3 ms·
>So he knew the polls were off. This is my thinking as well. Towards the end Nate Silver kept upping Trump's chances and other pollsters were baffled by Nate's
by drzaiusapelord 10y ago
>So he knew the polls were off.
This is my thinking as well. Towards the end Nate Silver kept upping Trump's chances and other pollsters were baffled by Nate's changes. Nate kept arguing that due to curring polling methods, current models, the lack of young turnout, and a lot of disenfranchised voters coming out that Trump was doing better than he seemed. There was no "liberal media conspiracy" but the model changes you needed to make in what was an extremely oddball election year. Note, the pollsters that got Trump wrong got the last few elections right or wrong but within a respectable margin of error, so from a question of bias, the conspiracy theories just don't hold. Same with the Brexit, the pollsters just didn't see that silent majority voting yes.
Nate obviously didn't change his model enough but he certainly saw the winds of change here and clearly other people like Thiel did too. Also, being a libertarian there's really no reason for him to support a Democrat anyway, so if he's going to play this game he's going to pick the GOP candidate.
>He also could probably see the media had gone 110% to shill for Clinton.
I didn't see this. Instead, I certainly saw the media play both sides as being less terrible than they truly were. I think there's a narrative of "oh they're just like each other so it'll be tight" and promoting a moral equivalence of "questionable email server use is just as bad as saying celebrity lets you grab women by their crotches with impunity" that sells ad impressions.
- ZeroGravitas 10y agoNate wasn't changing his model, the polls changed. His model just accounted for the high number of undecideds and the correlation of errors between states, and the possibility of systemic polling errors which accounts for his divergence from other predictions.