with whom Leavitt, J., agrees, concurring:
I agree with the majority’s analysis and conclusion, but concur to express my concern about the use of statistical evidence in criminal trials.
In this case, several experts testified about the probability of three infants in one family being stricken with SIDS. This testimony was based on the experts’ own investigations of unexplained infant deaths as well as literature discussing the subject. The experts explained that years of research and current thinking on the subject establish the statistical improbability of SIDS occurring three times in the same family.
The admission of statistical evidence to show the probability of an event occurring has long concerned courts throughout the United States. In State v. Sneed,1 the Supreme Court of New Mexico stated that “mathematical odds are not admissible as evidence to identify a defendant in a criminal proceeding so long as the odds are based on estimates, the validity of which have not been demonstrated.” In Sneed, a mathematics professor testified that the chances of the defendant committing the crime as opposed to some other person were 1 in 240 billion.2 The professor used the product rule to determine the probability that certain independent events would occur jointly.3 The product rule provides that the “probability of the joint occurrence of a number of mutually independent events is equal to the product of the individual probabilities that each of the events will occur.”4 The court determined that the factors used by the professor were unsubstantiated estimates and, thus, reversed the conviction and remanded the case for a new trial.5
In the landmark case of People v. Collins,6 the California Supreme Court concluded that testimony by a college mathematics *223professor on the probability of persons with the same distinctive characteristics as the defendants committing the crime was inadmissible. The court determined that the expert testimony lacked an adequate evidentiary foundation for the probability estimates and also lacked adequate proof of the statistical independence of the six factors used by the State’s witness when calculating the probabilities.7 On the other hand, other courts have concluded that statistics are admissible when based on objective, substantiated evidence and the number of variables is controlled.8
In cases dealing with multiple occurrences of SIDS, testimony of statistical probabilities has been received with caution. For example, in Johnson v. State,9 the Georgia Supreme Court concluded that evidence of the probability of multiple occurrences of SIDS was inadmissible due to the lack of information upon which it was based. Other courts, however, have upheld the admission of statistical evidence on the rarity of multiple occurrences of SIDS in a single household when a sufficient foundation is established.10
Here, there was substantial evidence, which the majority amply identifies, to establish that John and Jacob died of asphyxiation and not SIDS. It was the expert opinion of Dr. Clark and Dr. Ophoven that at least the second and third deaths were not caused by SIDS. Had there been no evidence other than the statistics, I would be very reluctant to affirm this conviction because a defendant should be convicted by the evidence, not exclusively by the numbers and probabilities. In this case, however, that did not happen.
Therefore, I respectfully concur.
414 P.2d 858, 862 (N.M. 1966).
Id. at 861.
Id.
People v. Collins, 438 P.2d 33, 36 (Cal. 1968).
Sneed, 414 P.2d at 861-62.
438 P.2d at 33.
Id. at 38-39.
See Rachals v. State, 361 S.E.2d 671, 675 (Ga. Ct. App. 1987) (admitting statistical evidence where the variables were controlled and the “statistics were not derived from a random sampling”); State v. Briggs, 776 P.2d 1347, 1358 (Wash. Ct. App. 1989) (noting that “[t]here is no prohibition against using well-founded statistics to establish some fact that will be useful to the trier of fact”).
405 S.E.2d 686, 688 (Ga. 1991).
See, e.g., State v. Pankow, 422 N.W.2d 913, 918-19 (Wis. Ct. App. 1988) (admitting statistical probability analysis because proper foundation was laid, and the State did not argue that the probability could be used to determine the defendant’s guilt).