IA-Selzer: Trump +7

[TiltsAreUnderrated]

Oct 26-29
814 likely voters
MoE: 3.4%
Changes with Sep 14-17

Trump 48% (+1)
Biden 41% (-6)
Don't want to tell 5% (not previously included)
Someone else 3% (-1)
Would not vote 0% (n/c from 0% at no votes)
Not sure/don't remember 2% (-1 from Not sure at 3%)

That "Don't want to tell" has taken a whole 5% of the sample. Probably not shy Biden voters, but interesting.

[TiltsAreUnderrated]

Greenfield and/or Biden need to release their statewide internals TONIGHT to rebut this. If they don’t, the next 48 hours are going to be a disaster.

Democrats being panicked is brilliant for the Biden campaign. Let everyone think it's a pure tossup election.

Exactly - preferably tilt or lean R, but competitive enough for the doomers to feel obligated to participate. The media pushing a horserace narrative warped public expectations in a frankly irresponsible manner, but it benefited Democrats in the general election phase (although not in the primaries).

[TiltsAreUnderrated]

Is it true that Selzer changed methodology since their last poll? That seems kind of ...important.

They do have a new option for respondents to pick - "prefer not to answer" - that has gotten 5% in the presidential race and 4% in the Senate race.

As with Marist, weighting by education is not the be-all-and-end-all.

[TiltsAreUnderrated]

It has been about 2 hours since the poll dropped and I intentionally muted "iowa" on twitter and didn't peek into this thread too often so I could gather some thoughts about this.

Ultimately, I think everyone here has been overhyping this poll.

In their last poll, it was a tie at 47% to 47%, in this poll, Trump leads 48% to 41%. Has Biden really lost 6% of his support in just a month? Well, no, of course not.

This poll differs from the last poll in which they included an option for people who've already voted: "don't want to tell." And that option is taking.. 5%! It's also presumably taking up a significant chunk of the people who say they've already voted, which is a group likely to benefit Biden.

It's not amazing news for the Democrats, but the jank crosstabs and the inclusion of the "I don't want to tell you who I voted for option" might make this poll look deceptively good to Trump.

that question is just straight up bizzare, wtf. No other pollster shows this drop

Plenty have a "refused to answer/pick" option which is analogous.

[TiltsAreUnderrated]

Its been a couple hours after this poll come out, anyone with a clear head want to make something of it or is it an outlier?

Most likely an outlier. Trump isn’t up by 15 in IA-01 lol

Considering the MoE on that crosstab (7.1), it doesn't imply the topline is junk.

[TiltsAreUnderrated]

It doesn’t weight for education, for god’s sake.

Why on earth are people giving this poll ANY consideration then, considering the big talking point is that was what caused 2016's error.

Also... I don't care who is in the lead. 11% undecided/other is absolutely ridiculous with 3 days to go. For any pollster.

Also, OMG at the R+15 result in IA-01. What in the actual hell is this poll

That undecided figure includes people who've already voted and don't recall how they voted.

[TiltsAreUnderrated]

So apparently this was the deleted text:

https://pastebin.com/raw/gYVALf47

I’m REALLY starting to think this poll was compromised. After a screw-up in the primaries too, Selzer is in real danger of losing “gold standard” status for good.

The way to lose the 'gold standard' status is to not publish your outliers. Selzer has done the right thing here even if the poll is way off and should be applauded for it.

This is better than being denied the final Iowa poll (thanks to a particular Buttigieg supporter).

[TiltsAreUnderrated]

T W E N T Y  P A G E S

It's like the CoD: Infinite Warfare trailer - we're on the road to 30!

[TiltsAreUnderrated]

Iowa is MAGA country.

Yep.

It's safe R in the long term because of trends but the thinking that its Senate race was more in reach than most other competitive seats beyond the core four was correct. It wasn't that bad relative to the national environment, although it will be next time.