The New Science of Sentencing

It may be subjective, but I don't think recidivism rate is so important that you would kill people who may not reoffend, or really might not be guilty in the first place. Personally I would prefer that at least in the case of non-violent crimes, it is better to parole almost everyone to reduce the suffering of innocents.

I also found the following quote weird because the article states people are worse at predictions and more biased than statistics:

> “You mean to tell me they’re using statistics to determine what’s going to happen to me?” he asked. “That ain’t right.”