There is some optimal standard for any class of cases. (At the very least varying the standard will produce better/worse results in relation to particular classes of cases.)

So, no—

> are you going to gather evidence and do cross-examinations before you'll believe them?

—that would be a very bad choice of standard.

I have not claimed to know what the optimal choice of standard is (I will claim that no one else knows it either though), but I do think that changing the standard would have a huge societal impact, and so it shouldn't be done on the basis of a hunch that the outcome would be better. I pointed out one possible complication (convicting more innocent people), though of course the possibilities there are endless.

So to be absolutely clear: I am not advocating for a tighter standard, I'm suggesting that there are complications entailed in lowering it, so it should only be done on a much firmer grounding than some vague notions about catching more bad guys.