It’s YouGov. The partisan split of the polling industry in the US is an unusual feature for me as a Briton. It comes up as a note of caution in political betting communities as it’s not something we really have here - all the major UK pollsters (including YouGov, who I assume subject their US operations to the same standards as they do at home) have been signed up to the British Polling Council for decades and have to adhere to various standards of transparency around their questions and methodologies. (Unusually, YouGov are the one UK polling outfit that sometimes get claims of partisanship thrown at them, but that’s because their founder later became a senior Conservative politician rather than because of any genuine evidence of partisan bias in their numbers!)

I was amazed by this for example:

But this thread is a reminder that without the equivalent of the British Polling Council some American pollster have a partisan skew which means when analysing the polls and betting on them that should be taken into consideration.

It is possible for an American pollster to ask this question

‘Are you planning on voting for the man God wishes was his son Donald Trump or the whore of Babylon Kamala Harris?’

and all we’d ever see from the pollster is ‘Trump 50%, Harris 50%’ as they don’t have to publish the question or data tables

Yes yes it is. And Nevada is weird on some polls in the other direction.

The polling error this year is a wild card. 2016 and 2020 both had errors in Trump’s favor(doubly impressive in 2020 since he was both a known element by that point and they had corrected from 2016 for the polls). 2016 was a 6 or 7 point error overall(Albeit closer to 5-6 if you only look at the last two weeks, there were small hints of correction) and 2020 was 3 or 4 points overall. But, 2022 was an error against the Republicans(1.5-2.5 points). BUT 2018 was dead on the money. So does the Roe V Wade factor (which is weaker now than it was in 2022 when it was dead fresh breaking news) from 2022 trump the ‘Trump on the Ballot’ factor that swung 2020 and 2016 away from the polls(which is also probably weaker than in 2020)? They’ve accounted for both, several polls tilted blue adjustments after 2022, or did red adjustments after 2020. The sample size says 'Republicans Always Over-preform if Trump is on the ballot" and “Democrats always over-preform post-Dobbs”, but the sample size is 2 and 1 respectively and we’ve never seen them go head to head, so…dunno. Also who got hurt by COVID more in 2020, Trump who was giving bad advice and getting torched by the media and losing his voters proportionally to death, or Biden who couldn’t campaign properly due to being one sniffle away from croaking which bogged down his outreach. Who got hurt more there?

Some states had different trends on the micro level though. Nevada was supposed to be relatively gettable in 2016(It was part of the ‘Only path to victory’ for Trump the media went with, he lost Nevada and New Hampshire which was supposed to be the end if not for him breaking the blue wall int he rust belt) and in 2020 it was supposed to be tighter than it was. So Nevada is over-reported Red? Well, yes, but it’s also the only swing state that went less blue in 2020 and it was red gains in the midterms against the pattern and current democrat polling has it Reder than NC or Georgia somehow. So which of those factors is at play here? Or Wisconsin. Jesus christ these guys massively over-preform for the Republicans. Michigan was polled pretty well in 2020 and Pennsylvania was about the national error margin, but Wisconsin was dead tight despite looking safer in the polls than other states. They also went very hard for Trump in 2016 while Michigan barely did. But Georgia was the opposite, one of only two states(Along with Nevada) that Trump under-preformed in compared to polls in 2020. And unlike Nevada(which was predicted to go blue, just way closer), Georgia flipped Blue against the polls. Could it be due to third party effect? Maybe(Trump 100% wins Georgia in a world without third parties and probably Arizona that year too, and in 2016 he wins another 5 states and the popular vote without them). Or Arizona. Went Trump in 2016 barely, went Biden in 2020. Is it trending Blue? Maybe? They got a blue Governor in 2022 against predictions, but it was also a state of bad House loses as the Republicans snagged several counties. Dominant bloc here is Hispanics who are blue, but getting less and less blue every year as Trump eases up the Mexico rhetoric. Also they don’t have RFK on the Ballot and Mark Kelly got publically turned down? Are those two things good or bad for the democrats? Zero clue, both could go either way.