4 ms·

538, admittedly at the end, gave him a 29% chance of victory. https://fivethirtyeight.com/features/why-fivethirtyeight-gave-trump-a-better-chance-than-almost-a

• rpmisms 7y ago
  Yes, I'm talking about the beginning of election day. I'm not claiming malice on the part of any pollsters, simply a tactic their models didn't account for.
  • cthalupa 7y ago
    538 consistently gave Trump a realistic shot at winning. https://projects.fivethirtyeight.com/2016-election-forecast/ https://projects.fivethirtyeight.com/2016-election-forecast/ You can see their chance of winning history in a line graph. October has a significant dip, to be sure, but it had recovered by November back to the ~30% it ultimately stayed at. Their 'polls plus' model also never dipped quite as far as the polls only model. Regardless, the initial comment was he 'easily' has 4 more years in him. Except now he's facing impeachment, has Republican allies breaking ranks, a long list of unfulfilled campaign promises, big name televangelists saying he's in danger of losing "the mandate of heaven", more and more evidence coming out about Russian interference in the election, and facing a party that knows it must focus on states that it largely ignored. Trump won in several states with fewer votes than Romney lost with - the Democrats aren't going to ignore those states again. The Russians are unlikely to be as effective again. I'm not saying he's going to lose - I could very well see him winning - but acting as if it is an easy thing is absolutely nuts. In a lot of ways it is a significantly worse landscape for him, with lots more room for it to get even worse.