It is a commonplace assertion that statistical evidence can cause miscarriages of justice by confusing lay triers of fact. In its celebrated opinion in People v. Collins, 438 P.2d 33 (Calif. 1968), the California Supreme Court asserted that in our “computerized society,” statistical testimony can be “a veritable sorcerer,” misleading jurors by “cast[ing] a spell” over them. To some extent the U.S. Supreme Court’s decision in Daubert v. Merrell Dow Pharmaceuticals Inc., 509 U.S. 579, 595 (1993), reflects the same concern. Near the end of its opinion, the Daubert Court commented that, even if proffered scientific evidence passes muster under Federal Rule of Evidence 702, it might be subject to exclusion under Rule 403; the Court approvingly quoted Eastern District of New York Judge Jack Weinstein’s remark that trial judges may exercise “more control over experts” because jurors can find their testimony “misleading” and “difficult [to] evaluat[e].”
Today, though, it is questionable whether the courts should maintain the same conservative attitude toward statistical evidence. The writings of leading scholars such as professors David Kaye and Richard Lempert have explicated many seemingly difficult statistical concepts for us and shown us that, in truth, some of the statistical concepts most relevant to legal proceedings are straightforward. Consider, for example, three concepts: the random-match probability, a likelihood ratio and a Bayesian posterior probability. These concepts are not only relatively simple but also closely related.
