teh speleeng polise

[teh speleeng polise @tehspeleeng]

@ChristianHeiens 11-3 once provisionals are counted in NC-3 most likely

[teh speleeng polise @tehspeleeng]

@baseballot Cook Political kept CA-22 at tossup up until the very end despite Dems winning the primary by 18 points. They kept all the NC house seats at safe weeks after the EV suggested they’d be competitive. They did not follow the data, and now look stupid as a result.

[teh speleeng polise @tehspeleeng]

@benwfreeman1 I wrote a memo for a client where I gave dems 25% odds on the saturday before the election

[teh speleeng polise @tehspeleeng]

@PokiJaciii it’s based off of 100k or so, idk how the 30k extra changed it but probably not enough. If dems were ahead I think they might’ve had a shot but when they’re trailing they need late ballots to be significantly bluer and I’m not seeing that

[teh speleeng polise @tehspeleeng]

@SemioticRivalry I suck at interpreting early vote and have gotten misled by it many times, so now I’m going to throw a temper tantrum and act like it is impossible for anyone to interpret it accurately, as the sole arbiter of whether it is legally allowable to interpret early vote

[teh speleeng polise @tehspeleeng]

@PoliticalKiwi 2019 Jesse would not have been surprised

[Jesse🔸⏹️ @PoliticalKiwi]

@AsInMarx @carterforva a lot of the major pundits too. half of the people who work at Cook Political Report shouldn't have a job in election analysis.

[teh speleeng polise @tehspeleeng]

@Nate_Cohn @rudnicknoah Are you able to publish your estimates for how Harris/Trump performed with voters who voted in 2022/24? Would be useful to gauge how much was differential turnout vs persuasion, because I think the biggest cause of the polling miss was modeling the wrong LV electorate

[teh speleeng polise @tehspeleeng]

@GaetenD you have just won the award for most pointless poll ever as there is not a single person on the planet who would be surprised by the results

[teh speleeng polise @tehspeleeng]

@lxeagle17 Unless you are the head of Quantus Insights or something

[teh speleeng polise @tehspeleeng]

@aggou Halloween may have depressed youth turnout is my guess, iirc 2024 Halloween also saw a similar effect

[teh speleeng polise @tehspeleeng]

@GaetenD Republicans have a very good shot in New Jersey after Governor Ciatarelli’s shock upset, which only right wing cope accounts saw coming

[teh speleeng polise @tehspeleeng]

@DjsokeSpeaking where are you able to see the 2024 turnout hub? I can only access 2026

[teh speleeng polise @tehspeleeng]

@ChazNuttycombe This is probably not even fully accounting for the fact that North Carolina is on pace to have a Harris+10 or above electorate lol

[teh speleeng polise @tehspeleeng]

@aggou Something I find particularly instructive is looking at the two way reg of 2026 primary voters who have showed up so far, and two way reg of those who did vote vote in the primary

[teh speleeng polise @tehspeleeng]

@bernardbulletin I am unable to analyze the early vote, therefore it is impossible for anyone to do so

[teh speleeng polise @tehspeleeng]

@PokiJaciii IPEV is a lot less Republican than years past due to fears among dems about USPS, and we are seeing this play out across the nation

[teh speleeng polise @tehspeleeng]

@GaetenD @Drexler771 @TonerousHyus Also I might add that Crockett winning was the good outcome for the GOP

[teh speleeng polise @tehspeleeng]

@AlbenbarkleyFan @gelliottmorris @50plus1news One red flag is that with so many responses, surely at least someone notices them doing a poll and posts about it on social media. For instance, you often see leaks about potential NYT polls bc they do real polls. I have yet to see someone claim they were contacted by trafalar

[teh speleeng polise @tehspeleeng]

@AlbenbarkleyFan @gelliottmorris @50plus1news I think their other polls were fake too, they just did so many, each with roughly the same response rate from their 75000 sample size, with no actual methodology or crosstabs posted. Just because the topline seems plausible doesn’t mean the poll is real