Election models megathread

[Terry the Fat Shark]

https://fivethirtyeight.com/methodology/how-fivethirtyeights-house-and-senate-models-work/

Found this page and it says (model tweak: June 30,2022) that talks about how their model is going to work this year, it also has a thing at the top that says 2022 forecast and if you click on it takes it to this page that's not quite live yet: https://fivethirtyeight.com/features/why-republicans-are-favored-to-win-the-house-but-not-the-senate/

[Terry the Fat Shark]

Also for those interested, JHK is releasing his own model late this week for Senate:

https://twitter.com/jhkersting/status/1542226445174542336/photo/1

[Senator Incitatus]

https://twitter.com/NateSilver538/status/1542228065123811329
It's confirmed.

[Terry the Fat Shark]

https://twitter.com/NateSilver538/status/1542228065123811329

It's confirmed.

Nice, seems like it's tomorrow then based on the evidence

[Senator Incitatus]

It is good that they're taking a more systematic approach to polling error:

We assume errors are more correlated from race to race. In presidential races, forecast and polling errors are highly correlated from state to state. This also holds in congressional races, though to a considerably lesser degree since there are different candidates on the ballot in each state. However, our analysis suggests that polling errors are becoming more correlated in congressional races, too. Therefore, we’ve increased the degree to which race-by-race errors are assumed to be correlated in the model.

Not a single mention of response bias though, which suggests they're going full-speed ahead on pretending the polling industry isn't in flames. Not a surprise, given how much motivated reasoning is necessary to keep people relying on statistical models at this point, but still disappointing. Maybe not quite the place for it.

This is also encouraging:

In evaluating fundraising for congressional candidates, the model now places more emphasis on contributions received within the candidate’s state. We found that a dollar received from within a candidate’s state is about five times as valuable as a dollar received outside of it in predicting the eventual election outcome.1 That’s why the model now applies that ratio, multiplying any funds raised from voters within the candidate’s state by 5, when assessing which candidate has the fundraising edge.

[SnowLabrador]

I won't take it literally, but I'm still quite excited.

[Buffalo Mayor Young Kim]

538 has been crap ever since Nate Silver replaced the real model with that polls plus crap.

Though Nate Silver is a funny story. He got famous showing that a simple weighted aggregate of polls was more predictive than random pundit thoughts. Then everyone decided he was a wizard so he started doing pundit hot takes.

[ProgressiveModerate]

My guess is the model has the GOP doing significantly better in the Senate than one expects, ah least relative to the House forecast.

Statistically, the senate and house predictions by pundits don’t rlly align because many pundits are treating races in Biden + 0/1 states as true tossups while Biden + 1 House seat is immediately written off as an R win. I suspect the model for instance will have a similar rating for Senate seats like AZ/GA/PA with House seats like IA-03/TX-15/VA-02

[Amenhotep Bakari-Sellers]

Twitter project 230RH and 52/48 D S with WI and PA and DeSANTIS losing in FL but they are cautious about projecting a D H because the Environment, D's goal is to win OH, NC and FL Senate seats and 218H seats Crist and Ryan are leading and Beasley, Demings and Kunce are down 5/6 points and last Change poll in April had Grassley only up 3, MOE and if  Crist wins in FL, Demings will win

[wbrocks67]

My guess is the model has the GOP doing significantly better in the Senate than one expects, ah least relative to the House forecast.

Statistically, the senate and house predictions by pundits don’t rlly align because many pundits are treating races in Biden + 0/1 states as true tossups while Biden + 1 House seat is immediately written off as an R win. I suspect the model for instance will have a similar rating for Senate seats like AZ/GA/PA with House seats like IA-03/TX-15/VA-02

the URL literally says - why-republicans-are-favored-to-win-the-house-but-not-the-senate/

[Sir Mohamed]

They're providing chances for majorties, right? What I expect:

- Chance for GOP majority in the House: 93%

- Chance for GOP majority in the Senate: 64%

[Shilly]

It's up.
https://projects.fivethirtyeight.com/2022-election-forecast/
Republicans have an 87% chance to take the House, 53% for the Senate.

