-17, infra), the ten extrinsic offenses show, at best, a “plan” with no definable goal *813 except a desire to randomly kill mostly young, poor black males. In such a “plan,” all the crimes were independent and separate, not mutally dependent, they occurred at different times over a 2-year period, and no single crime was necessary to bring any goal to fruition. The deaths of Payne and Cater are not logically explained as a result of any larger “plan” of Williams, and the only inference that can be drawn in this regard from the admission of the other crimes evidence is the forbidden one; namely that, because Williams had a propensity to kill young, poor black males, he acted consistently with this propensity in killing Payne and Cater. See Walraven v. State, supra, at 407; Wright & Graham, supra, at 500. Because “plan or scheme” is simply inapplicable here, I shall not further discuss whether the state’s evidence on this point met the three-part test for admissibility of other crimes evidence.

(b) The only remaining legitimate purpose for which the state could have offered the other crimes evidence was to show that Williams employed a singularly unique modus operandi 2 when committing all twelve murders. Examination of the record indicates that proving Williams’ modus operandi (and hence his identity) was in fact the state’s main purpose in introducing the other crimes evidence at trial. This issue, then, should be carefully analyzed in light of the state’s burden of showing similarity or logical connection, perpetration, and legal relevance.

(1) As to similarity, it is a longstanding premise of evidence law that in order to prove the accused’s identity by use of modus operandi evidence, the extrinsic offense and the charged crime must bear an extremely high degree of similarity. See Sweeny v. State, 152 Ga. App. 765, 767 ( 264 SE2d 260 ) (1979) (extrinsic offense and charged crime “sufficiently bizarre and analogous” to allow admission of other crimes evidence); United States v. Beechum, 582 F2d 898, 912 n. 15 (5th Cir. 1978); United States v. Myers, supra, at 1045; Wright & Graham, supra, § 5240, p. 513. McCormick states that the other crimes must be “so nearly identical in method as to earmark them 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 device used must be so unusual and distinctive as to be like a signature.” McCormick on Evidence, § 190, p. 449.

Examining the majority opinion’s factual review of the Payne *814 and Cater murders and the ten uncharged offenses, one is indeed struck by several similarities among the twelve crimes. Each of the victims was a low-income black male, slightly built, who was often seen alone in the streets of Atlanta. Payne, Cater, and five of the ten other crimes victims were seen with Williams sometime prior to their death. All but two of the victims, Porter and Middlebrooks, were killed by some form of asphyxiation.

Just as important as these similarities, in my opinion, however, are substantial dissimilarities between the two charged offenses and the ten extrinsic crimes. Payne and Cater, age 21 and 28, respectively, were adults; the ages of the victims of the uncharged crimes ranged from 11 years to 28 years and averaged only 15.7 years. With the exception of 28-year-old John Porter, the extrinsic offense victims were essentially children. Another striking dissimilarity between the Payne and Cater killings, on the one hand, and the ten extrinsic offenses, on the other, is that while the bodies of Payne and Cater were both apparently thrown into the Chattahoochee River near the 1-285 overpass, only one of the ten extrinsic offense victims’ bodies, that of Joseph Bell, was found in a river. Bell’s body was discovered in the South River near Rockdale County, miles from where Payne and Cater were found. The remaining nine were deposited on land. Although there was evidence tending to show that the Cater killing was sexually motivated, there was absolutely no medical evidence showing sexual abuse of any of the victims. In addition, it is critical to note that the state’s fiber evidence allegedly linking Williams to all twelve victims, while slightly probative on the issue of whether Williams actually perpetrated the ten other crimes, see pp. 816-17, infra, has no relevance to the modus operandi issue, for the simple reason that the fiber evidence in this case provides no information as to the murderer’s technique in killing or disposing of his victims. The state’s own experts testified that they could not determine the exact mechanism of the alleged transfer of fibers from Williams to the victims. Thus, the sole implication of this type of trace evidence is that each of the victims possibly was in contact with Williams, his house, or his car sometime before his death. Although this inference may be probative of the identity of the killer of the ten extrinsic victims, it does not establish a unique modus operandi, since it would be possible for the murderer to apprehend, kill, and dispose of his twelve victims in dissimilar ways, yet transfer fibers to them in each case. See United States v. Myers, 550 F2d 1036, 1045 (5th Cir. 1977). Thus the presence or absence of fiber evidence has no relevance in the case before us to the narrow issue of modus operandi. Finally, the evidence that Wayne Williams had been sighted at the funerals of Pue, Geter and Rogers, while an interesting sidelight, has no *815 relevance here, since this is not a feature shared by either the Payne or Cater cases.

Comparing the relative similarities and differences of the ten extrinsic offenses, it is evident to me that only one — the Joseph Bell killing — is “so nearly identical in method” to the charged crimes as to possibly be admissible as evidence of Williams’ modus operandi. Like Payne and Cater, Bell was seen with Williams shortly before his disappearance. He was killed by asphyxiation, a relatively rare form of homicide. Although dissimilar to Payne and Cater in the sense that he was only 15 years old when killed, he was, like them, a black male with a “street kid” lifestyle whose body was dumped in a river by the killer. These circumstances make the Bell killing substantially unique, yet similar enough to the Payne and Cater homicides, so as to be probative of modus operandi.

