2022 Generic Ballot / Recruitment / Fundraising / Ratings Megathread

[GALeftist]

>enter discussion of possible Dobbs bump for Democrats on GCB
>post good GCB poll for Rs
>partisan R pollster
>conducted before Dobbs
>exact same GCB spread as their poll from two months ago even though Biden's approval dropped by 10 points since then
>refuse to elaborate
>leave

[GALeftist]

Rasmussen making sure that R+0.9 538 average didn't stay for long!

https://twitter.com/stella2020woof/status/1550490139956285443

I increasingly think that these polls are just straight up fake.

[GALeftist]

WRT Youngkin winning/Murphy almost losing when the generic ballot was ~D+3, it's worth noting that polling at the time had, on average:

- Youngkin 48.2, McAuliffe 46.9, undecided/other 5.0, Youngkin+1.4
- Murphy 50.5, Ciattarelli 42.7, undecided/other 6.8, Murphy+7.8

Final results were:

- Youngkin 50.6, McAuliffe 48.6, other 0.8, Youngkin+2.0
- Murphy 51.2, Ciattarelli 48.0, other 0.8, Murphy+3.2

These were the same polls that were showing Democrats leading the GCB and they were not horrible misses. VA was obviously very close, but even NJ wasn't that bad; taking the polls at face value, A. Ciattarelli did better with the undecideds than Youngkin, and B. there were more undecideds in NJ than VA. It's not like the polls had Murphy at 55 or something like that. So, if the polls are consistent with both reality and with a D+3 GCB, why are we using 2021 as evidence that they are faulty? I can think of a number of reasons why there might be a split between the numbers. For one thing, both these states are trending left, so it's unsurprising that they'd have some residual strength for opposite party gubernatorial candidates, same as Kentucky or whatever. For another, voters might prioritize things in Congress that wouldn't make sense to prioritize in gubernatorial races or vice-versa. Honestly, if these states had had Senate races in 2021, I think Democrats would have won both and NJ would not have been close. Thirdly, a LV screen probably makes the GCB about even. Finally, and perhaps most importantly, the GCB had a much higher number of other/undecideds – 15.6% – than either set of gubernatorial polls did. If the midterms had been held then, my guess is that undecideds would have broken for Republicans and they would have won the house popular vote by maybe 3 points or so.

All this is to say that I do not think there is any good evidence in favor of there having been a big polling miss in 2021 gubernatorial races.

[GALeftist]

Rasmussen will Rasmussen.... their GCB is R+3 today, 46-43.

IT has gone from R+10 to R+8 to R+5 and now R+3

itshappening.gif

Between this and the Washington primary results I wouldn't be surprised if Democrats narrowly lead the GCB right now. We'll see what happens in the next hundred or so days.

[GALeftist]

So we’ve gone back from “GCB will be far more Republican than 2010” to “atypical midterm/2002 redux”, then? Got it.

To be honest I didn't believe all the "atypical midterm" stuff either until this week. I thought a 6 point popular vote victory was probably the best possible outcome for Republicans and therefore unlikely, but I also thought all the counterarguments to the fundamentals were mostly cope, we should trust empirics over intuition, and that polling (while obviously lightyears better than intuition, which is basically useless in the presence of real data imo) should count for less than empirics after recent high-profile misses. This week what is really convincing me though are the Washington primaries, which are looking pretty trash for Republicans, like in the neighborhood of an even GCB. If that keeps being the case as more votes are counted I think we are probably looking at an R+2 November popular vote margin, and frankly that's probably erring on the side of caution.

[GALeftist]

https://twitter.com/alivitali/status/1556245847721082883?s=21&t=MalYBwgA5E0i...

It's pretty safe to say that the "Americans don't care about Dobbs" narrative has been pounded into the ground.

C'mon bro just one more month bro just one more month and everyone will move on bro our position isn't actually that unpopular bro out voters are just too stupid to understand referenda bro trust me

[GALeftist]

A D House is a lot like a Trump win in 2020. 538 has it at a 1 in 4 chance, but that's not zero; sometimes you flip a coin twice and get two heads. Not often enough where you'd guess it would happen in any individual case, sure, but not so infrequently that the possibility should be discounted either.

[GALeftist]

I was reliably informed that the magic of after-labor-day-ness would cause a Republican tidal wave of epic proportions, yet this morning I woke up and found to my horror that Democrats still led the GCB. What gives??

[GALeftist]

George Will’s famous quote :

Will said that President Barack Obama is "the most crystalline" representation today of Woodrow Wilson's philosophy, while James Madison's ideas are represented by conservative movements such as the Tea Party. "In the Madisonian persuasion, of which I am acolyte, gridlock is not an American problem, it is an American achievement,

https://www.theatlantic.com/politics/archive/2010/09/george-will-in-defense-of-partisanship/63856/

Famously conservative James Madison. Lel

[GALeftist]

I just googled abortion 100 times in a row, doing my part o7

[GALeftist]

Cook and Sabato put TX-34 in tossup today. Why lol. The district is BIDEN+16 people. It's basically identical to New Jersey in terms of partisanship. Incumbency shmincumbency, Gonzalez is heavily favored.

