General Discussion
Related: Editorials & Other Articles, Issue Forums, Alliance Forums, Region ForumsWill pollsters adjust their likely voter models based on yesterday's election?
Clearly turnout cannot be expected to be the same levels as a normal midterm. Data for progress' last LV poll still seems like it might have a poor LV model. Democrats seem to be pretty solid on all of the RV polls taken, but LVs is where they struggle the most. Will do you think those models will change before the November election?
I think they were also substantially off in 2020 when they seemed to assume larger turnout for Democrats than actually as the case.
mopinko
(70,222 posts)if they dont, they should go out of business.
ColinC
(8,330 posts)So we might see some dramatically more favorable polling for Democrats in the near future. I'm almost worried about how that could suppress or help turnout in November.
PortTack
(32,794 posts)Cycles. Most, not everyone but most including 538, RCP are worthless.
Yeah. I mean probability/ data science, is a science and all, but their predictive qualities are super overrated.