The Pue, Geter, Rogers and Barrett crimes, on the other hand, are of questionable relevance and should have been excluded by the trial judge. While sharing many characteristics of the charged offenses (particularly asphyxiation), all four of these victims were dumped by the killer on land, not in water. In my view this fact distinguishes them from the Payne and Cater murders, making them less relevant on the modus operandi issue.

A third group of victims, consisting of Evans, Stephens, and Baltazar, are even less relevant to the issue of modus operandi than the Pue, Geter, Rogers and Barrett crimes. Although each was a young black male who was killed by asphyxiation, substantial dissimilarities exist between these cases and the Payne and Cater murders. The bodies of Evans and Stephens were discovered partially clothed, and Baltazar was found fully clothed. Each was deposited on land, not in water. In addition, neither Evans, Stephens, nor Baltazar was ever seen with Wayne Williams. Weighing these factors, while considering the highly prejudicial impact of such testimony, I conclude that evidence of the Evans, Stephens, and Baltazar killings should have been kept from the jury.

As to the Middlebrooks and Porter extrinsic offenses, it is clear that neither bore the requisite degree of similarity to the charged crimes. Both Middlebrooks and Porter were found on land, fully clothed, but, more significantly, neither was ever seen with Williams and neither was killed by asphyxiation. Accordingly, neither is even slightly probative of modus operandi, and their admission into evidence was error.

Therefore, I would reverse the conviction on the ground that nine of the ten extrinsic offenses were not sufficiently similar or logically connected to the two charged offenses and were thus inadmissible to show modus operandi. Although this ground is *816 sufficient for reversal, for purposes of clarity I will also address the two remaining prongs of the test for relevancy of other crimes evidence — whether Williams was the perpetrator of the extrinsic offenses, and whether the probative value of the other crimes evidence outweighed its prejudicial impact.

(2) The evidence tending to show that Williams perpetrated the ten extrinsic offenses was purely circumstantial and can only be characterized as weak. Apart from the sightings of five of the ten extrinsic offense victims with Williams prior to their deaths, the only evidence linking Williams to any of the ten other crimes consisted of two blood stains and the state’s dog hair and fiber evidence, which was of questionable reliability and probative worth (see Division 2, infra). The majority opinion admits that the evidence showing that Williams perpetrated the extrinsic crimes “was stronger in some cases than others,” but maintains that “in each case there was some evidence connecting Williams with the victim.” (Emphasis supplied.) I submit that a mere showing of “some evidence” that the accused committed the other crimes sought to be used against him has never been sufficient (at least until today’s decision) for admission of evidence of extrinsic offenses.

While this court has never squarely addressed the issue of the proper standard of proof for “other crimes” evidence, 3 I would require introduction of “clear and convincing” evidence that the accused committed the extrinsic offenses. See McCormick, supra, § 190, p. 452; Wright & Graham, supra, § 5249, p. 532; Sweeny v. State, supra, (“clear proof”). This appears to be the predominant view in other jurisdictions. See State v. Botham, 629 P2d 589 (Colo. 1981); Page v. United States, 438 A2d 195 (D. C. App. 1981); Chapman v. State, 417 S2d 1028 (Fla. App. 1982); State v. Lee, 381 S2d 792 (La. 1980); Cross v. State, 386 A2d 757 (Md. App. 1978); State v. McAdoo, 330 NW2d 104 (Minn. 1983); State v. Wilson, 385 A2d 304 (N. J. Super. 1978); State v. Conyers, 233 SE2d 95 (S. C. 1977); Sanders v. State, 604 SW2d 108 (Tex. Ct. App. 1980); United States v. O’Brien, 618 F2d 1234 (7th Cir. 1980); United States v. Engleman, 648 F2d 473 (8th Cir. 1981); United States v. Bailleaux, 685 F2d 1105 (9th Cir. 1982). A relatively high standard of proof is necessary because of the questionable relevancy of other crimes evidence and its devastating *817 effect on a defendant’s case. 4 Such a standard is particularly appropriate where, as here, the alleged perpetrator was never charged with or convicted of the other crimes, and the evidence of his commission of them is purely circumstantial. Introduction of a prior conviction would, of course, satisfy this standard.

The fragmentary and wholly circumstantial evidence introduced by the state on this point can hardly be characterized as “clear and convincing.” I must take issue with the majority’s approach to this issue, which seems to be that, although the evidence showing that Williams was the perpetrator of several of the extrinsic offenses was quite weak, the ten crimes as a group were admissible because of their “sheer number,” and because the existence of a perceived “pattern” of killings obviates the necessity of proving that Williams was, in fact, the perpetrator of each extrinsic offense. See, e.g., United States v. Woods, 484 F2d 127 (4th Cir. 1973). This “cumulation” theory of admitting other crimes evidence is at odds with the rule of evidence which requires clear proof of the accused’s commission of each and every extrinsic offense. Wright and Graham note that “[s]ome cases seem to admit evidence of other crimes on a slender showing of the defendant’s connection with the crime where the crimes are numerous, apparently on some sort of cumulation theory. This approach should be used sparingly, as it seems capable of producing unjust results; defendant’s suspicious connection with a number of crimes does not necessarily make him guilty of any of them.” Wright & Graham, supra, § 5246, p. 514.

