[dralley]

>Polls are weird. In 2016, the polls strongly suggested Trump would not win.

None of the polling suggested that. They suggested Clinton had a 2-3 point popular vote lead. She did in fact have a 2-3 point popular vote lead. It's not the polling's fault that it's measuring something that the US doesn't actually use to select the President.

Outlets that did take state polls into consideration and tried to simulate the electoral vote, namely 538, gave Trump a basically 1/3 chance to win.