LogFAQs > #969250111

LightningStrikes posted...

I was thinking this before seeing that:

Why are polls in the US so bad? Like, polls miss occasionally in other countries (famous example being the UK in 2015 though even that was only 3%) but 95% of the time theyre accurate. In the US it seems like everybody expects an error its just a matter of which way it goes.

The great irony is that the example everybody cites is 2016, where polls were actually bang on at a national level and only off in Wisconsin and Pennsylvania at a state level which didnt actually matter as Trump still would have won either way. People were just unreasonably confident relative to the data.

That is btw the Hill a leading liberal websites take on one singular poll.

Now that said, all through this midterm I have said it is odd how little polling we have. A lot of those polls we see are ones we would have derided as Republican biased polls in the past. As Silver does point out, these Republican biased polls in the Trafalgar and the Rasmussen and InsiderAdvantage did end up being the most accurate in 2020. So, there is that. But, where is the other polls? It's hard to argue the polls are being flooded by Republican polls when very little polls from anyone else are coming out.

Yougov and surveyusa polls are unreliable because I believe they are just online surveys.

The Sienna appears to be a slanted outlier (clearly about 4 points to the favor of the democrat in almost every poll across the board). We are barely getting consistent polling though for the Senate races that you would think you would get though. Is it flooding or lack of polling?

We know that Georgia will go to a runoff. So the polling there doesn't really matter. But, Nevada and PA do matter.

My predictions are.

Republicans take Nevada and PA senate seats. Georgia goes to runoff where it doesn't matter.

Everyone else holds. NH, Arizona, etc stay Dem. Wisconsin, Ohio etc Repub.