As to the Porter and Middlebrooks extrinsic offenses, the evidence tending to show Williams was the perpetrator was weak, consisting of a very few matching fibers and dog hairs, and a blood stain consistent with Porter’s blood type found in Williams’ car. The fact that these two crimes could be admitted on such scant evidence illustrates the basic unfairness of this trial and Williams’ unenviable position as a defendant who, charged with two murders, was forced to defend himself as to twelve separate killings. The state’s case on other crimes evidence was built on innuendo, suspicion and guilt by association, not on clear and convincing proof, and the admission of the other crimes evidence in this case was tantamount to a “bootstrap operation” which allowed the state to bypass the requirement that it prove each of the twelve offenses sought to be used against appellant. *818 See United States v. Woods, 484 F2d 127, supra (Widener, J., dissenting).

(3) The final prong of the three-part standard for admissibility of other crimes evidence is the balancing of the proferred evidence’s probative value against its potential for prejudice to the defendant. The majority opinion completely sidesteps this issue, which is considered crucial by most courts and commentators. See Robinson v. State, 246 Ga. 469, 470 ( 271 SE2d 786 ) (1980); Tuzman v. State, supra; Carroll v. State, supra; United States v. Beechum, supra, 582 F2d at 913-14; McHale v. United States, 398 F2d 757 (D. C. Cir. 1968); Wright & Graham, supra, § 5249, pp. 533-38, and § 5250, pp. 541-551; McCormick, supra, § 190, pp. 452-454; Note, Other Crimes Evidence at Trial: Of Balancing and Other Matters, 70 Yale L. J. 763 (1961).

In Carroll v. State, supra, the Court of Appeals stated: “The prejudicial effect of evidence concerning independent crimes is the paramount consideration behind the general rule of inadmissibility of such evidence... Even [if independently relevant to an issue in the case], the evidence will not be admissible unless its relevance to the issue outweighs its prejudicial impact. [Cit.] If the evidence tends to show a general criminal propensity more than it tends to prove an issue in the case, it should not be introduced to the jury.” Id. at 797. McCormick’s advice bears repeating here: “[T]he problem is not merely one of pigeonholing, but one of balancing, on the one side, the actual need for the other-crimes evidence in the light of the issues and the other evidence available to the prosecution, the convincingness of the evidence that the other crimes were committed and that the accused was the actor, and the strength or weakness of the other-crimes evidence in supporting the issue, and on the other, the degree to which the jury will probably be roused by the evidence to overmastering hostility.” McCormick, supra, § 190, p. 453. In this regard, the trial court must weigh all these elements and exercise its discretion in deciding whether to admit or exclude the proferred other crimes evidence. “It should be recognized, however, that this is not a discretion to depart from the principle that evidence of other crimes, having no substantial relevancy except to ground the inference that [the] accused is a bad man and hence probably committed this crime, must be excluded. The leeway of discretion lies rather in the opposite direction, empowering the judge to exclude the other-crimes evidence, even when it has substantial independent relevancy, if in his judgment its probative value for this purpose is outweighed by the danger that it will stir such passion in the jury as to sweep them beyond a rational consideration of guilt or innocence of the crime on trial. Discretion implies not only leeway but *819 responsibility.” Id. at 453-54.

The record does not show that the trial judge responsibly exercised his discretion in this case either by assessing the relative probative strength of the ten extrinsic offenses, or by determining whether the other crimes evidence was reasonably necessary to the state’s case. See United States v. Kirk, 528 F2d 1057, 1060 (5th Cir. 1976). Instead he allowed into evidence all ten extrinsic crimes sought to be introduced by the state, regardless of the factual dissimilarities between the ten crimes, the state’s sketchy proof that the defendant committed each crime, and the extremely prejudicial impact of admitting evidence of the extrinsic offenses. The majority opinion concludes that the “sheer number” of other crimes, all with some degree of similarity to the Payne and Cater killings, makes them admissible as evidence of a “pattern” of murders. But the “sheer number” of other crimes introduced in this case and their prejudicial weight are the very reasons the evidence should have been excluded. Each additional crime had an incremental prejudicial effect and contributed to confusion of the issues and undue consumption of trial time, with very little countervailing probative worth. See United States v. Gaus, 471 F2d 495, 499 (7th Cir. 1973) (proof of eleven similar offenses held inadmissible because not necessary to the state’s case); People v. Schlatter, 390 NYS2d 441 (1977) (proof of 40 similar offenses inadmissible because of extreme prejudicial effect). In this case, several factors, including the questionable probative worth of the fiber and dog hair evidence; the availability of other proof as to the Payne and Cater killings; the weakness of the proof of other crimes, especially as to Porter and Middlebrooks; the heinous nature of the other crimes proved; and the inordinate amount of trial time (approximately one half) devoted to proof of the extrinsic offenses, all militate toward the conclusion that the prejudicial impact of the other crimes evidence outweighed its probative worth. See United States v. O’Connor, supra, at 43 (other crimes evidence “distorted the emphasis at trial away from the crimes covered by the indictment to those not so charged.”)

