The Economist Senate and House Models are up

[Stuart98]

Senate model
House model

Gives Dems a 67% chance of controlling the senate, 99% chance of controlling the house.

[Astatine]

Preferable to 538's model imho. Numbers seem quite realistic in general (it still annoys me that AL has a 30 % chance of going Democratic in the 538 Classic model, here it's 6 %).

[Amenhotep Bakari-Sellers]

AZ, CO, GA, IA, KS, ME, MT, and SC are 8 to 9 seats the max Dems are gonna get, Dems aren't gonna win 12 or 13 seats

AK, ID, KY, MS, and TX need to be triaged, like they are supposed to.

[MT Treasurer]

John Hickenlooper has a 78% chance to gain Cory Gardner's Senate seat.
Tommy Tuberville has a 94% chance to gain Doug Jones' Senate seat.
Gary Peters has a 95% chance to win reelection.

Wow Senator Peters is an even stronger incumbent than I thought if his seat is less competitive than CO and AL.

John Hickenlooper has a 78% chance to gain Cory Gardner's Senate seat.
Cal Cunningham has a 78% chance to gain Thom Tillis' Senate seat.

Yikes, imagine how well Hickenlooper could be doing if he was like that Cunningham guy.

[Sorenroy]

Democrats — 236 (53.1%)
Republicans — 199 (46.9%)

The Economist seems to think that third parties/independents will get exactly 0% of the vote in every seat in the country.

[GeorgiaModerate]

The Economist seems to think that third parties/independents will get exactly 0% of the vote in every seat in the country.

In the presidential model, the percentages are the modeled shares of the two-party vote, not the all-party vote.  I assume this model is doing the same thing.

[Penn_Quaker_Girl]

Interesting that there's a non-near-zero chance of Arizona having Democratic senator AND a Republican senator in Colorado. 

[President Johnson]

Why has Hick only a 78% chance? Mark Kelly has 90%. Also think 70% chance for Daines is a bit too much.

House model looks about correct.

[Xing]

Better than 538's model, though some (CO and MI, for example) are pretty laughable. Also, what's up with Republicans winning the Georgia Special election 73-27?

[GeorgiaModerate]

Better than 538's model, though some (CO and MI, for example) are pretty laughable. Also, what's up with Republicans winning the Georgia Special election 73-27?

That's not unreasonable when you consider the nontrivial chance of a Collins vs. Loeffler runoff.

[ProgressiveModerate]

Like these models a lot more than 538. I don't buy that Peters is at a 95% chance of winning though; seems mostly polling based with a small incumbency boost and that's it.

[Amenhotep Bakari-Sellers]

Daines, Marshall, Graham and Cornyn is a blue wave, will lose, in a Recession like this, as well as Sullivan. Both Bollier and Gross are Surgeon Generals and Drs in a time of a Pandemic, thats why they can win upsets

[OneJ]

I think they're being pretty bearish on Valenzuela's chances in TX-24.

[Stuart98]

I think they're being pretty bearish on Valenzuela's chances in TX-24.

They're also bullish on Peterson's chances in MN-07. What's most surprising to me though is how bearish they are on Gimenez' chances in FL-26th. They give him a 2% chance in a race everyone else has as a tossup and that I think he's narrowly favored in. I don't get it.

[ProgressiveModerate]

I think they're being pretty bearish on Valenzuela's chances in TX-24.

They're also bullish on Peterson's chances in MN-07. What's most surprising to me though is how bearish they are on Gimenez' chances in FL-26th. They give him a 2% chance in a race everyone else has as a tossup and that I think he's narrowly favored in. I don't get it.

They're model has partisanship winning out heavily and Hillary won FL-26 by like 16%, and this cycle, the national environment looks better for Ds. This is also why Petersen's chances in MN are very low. They expect split-ticket voting to be quite low, but there are certain parts of the country where there are interesting dynamics at play

[Vosem]

The Senate model seems oddly bullish on Democrats in the Southeast; there's no way Ossoff has a 40% chance of winning, or Harrison at 30%. I'd guess that explains how they're predicting 51.4 Democratic Senate seats when their own likeliest scenario is the same as everyone else's (50-50).

James having only a 5% chance is odd; it's not like he's down by all that much more than (say) Bullock, and MT is far redder than MI is blue. It feels like there's an underlying prediction here that undecideds break Democratic. (Though not everywhere; I'd buy Gideon at 70%.)

The House model feels like it has some similar issues, but I'll note that I find it really weird that they're confident enough to call four seats as Democratic pickups, and one of them is...the Schweikert/Tipirneni seat? I would by no means be shocked if Democrats gain it, but it's nowhere near their fourth likeliest pickup. What happened to GA-7 or TX-24 or like, any of the Hillary seats held by Republicans?

[Water Hazard]

Am I the only one who misses the plain and simple interface with states shaded fully on the maps (ala 538's 2016 PRES/SEN or 2018 House? Not a fan of the boxes and cartograms that seem to be popping up in all the models this year.

[Amenhotep Bakari-Sellers]

These models don't predict a  blue wave, AZ, CO, GA, IA, KS, ME, MT, NC, SC and TX will go D in a blue wave, Bullock and Bollier will win in a blue wave. These models underestimate Bullock's chances.

[GeorgiaModerate]

The Senate model seems oddly bullish on Democrats in the Southeast; there's no way Ossoff has a 40% chance of winning...

40% for Ossoff is in the right ballpark.  He's even with or very slightly behind Perdue at the moment.

[Penn_Quaker_Girl]

Am I the only one who misses the plain and simple interface with states shaded fully on the maps (ala 538's 2016 PRES/SEN or 2018 House? Not a fan of the boxes and cartograms that seem to be popping up in all the models this year.

Oh, but it needs to be "shiny" and "sleek" and "trendy". 

A Senate map in the shape of Chuck Schumer's head? Sure!