He seems to have become more of a pundit now. He was more about data & models than punditry before. Even in 2016, his final forecast gave Trump a little less than a 30% chance of winning - which is a real chance. Most other models/polls gave Trump a much smaller chance of winning.