David Crump


The California Supreme Court’s decision in People v. Collins is a staple in Evidence casebooks. An innovative assistant district attorney in the trial court had presented a mathematician who applied probabilities to questions about the perpetrators’ characteristics. The state supreme court disapproved the injection of an equation featuring what mathematicians call the “product rule.” The opinion contains thank-goodness-we-escaped-that-disaster reasoning and condemnation of this use of mathematics with probabilities. But the court’s analysis probably would be different if the case were decided today, as the “new” People v. Collins. Therefore, this Article considers what the author calls the new People v. Collins: that is, the Collins analysis as it would be presented now, as the Collins of the present day. The Article concludes that the California court’s reasoning was wrong as viewed from today, even if the result is defensible. Its opinion relied on a one-sided characterization of the ADA’s evidence and argument. The court’s conclusions would have been better presented if they had included balancing in the manner of Evidence Rule 403, of the value of probabilistic reasoning against its tendency to mislead as weighed by the court. And the court declined to consider the principle that no one piece of evidence is required to prove the entire case, by its indicating that the mathematics could not by itself prove guilt.

First Page