2022 Generic Ballot / Recruitment / Fundraising / Ratings Megathread

[Person Man]

Echelon has D+2 on GCB, 48-46

http://echelonin.wpenginepowered.com/wp-content/uploads/October-2022-Omnibus-FULL-EXTERNAL-Topline.pdf

Where were they before?

[wbrocks67]

Echelon has D+2 on GCB, 48-46

http://echelonin.wpenginepowered.com/wp-content/uploads/October-2022-Omnibus-FULL-EXTERNAL-Topline.pdf

Where were they before?

I believe they were D+5 in September, not sure if it was LV or RV though.

[Ferguson97]

https://twitter.com/politics_polls/status/1585648868133412864

[Ferguson97]

D+3, D+2, R+4, R+1

Lol so we’ve just got no clue what Election Day is gonna look like, that rocks

[GeorgiaModerate]

D+3, D+2, R+4, R+1

Lol so we’ve just got no clue what Election Day is gonna look like, that rocks

I don't see what the problem is with those.  A range of results like that is completely consistent with a GCB that is close to tied.  Expecting the polls to show results much closer to each other is unrealistic just due to margin of error, without even considering other possible error factors.  Now if they were showing R+10, D+10 then we'd have a problem.

[UncleSam]

https://twitter.com/politics_polls/status/1585648868133412864

At this point RV is just not something that should even be considered in averages. We are within two weeks of the election, people know whether they’re gonna vote or not.

[GeorgiaModerate]

https://twitter.com/politics_polls/status/1585648868133412864

At this point RV is just not something that should even be considered in averages. We are within two weeks of the election, people know whether they’re gonna vote or not.

Except that likely voter models are more than just asking people whether they're gonna vote.  The pollster makes certain assumptions about turnout in their LV models, which may or may not turn out to be true.  Since this looks like a high-turnout election for both parties at this point, using an RV model is probably just as reasonable.

Also, IIRC there have been studies showing that voter intention actually isn't that predictive.  That is, following up with such voters after the election showed that a surprisingly low percentage (68% from memory, but that may be inaccurate) of the "definitely will vote" actually did vote, while a surprisingly high percentage of the "maybe" or "possibly" groups voted.

[Kahane's Grave Is A Gender-Neutral Bathroom]

Even some of the GOP hacks are admitting only a mildly R or neutral environment.

[Dani Rose]

I'm beginning to suspect that the GOP surge last week might have been certain pollsters overreacting to a relatively lackluster week of early voting for us.

[kwabbit]

https://twitter.com/politics_polls/status/1585648868133412864

At this point RV is just not something that should even be considered in averages. We are within two weeks of the election, people know whether they’re gonna vote or not.

Except that likely voter models are more than just asking people whether they're gonna vote.  The pollster makes certain assumptions about turnout in their LV models, which may or may not turn out to be true.  Since this looks like a high-turnout election for both parties at this point, using an RV model is probably just as reasonable.

Also, IIRC there have been studies showing that voter intention actually isn't that predictive.  That is, following up with such voters after the election showed that a surprisingly low percentage (68% from memory, but that may be inaccurate) of the "definitely will vote" actually did vote, while a surprisingly high percentage of the "maybe" or "possibly" groups voted.

I’m not opposed to just releasing RV numbers and letting people apply their own LV ideas, but doing so doesn’t make the poll more accurate to the final results. The GOP is going to have a 0-4pt gain from RV to LV. Just releasing RV implies that the typical gap is 0, when a gap of zero is a great result for Democrats.

I believe the NYT has voter history so I generally trust their LV model, which had a 2.5pt gain for the GOP. In conjunction with voter history vote intention is useful, but on its own it’s basically useless. Not all pollsters invest in the voter files.

[MRS DONNA SHALALA]

https://twitter.com/politics_polls/status/1585648868133412864

At this point RV is just not something that should even be considered in averages. We are within two weeks of the election, people know whether they’re gonna vote or not.

Except that likely voter models are more than just asking people whether they're gonna vote.  The pollster makes certain assumptions about turnout in their LV models, which may or may not turn out to be true.  Since this looks like a high-turnout election for both parties at this point, using an RV model is probably just as reasonable.

