Not when she’s nailing midterms and primaries as well. I just didn’t mention them
This is based on asking them, not a statistical analysis of past voting history.
Gonna be hilarious if Harris wins a blowout but the Democrats lose the Senate.
From Selzer? No. From Rasmussen? Sure.
Nah we’d be dooming it UP.
I mean we get a set of only decent polls for NYT/Sienna that show Harris winning a close race, and posters are ready to go back to their normal despair.
First rule of data science, if you get a significantly outlier result treat it as such until you have more data that confirms it.
Apropos of nothing, but I mention itt a few days ago that I was polled by Gallup before the 1996 election. After thinking of it I went back and looked up my results from the Gallup microdata and found out that they miscoded at least a couple of questions - it indicated that I hadn’t voted in 1992 (I had) and that I voted for Perot in 1992 (I did not). So if they used voting history as a weighting strategy I don’t know what would have happened but they would have definitely gotten it wrong.
Is anyone else polling Iowa?
If not then it’s even less worth believing no?
These aren’t really “crosstabs” which they don’t publish, they’re highlights from the DM Register’s explainer article. Also I don’t think they’re entirely correct.
I have to say that reading the DM Register’s piece is… eyebrow-raising, to say the least:
Similarly, senior voters who are 65 and older favor Harris. But senior women support her by a more than 2-to-1 margin, 63% to 28%, while senior men favor her by just 2 percentage points, 47% to 45%.
In the tweet you linked, he has Selzer polling senior men at R+2, but the article says the result was D+2. Which is… insane. I really don’t think men over 65 have had an en masse Coconut Conversion experience. Combine that with this:
The poll shows 62% of Iowans younger than 35 are likely voters, down from 73% in September.
But 93% of seniors say they are likely voters — even higher than the 84% who said so in September.
So essentially, the poll is showing young people not voting and old people voting en masse, but this is… GOOD for Kamala, because don’t worry, old men are totally not voting Trump.
I bought a bunch of No to Trump winning by 10+ shares on Polymarket at 80c on the theory that although the Selzer poll is obviously wrong, it couldn’t be THAT wrong. I’m not so sure about that after reading this, I’m going to profit take at 83c.
If Seltzer is even in the ballpark, the most interesting part will be what she says at the end - there is no paid campaign for either party in Iowa. Which means people would be choosing based on that they see directly from the candidates, not glossy ads. facebook memes.
That’s what I assumed too. Do you think it makes the choice less interesting? Still seems relevant to me. I would assume most pollsters include “maybe” voters and weigh them less.
I’m not saying it’s a bad idea. It could absolutely be better, especially this late when a lot of people have already voted.
It appears the garbageman gambit may have not been as shrewd of a maneuver as originally thought
Important article here from NBC Decision Desk about “mirage” effects in the swing states caused by which votes are reported first.
Summary:
NC: Blue mirage
GA: Blue mirage
PA: Early blue mirage, then red mirage as day progresses
MI: Small red mirage
WI: Small red mirage
AZ: Blue mirage
NV: Possible red mirage
Sounds great!
I kind of feel like I should try to put down a huge (for me, I’m poor, don’t judge) bet on Trump just so I don’t kms on Wednesday
Is anyone here interested in a few thousand, I’ll escrow if it’s not completely insane rake wise to escrow with Venmo (I think any other service there would be a delay since I’ve not used them before, only venmo, but maybe I’m wrong)
Yeah, I read the explainer article earlier and I have no idea what to make of it. You’d expect some weird results in a sample of this size that hasn’t been stat-tortured. But a couple of these numbers seem totally crazy.
I don’t know if it’s a sample size issue or nonresponse bias or what.