2018 Congressional Generic Ballot/House Polls Megathread 2 - no debates please

[LimoLiberal]

Reuters Ipsos 7/15-19

Democrats - 41 (-5)
Republicans - 37 (+4)

compared to 7/10-14 poll

Another huge swing from Ipsos for absolutely no reason. More reason to ignore online trackers.

Edit: Although I'm not sure where said results are.

http://polling.reuters.com/#poll/TM1212Y17/filters/PD1:1

I'm just transcribing it from the 538 aggregate...

[Ebsy]

Can we just post the Reuters/Ipsos weekly report and not report on the ridiculous fluctuations?

[KingSweden]

Can we just post the Reuters/Ipsos weekly report and not report on the ridiculous fluctuations?

Yeah I think thatd be best

[Devout Centrist]

I tried looking everywhere for this poll and I couldn't find it. This is the best I could find and this is D+9:

https://www.ipsos.com/sites/default/files/ct/news/documents/2018-07/2018_reuters_tracking_-_core_political_07_18_2018.pdf

[LimoLiberal]

I tried looking everywhere for this poll and I couldn't find it. This is the best I could find and this is D+9:

https://www.ipsos.com/sites/default/files/ct/news/documents/2018-07/2018_reuters_tracking_-_core_political_07_18_2018.pdf

Again, I'm just taking it off the 538 website: https://projects.fivethirtyeight.com/congress-generic-ballot-polls/

The link they used for the poll is the same link superbudgie had above, so maybe there was just an input error.

[GeorgiaModerate]

The link given above by superbudgie (http://polling.reuters.com/#poll/TM1212Y17/filters/PD1:1) is the one that 538 lists.  This just takes you to the overall Ipsos polling site.  To get the GCB, you need to click on "Midterms: how will you vote?"  That puts you into this chart with a date range up to the present, e.g. today's is http://polling.reuters.com/#!response/TM1212Y17/type/day/filters/PD1:1/dates/20180301-20180720/collapsed/true.

If you change the Time Frame selection to daily, you will get the 5-day rolling average.  This is what 538 puts in the database (every few days).  So this D+4 result is in fact a legitimate comparison with 538's previous entries for the Ipsos tracker.  As others have noted, this is quite noisy (even with a 5-day average) and the weekly is smoother.  If you select the weekly time frame, the most recent is D+12.7; but that's for the week ending July 15th, so it's not as current as the daily.  Take your pick: currency vs. noise.

[LimoLiberal]

NBC/WSJ high quality live-caller poll taken partially after the Helsinki summit

https://twitter.com/SteveKornacki/status/1021138283412287488

Last poll was in early June.

Democrats - 49 (-1)
Republicans - 43 (+3)

[BudgieForce]

The generic ballot aggregate will fluctuate. Try not to freak everybody.

[Hindsight was 2020]

The generic ballot aggregate will fluctuate. Try not to freak everybody.

BURN EVERYTHING DOWN

[Doimper]

The generic ballot aggregate will fluctuate. Try not to freak everybody.

BURN EVERYTHING DOWN

[Zaybay]

NBC/WSJ high quality live-caller poll taken partially after the Helsinki summit

https://twitter.com/SteveKornacki/status/1021138283412287488

Last poll was in early June.

Democrats - 49 (-1)
Republicans - 43 (+3)

Actually, looking at some stats in the poll
Overall approval:45-52

Indie approval:36-58
Dem approval: 9-90
Rep approval: 89-9

GCB: 49/43 D+6

Indie GCB 48/26 D+22

There is something fishy with these totals. With the Dems and Reps cancelling each other out, the Indies should swing it more D, but it doesnt. It seems like Rs were oversampled.

Edit: Accidentally put D+20 instead of D+22. Fixed.

[Bandit3 the Worker]

Indie GCB 48/26 D+20

Whoa!

I was thinking the GOP might still keep the House, but do people even realize just how bad this is for them? This is disastrous for them. Never seen anything like it.

