AZ, MI, NC - Redfield & Wilton Strategies: Kelly +17%, Peters +16%, Cunningham +11%

[TiltsAreUnderrated]

AZ
September 12-16
855 likely voters
MoE: 3.35%
Changes with August 30 - September 4 poll

Kelly 52% (-1)
McSally 35% (-3)
Another Third Party/Write-in 2% (+1)
Don't know 11% (+4)

MI
September 12-14
930 likely voters
MoE: 3.21%
Changes with August 30 - September 3 poll

Peters 51% (+1)
James 35% (-3)
Another Third Party/Write-in 2% (+1)
Don't know 11% (n/c)

NC
September 12-15
1092 likely voters
MoE: 2.97%
Changes with August 30 - September 3 poll

Cunningham 49% (+2)
Tillis 38% (+1)
Another Third Party/Write-in 4% (-1)
Don't know 9% (-5)

[MT Treasurer]

Hard to take a pollster that publishes numbers like that seriously.

[President Johnson]

I want to believe, but no. That's a bit too much my friends.

[Landslide Lyndon]

Their presidential numbers are plausible, so what explanation might be for the senate ones?

[Xing]

Subtract 8-10 points from each (maybe a bit more for MI) and it's believable.

[SenatorCouzens]

I'm sure this won't be a popular take, but if an apparently valid pollster with an actual staff and methodologists get numbers that are so obviously wrong, this is concerning for Biden that apparently it is actually very hard to do a decent poll these days and perhaps there is a big bias if not done very very very carefully.

[SenatorCouzens]

Subtract 8-10 points from each (maybe a bit more for MI) and it's believable.

When you have to subtract TEN off every poll, isn't their methodology so sloppy as to be completely dismissed? It's like saying poll the customers of the Whole Foods in Ann Arbor, subtract 10, and that's the real Biden number. Doesn't there come a point when the methodology, or the people that are actually being reached, are so off that it can't be taken seriously? Honest question from me.

[Alben Barkley]

This poll is certainly a counter to the Peters +4 from earlier.

[WD]

Hard to take a pollster that publishes numbers like that seriously.

Yeah, what the hell? Senator Peters only up 16? He’s at the very minimum cracking 60%. This is probably an R internal tbh imo tbh