[Amenhotep Bakari-Sellers]

They're providing chances for majorties, right? What I expect:

- Chance for GOP majority in the House: 93%

- Chance for GOP majority in the Senate: 64%

No, have you seen the polls Barnes and Fetterman, CCM, Hassan, Warnock and Bennett are tied or leading and Kellyits a 40 percent chance of Rs winning S the Hits 60/40

[TiltsAreUnderrated]

>The Senate is a toss-up

F**k off, Nate.

It gets worse at the individual level, where they project that Democrats are narrowly ahead in AZ, NV, PA and perhaps GA (in that last state, they show Warnock with a positive margin of victory but have his odds of winning in Republican colours, which could be a formatting error but is more likely a prediction of a Republican victory distorted by potential Warnock NUT maps).

[Amenhotep Bakari-Sellers]

Barnes is two pts ahead of Johnson, Rs haven't lead in a single AZ poll


Nate Silver follows the 303 blue wall that's why he gives this forcast and it's still July not Labor Day it's still 4 months away

[wbrocks67]

>The Senate is a toss-up

F**k off, Nate.

It gets worse at the individual level, where they project that Democrats are narrowly ahead in AZ, NV, PA and perhaps GA (in that last state, they show Warnock with a positive margin of victory but have his odds of winning in Republican colours, which could be a formatting error but is more likely a prediction of a Republican victory distorted by potential Warnock NUT maps).

I mean, I don't see this as totally unbelievable? Dems have strong chances in all of those states and I think Rs at a 53% chance of taking the senate (and 87% in the house) both sound pretty correct at this point in time...

[Amenhotep Bakari-Sellers]

D's are gonna have a 52/48 S since PA and WI and GA are gonna follow the Gov races but we are targeting OH, IA, NC, FL, LA and MO too and Ryan and Crist are leading, if Crist wins I find it hard to believe that Demings whom is only 5 down won't win


Be advised this a model and models can be wrong I expect D's to keep their majority, this model is solely going off of Biden low Approvals by Historical trends and it's Turnout that matters

[Sir Mohamed]

They're providing chances for majorties, right? What I expect:

- Chance for GOP majority in the House: 93%

- Chance for GOP majority in the Senate: 64%

No, have you seen the polls Barnes and Fetterman, CCM, Hassan, Warnock and Bennett are tied or leading and Kellyits a 40 percent chance of Rs winning S the Hits 60/40

Its 150 days to election, Rs can still win most tossup races due to fact Prez Biden is at 40s approvals.

[Amenhotep Bakari-Sellers]

They're providing chances for majorties, right? What I expect:

- Chance for GOP majority in the House: 93%

- Chance for GOP majority in the Senate: 64%

No, have you seen the polls Barnes and Fetterman, CCM, Hassan, Warnock and Bennett are tied or leading and Kellyits a 40 percent chance of Rs winning S the Hits 60/40

Its 150 days to election, Rs can still win most tossup races due to fact Prez Biden is at 40s approvals.

They're losing in WI, PA

[Alben Barkley]

Toss-up Senate, you love to see it!

Really could be a reverse 2018 in the end thanks to the terrible candidates the GOP recruited this cycle.

[LabourJersey]

Interesting how the model gives Kelly a 62% chance of winning right now, which seems quite generous compared to 53% for Warnock

[Roll Roons]

Interesting how the model gives Kelly a 62% chance of winning right now, which seems quite generous compared to 53% for Warnock

There hasn't been much GE polling in Arizona because the primary isn't fully settled, but there were some earlier polls that showed Walker leading by a bit.

[Amenhotep Bakari-Sellers]

EDay is after Labor Day it doesn't end on the 4th of July to my R friends

[MT Treasurer]

MI-3 (Meijer, Biden +9): "Likely R", 90% chance of GOP hold
MI-10 (open, Trump +1): "Lean R", 73% chance of GOP hold

WI-GOV (Biden +<1): "Lean D", 64% chance of DEM hold
OR-GOV (Biden +16): "Lean D", 63% chance of DEM hold

Ok.

[Amenhotep Bakari-Sellers]

OR is a D hold and NH Gov is not because Hassan is not gonna win by as much as Ron Wyden if Ron Wyden whom is gonna get 70 percent wasnt on the ballot Tina K would be in trouble but he is but never say never in NH in a big tsunami Sununu can lose it's NH and Scott and Sununu always start out  70/30 and it gets close at end