> Humans just aren't very good at ranking things on a normal distribution Anecdotally this feels correct, the more common user rating distributions I see are a

> Humans just aren't very good at ranking things on a normal distribution Anecdotally this feels correct, the more common user rating distributions I see are an inverted bell curve - lots of 5's and 1's, not too many 3's. But as far as the same problem that plagues every other review system goes, I would say the paid/fake reviews are the far bigger problem.