Interesting : in a few courts (not many but a few), jurors were given quick instructions on Bayesian inference methods and an easy to use sheet for accumulation of evidence.
Might seem ideal or at least an improvement and it *could* be if it’s done properly. But it can also be misleading in several ways. One of which is population: the likelihood of guilt factoring in the size of the population living in the area.
But: there’s another problem. Innocent until proven guilty right?
Rather than probability of guilt, it’s more properly used as a probability of the evidence, given the defendant is innocent.
Sounds like the same but it’s not.
In any case it’s an interesting idea. One of the test cases using it was challenged and won the challenge with the judge concluding that requiring the Bayesian system for jurors was an unnecessary level of complexity which interfered with their duties.
Still though, interesting use of Bayes.
