> Or when a trial was unfair due to a judge's mistake. Effectively, it is meant to protect you against judge's mistakes and unscrupulous prosecutors. There is m

> Or when a trial was unfair due to a judge's mistake. Effectively, it is meant to protect you against judge's mistakes and unscrupulous prosecutors. There is much less protection against experts that happen to be wrong. IANAL, but I understand certain kind of "evidence" are inadmissible in court due to unreliability/being junk science (e.g. lie detector results). I think it should be possible to classify the admission of unreliable evidence as the judge's mistake, even if that evidence (or type of evidence) was found to be unreliable after conviction. Basically, the standard should be that judges are supposed to be perfectly reliable when evaluating evidence, and should have been expected to catch any issue that was later discovered, so a later standard of inadmissibility is retroactive. If it isn't the, the rules ought to be changed (and this change seems narrow enough to me that it shouldn't cause any problems).