Will more Republicans be elected to the House than 538 predicts?
29
4
แน€360
resolved Nov 17
Resolved
NO

This question resolves to YES if the number of Republicans elected to the house is more than the average number of seats 538 predicts they will hold, as given by [their model](https://projects.fivethirtyeight.com/2022-election-forecast/house/) on the day before the election.

Close date updated to 2022-11-30 1:59 am

Get แน€200 play money

๐Ÿ… Top traders

#NameTotal profit
1แน€380
2แน€174
3แน€27
4แน€24
5แน€19
Sort by:
predicted YES

Looking at the Forecast at 12:30 Central Time, looks like the Republicans are expected to hold 230 seats.

bought แน€10 of NO

This is a market where I feel unusually safe placing limit orders because it's hard to think of the type of news that would really impact my confidence.

I feel like I see a lot of comments saying they think 538 predictions are wrong, but not explaining why. This market is a test to see if people think in general that 538 has a bias.

predicted YES

@BoltonBailey in my case it's not 538, it's polling as a tool in general. By nature it's heavily influenced by backwards poll accuracy; polling response rates are ridiculously low, to the point where the lizardman constant swamps a lot of the signal. In political polling in particular you see one tribe which is much less likely to answer polls than the other.

Pollsters counter this with fudge factors in order to try to account for that, but the fudge factors are based on prior poll results as compared to the events they tried to predict; in this past 2 years I believe the underlying likelihood of answering a pollster has shifted enough that all the fudge factors are severely out of date, in the direction of under-predicting red tribe success.

Therefore I'm pretty sure that whatever 538 predicts will be skewed blue. Not because they're not trying to be accurate, but because of the nature of their knowledge and their models.