There is actually a very good reason: given that it is unlikely that we will miraculously find a way to change the ratio of wrong to right verdicts, you can only lower the amount of people wrongfully convicted by lowering the number of convictions. As most convicted people are actually guilty, that means that most of the newly unconvicted people are, in fact, guilty.
Assume that currently 1 person is wrongly convicted for every 10 guilty ones that go free and try. Do the math to change that to 1 in a 100 and be affraid of the consequences of putting this principle into practice.
And again, note that you are not talking about changing the ratio of wrongful convictions to right convictions, because that is highly unlikely: you are talking about changing the absolute amount of wrongful convictions by allowing a larger amount of wrongful exonerations. Because of the ratio of the amount of people involved, the amount of guilty people that would need to go free is staggering. You need thorough Bayesian reasoning to understand this one.
Assume that currently 1 person is wrongly convicted for every 10 guilty ones that go free and try. Do the math to change that to 1 in a 100 and be affraid of the consequences of putting this principle into practice.
And again, note that you are not talking about changing the ratio of wrongful convictions to right convictions, because that is highly unlikely: you are talking about changing the absolute amount of wrongful convictions by allowing a larger amount of wrongful exonerations. Because of the ratio of the amount of people involved, the amount of guilty people that would need to go free is staggering. You need thorough Bayesian reasoning to understand this one.