Too much attention has been placed on polling predictors. It wasn't just polling that was off, it was other predictors as well. PollyVote, for examples, aggre

Too much attention has been placed on polling predictors. It wasn't just polling that was off, it was other predictors as well. PollyVote, for examples, aggregates over many different types of predictors, and non-polling predictors still predicted a Clinton win. So whatever you'd have to use to explain how the polls were off would also have to account for why economic predictors were off, index models... If this were an undeveloped country, my first thought would have been election tampering. However, people also seem to be forgetting that at some level, the polls were correct: Clinton did win the popular vote. The polls were just off about where the votes were won. Important, but a harder aim, prediction-wise.