4 ms·

That's a good point. I didn't consider the fact that possibly many discoveries have a very low prior expectation of truthood.