It is no news that eyewitness identification in criminal cases is a problem; it is an old and famous problem. Judges and lawyers have long known that the identification of strangers is a chancy matter, and nearly a century of psychological research has confirmed this skeptical view. In 1967 the Supreme Court attempted to mitigate the problem by regulating the use of eyewitness identification evidence in criminal trials; since then it has retreated part way from that effort. Legal scholars have written a small library of books and articles on this problem, the courts' response to it, and various proposed solutions.5 But all this extensive attention has focused on one side of the problem of eyewitness identification: Why are mistakes so common, and how can they be prevented? I start from a different question: Why are eyewitness misidentifications so rare?
Gross, Samuel R. "Loss of Innocence: Eyewitness Identification and Proof of Guilt." J. Legal Stud. 16 (1987): 395-453.