(c) A wholly independent ground on which the other crimes evidence should have been excluded is that its introduction, in the fourth full week of Williams’ trial, unfairly surprised appellant and deprived him of an opportunity to adequately prepare his defense as to the ten extrinsic offenses. The record shows that Williams contemplated that evidence of some 27 extrinsic offenses might be used against him at trial and that he asked for and received Brady materials on each of the 27 other crimes prior to trial. The trial began December 28, 1982. It was not until January 22, 1982, midway through the trial, that Williams was informed which of the 27 *820 extrinsic offenses would be introduced into evidence against him. During this entire period, the state was conducting scientific tests which were used against Williams at trial. Thus, from the time of his arrest on June 21, 1981 until January 22, 1982, Williams was for all practical purposes defending himself against 29 murder charges. Such a procedure flies in the face of any concept of an orderly, fair trial process in which a defendant is put on notice of the charges against him and allowed time to adequately prepare a defense. This court should, at a minimum, require the state to give pre-trial notice of its intention to introduce other crimes evidence against a defendant. Other state courts have recognized the need for such a pre-trial notice requirement. See State v. Acquin, 381 A2d 239 (Conn. Super. 1977); State v. Germain, 433 S2d 110 (La. 1983); State v. Just, 602 P2d 957 (Mont. 1979); Burks v. State, 594 P2d 771 (Okla. Cr. App. 1979); Gipson v. State, 619 SW2d 169 (Tex. Cr. App. 1981); State v. Nicholson, 252 SE2d 894 (W. Va. 1979). An acceptable procedure was outlined in Burks v. State, supra: (1) 10 days prior to trial, the state must furnish the defendant with a written statement of the extrinsic offenses it intends to prove at trial; (2) at the time the evidence is offered, the state must specify the exception under which the evidence is sought to be admitted; (3) there must exist a strong similarity or connection between the charged offenses and the other crimes; (4) the state must make a showing of necessity and that the evidence is not merely cumulative; (5) the evidence that the defendant committed the other crimes must be clear and convincing; and (6) limiting instructions must be given. In my view a pre-trial notice procedure is essential to the fairness of trials in which the state introduces evidence of other crimes.

The rule excluding other crimes evidence is not a mere technical rule of evidence, violation of which can be dismissed by incantation of the phrases “pattern” or “harmless error.” Rather, the rule implicates basic constitutional rights to due process and a fundamentally fair trial. See Lovely v. United States, 169 F2d 386, 389 (4th Cir. 1948); Bray, Evidence of Prior Uncharged Offenses and the Growth of Constitutional Restrictions, 28 U. Miami L. Rev. 489 (1974). As the Supreme Court stated in Brinegar v. United States, 338 U. S. 160, 174 (69 SC 1302, 93 LE 1879) (1949): “Guilt in a criminal case must be proved beyond a reasonable doubt and by evidence confined to that which long experience in the common-law tradition, to some extent embodied in the Constitution, has crystallized into rules of evidence consistent with that standard. These rules are historically grounded rights of our system, developed to safeguard men from dubious and unjust convictions, with resulting forfeitures of life, liberty and property.” These safeguards were denied here.

*821 2. I cannot join Division 1 of the majority opinion, which purports to deal with the issue of the scientific reliability of the fiber evidence introduced at trial. Despite the novelty of this issue and the central role this type of evidence played in the state’s case, a majority of the court today approves of the admission of this evidence almost without discussion, and apparently holds that fiber evidence has reached a stage of scientifically verifiable certainty. 5 do not agree. At trial, various state experts testified concerning their recovery and analysis of certain fibers discovered on the bodies of the twelve victims and in the Williams home and car. Based on their analyses, they opined that some of the fibers found on the victims displayed characteristics which were consistent with the characteristics of fibers from the Williams environment, and further inferred that the fibers associated with the victims could have been transferred from the Williams home and car. The experts described criteria they used to assess the significance of the matches resulting from the fiber comparisons and, applying these criteria, concluded that it was “virtually impossible” for the fibers obtained from the bodies of eleven of the twelve victims 6 to have originated anywhere other than the Williams environment. My review of the record, however, indicates that the state failed to lay a foundation sufficient to establish that the methodologies its experts used to draw their inferences of significance are scientifically valid. For that reason I would hold that the trial court erred when it allowed the introduction of the fiber evidence.

In Harper v. State, 249 Ga. 519, 525 ( 292 SE2d 389 ) (1982), we stated that “it is proper for the trial judge to decide whether the procedure or technique in question has reached a scientific stage of verifiable certainty, or . . . ‘rests upon the laws of nature.’ The trial court may make this determination from evidence presented to it at trial by the parties; in this regard expert testimony may be of value. Or the trial court may base its determination on exhibits, treatises or the rationale of cases in other jurisdictions. [Cits.] The significant point is that the trial court makes this determination based on the evidence available to him rather than by simply calculating the *822 consensus in the scientific community. Once a procedure has been recognized in a substantial number of courts, a trial judge may judicially notice, without receiving evidence, that the procedure has been established with verifiable certainty, or that it rests upon the laws of nature.”

In this case it does not appear that any decisions of other jurisdictions were brought to the trial judge’s attention during the trial or that the judge ever made a specific finding that the techniques of assessing the significance of fiber comparisons used by the state’s experts have reached a “scientific stage of verifiable certainty,” Harper v. State, supra at 525. On appeal the state has cited cases to support its contention that fiber comparisons are competent evidence, but these cases are not on point. Although some of them support the proposition that present techniques and theories of fiber recovery and comparison are sufficiently verifiable to permit the drawing of an opinion that a certain fiber may have originated from a particular source, none addresses the problem of the verifiability of the methods used by experts to venture beyond those assessments of significance and state that there is a high degree of certainty that a particular fiber did in fact come from a particular source. Furthermore, my research reveals no cases which specifically address this problem. For this reason it was not possible for the trial court (or for this court) to judicially notice the verifiable certainty of the procedures used to estimate the significance of fiber matches testified to by experts at trial. Instead, the competency of the expert estimates of significance in this case must be proven, if at all, by the evidence received at trial.

