So its a subjective claim then not something that can be looked quantitatively? There are around 13,000 murders per year in the US, how many of these murders per year result in the wrong person being convicted? I would be surprised if its greater than 4%