Sure, it’s a ballpark, but in the context of a conversation about how many innocent people are being executed, I think the wrongful conviction rate of the entir

Sure, it’s a ballpark, but in the context of a conversation about how many innocent people are being executed, I think the wrongful conviction rate of the entire prison population is less relevant than a ballpark rate of those wrongfully sentenced to life/death. For example, if the wrongful conviction rate for minor offenses or white collar crimes was as high as 10%, it wouldn’t necessarily indicate that 10% of people sentenced to death were also wrongfully convicted. Its very possible that juries dealing with more serious crimes and more serious punishments are more careful in their decisions or innately require a higher burden of proof to convict. I assumed that most of the letters received by the IP are for life/death sentences or at least major crimes with multi-decade long sentences, which I think would probably have a more consistent wrongful conviction rate in the group vs. the entire prison population. And while you’re right that the IP might only have the resources to attempt to overturn the top X% cases, one could assume that they don’t win all their cases and that the next tier of cases they don’t accept would have diminishing returns. So I think it’s fair to treat this estimate as closer to a lower bound with some significant error bars, but I’d be surprised if the reality is 5x or 10x higher.