Maybe Wang lurks here because today’s PEC entry directly addresses this issue. For any who want to get into the weeds he explains the basis behind his choice of prior and S.D. of current polling to final results and compares/contrasts with 538’s in, I think, a fair manner. He explains the uncertainty in determining what the uncertainty is and that he for now is using the larger uncertainty assumption (a less polarized electorate with more who are swingable, as in from 2000 and before rather than 2004 and after). He gives using econometrically-based priors their props even as he prefers his model that eschews adding in “fundamentals” but does end up with this comment:
I know. I thought I’d hit the Nowcast first, since that’s about where it was yesterday. Even Polls-Plus has changed significantly (77.7%), and it’s the one that’s supposed to be slowest to react.
Well, it could go to 99 or higher. If you clearly establish one candidate has a lead of, say, ten points on Election Day, there is really no doubt they’ll win the election.
Georgia flipping in a Clinton easy victory makes sense. Large African-American population, and an electorate that’s increasingly urban, rather than rural.
Yeah, Georgia isn’t nearly as deep red as most of its neighbors, it’s a state that will probably be genuinely purple in a few more election cycles and could easily be snagged by a Democrat in a down year for the GOP.
Polls plus shows Clinton at near 75%. This race looks over to me. Unless the economy crashes soon (I’m doubtful), or something damning happens, there will be no party change on the executive branch of the government.
I get the feeling that, if there’s other hacked documents to leak, it’ll happen sooner rather than later. You can’t just let someone sit at +10 for weeks, let everyone assume that the other guy is a big loser, and then hope that a news story drops them into second place at the last minute.
If accurate, a 1-in-4 chance of Trump winning is still not something I’m comfortable with. That’s still scary high for me. Those were the Brexit odds, at least according to the betting markets, at 8 pm on June 22, the night before the vote.
That’s one reason I’ve supported Clinton: They’ve been trying to smear her for decades and nothing’s stuck. They’ve had an eon in political terms to invent some scandal and they haven’t been able to.
She has good ideas and is a good candidate, but neither of those things matter if she can be Swift-Boated before she has a chance to do anything.
I just picked the first one I found. Do you not agree that a 25% chance is high for something like a Trump presidency? The betting markets today have Hillary at 76%, just like they had on Brexit the day before. I am not cherry picking here. The links to polls do not translate to %age chance of winning. I am fully aware (as I have posted about this before) that some polls had shifted by the end to show that Brexit had a true chance of passing. But what were the odds? The best I can find is what the betting markets thought.
And, remember, too, that was from the day before. There’s still around 100 days to go, and while I can’t expect much higher odds than 75% right now, there’s still plenty of room for Trump to come back. If we’re still at 75% the night before the election, I’ll be worried.
Right now it shows Clinton with a 50.2% chance of winning Georgia.