2020 Poll Hype Thread

[n1240]

https://twitter.com/gravismarketing/status/1274351164490907648?s=21

[Jopow]

No way Biden's up 16 in MN. Most I would say is 4.

[GeorgiaModerate]

No way Biden's up 16 in MN. Most I would say is 4.

"Poll disagrees with my priors, therefore it is wrong."

[The Mikado]

No way Biden's up 16 in MN. Most I would say is 4.

"Poll disagrees with my priors, therefore it is wrong."

I mean, I'd gladly take the "under" on a Biden +10 over/under in MN, and a Biden +17 from Gravis isn't changing my mind.

[GeorgiaModerate]

No way Biden's up 16 in MN. Most I would say is 4.

"Poll disagrees with my priors, therefore it is wrong."

I mean, I'd gladly take the "under" on a Biden +10 over/under in MN, and a Biden +17 from Gravis isn't changing my mind.

Looking at the 538 and Economist models, that looks like about an even bet at the moment.
TBH, I don't think Biden's up by 16 in MI either.  But a ceiling of +4 there, which is what Jopow claims, is also ridiculous.

[SingingAnalyst]

It's good that there are so many polling companies, as polls from the same week or fortnight can be aggregated into a single, large-sample result, with a lower margin of error.

Keep in mind, if a poll showing Biden+5 has a 3% MoE (about 1,100 respondents), that means it could be anything from Biden+11 to Trump+1.

[Gustaf]

No way Biden's up 16 in MN. Most I would say is 4.

"Poll disagrees with my priors, therefore it is wrong."

I mean, I'd gladly take the "under" on a Biden +10 over/under in MN, and a Biden +17 from Gravis isn't changing my mind.

Looking at the 538 and Economist models, that looks like about an even bet at the moment.
TBH, I don't think Biden's up by 16 in MI either.  But a ceiling of +4 there, which is what Jopow claims, is also ridiculous.

MN was R+1 in 2016, so if nothing changed and Biden is up by like 9 now you'd expect him to be ahead by 8 in MN. My guess is Minnesota isn't trending much either way but who knows.

[Beefalow and the Consumer]

No way Biden's up 16 in MN. Most I would say is 4.

"Poll disagrees with my priors, therefore it is wrong."

I mean, I'd gladly take the "under" on a Biden +10 over/under in MN, and a Biden +17 from Gravis isn't changing my mind.

Looking at the 538 and Economist models, that looks like about an even bet at the moment.
TBH, I don't think Biden's up by 16 in MI either.  But a ceiling of +4 there, which is what Jopow claims, is also ridiculous.

MN was R+1 in 2016, so if nothing changed and Biden is up by like 9 now you'd expect him to be ahead by 8 in MN. My guess is Minnesota isn't trending much either way but who knows.

Oddly enough, 538 has Minnesota's elasticity coefficient slightly higher than 1. I always thought they were inelastic.

So, this seems reasonable.

[pbrower2a]

49th best state (and barest win) for Nixon in 1972, and 50th best state (and a bare loss) for Reagan in 1984.  But only 10% for Obama in 2008? It's hard to ignore such amazing coincidences.

In any event, Minnesota looks to be swinging wildly away from Trump. Trump made wild promises of improved infrastructure, and iron miners figured that that meant much more iron for rebars for new, rebuilt, and improved highways.  Instead it was simply to add tolls to highways cthat people were already using

Make promises and win votes. Disappoint people and lose those votes.

[Landslide Lyndon]

The Daily Kos poster was right.

https://twitter.com/Nate_Cohn/status/1275498594519470080

[JRP1994]

https://twitter.com/Nate_Cohn/status/1275498594519470080

[LimoLiberal]

The Daily Kos poster was right.

https://twitter.com/Nate_Cohn/status/1275498594519470080

My prediction:

National: Biden +7
Arizona: Biden +3
Florida: Trump +1
Michigan: Biden +5
Pennsylvania: Biden +4
Wisconsin: Biden +2
North Carolina: Trump +1

AZ-Sen: Kelly +7
NC-Sen: Cunningham +1

[Buzz]

Bless up give thanks

[GeorgiaModerate]

https://twitter.com/Nate_Cohn/status/1275499205738586114

[xavier110]

My guess is that these will be pretty good for Biden, as most polling has been. Otherwise I don't know why they need to preemptively address the un-skewers. lol

Predictions:
- National: Biden +8
- Michigan: Biden +7
- Pennsylvania: Biden +7
- Wisconsin: Biden +5
- Florida: Biden +3
- North Carolina: Trump +1
- Arizona: Biden +2

[soundchaser]

https://twitter.com/Nate_Cohn/status/1275499205738586114

Huh, looks like these'll be pretty favorable to Trump unless the NYT does some serious weighting. Unless I'm misunderstanding the chart!

[Ronnie]

https://twitter.com/Nate_Cohn/status/1275499205738586114

Hmm, if the Arizona sample voted for Trump by 12 points over Clinton, then I think that's a pretty good hint that Trump will be leading in the poll.  I guess we'll find out, though.

[Wiz in Wis]

Obviously they could be good for Trump, or Biden, or a mixed bag, but the weighting looks very reasonable outside AZ, and even then, that might just be a bad sample.

[Tender Branson]

The AZ sample seems kinda pro-Trump (Trump+12), but the other samples seem ok.

[Wiz in Wis]

Also, as a reminder, we get a Marquette poll tomorrow too!

https://twitter.com/MarquetteMedia/status/1273367468447318016

[Tender Branson]

The poll of WI was fielded from June 8-15.

PA and AZ were fielded from June 8-16.

MI was fielded from June 8-17.

NC and FL were fielded from June 8-18.

The national survey was fielded from June 17-22.

[President Johnson]

Go Uncle Joe!

Predictions:
- National: Biden +9
- Michigan: Biden +9
- Pennsylvania: Biden +7
- Wisconsin: Biden +5
- Florida: Biden +2
- North Carolina: Biden +1
- Arizona: Biden +3

[Buzz]

Predictions:

MI: Biden +8
PA: Biden +6
WI: Biden +6
FL: Biden +3
NC: Tied
AZ: Biden +2

[Alben Barkley]

I'm just glad we're finally going to have a new high quality PA poll. For some reason it hasn't been polled as much as the others lately.

[Gass3268]

That Arizona sample is laughable. If Biden is within 5 points he should be happy.