Also, IIRC there have been studies showing that voter intention actually isn't that predictive.  That is, following up with such voters after the election showed that a surprisingly low percentage (68% from memory, but that may be inaccurate) of the "definitely will vote" actually did vote, while a surprisingly high percentage of the "maybe" or "possibly" groups voted.

I think Pollster said in another thread that those models skew towards people who tend to answer questions with more certainty, which is often older men (someone correct me if i'm not remembering correctly).

[Calthrina950]

It would be something if this ended up being a neutral or Democratic year, and would be unprecedented. Over the last century, there have only been four midterms in which the President's party did well - 1934 (New Deal realignment), 1962 (Cuban Missile Crisis), 1998 (Clinton impeachment), and 2002 (9/11 effect). What distinguishes 2022 from these preceding years is that the incumbent President in those years enjoyed strong approval ratings, while Biden's approval (according to RCP) is at around 43%, similar to Obama's in 2010. It would be extraordinary if Democrats managed to hold their ground with the President underwater by 10 points in approvals.

[2016]

It would be something if this ended up being a neutral or Democratic year, and would be unprecedented. Over the last century, there have only been four midterms in which the President's party did well - 1934 (New Deal realignment), 1962 (Cuban Missile Crisis), 1998 (Clinton impeachment), and 2002 (9/11 effect). What distinguishes 2022 from these preceding years is that the incumbent President in those years enjoyed strong approval ratings, while Biden's approval (according to RCP) is at around 43%, similar to Obama's in 2010. It would be extraordinary if Democrats managed to hold their ground with the President underwater by 10 points in approvals.

Most people who don't answer Polls are Trump America First Folks because Biden declared them enemies of the Republic.

[Devils30]

It would be something if this ended up being a neutral or Democratic year, and would be unprecedented. Over the last century, there have only been four midterms in which the President's party did well - 1934 (New Deal realignment), 1962 (Cuban Missile Crisis), 1998 (Clinton impeachment), and 2002 (9/11 effect). What distinguishes 2022 from these preceding years is that the incumbent President in those years enjoyed strong approval ratings, while Biden's approval (according to RCP) is at around 43%, similar to Obama's in 2010. It would be extraordinary if Democrats managed to hold their ground with the President underwater by 10 points in approvals.

One thing we are missing is polls from the 1876-1900 era which is probably the most similar to today's. At some point one side will breakthrough but the activists for now are content to only play to the base.

[Calthrina950]

It would be something if this ended up being a neutral or Democratic year, and would be unprecedented. Over the last century, there have only been four midterms in which the President's party did well - 1934 (New Deal realignment), 1962 (Cuban Missile Crisis), 1998 (Clinton impeachment), and 2002 (9/11 effect). What distinguishes 2022 from these preceding years is that the incumbent President in those years enjoyed strong approval ratings, while Biden's approval (according to RCP) is at around 43%, similar to Obama's in 2010. It would be extraordinary if Democrats managed to hold their ground with the President underwater by 10 points in approvals.

One thing we are missing is polls from the 1876-1900 era which is probably the most similar to today's. At some point one side will breakthrough but the activists for now are content to only play to the base.

True, although polling didn't truly exist until the early 20th century. This country is as polarized now as it was during the Gilded Age, and I've seen that comparison before. On balance, I still think this year is going to be a Republican wave. Many pollsters will be deeply discredited, even more then they have been, if Democrats overperform in a few weeks and either minimize Republican gains or gain outright.

[Mr.Phips]

It would be something if this ended up being a neutral or Democratic year, and would be unprecedented. Over the last century, there have only been four midterms in which the President's party did well - 1934 (New Deal realignment), 1962 (Cuban Missile Crisis), 1998 (Clinton impeachment), and 2002 (9/11 effect). What distinguishes 2022 from these preceding years is that the incumbent President in those years enjoyed strong approval ratings, while Biden's approval (according to RCP) is at around 43%, similar to Obama's in 2010. It would be extraordinary if Democrats managed to hold their ground with the President underwater by 10 points in approvals.

