Jerry Coyne weights in on surveys:
1 Like
Why would that ruin its record? Does Coyne not understand probabilities?
2 Likes
Clinton won popular vote by 3 million. But it is how the 538 electoral votes are apportioned that determines the election. Without good state polling, it is hard to predict the electoral college vote and thus the winner.
You’re missing my point. The prediction was based on state polling.
If one predicts a 71% probability of A over B, that doesn’t mean that B represents any failure whatsoever. It’s expected to happen in a third of the trials.
2 Likes
Jerry like me can’t fathom how the unexpected happened. We shouldn’t blame Nate, he didn’t cause it to happen. I am worried that the unfathomable will happen again - reelection.
Me too, but calling a 71% prediction a failure only makes things worse, in addition to being objectively false.
3 Likes
It wasn’t that unexpected.
Nate did not predict that Hillary would get 71% of the vote (which would be a landslide). He only predicted a 71% probability that Hillary would edge out Trump by a very narrow margin. In some ways, that “71%” tends to mislead us because it is easy to confuse it with the percent of the vote. But, when viewed properly, the 71% falls well short of the 90% or 95% typically used in significance testing. So Hillary had a lead that was not enough to count as significant.
4 Likes
I wish I had a dollar for every time I’ve heard someone try to use this against 538. Just before the election, I read a vociferous attack against him, based on the fact that other aggregators were calling Clinton at around 95%. They didn’t come out looking so good. I don’t think anyone did better than 538 as far as aggregating goes. Surprised to hear Coyne say this as well.
And if even more Democrats lose their faith, that will only exacerbate the acrimonious rift between secular liberals and religious conservatives.
Ask me if I care! Are we supposed to engage in superstitious delusions just so Democrats and Republicans can be friends?
Well, yeah 
Other that that though (not that I would put it that way ), why does it have to widen any “acrimonious rift” in itself, at all?
As for the larger picture, it also occurs to me that if community engagement etc. are desirable ends, then it behooves us to pursue them directly. Intermediate levels of belief are all well and good for those who believe them, but no one has to in the end.
3 Likes