538 senate model now out

[Gustaf]

They give Democrats 58% chance at the senate with AZ, CO, NC and ME the 4 most likely pickups in that order. And it has a map!

https://projects.fivethirtyeight.com/2020-election-forecast/senate/

[n1240]

Seems like they don't have detailed forecast information yet but looks like they give too much of an advantage to some incumbents, particularly Jones, Gardner, and Collins.

[Pericles]

Seems decent, though I'm not sure what non polling factors they are including in there. Gideon is consistently getting comfortable leads, but it's only a 51% chance for her in the forecast.

[Gustaf]

Seems decent, though I'm not sure what non polling factors they are including in there. Gideon is consistently getting comfortable leads, but it's only a 51% chance for her in the forecast.

The polls-only model gives Gideon like 75%, the Deluxe one drops it to 51%.

[Skye]

That's... very favorable for incumbents.

[Pericles]

Seems decent, though I'm not sure what non polling factors they are including in there. Gideon is consistently getting comfortable leads, but it's only a 51% chance for her in the forecast.

The polls-only model gives Gideon like 75%, the Deluxe one drops it to 51%.

Oh I didn't realise there was a polls-only model, thank you. I like this one.

[Stuart98]

Polls only model seems fine, polls+ or whatever they call it is ludicrous. Doug Jones does not have a 28% chance of winning.

[ProgressiveModerate]

Incumbency is weighted too heavily. after that, the model seems pretty fair and decent though. I like how they actually have a map. Partisanship should be weighted more. Part of the reason why Jones chances in AL are so high in the Deluxe model is because the "experts" tend to be very cautious about calling flips (most of them still say CO is tossup!)

[Figueira]

Yeah, really not buying that Democrats have a higher chance of holding Alabama than flipping either of the Georgia seats.

[Alben Barkley]

Like with the presidential forecast, they are being too conservative, still pretending things like incumbency really matter.

Nate is always the one going on about looking at the polls and data, don’t let pundit type speculation interfere, yet he stuffs his models with all sorts of dubious assumptions that obscure the data. Yes, there might technically be an empirical justification for some of it, but that assumes we don’t live in a country with a rapidly changed and highly polarized political climate. But we do. Things that mattered a lot a decade or two ago just don’t anymore. For instance, in 2016 there were NO states that split their party vote between president and senate. That had never happened before, and it’s a sign of just how polarized we are. That has not changed in the last four years, and thus it should be assumed that since Trump is the heavy favorite in Alabama, Doug Jones is the heavy underdog. The presidential forecast actually should affect the senate forecast in that way. It matters much more than whether Jones is the incumbent or not.

[Amenhotep Bakari-Sellers]

CO  46th. Road to 51
ME 47th
AZ. 48th
NC 49th
Ga R 50th
GA S 51th

SC Road Filibuster wishlist 53td
IA 54th
KS 55th
AK 56th
MT 57th
TX  58th
ID 59TH
AL/KY TIED 60TH

[Xing]

Not at all as bad as I was expecting, but it does seem too friendly to incumbents in several races, especially AL and CO.

[TiltsAreUnderrated]

Tfw ALSEN is more competitive than GASEN

[Amenhotep Bakari-Sellers]

This is a good model, but we are dealing with Trump, not a typical R, like Rubio, SC poll show filibuster proof wave status

In a Covid environment, Ds would only get 52 seats with a typical R other than Palin like Trump , that's the mistake that Rs and Fox news make. Also, Trump made fun of every R that he ran against, that's why Fiorina, Boehner, Kasich, Hogan,, and Hagel have failed to endorse Trump

[Landslide Lyndon]

Silver once again shows his hand. He says that this year his default will be the deluxe model for seemingly no other reason other than that it's the most favorable to Republicans.

[Nyvin]

Far too heavily weighted for incumbents.

[ProgressiveModerate]

Far too heavily weighted for incumbents.

A lot of that is because when they weight expert ratings (who generally are cautious about calling flips) it gives incumbents huge advantages, especially in races where the challenger is a clear favorite to win based off all other available data.

[Left Wing]

Projecting Cornyn winning by more than 10 is laughable.

[Left Wing]

Also having several other Ds getting a higher vote share than Adrian Perkins? wtf

[Kuumo]

Even though the model is overly bullish on incumbents, I'm relieved that it looks a lot like the 2018 model and doesn't show 100 maps on a graph that include stuff like Republican Minnesota and Democratic Kentucky on the same map.

[President Johnson]

32% chance of winning for Cory Gardner? Lmao.

[Antonio the Sixth]

Silver once again shows his hand. He says that this year his default will be the deluxe model for seemingly no other reason other than that it's the most favorable to Republicans.

Don't be such a hack. 538 always uses the fundamentals-inclusive model as its default. It did the same thing in 2018 and in fact that time it led to overestimating the Democrats (since most incumbents were Democrats that cycle). There's a legitimate criticism that his model weighs incumbency too heavily, but that's different from accusing Nate ing Silver of having a partisan agenda.

[Amenhotep Bakari-Sellers]

Ds are now investigating in AK, AZ, CO, GA, ME, NC and SC

[Left Wing]

Ds are now investigating in AK, AZ, CO, GA, ME, NC and SC

What are they investigating?

[n1240]

Personally might stick with using the Lite model myself, and I would like for 538 to publish more detailed forecast information like they do for the president race. The Deluxe model seems to overweigh Expert ratings and the Classic model seems to overweigh incumbency. Jones has a even higher probability in the Classic model to win at 35%, which just doesn't make sense to me (also possible they're taking into account his significant fundraising advantage).

Would also like for them to publish runoff advance probabilities for GA special, currently going through all the bubbles it seems like it's about 50/50 chance of two R runoff in their Deluxe forecast and 60/40 in their Lite forecast.