[GALeftist]

Check out RCP's new "projection" methodology. This is just sad bruh. I got a particular kick out of the fact that there was no possible way for them to cook/unskew the numbers enough to get Walker ahead so they just arbitrarily decided that he would win lol.

[GALeftist]

Check out RCP's new "projection" methodology. This is just sad bruh. I got a particular kick out of the fact that there was no possible way for them to cook/unskew the numbers enough to get Walker ahead so they just arbitrarily decided that he would win lol.

Most of this isn't completely ridiculous - it's not exactly how I'm looking at things, but trying to account for prior polling error isn't the worst idea - but the Georgia s*** is just flat-out partisan hackery.

I'd have a lot more sympathy for their trying to account for a possible polling miss if they hadn't obviously cooked various things to get Oz just over the hump.

34 days to election day in 2020, RCP had Biden+6.3; 538 had Biden+5.6. Essentially the same. 34 days to election day in 2022, RCP has Fetterman+4.3; 538 has Fetterman+6.6. That's a >2 point gap. So essentially what this chart is showing, at least in PA, is that they took the error between their polling averages and the actual results in 2020 when they were more or less an honest broker (things really went downhill only in the latter days of 2020 to my memory) and then carried it over to now, when they are obviously putting their thumb on the scale, and got like a 0.7 point lead for Oz. Yeah, no kidding you're going to come out with an Oz win if you do that.

At the end of the day, it seems obvious that the good folks at that site want to dabble in punditry because they disagree with their own polling averages even after they've significantly altered them to be more R favorable. That's perfectly defensible. What is not defensible, in my opinion, is cloaking what are essentially their educated guesses in this pseudoscientific nonsense.

EDIT: Added parenthetical

[GALeftist]

An update from RepublicanCuratedPolls – Tom Bevan has now (tearfully, one assumes) accepted that Warnock is favored, but to compensate, he has for some reason decided that a 2 point (I can't put enough quotation marks around the following word so I won't try) "adjusted" Kelly lead is actually a Masters lead.

[GALeftist]

Just put it in the average people good lord

[GALeftist]

RCP unironically moved NY-SEN to Likely.

[GALeftist]

All the dooming/gloating is annoying. Behave like adults, people. Make your prediction and move on. It will speak for itself come November 8.

[GALeftist]

https://twitter.com/AndrewSolender/status/1586804261874188288?t=1b9fNTpBNAjtUx4e...

Please don't throw me in the briar patch br'er NRCC

[GALeftist]

Boy, this place is getting worse and worse the closer election day gets. I understand emotions run high, but not everyone who disagrees with you is besmirching your honor or whatever. Chill out. We'll see what happens in under a week now.

[GALeftist]

I'm pretty much done caring about polls for now, but one thing I will say is that weighting polls more heavily based on how closely they hew to your priors is pretty asinine and no better than just ignoring polls altogether. It's one thing if it contradicts the fundamentals by some ludicrous amount, like that rumored poll that showed that Thune was in danger or whatever, but you shouldn't just chuck polls like that YouGov one.

[GALeftist]

I'm pretty much done caring about polls for now, but one thing I will say is that weighting polls more heavily based on how closely they hew to your priors is pretty asinine and no better than just ignoring polls altogether. It's one thing if it contradicts the fundamentals by some ludicrous amount, like that rumored poll that showed that Thune was in danger or whatever, but you shouldn't just chuck polls like that YouGov one.

Hard disagree. There’s no good poll that has a D +14 electorate and Biden +6 recall. That’s not about priors, that just makes no sense. It’s false equivalence to just throw these in the average and frankly no better than junk GCB polls from Trafalgar and Rasmussen.

How is Biden+6 unreasonable? That's like 2 points off from 2020, not terribly ludicrous for an RV poll. (Parenthetically, I've always found the party registration metric pretty useless; I've seen it used more to discount polls people don't like than to provide any actual insights.)

[GALeftist]

https://twitter.com/NateSilver538/status/1588916015869878272?t=J9Wz9Bfa732UFxcf6...

Question I have about this: hasn't there also been a dearth of R internal polling (except perhaps from the Walker campaign) and nonpartisan polling? What's the incentive structure there? If you're an R, wouldn't you want to be releasing all the absurd leads you're finding in your polls? If you're a nonpartisan pollster finding R leads, why wouldn't you release them?

[GALeftist]

I'm quite baffled by the PA change. It sure is close, but Oz has led in *1* nonpartisan poll, and it was by a single point. He hasn't been able to top 48% in any poll besides one (and that was the CBS/YouGov where people were essentially forced to pick).

Even the GOP polls that have him winning only have him at 47-48%, which certainly seems like his ceiling.

Seems like their just going off of the "PA polls overestimate Democrats" usual thing.

The actual reason is that A. no one wants to bet against Ralston and B. no one wants to say Democrats are favored to take the Senate which C. necessitates giving GA and PA to the GOP. It's shoddy reasoning, but might be rewarded. IDK

[GALeftist]

Feeling really pleased with my predictions but honestly should have gone with my gut which was even more bullish for Dems. Hope some of the conservatives on this board learned a valuable lesson last night. I think watching the House, which I thought was Safe D, turn into a slow knife fight for a bare majority in 2020 was really valuable for my ability to think critically.