[Hindsight was 2020]

NBC/WSJ high quality live-caller poll taken partially after the Helsinki summit

https://twitter.com/SteveKornacki/status/1021138283412287488

Last poll was in early June.

Democrats - 49 (-1)
Republicans - 43 (+3)

Actually, looking at some stats in the poll
Overall approval:45-52

Indie approval:36-58
Dem approval: 9-90
Rep approval: 89-9

GCB: 49/43 D+6

Indie GCB 48/26 D+20

There is something fishy with these totals. With the Dems and Reps cancelling each other out, the Indies should swing it more D, but it doesnt. It seems like Rs were oversampled.

Yeah how can Dems be winning indies this much but be up only 6?

[GeorgiaModerate]

NBC/WSJ high quality live-caller poll taken partially after the Helsinki summit

https://twitter.com/SteveKornacki/status/1021138283412287488

Last poll was in early June.

Democrats - 49 (-1)
Republicans - 43 (+3)

Actually, looking at some stats in the poll
Overall approval:45-52

Indie approval:36-58
Dem approval: 9-90
Rep approval: 89-9

GCB: 49/43 D+6

Indie GCB 48/26 D+20

There is something fishy with these totals. With the Dems and Reps cancelling each other out, the Indies should swing it more D, but it doesnt. It seems like Rs were oversampled.

Not at all.  If D, R, and I are approximately the same percentage of the sample (i.e., if I is 1/3 of the total) and D & R cancel each other out exactly, then the I proportion of D+20 should swing the entire sample to about 1/3 of that, or around D+6 or 7.

[GeorgiaModerate]

Here are the crosstabs: http://media1.s-nbcnews.com/i/today/z_creative/18570NBCWSJJulyPoll7-22-18Release.pdf

Party ID:

Strong Democrat ............................... 24
Not very strong Democrat ................. 7
Independent/lean Democrat .............. 9
Strictly Independent........................... 15
Independent/lean Republican ............ 11
Not very strong Republican ............... 6
Strong Republican ............................. 19
 Other (VOL) ..................................... 7
 Not sure ........................................... 2

Racial breakdown:

White............................................................... 77
Black ............................................................... 12
Asian ............................................................... 1
Other ............................................................... 3
 Hispanic (VOL) ............................................. 6
 Not sure/refused ........................................... 1

White, Non-Hispanic ....................................... 74

[Zaybay]

NBC/WSJ high quality live-caller poll taken partially after the Helsinki summit

https://twitter.com/SteveKornacki/status/1021138283412287488

Last poll was in early June.

Democrats - 49 (-1)
Republicans - 43 (+3)

Actually, looking at some stats in the poll
Overall approval:45-52

Indie approval:36-58
Dem approval: 9-90
Rep approval: 89-9

GCB: 49/43 D+6

Indie GCB 48/26 D+22

There is something fishy with these totals. With the Dems and Reps cancelling each other out, the Indies should swing it more D, but it doesnt. It seems like Rs were oversampled.

Whoops, made a mistake. The Indie vote is not D+20.....its D+22. Fixed it

NBC/WSJ high quality live-caller poll taken partially after the Helsinki summit

https://twitter.com/SteveKornacki/status/1021138283412287488

Last poll was in early June.

Democrats - 49 (-1)
Republicans - 43 (+3)

Actually, looking at some stats in the poll
Overall approval:45-52

Indie approval:36-58
Dem approval: 9-90
Rep approval: 89-9

GCB: 49/43 D+6

Indie GCB 48/26 D+20

There is something fishy with these totals. With the Dems and Reps cancelling each other out, the Indies should swing it more D, but it doesnt. It seems like Rs were oversampled.

Not at all.  If D, R, and I are approximately the same percentage of the sample (i.e., if I is 1/3 of the total) and D & R cancel each other out exactly, then the I proportion of D+20 should swing the entire sample to about 1/3 of that, or around D+6 or 7.

