Election models megathread

[Ferguson97]

Something random I noticed - since 2016, the fourth most likely outcome in the Classic model was the outcome that ultimately occurred.

If this happened again, Republicans would have either 240/238/223 House seats (3-way tie for 2.2%) and the Senate would remain 50/50.

[Ferguson97]

https://twitter.com/ronbrownstein/status/1570178082052657152?s=46&t=SzRO0yVO...

[Ferguson97]

Interesting 538 article on how betting markets have diverged from the major election models recently, specifically that the betting markets seem to have priced in a much larger movement toward Republicans than the models have.

https://fivethirtyeight.com/features/betting-markets-are-treating-the-midterm-elections-like-its-a-presidential-election/

Historically how accurately do betting markets perform vs models?

[Ferguson97]

R+2 on the Generic Ballot (47/45)

Looks to me a lot more believeable than FOX.

https://www.cbsnews.com/news/republicans-lead-house-control-opinion-poll-2022-10-16/

This would be a historically bad result for the GOP

[Ferguson97]

This would be a historically bad result for the GOP

You are pretending like an R+2 ballot would not still be enough to net the republicans a good number of house seats and the senate and a bunch of governorships.

No I'm not.

All I am saying is that a 2% lead in the house popular vote would not be as strong of a result for the challenging party as other midterms have been.

[Ferguson97]

https://projects.fivethirtyeight.com/2022-flip-senate-house/

I made the following calls

House

Democrat: AK-AL, MI-03, PA-08, VA-02

Republican: CA-22, CA-27, IA-03, OR-05, PA-07, RI-02, TX-15, TX-34

Uncalled: KS-03, NY-19, PA-17

House odds: 98 GOP / 2 Dem


Senate

Democrat: AZ, CO, GA, NH, WA

Republican: FL, IA, IN, LA, MO, NC, OH, WI

Uncalled: NV, PA

Senate odds: 76 Dem / 24 GOP

[Ferguson97]

CBS Poll this Morning showed a close Race for the House with R's leading D's 47-45.

However Republicans have a WHOPPING 16-PERCENTAGE POINT LEAD among Independent Voters. How are the Democrats going to offset that? It's virtually impossible.

BTW TS Clown Tom Bonior still thinks Democrats will hold the House 218-217 based on his Trash Model, LOL

"I believe the crosstabs but not the topline" is an... interesting choice, to say the least.

[Ferguson97]

https://twitter.com/RalstonReports/status/1587176269728260096

[Ferguson97]

There's basically no difference between 51% GOP odds and 49% GOP odds, so it's gonna be funny to see people react hysterically to this, and then again when it switches back to 52% Dem in three days, back to 51% GOP in five days, and stuck at 50% the day of the Election.

[Ferguson97]

https://fivethirtyeight.com/features/the-case-for-a-republican-sweep-on-election-night/

This is a completely separate article from the model, indeed he plays Devil's Advocate against his own model and talks about many other factors besides the model, so it deserves its own thread. Stupid to shove every article Silver publishes in the model thread. And didn't used to happen, so WTF mods? Are you just trying to bury this for some reason?

I read that without realizing he was playing devil's advocate with a fictional person, so I kept thinking "oh my god, this Redd guy is insufferable, why is Nate talking with him?" And then I read the premise description.

I look forward to his discussion with Nathaniel Bleu.

[Ferguson97]

https://mobile.twitter.com/gelliottmorris/status/1589613622241947649

With such a wide margin of error, I think that barely qualifies as a prediction.

I'm pretty sure any random shmuck off the street could tell you that Ohio will probably be somewhere between a very comfortable Vance victory and a squeaker for Ryan.