538 model & poll tracker thread

[Brittain33]

This seems pretty accurate, I don’t know why Atlas is so upset.

7% chance of Biden winning AR. 13% chance in MS. Let that settle in.

That’s because of the crazy uncertainty around COVID and economic collapse, right? I don’t really believe he’s right, but I can see someone saying “there’s a 13% chance that current conditions are experienced as such a break from normality that big chunks of the Republican base don’t show up.” I wouldn’t bet on it, but it’s not crazy.

[Brittain33]

This is just embarrassing.

40,000 simulations or not, this shouldn't show up if you do 40,000,000 simulations.

Titanium Tilt R Florida and "Just Not There Yet" Georgia.

Titanium R Oregon. Fivey Fox took drugs again.

That one poll showing Oregon as a 12-point gap has polluted the data.

[Brittain33]

Right now Biden and Cornyn have equal odds of winning - 87%.

[Brittain33]

I'm a bit confused how Biden is only leading by 8.8 in the aggregate after today's polls.

Previous CNN poll was an even bigger margin than today's.

[Brittain33]

So here's some weirdness from the negative state correlations side of things:

If Biden wins Georgia and North Carolina, he has a 16% chance of winning Mississippi.

If Biden wins Georgia, North Carolina, and Minnesota, he still has a 16% chance of winning Mississippi.

However, if Biden wins Georgia and North Carolina, but loses Minnesota, he suddenly has a 53% chance of winning Mississippi.

This one actually kind of makes sense — if Biden’s winning GA and NC but losing MN, something must have turbocharged black turnout nationwide.

Also, isn’t there a latitude factor in the model?