I think thats the problem that many are missing in this election. The D and R vote are not going to be equal. The D vote is always larger than the R vote, in almost every election. In the 2010 and 2014 midterms, D voter turnout was around the same as R voter turnout, except the Indies sided with the Rs. In 2016, the same dynamics occurred, except D turnout was much higher, but not high enough. If this is a midterm where the Indie vote is extremely pro D, and D turnout will be much higher than R turnout, then, lets just say many pundits and strategists will lose their jobs.

[GeorgiaModerate]

NBC/WSJ high quality live-caller poll taken partially after the Helsinki summit

https://twitter.com/SteveKornacki/status/1021138283412287488

Last poll was in early June.

Democrats - 49 (-1)
Republicans - 43 (+3)

Actually, looking at some stats in the poll
Overall approval:45-52

Indie approval:36-58
Dem approval: 9-90
Rep approval: 89-9

GCB: 49/43 D+6

Indie GCB 48/26 D+22

There is something fishy with these totals. With the Dems and Reps cancelling each other out, the Indies should swing it more D, but it doesnt. It seems like Rs were oversampled.

Whoops, made a mistake. The Indie vote is not D+20.....its D+22. Fixed it

NBC/WSJ high quality live-caller poll taken partially after the Helsinki summit

https://twitter.com/SteveKornacki/status/1021138283412287488

Last poll was in early June.

Democrats - 49 (-1)
Republicans - 43 (+3)

Actually, looking at some stats in the poll
Overall approval:45-52

Indie approval:36-58
Dem approval: 9-90
Rep approval: 89-9

GCB: 49/43 D+6

Indie GCB 48/26 D+20

There is something fishy with these totals. With the Dems and Reps cancelling each other out, the Indies should swing it more D, but it doesnt. It seems like Rs were oversampled.

Not at all.  If D, R, and I are approximately the same percentage of the sample (i.e., if I is 1/3 of the total) and D & R cancel each other out exactly, then the I proportion of D+20 should swing the entire sample to about 1/3 of that, or around D+6 or 7.

I think thats the problem that many are missing in this election. The D and R vote are not going to be equal. The D vote is always larger than the R vote, in almost every election. In the 2010 and 2014 midterms, D voter turnout was around the same as R voter turnout, except the Indies sided with the Rs. In 2016, the same dynamics occurred, except D turnout was much higher, but not high enough. If this is a midterm where the Indie vote is extremely pro D, and D turnout will be much higher than R turnout, then, lets just say many pundits and strategists will lose their jobs.

I agree with you that there is likely to be much greater D enthusiasm, and thus turnout, in November.  It will be interesting to see what how the GCB will be affected when the pollsters switch to likely voter models.

[Hindsight was 2020]

Nate just announced on Twitter the poll got scrapped due to a method issue

[GeorgiaModerate]

Nate just announced on Twitter the poll got scrapped due to a method issue

We've been stood up.

[Zaybay]

Nate just announced on Twitter the poll got scrapped due to a method issue

I think I found that method issue, BTW. I knew there was something wrong with those numbers.

[GeorgiaModerate]

Nate just announced on Twitter the poll got scrapped due to a method issue

I think I found that method issue, BTW. I knew there was something wrong with those numbers.

Different poll.  He's talking about the poll of 60 battleground districts that Upshot was working on.

[Zaybay]

Nate just announced on Twitter the poll got scrapped due to a method issue

I think I found that method issue, BTW. I knew there was something wrong with those numbers.

Different poll.  He's talking about the poll of 60 battleground districts that Upshot was working on.

ah, nvm.

[Dr Oz Lost Party!]

Half of that NBC/WSJ poll was taken before the summit.

[Bandit3 the Worker]

I can't believe 538 is actually using the discredited NBC poll.

[Gass3268]

It's on, like Donkey Kong:

https://twitter.com/MonmouthPoll/status/1021412755700682752