I had hope for this in 2014, but the fundamentals won out unfortunately.  At this point the senate races (PA and AZ) where Dems were doing better than the fundamentals are now lining up more closely with an environment that is around R+1 or R+2.  

[Tekken_Guy]

It would be something if this ended up being a neutral or Democratic year, and would be unprecedented. Over the last century, there have only been four midterms in which the President's party did well - 1934 (New Deal realignment), 1962 (Cuban Missile Crisis), 1998 (Clinton impeachment), and 2002 (9/11 effect). What distinguishes 2022 from these preceding years is that the incumbent President in those years enjoyed strong approval ratings, while Biden's approval (according to RCP) is at around 43%, similar to Obama's in 2010. It would be extraordinary if Democrats managed to hold their ground with the President underwater by 10 points in approvals.

I had hope for this in 2014, but the fundamentals won out unfortunately.  At this point the senate races (PA and AZ) where Dems were doing better than the fundamentals are now lining up more closely with an environment that is around R+1 or R+2.  

GA and WI too.

[Mr.Phips]

It would be something if this ended up being a neutral or Democratic year, and would be unprecedented. Over the last century, there have only been four midterms in which the President's party did well - 1934 (New Deal realignment), 1962 (Cuban Missile Crisis), 1998 (Clinton impeachment), and 2002 (9/11 effect). What distinguishes 2022 from these preceding years is that the incumbent President in those years enjoyed strong approval ratings, while Biden's approval (according to RCP) is at around 43%, similar to Obama's in 2010. It would be extraordinary if Democrats managed to hold their ground with the President underwater by 10 points in approvals.

I had hope for this in 2014, but the fundamentals won out unfortunately.  At this point the senate races (PA and AZ) where Dems were doing better than the fundamentals are now lining up more closely with an environment that is around R+1 or R+2.  

GA and WI too.

WI yes, but GA is a bit different due to the runoff.  I don’t see either candidate getting to 50% there.

[Ferguson97]

It would be something if this ended up being a neutral or Democratic year, and would be unprecedented. Over the last century, there have only been four midterms in which the President's party did well - 1934 (New Deal realignment), 1962 (Cuban Missile Crisis), 1998 (Clinton impeachment), and 2002 (9/11 effect). What distinguishes 2022 from these preceding years is that the incumbent President in those years enjoyed strong approval ratings, while Biden's approval (according to RCP) is at around 43%, similar to Obama's in 2010. It would be extraordinary if Democrats managed to hold their ground with the President underwater by 10 points in approvals.

I think the difference here is that whereas 1934/1962/1998 were specifically due to the actions taken by/against the President and 2002 was a result of 'rally around the flag', the Dobbs decision is completely independent of Biden's actions, so people may vote for Democrats even if they're not improving their opinion of him.

[wbrocks67]

It's really a choose your own adventure!

InsiderAdvantage: Biden 45/54 approval, R+5 on GCB
Echelon: Biden 45/54 approval, D+2/3 on GCB

[wbrocks67]

FWIW, Civiqs has remained steady

https://twitter.com/DrewLinzer/status/1585671635851280385

[wbrocks67]

Damn, no wonder Scheller dropped $700K of her own money into the race. Wild raised $500K+ in just 18 days!

https://twitter.com/CATargetAlt/status/1585699342131859475

[Devils30]

It's really a choose your own adventure!

InsiderAdvantage: Biden 45/54 approval, R+5 on GCB
Echelon: Biden 45/54 approval, D+2/3 on GCB

It is really tempting to just block RCP, Twitter and even this site for 12 days. Not sure what looking at the polls is going to accomplish.

[Ferguson97]

FWIW, Civiqs has remained steady

https://twitter.com/DrewLinzer/status/1585671635851280385

How did their 2018 tracker look?

[soundchaser]

FWIW, Civiqs has remained steady

https://twitter.com/DrewLinzer/status/1585671635851280385

How did their 2018 tracker look?

Hard to tell because they've scrubbed their results page, but this suggests it underestimated Dems a bit. (Although I remember them being one of the few to nail the Texas Senate race.)

https://twitter.com/markos/status/1265329307167014918