The record shows that the key to the state’s fiber evidence case was expert testimony concerning the alleged uniqueness of two types of carpet fibers recovered and analyzed by the state’s experts. These “unique” fibers were identified by the state’s experts as (1) the green nylon carpet in appellant’s bedroom, and (2) the green-black rayon floorboard carpet of the 1970 Chevrolet station wagon Williams was driving on May 22, 1981, the night of the Jackson Parkway bridge incident.

As to the bedroom carpet, there was testimony to establish that it was composed of fibers which were produced by a Boston, Massachusetts, company named Wellman, and that this type of fiber was sold to a Dalton carpet manufacturer named West Point Pepperell, and to as many as six other Georgia carpet manufacturers during the period in question. A Wellman employee testified that the fiber, known as 181-b, had an unusual shape (trilobal, with two long lobes and one short lqbe), which was designed to avoid infringing upon a patented DuPont equilateral trilobal shape. *823 This employee also identified an exhibit which he said was a summary of Wellman’s sales of the 181-b during the late 1960’s to early 1970’s.

A West Point Pepperell employee named Baggett testified that most of West Point’s manufacturing of carpets using the 181-b had taken place in 1971, but that because the company would have continued to use any remaining inventory of that fiber until it was exhausted, production with that fiber would have continued for some undetermined time subsequent to 1971, and sales of carpets fabricated with the fiber could have occurred as late as 1973. The West Point employee said that his company had manufactured several carpet lines with the 181-b but that he had experienced considerable difficulty in ascertaining sales figures for those lines because records of sales were usually discarded after seven years. However, although total sales for all lines using the 181-b could not be established, he had located the records for several lines, including two known as Luxaire and Dreamer.

Baggett said that Luxaire was produced in several colors, including one using a dye formulation named English Olive, and that Dreamer was also colored with this dye. He identified two pieces of carpet as samples from the company’s records of Luxaire dyed to the English Olive hue, and also identified a copy of his company’s sales records for Luxaire and Dreamer English Olive in the years 1971 and 1972. Baggett also gave his opinion whether another exhibit identified as a piece of the green carpet from the Williams bedroom was similar to the samples of Luxaire English Olive from his company’s records. Although he admitted he was not qualified to perform microscopic analysis of single fibers, he testified that based upon his visual inspection of aggregate physical characteristics such as height of pile, weight of carpet, and type of backing, the carpets appeared to be similar.

Harold Deadman, an FBI microanalyst, identified another piece of carpet as a sample the FBI had obtained from West Point Pepperell. Deadman testified that West Point Pepperell had informed the FBI that the carpet sample was a piece of the Luxaire English Olive. He compared the gross physical characteristics of this exhibit with another piece of the Williams bedroom carpet and concluded that there were no significant differences in their construction and that they were probably manufactured by the same company. Then, building on the Luxaire and Dreamer sales records of West Point Pepperell, information orally supplied him by Baggett, and housing statistics provided by the Atlanta Regional Commission, as well as a number of wholly speculative assumptions (chief of which was that the Williams carpet was in fact a West Point Luxaire or *824 Dreamer English Olive carpet), 7 Deadman attempted to use the calculus of compound probabilities to perform a series of calculations to establish the rarity of that type of carpet in the Atlanta metropolitan area. See McCormick, supra, § 204. He finally concluded that there was a one in 7792 chance of randomly selecting a home in the Atlanta area and finding a room containing carpet similar to the Williams bedroom carpet.

Regarding the green-black 1970 Chevrolet carpet, both Deadman and Peterson testified that they had information indicating that in the Atlanta area only 620 out of over two million cars had that type of carpet. Deadman explained that this data had been supplied by General Motors.

During arguments to the jury an assistant district attorney summarized the foregoing testimony and embellished his summary with his own attempt to quantify the probative force of the fiber evidence. Accordingly, he rounded off the figures for the 181-b bedroom carpet and the green-black floorboard carpet and multiplied them together in order to calculate the chances “that there is another house in Atlanta that has the same kind of carpet as the Williams house and that the people who live in that house have the same type station wagon as the Williamses do . . .,” arriving at a probability of one in forty million. Adjusting this figure to account for an additional assumption of his own, the prosecutor argued that the appropriate figure was actually one in one hundred fifty million.

Taken at face value, the testimony establishing the rarity of the two fiber types might seem to be a sufficient underpinning for the ultimate conclusion of the experts that the fibers of those types found on the bodies were probably transferred from the Williams home or car. Closer examination reveals, however, that the testimony suffers from fatal evidentiary and constitutional defects, and is in fact worthless as evidence on the issue of scientific admissibility of the expert’s opinions of significance.

