538 model & poll tracker thread

[TiltsAreUnderrated]

National: Biden +9.2%

California: Biden +29.4%
New York: Biden +26.8%
Washington: Biden +24.7%
New Jersey: Biden +21.2%
Colorado: Biden +17%
Virginia: Biden +10.6%
Michigan: Biden +10.2%
Nevada: Biden +8%
New Hampshire: Biden +7.7%
Florida: Biden +6.9%
Wisconsin: Biden +6.6%
Pennsylvania: Biden +5.3%
Arizona: Biden +4%
North Carolina: Biden +3%
Ohio: Biden +2.7%
Georgia: Biden +1%
Iowa: Trump +0.6%
Texas: Trump +0.7%
Missouri: Trump +4.9%
Utah: Trump +6.7%
South Carolina: Trump +7.8%
Kansas: Trump +9.5%
Tennessee: Trump +11.4%

[TiltsAreUnderrated]

Hmm...

I thought that it's going to be more like the Economist and JHK version (i.e. maps with chances of winning, electoral votes, etc.)

These are just weighted poll averages. What you’ve mentioned comes out with the (as of yet unreleased) full model. Is there an ETA on when that’s coming out?

[TiltsAreUnderrated]

A KY average has just been generated: Trump +17.4%.

[TiltsAreUnderrated]

A KY average has just been generated: Trump +17.4%.

Taking a glance at their 2016 forecast, which had Trump leading by 12.3% at this time in 2016.

Trump ended up winning by 29.8%.

N.B. this isn't their forecast. I expect Trump to perform a bit better in that because of undecided voters' Republican skew.

[TiltsAreUnderrated]

538 now has its first trendline up for Alabama:

Trump 55%
Biden 39.1%

[TiltsAreUnderrated]

Biden has now pulled ahead in IA. I believe that's the first state to flip within the average since the tracker was launched.

[TiltsAreUnderrated]

20% Polling
30% Voting patterns
50% BS

5% data
50% pain
And 100% reason to regard with disdain

N.B. I don't think rating the election as Lean Biden is unreasonable, but the MS/AR numbers made me LOL.

[TiltsAreUnderrated]

Have they confirmed whether or not they're doing a Senate model this year yet? I haven't found mention of one.

[TiltsAreUnderrated]

https://twitter.com/gelliottmorris/status/1298797791523868673?s=20

Isn't the fact that the model was "wrong" in 2016 mean that they adjusted it in some way - maybe it was too "confident" based on early polls and now it is more conservative?
 

Somewhat wider confidence intervals are appropriate because of shifts in VBM and COVID-19’s potential effects on differential turnout.

[TiltsAreUnderrated]

https://twitter.com/baseballot/status/1299525104842543104

[TiltsAreUnderrated]

346: the number of electoral votes Biden will get?

(I know Arizona going Republican is a little unlikely given the rest of the map, but maybe Biden underperforms with Hispanics.)

https://twitter.com/AndrewSolender/status/1304979308822265857

[TiltsAreUnderrated]

346: the number of electoral votes Biden will get?

(I know Arizona going Republican is a little unlikely given the rest of the map, but maybe Biden underperforms with Hispanics.)

https://twitter.com/AndrewSolender/status/1304979308822265857

What did you think I was responding to?

Oof. I should read threads more carefully.

[TiltsAreUnderrated]

It’s getting a little bit creepy how stable the model is right now.  Biden has been at 77 for 7 straight days.  It seems like there should be a little movement just by random chance.  I’m surprised Nate isn’t tweeting worries about herding yet.

The model assumes pretty strong convention bounces that last for up to a month.  Next week could be significant as I believe that's when Trump's assumed bounce finally goes to 0.  The narrowing to Biden being only a 2:1 favorite in late August was because the model assumed a DNC polling bounce that didn't really materialize.  I have a feeling it's overweighting the RNC as well and perhaps giving Biden too much credit right now.

I don't think the bounces are meant to last that long, where have you seen that? The tightening in mid-August was because Trump was getting an RNC bounce and Biden wasn't getting a DNC bounce.

538 has written several articles on how they incorporate convention bounces into previous iterations of the model and how they expect them to last 2-3 weeks if not interrupted by another party's convention. If you look at the "nowcast" element of their model (they're not calling it that, but it exists, the adjustments made to Trump's and Biden's polling numbers seem to come from "current events" which I would assume is them attempting to measure the remainders of the convention bounces in the polls they're using.

Adjustments for the RNC bounce will disappear on a state-by-state basis, depending on how quickly state-level polls come out.

[TiltsAreUnderrated]

538 has started to remove the SurveyMonkey rolling averages that overlap from their poll trackers but their forecast pages still use these removed polls.

I'm pretty sure this is screwing with their model as their average's "X most important polls" for specific states now have up to 4 of the most surveymonkey polls despite the fact that their samples overlap.

[TiltsAreUnderrated]

Bump: they only started removing these overlapping SurveyMonkey polls recently and it seems that whenever they add a new poll for a given state, that state loses its overlapping SurveyMonkey polls problem. That's happened with MN and NV today.

[TiltsAreUnderrated]

https://twitter.com/NateSilver538/status/1321466975302963205

Later in the thread he gives himself some wiggle room w.r.t. reversing that decision if there's a Comey Letter II.