Like, I hate saying it too. It is terrifying to be going through this cycle thinking one thing and then having this come. It should lead to introspection, not denialism, and there’s little to be gained from going out on such a limb. This is the person who says the Obama breakthrough, the one who always says she listens to what the data says, all that. This is what a good pollster is supposed to do, too – release the final poll, even if you think it’s barking mad. And yet…
It’s one thing to get a weird result, but entirely another to do so in a way that makes no sense. Apparently Democrats are losing the battle for the 1st District by 15%, which is not happening, and Trump is winning young voters, which is also not happening. Nor, for the record, is the GOP winning the 2nd District.
Amongst the things that are wrong about this poll is the fact that Selzer doesn’t weight by education, which I either knew and discarded because I trust Selzer or never learned, but either way, it’s bad. I suspect that this is a fairly leading cause of the error – an overly rural sample causes an error. Selzer missed in 2018 by overstating Democrats in the Governor’s race, and it is frequently the case that pollsters over adjust after a miss. The other main thing that causes alarm is that Selzer has basically given up on actual polling by letting 4% of the sample of her poll say they voted without saying who for. What fucking value is there to those people?