the-pi-guy said:
It's still stunning how quickly the model changed. In like a week, Bernie pretty much doubled his chances. |
Before Iowa the model had quite a lot of unsures baked in. Look at the range, they write something like: in 80% of simulation he has between x% and y% of the vote. That range was ridiculous broad before Iowa and now certainly got more concrete. The model had a big Iowa bump calculated in, so the winner of Iowa (which was close) was shaping the whole race. That is why the model flipped so fast.







