Apply Bayesian methods to trial evidence in criminal trials would make explicit a conflict that currently goes unstated. Unlike the standard “a preponderance of the evidence,” for which there is both folk and professional consensus that a probability of 0.51 is required, “beyond a reasonable doubt” does not, to my knowledge, have an associated mathematical probability. At a folk level people in the US claim to believe “It is better to let ten guilty men go free than to send one innocent man to prison” but there is ample circumstantial evidence that this not always a true preference, and I highly doubt there would be anything remotely approaching consensus for a standard of 0.9.
Trying establish a numerical standard would devolve into mindkilling politics fairly quickly, I suspect. It might break along party lines, or it might break along lines of “people more likely to know someone who was the victim of a previously-acquitted criminal” vs. “people more likely to know someone who was wrongfully prosecuted”, but either way, it would just be something new to argue about.
Apply Bayesian methods to trial evidence in criminal trials would make explicit a conflict that currently goes unstated. Unlike the standard “a preponderance of the evidence,” for which there is both folk and professional consensus that a probability of 0.51 is required, “beyond a reasonable doubt” does not, to my knowledge, have an associated mathematical probability. At a folk level people in the US claim to believe “It is better to let ten guilty men go free than to send one innocent man to prison” but there is ample circumstantial evidence that this not always a true preference, and I highly doubt there would be anything remotely approaching consensus for a standard of 0.9.
Trying establish a numerical standard would devolve into mindkilling politics fairly quickly, I suspect. It might break along party lines, or it might break along lines of “people more likely to know someone who was the victim of a previously-acquitted criminal” vs. “people more likely to know someone who was wrongfully prosecuted”, but either way, it would just be something new to argue about.