Also, as far as I know, nobody measures false negatives in hiring. How would you even do it? Keep track of everyone you rejected, and then 5, 10, and 20 years later check on their career? I'd be fascinated to see the results of such a study. I'm sure it would be super valuable if you could find some kind of pattern where some filter is falsely excluding candidates that are actually great.