[olliej]

It seems like the behavior between WA and MS could just be statistics saying that WA and MS always[1] vote for the opposite candidate, rather than considering a massive sudden change in the direction that one of them votes in. E.g. it's not reflecting who they vote, just who they most vehemently disagree with.

I'm not sure why that kind of interstate correlation should impact predictions?

<incoherent rambling :D> IANAS but it feels like these correlations were added to compensate for the failure in 2016 to recognize that state A going one way implied that state B would also go that way. It "feels" like a more correct approach would be to compute some kind of error/weakness measure in a states polls by bringing in those of its geographical neighbors and incorporating the polling error of that entire block vs prior years. Or something.

The intuition I'm having difficulty conveying is that actual voting correlation is based on neighboring states only because you've got bubbles of ideology that aren't strictly cut along state lines. If strength of opinion in a bubble is going one way, then you'll see that mostly in the state at the center of the bubble, but the bubble still spreads into neighboring states, and a "stronger" bubble could push it geographically further into those neighbouring states, and/or could increase the bias in areas inside the bubble. </rambling>

[1] "Always" == most recent history

[RivieraKid]

> I'm not sure why that kind of interstate correlation should impact predictions?

538 has low positive correlations between states on average, which actually has a big impact, it increases overall uncertainty (and therefore Trump's win probability). Why? If the states are not correlated, you usually end up with a few states going off the rails, like Trump winning Colorado without any nationwide swing.

[DavidSJ]

Other way around: uncorrelated errors tend to cancel each other, correlated errors tend to reinforce each other.

[RivieraKid]

Well Gelman claims the opposite.

[DavidSJ]

No, what Gelman says is that he suspects that to compensate for the fat tails in 538's state distributions, they had to reduce the between-state correlations, to get a desired overall level of uncertainty.

This implies that correlations increase, rather than decrease, the overall level of uncertainty.

This is also easy to see from a basic probability perspective, using the concept of variance. For example, if you have two coin flips, with outcomes {-1, +1} chosen uniformly at random, then the sum has variance 2 if the flips are independent, but variance 4 if the flips are perfectly dependent.

[RivieraKid]

No, here's what Gelman says:

> the lower the correlation between states, the more uncertainty you need for each individual state forecast to get a desired national uncertainty