Much of the state’s failure of proof in this respect resulted from its experts’ insistence on incorporating hearsay into their testimony setting out the factual bases for their opinions. Although an expert may base his opinion on facts he himself has observed or on facts testified to by other witnesses, in Georgia the rule is that hearsay, even if admitted without objection, has no probative value and cannot be considered to sustain a verdict. Green, Ga. Law of Evidence *825 (2d ed. 1983) §§ 111 & 222; Agnor’s Ga. Evidence (1976) §§ 9-7 & 11-3. In recent criminal cases, we have reaffirmed the principles that it is improper for an expert witness to repeat the hearsay opinion of another expert, Dean v. State, 250 Ga. 77 ( 295 SE2d 306 ) (1982), and that an expert may not base his opinion upon hearsay facts which have not been admitted into evidence, Vaughn v. State, 249 Ga. 803, 805 ( 294 SE2d 504 ) (1982). Only one of our decisions appears to run contrary to this rule. In King v. Browning, 246 Ga. 46 ( 268 SE2d 653 ) (1980), a land line case, we stated that “[a]n expert may base his opinion on hearsay and may be allowed to testify as to the basis for his findings . ..” Id. at 47 . Whatever the impact of the ruling in King v. Browning on civil cases, it is clear that in criminal cases allowing an expert to testify based solely on hearsay violates the constitutional right of a criminal defendant to confront the witnesses against him. See Stewart v. State, 246 Ga. 70 (4) (c) ( 268 SE2d 906 ) (1980); Harrell v. State, 241 Ga. 181 ( 243 SE2d 890 ) (1978); Ohio v. Roberts, 448 U. S. 56 (100 SC 2531, 65 LE2d 597) (1980); Agnor, supra, at § 11-4; Green, supra, at § 224.

Applying the well-established rule which disallows the probative value of hearsay evidence, it becomes clear that the state’s attempts to prove uniqueness of the Wellman fibers and the automobile carpeting were totally inadequate. First, the sales figures from Wellman and West Point Pepperell are obvious hearsay which were not brought within the business records exception to the hearsay rule, OCGA § 24-3-14 (Code Ann. § 38-711); see Green, supra, Ch. XX, and thus there is no competent evidence about the total amount of Wellman 181-b fiber or West Point Pepperell Luxaire and Dreamer English Olive carpet produced. Second, the two Luxaire English Olive carpet samples, as well as the sample previously supplied to the FBI and identified by Deadman, were not admissible into evidence because each one was in fact a business record which incorporated not just the physical piece of carpet, but also the attendant written corporate record identifying it, with none of the three being qualified as a business record exception to the hearsay rule. Accordingly, there were no samples of the Luxaire properly in evidence with which to compare the Williams carpet, and no probative inferences could be drawn identifying the Williams carpet as Luxaire English Olive, or for that matter as a West Point Pepperell product. These gaps in the state’s proof, together with the hearsay nature of the statistics from the Atlanta Regional Commission, made a sham of the state’s theory of fiber uniqueness. Deadman’s mathematical calculations were particularly worthless, in light of the fact that they were in several instances based not only upon hearsay evidence but also upon his own unproven assumptions. The weakness *826 of the Wellman fiber evidentiary prop, coupled with the hearsay character of the General Motors data as to the commonness of the green-black floorboard carpet, destroys any basis for multiplying probabilities in the manner of the prosecutor’s closing argument to arrive at a fictional one-in-millions probability that any other Atlanta household had the same type of bedroom and auto floorboard carpets. This evidence, which was extremely prejudicial to Williams, should never have been admitted.

Thus the only competent evidence pertaining to fibers was the proof of the recovery and comparison of fibers, and the testimony that fibers found on the victims appeared similar to fibers found in the Williams home and car and could have had a common origin. The remaining facts and inferences were rank hearsay, unproven assumptions, and guesswork, and should not have been admitted by the trial court. Since the evidence of guilt in this case was solely circumstantial and because the ultimate conclusion of significance of the fiber matches was repeatedly placed before the jury in terms expressing a very high degree of confidence, the error in admitting the fiber evidence and testimony in all probability contributed to the jury’s verdict.

3. I also disagree with the conclusion reached in Division 3 of the majority opinion involving appellant’s request for scientific reports. The majority relies on Law v. State, 251 Ga. 525 ( 307 SE2d 904 ) (1983), for the proposition that only written scientific reports must be given over to a defendant who makes a timely request pursuant to OCGA § 17-7-211 (Code Ann. § 27-1303). However, the dissents of Justice Weltner and myself in Law point out the precise problem presented here. Whereas in Law the state contended that exigent circumstances prevented introduction of written scientific report evidence due to time limitations, the present case is clearly an example of the kind of manipulation warned against in the Law dissents.

On August 14,1981, appellant made a proper and timely request under OCGA § 17-7-211(a) (Code Ann. § 27-1303) for production of scientific reports. Barry Gaudette, a fiber and hair analysis expert from the Royal Canadian Mounted Police, visited Atlanta from December 7 to December 18, 1981. He conducted fiber and hair comparisons in three of the cases then under investigation, including those of Jimmy Ray Payne and Nathaniel Cater, for whose murders the appellant was ultimately convicted. He did not prepare written reports of his findings and no information regarding his research was provided to the defendant. Trial began on December 28, 1981, and during the trial Gaudette testified from personal notes as to the results of the fiber tests he had conducted. Defendant objected on the *827 ground that he had not been provided a copy of these test results pursuant to his request under OCGA § 17-7-211 (Code Ann. § 27-1303). Unlike Law , there is no indication here that Gaudette was rushed or hurried in completing his tests or preparations for testifying. The only similarity with Law is that the state and its expert witness have openly and clearly been allowed to flaunt the provisions and purpose of OCGA § 17-7-211 (Code Ann. § 27-1303). We are now faced with exactly the evil I anticipated Law would spawn, and the majority continues to encourage rather than eradicate this evil.

