Probabilities are predictions. If Nate’s model says “x has an 80%” chance of winning, then in he is predicting that if 5 elections were held, x would win 4 of them. The fact that only one election is going to be held makes ascertaining the accuracy of his predictions more complicated, but it doesn’t make them not predictions.
I’m liking the electoral distribution curve from the Princeton Election Consortium.. Also check out their electoral maps- even if Putin’s Bitch outperforms polling data by 3%, he still loses to Biden 319-169.
IMO, it’s only a racket to people who don’t understand what polling analysis is. If you do, then it’s obvious that 538 is pursuing the only possible respectable path. Anything else is pundit blather. Or worse, internet blather.
Biden is at least even if not leading big in every poll for every battleground state, and he’s even or slightly ahead in several larger traditionally red states. The only places Trump leads is in smaller shithole states (to use Trumps adjective).
Not getting confident, but sure is nice to see. Hope the trend continues over the next 3 months. The news in the local paper consistently described Trump’s campaign w/ words like “struggling”, “disarray”, etc. Hopefully, that is not only accurate, but will persuade undecideds to go w/ a likely winner - Biden (as I understand many undecideds tend to do.)
Also good as early voting nears. Folk will vote NOW, and can’t change their vote no matter what happens later.
One thing that concerns me is that 538’s national conglomerate poll shows Trump gaining on Biden recently, albeit just a bit. Given how bad Trump has been recently I can’t fathom why that would happen. If Trump is somehow able to (slightly) reign himself in will we have a repeat of 2016?
I think part of this is that it’s August. People, even during a pandemic, are doing August-type things, and that doesn’t necessarily mean paying attention to politics or current events. In a few weeks, when school is set to start, look out. Because when the first kids or teachers are diagnosed with COVID after school starts…Then when the first district-wide outbreak happens…Then when the first several districts shut down in-person learning, and parents have no way to teach/watch their kids while also working…Then when the first teacher dies…Then when the first students die…Let’s just say people will remember how Trump and DeVos were pushing hard for in-person schooling, while offering no federal support to make it happen effectively and safely.
The Trump campaign is doing a lot of door-to-door outreach, whereas the Biden campaign is doing everything online or via phone. It’s possible that is having/will have at least some effect on the polls.
Didn’t anyone in the campaign think through the implications of door-to-door outreach during a pandemic? Your supporters are likely to talk to you longer, thus increasing the risk of exposure to the virus, so you might actually be killing off the people most likely to vote your way.
The irony of it is that back in 2016, lots of people (including me) noted the inferiority of Trump’s ground game with respect to Clinton, and deemed it yet one more reason he couldn’t possibly win. Weird to see the situation reversed now.
Perhaps the biggest state-level change in Aug has been a dramatic tightening in NC from a Biden lead of 4-5 points to even. Arizona has also tightened a bit. Not a huge deal but these are two nice bonus states which would give Biden more routes to victory.
The good news for him is that his leads in Florida and PA remain solid.
NC and AZ tightening doesn’t surprise me, as my monthly prediction lists to this thread (it is that a different thread?) indicate.
What does surprise me is Minnesota leaning a little more Republican than Wisconsin, according to the latest 538 aggregations. (They’re both very much swing states). I live in Wisconsin, just two miles from the Minnesota border, and there’s a bit of hipper-than-thou good-natured (and partly deserved) smugness among some in Minnesota that they’ve voted reliably Dem in recent elections, while Wisconsin has been notoriously…unreliable.
So, a part of me is happy to see Wisconsin as slightly bluer than Minnesota (this week, anyway).
But that joy is quickly squelched by a sudden fear that Minnesota could actually go red in November (!!!)…and of that happens, we are all in deep doo-doo.
(Since Minnesota is ground zero for George Floyd’s death and its aftermath, might this reflect some traction of the “support the police — law and order” theme we’ve heard from Republican supporters and candidates? I have no idea.)
Nate Silver’s model is up! And he gives Trump a 28% chance of winning which is terrifyingly close to his model’s final percentage in 2016.
This piece provides some more analysis and it makes a pretty persuasive case that there is still a way to go and Covid adds a lot of uncertainty.
My own subjective sense of the election is still closer to the Economist model which now predicts 11% for Trump. Trump’s numbers have remained remarkably stable through his Presidency and also his head-to-head against Biden in marked contrast to 2016. I am skeptical that there are enough persuadable voters left for him to win. The Harris choice also cuts some of Biden’s risk and provides him additional insurance if there is more rioting before the election.