i would agree that the n=79 was small, made smaller by the fact that many of those 79 are inherently correlated and made even smaller by the fact that many of these were, for lack of a better word, obvious. think that our main focus was showing there was no explicit/statistically significant pro-red bias (though perhaps I could’ve added a hypothesis test or two). thanks for the comment and i’m looking forward to doing a better version after 2026!
I...am not sold on this analysis. Mostly because *everyone* was overrating Dems in '24. Trump overperformed polls. Also "which presidential candidate won in KY?" is useless in predicting accuracy. It's *Kentucky*, we know who will win. Though I suppose using only swing states makes for a poor sample size.
Effort appreciated, but this is a bit half-baked. On a kinder note, it is a good jumping-off point for additional analysis.
3 such cases in 2022 at the 7 day time horizon, specifically Warnock, Fetterman, and Masto. Unsure about the 1 day time horizon (we were only able to source 7 day data through an index of the event page on the wayback machine)
i would agree that the n=79 was small, made smaller by the fact that many of those 79 are inherently correlated and made even smaller by the fact that many of these were, for lack of a better word, obvious. think that our main focus was showing there was no explicit/statistically significant pro-red bias (though perhaps I could’ve added a hypothesis test or two). thanks for the comment and i’m looking forward to doing a better version after 2026!
I...am not sold on this analysis. Mostly because *everyone* was overrating Dems in '24. Trump overperformed polls. Also "which presidential candidate won in KY?" is useless in predicting accuracy. It's *Kentucky*, we know who will win. Though I suppose using only swing states makes for a poor sample size.
Effort appreciated, but this is a bit half-baked. On a kinder note, it is a good jumping-off point for additional analysis.
there was an example of predicted GOP victory for Senate in midterms 2022, won by Dem?
3 such cases in 2022 at the 7 day time horizon, specifically Warnock, Fetterman, and Masto. Unsure about the 1 day time horizon (we were only able to source 7 day data through an index of the event page on the wayback machine)
I mean entire Senate control victory? I think GOP favs, DEM victory?