[CrazyStat]

By "other models" I think he means models other than 538's, several of which did put the probability of Clinton winning at over 95%.