AZ, FL - ABC/WaPo: Trump +1%, Trump +4%

[wbrocks67]

What in the world...

In Florida, Biden +1 among RVs, but Trump +4 among LV? There's a 5% shift in LV? And 3% LV shift in AZ?

EDIT: McSally winning suburban voters by 20%? Huh? Biden winning Indies by 20% in AZ but down in the overall LV race?

ABC/Wapo is well trusted, but these are just... odd

[wbrocks67]

Never thought I'd say this but it looks like Georgia may be more likely to flip than Florida.

AZ looks like an outlier, I find it very hard to believe Kelly is only up one. Though of course if those numbers are corroborated Biden is in deep trouble.

I mean, the FL RV # of Biden +1 looks fine, but the LV # is wildly ridiculous.

[wbrocks67]

Even Harry is skeptical

https://twitter.com/ForecasterEnten/status/1308718660882706432

[wbrocks67]

Well, as much as the 16 pt Minnesota lead was likely an outlier, so are these.

[wbrocks67]

Given the weird LV screens, I think a major problem this year for pollsters is turnout and figuring out who's coming out to vote. Turnout is going to be sky high and I'm not necessarily convinced that the electorate can be properly predicted.

Like the FL poll clearly predicts that Democrats are going to have low, low turnout, which seems... unlikely

[wbrocks67]

THANK YOU!! The greatest of honors!!

On a more serious note, it’s a outlier for now.  The AZ number especially.   FL doesn’t surprise me, I still don’t buy Trump is gonna do significantly worse with 65+ then 2016.

It also has Trump winning 18-49 year olds, a demographic that Clinton took by about 15% in 2016, so yeah... these crosstabs are very weird to say the least.

[wbrocks67]

https://twitter.com/peterschorschfl/status/1308737954513387522?s=21

I feel so god damn exhausted at the "Trump vs Biden supporter enthusiasm" question and why it's so flawed and yet pundits keep pushing it

[wbrocks67]

https://twitter.com/nate_cohn/status/1308730492217430018?s=21

So it's very clear a strict likely voter model is at play when it probably shouldn't be. Also, not surprising that Nate is defending it considering every single NYT/Siena poll has oddly had an R ID edge, for whatever reason, regardless of the end outcome.

[wbrocks67]

Trump approval (RV)
FL 47/51 (-4)
AZ 47/51 (-4)

Honestly, the RV matchup in FL looks normal. Trump -4 approval, Biden +1 edge among RV. However, the entire AZ sample and FL LV sample on the other hand...

[wbrocks67]

Trump approval (RV)
FL 47/51 (-4)
AZ 47/51 (-4)

Honestly, the RV matchup in FL looks normal. Trump -4 approval, Biden +1 edge among RV. However, the entire AZ sample and FL LV sample on the other hand...

Well, of course when it doesn't suit you. I mean how you don't get tired unskewing every single poll on this board which doesn't go with your way of thinking where the race is. It's just plain annoying. You unskew this, but you don't say a word for change polls you posted with less than 300 people in AZ sample from a known bad pollster, you're silent there. But hey this sample looks bad.

I'm literally going off of nearly every other poll + the polling averages we've gotten. Not opinion. Trump leading in both states IS an outlier, as well as many of the crosstabs. You can see yourself out.

[wbrocks67]

No, you didn't say a word for those Change polls. No unskewing, no to looking in samples because you like those results.

I literally haven't commented on Change at all. I just reported the results. Stop. I've already made it clear my feelings on Change.

[wbrocks67]

Chill. These are obvious outliers.

So fast do people forget that we just got another A+ poll barely a week ago with Biden +9 in AZ.

[wbrocks67]

Since we like to pick on 538, I'd like to point out that, even though the state probabilities for AZ and FL have (understandably) dipped, their topline probability is completely unchanged as a result of this poll - still 77-23 Biden.

I get the opposite impression, as most people on here seem to have a more favorable opinion of 538 than they do of RCP-even though both websites have very similar poll averages, and very reasonable predictions for the presidential election, as things stand now.

I think a lot of people think highly of their poll aggregation efforts and much less highly of the non-polling aspects of their model.

FWIW, the fact that Biden's chance of winning in the model has been so stable (he's been at 76 or 77 for the last 9 days straight) does suggest the race is tightening very slightly.  If his polling lead were to stay exactly where it is, his winning probability should increase as the election gets closer.

and it has though. He was down to like 67% a few weeks ago.