PredictIt is somewhat volatile, so the odds may have changed (I also rounded a bit). I tend to think that we have some problems with lagging forecasts (in comparison to prediction markets), even if our overall forecasts have about the same (or higher) quality.
Yea, one advantage of pred mkts is their graphs jump around a lot more. While ours is overly stable (conservative, because it includes old forecasts from a lot of people). Gotta to find a middle ground.
@GJDrew, the middle ground is to exponentially decay stale forecasts over the lifetime of the question, i.e. apply a form of discounting. I'll make a note of this and try it out when I have time for my personal super-picking process. [1]
PredictIt [1] gives even odds to Clinton+Sanders.
[1] https://www.predictit.org/Market/1460/Who-will-win-the-2016-New-Hampshire-Democratic-primary
I see 54-49, Bernie. On the plus side, I see they have 141 comments, while we have 1227.
PredictIt is somewhat volatile, so the odds may have changed (I also rounded a bit). I tend to think that we have some problems with lagging forecasts (in comparison to prediction markets), even if our overall forecasts have about the same (or higher) quality.
Yea, one advantage of pred mkts is their graphs jump around a lot more. While ours is overly stable (conservative, because it includes old forecasts from a lot of people). Gotta to find a middle ground.
@GJDrew, the middle ground is to exponentially decay stale forecasts over the lifetime of the question, i.e. apply a form of discounting. I'll make a note of this and try it out when I have time for my personal super-picking process. [1]
[1] https://larswericson.wordpress.com/2015/12/12/forecasting-tournament-probabilities-are-overly-stable-compared-to-prediction-market-probabilities-due-to-stale-predictions/