Appellant also contends that the personal notes and test results compiled by an expert witness should be included within the scope of “scientific evidence” under OCGA § 17-7-211 (Code Ann. § 27-1303). The majority disagrees and follows decisions of the Court of Appeals which conclude that a witness’ work product is not discoverable. We need not overrule those cases, but we do need to add that where only notes and test results exist and are used to support and refresh the expert’s memory during trial, then those materials should be discoverable under OCGA § 17-7-211 (Code Ann. § 27-1303) to the extent that they embody the results of scientific tests. By allowing an expert to forgo delivery of a full written report and to later testify orally where, as here, he had ample time to prepare such a written report and conducted tests too complex to remember unaided, we permit ever more egregious injustice and violation of the intent of OCGA § 17-7-211 (Code Ann. § 27-1303), which is to put into the defendant’s hands these reports with sufficient time before trial to enable him to check and challenge their content.

4. I turn now to appellant’s claim that he was denied effective assistance of counsel. Although the majority declines to review this issue, I find no justifiable reason for doing so where, as in this case, the ineffectiveness of trial counsel is plain and egregious and can be discerned from the record and transcripts.

This court has interpreted effectiveness of counsel to mean “ ‘not errorless counsel, and not counsel judged ineffective by hindsight, but counsel reasonably likely to render and rendering reasonably effective assistance.’ ” Johnson v. Zant, 249 Ga. 812 ( 295 SE2d 63 ) (1982) (quoting MacKenna v. Ellis, 280 F2d 592, 599 (5th Cir. 1960)); Pitts v. Glass, 231 Ga. 638, 639 ( 203 SE2d 515 ) (1974). Moreover, we have noted that when inadequate representation is alleged the crucial inquiries are ordinarily whether the defendant had a defense which was not presented; whether trial counsel adequately consulted with the accused and adequately investigated the facts and the law; and whether the omissions charged to trial counsel resulted from inadequate preparation rather than from *828 unwise choices of trial tactics and strategy. Hudson v. State, 250 Ga. 479 (8) ( 299 SE2d 531 ) (1983). The amount of investigation that is adequate “will necessarily depend upon a variety of factors including the number of issues in the case, the relative complexity of those issues, the strength of the government’s case, and the overall strategy of trial counsel.” Washington v. Strickland, 693 F2d 1243, 1251 (5th Cir. 1982), cert. granted, 51 USLW 3871 (6/6/83). However, trial counsel must “ ‘conduct appropriate investigations, both factual and legal, to determine what matters of defense can be developed.’ ” Id. at 1254.

At the outset I note that Williams had the services of many attorneys during the course of his trial, and that they were confronted with many difficult tasks in investigating his case and preparing defenses for trial, including deciding which of hundreds of witnesses to interview and then interviewing them, and investigating and examining the voluminous, complicated, and novel fiber evidence accumulated by the state. Nevertheless, for the reasons which follow I find that Williams’ defense counsel rendered ineffective assistance because of their failure to adequately investigate the law and facts concerning several lines of defense which they pursued at trial.

Perhaps most importantly, although Williams was entitled, pursuant to Sabel v. State, 248 Ga. 10 (6) ( 282 SE2d 61 ) (1981), to have an expert of his own choosing examine the critical fiber evidence used by the state to implicate him in the two charged and ten uncharged murders, many actions of defense counsel emasculated Williams’ rights under Sabel and rendered inadequate the defense’s investigation of the fiber evidence and its development of a defense thereto.

First, Williams’ attorneys were dilatory in choosing an expert to examine this evidence. Despite knowledge of the important role an expert would play in Williams’ defense, and despite several promptings by the trial court in September and the early part of October of 1981 to quickly obtain and submit an expert for its approval, defense counsel did not submit their chosen expert, a Dr. Morton, for the trial court’s approval until November 12, 1981. Second, after choosing Dr. Morton, defense counsel failed to coordinate his schedule with that of the crime lab, with the result being that on several occasions he came to Atlanta from California and was unable to gain access to the crime lab and the evidence and equipment therein. Third, despite the pre-trial knowledge that the state intended to introduce evidence concerning at least a portion of the other twenty-seven murders on the task force list, the defense did not request, either in its August 14,1981 Sabel motion or at any other time prior to trial, that the court require the state to disclose which of *829 these crimes it intended to introduce and then allow the defense access to the fiber evidence pertaining to each of them. (The trial court’s November 12, 1981 Sabel order only granted the defense access to the fiber evidence pertaining to the Payne and Cater cases.) Furthermore, after the state, on January 22, 1982, pinpointed the other crimes it wished to introduce, and after the court, on January 25,1982, granted the state’s motion to be allowed to present evidence on these murders, Williams’ attorneys did not move for a continuance to enable Dr. Morton to examine the fiber evidence in those cases. In fact, it was not until February 1, 1982, when the defense made a motion to suppress evidence relating to the ten independent murders, that the defense expressed dissatisfaction with the trial court’s failure to authorize its expert to inspect the fiber evidence pertaining to the ten extrinsic murders. Additionally, the defense failed to take advantage of two opportunities given to it by the state to examine the fiber evidence in the independent offenses.

Although all of the above omissions rendered trial counsel’s preparation of an adequate defense to the state’s fiber evidence practically hopeless, the fatal blow to defense counsel’s effectiveness in this area was their failure to have Dr. Morton testify at trial. Morton apparently left Atlanta towards the end of January or early February 1982 and refused to return, thus forcing the defense to obtain a second expert in the twilight of the trial. The result was that this expert spent less than five hours examining the state’s voluminous collection of fiber evidence. Regardless of the reasons why Morton refused to return to Atlanta, defense counsel’s failure to take the necessary measures to assure that he would testify is inexcusable — Morton was the only expert Williams had who had had enough time to review the fiber evidence, the linchpin of the state’s case, and help prepare a defense thereto.

