Prosecutor’s comments on math lead to overturned conviction

By Martin Finucane, Globe Staff

A Bristol County prosecutor overstepped his bounds when he offered in his closing argument a mathematical analysis of the probability that a crime victim had picked out the right suspects from a selection of police photos, the state’s highest court ruled today.

The Supreme Judicial Court threw out the conviction of Kris N. Ferreira of unarmed robbery of a person 60 or older for a Nov. 5, 2006, robbery of a pizza deliveryman in Fall River. The high court also threw out the Superior Court’s determination that he was a habitual offender, which had resulted in a life sentence.

The victim picked out two men from 12 photos shown him by police. He picked out one suspect quickly, but then picked out a photo of Ferreira with less confidence. The first suspect and Ferreira, according to evidence introduced at the trial, were friends.

The court, in an opinion written by Justice Ralph Gants, said the prosecutor’s calculation of the odds that the victim would randomly pick out two men who were friends -- and alleged accomplices -- was flawed, first of all, because the victim was not supposed to be making a random selection.

“The victim’s eyewitness identification has potential evidentiary consequence only to the extent that it is not random, but reflects his recognition of two persons who had robbed him approximately ninety minutes before the identification procedure,” the opinion said.

The opinion also said no evidence was presented as to how many of the other people in the photos shown to the witness might also have been friends with the first suspect. The court also said the evidence should have been offered through a mathematical expert.

The prosecutor also erred in equating mathematical probability of guilt with proof beyond a reasonable doubt, the court said.

The court cited previous decisions that said the concept of reasonable doubt is not mathematical, it is “inherently qualitative” and a concept that seeks “abiding conviction” or “moral certainty” rather than statistical probability.