Related to the above ineffectiveness of defense counsel was their failure to challenge the scientific reliability of the principles and techniques used by the state’s experts in conducting their fiber comparisons. See Division 2 of this dissent. This challenge was crucial to an adequate defense to the fiber evidence, and its omission demonstrates inadequate preparation and investigation by defense counsel and is a significant example of counsel’s ineffectiveness.

Another glaring failure by defense counsel to investigate the law and facts concerns their failure to adequately raise the issue of or invoke a ruling by the trial court on the state’s failure to advise Wayne Williams of his Miranda rights on May 22,1981, a night on which as many as ten police officers and FBI agents interrogated Williams for approximately ninety minutes after stopping him on Interstate 285.

In addition to the above failures by defense counsel to ade *830 quately investigate the law and facts relative to certain defenses, the trial transcript reveals that in many instances defense counsel waived the right to object to or failed to object to certain prejudicial evidence introduced at trial and to certain improper conduct which occurred at trial. These omissions include: 1) the failure to object to the highly questionable and prejudicial use by a state’s witness of the calculus of compound probabilities to explain the probative force of certain fiber evidence (this includes the use of hearsay by the state’s expert, to which defense counsel did not interpose an objection concerning Williams’ right of confrontation) (see Division 1 of the majority opinion and Division 2 of this dissent); 2) the failure to object to a prosecutor’s use of these probabilities, supplemented with mathematical odds based upon his own calculations, during closing argument; 3) the waiver of the defense’s presence on certain occasions when the trial court talked with the jury (see Division 11 of the majority opinion); 4) the failure to object to improper closing argument (see Division 14 of the majority opinion); 5) the waiver of the right to object to Dr. Zaki’s references to other crimes (see Division 4 of the majority opinion); 6) the waiver of the right to object to the prosecution’s question to Tilbert Boynham concerning whether he had been given a lie detector test (see Division 15 of the majority opinion); and 7) the defense’s failure to object to the trial court’s failure to preserve for appellate review the Brady material it inspected. It is clear that these omissions, when considered together, were prejudicial to Wayne Williams.

For the above reasons, I find that Wayne Williams’ attorneys, due to inadequate preparation and inadequate investigation into the law and facts of their client’s case, rendered him ineffective assistance of counsel. 8 I would therefore reverse his convictions and remand the case for a new trial.

The meaning of the majority’s references to “pattern” as a purpose for which the other crimes evidence was admissible is unclear. “Pattern” is not a recognized exception to the general rule excluding evidence of other crimes. McCormick, supra, § 190, pp. 447-451. Although the trial judge in this case ostensibly admitted the evidence of 10 other crimes to show Williams’ “plan, scheme, pattern, bent of mind, and identity,” plan, scheme, and pattern are properly thought of as components of the traditional “plan or scheme” exception, which is relevant to the ultimate issue of *812 identity. See McCormick, supra, at 448. “Bent of mind” has no application here, since intent was not a contested issue at trial. I conclude that identity (and particularly, modus operandi) was the only relevant purpose for introduction of this evidence. Identity may be inferred from other crimes evidence showing a plan or scheme, motive, or modus operandi. See pp. 4-6, infra; McCormick, supra, § 190, pp. 448-451; Wright & Graham, Federal Practice and Procedure § 5246, pp. 511-512 (1978); United States v. Myers, 550 F2d 1036, 1045 (5th Cir. 1977).

Webster’s Third New International Dictionary (1971) defines “modus operandi” as “a distinct pattern or method of procedure thought to be characteristic of an individual criminal and habitually followed by him . . .” Id. at 1453.

The majority opinion again fails to articulate a clear standard in this case, stating only that the accused’s perpetration of the extrinsic offenses need not be proven “beyond a reasonable doubt.” I find it interesting that this court, which sua sponte created a “clear and convincing” standard of appellate review in child custody cases, see Blackburn v. Blackburn, 249 Ga. 689 ( 292 SE2d 821 ) (1982), does not recognize the need for a clear standard of proof to guide judges and juries in criminal trials involving extrinsic offenses.

Studies have shown that introduction of other crimes evidence has a decided impact on the outcome of criminal trials. See Note, 76 Colum. L. Rev. 327 , 343 n. 95 (1976) (“[W]hen a defendant’s criminal record is known and the prosecution’s case has contradictions, the defendant’s chances of acquittal are 38 %, compared to 65 % otherwise.”)

The record indicates that defense counsel failed to object at trial to the state’s fiber evidence on the grounds of scientific unreliability. Despite this fact, the majority opinion appears to address the merits of the reliability issue. Therefore I feel compelled to state my views on this issue.

The experts declined to attribute this level of significance to the fibers found on Joseph Bell, apparently because so few were found which compared favorably with the Williams fibers.

This fact, critical to the state’s “uniqueness” argument, was never conclusively established and was indeed questionable in light of testimony (based on Wellman sales records) that Wellman fiber was sold to a number of Georgia and southeastern manufacturers during the period in question.

I believe that, in light of the compulsory legal education requirements adopted by this court and effective January 1, 1984, this court should establish a procedure whereby any attorney found to be ineffective would automatically be referred to the State Disciplinary Board for investigation and appropriate sanctions.