trump’s rallies are super spreader events. the areas he has them tend to nearly double in covid cases after one. he has been having quite a few of them in pa. I wonder if that will have an effect on voting in those areas as most trump supporters are planning on nov 3 rd voting.
Ultimately, I think Trump will lose because of the WOMEN (Wisconsin, Ohio, Michigan, ‘Ennsylvania and North Carolina).
I’d say it’s tightening. Last week 538 had Biden 51-49 in Ohio. It’s now Trump 57-43.
You mean Biden will POWN’M? Trump that is.
Could you point to where you are seeing that?
EDIT: Ah, you are seeing the favorability forecast, rather than the vote results prediction.
Right, let’s let them go to the hospital and run up thousands of dollars in bills, be sicker than they have ever been, and possibly have organ damage. These people are unbelievable, and Biden should be talking about this and not just the number of deaths
Right. Please, people, let’s be clear what we mean: “probability of winning the state” (or, equivalently, “percent of recent model runs where that state was won”)
VERSUS…
“Among recent polls, average percent of likely voters indicating their intention to vote for that candidate.”
Only in the truly toss up states (Ohio and three or four others) is it possible to confuse these two very different figures now…but let’s try to be clear.
The fact that Ohio (along with Florida, Georgia, Texas, Iowa, NC) are still considered toss-ups with 9 days until Election Day, tells me those states aren’t tightening, even if polls and forecasts slightly rise and dip from day-to-day or even week-to-week. Ohio wasn’t even supposed to be a swing state this year. The fact that it’s this close so close to the election is a positive for Biden.
Here’s a map I made on 538’s interactive state-by-state forecast tool. I’ve plugged in a very Trump-friendly scenario: If Biden is up by 7.5+% in a state, I’ve given it to Biden. If Trump is up by a mere 1+% in a state, I’ve given it to Trump. That leaves 9 states, plus Maine’s and Nebraska’s 2nd districts, falling somewhere in between. So with all that entered, Nate’s forecast tool gives Biden a 92% chance of winning the election, with an average EC count of 319. Pretty close to what the Economist’s model shows.
I’m not seeing any worrisome trends appearing.
That’s very encouraging, thanks!
And see:
Just to be clear, I’m not particularly worried. As we all know, there are several reasons this isn’t 2016: fewer undecideds; Trump as a more known quantity that many dislike; on balance, improved access to voting (despite myriad Republican-imposed stumbling blocks); less third-party interest; eliminating the misogyny factor; polling weighted for education; less complacency among Dem supporters…
But the fact is, we are undergoing a slight tightening, as many expected. All other things being equal, Biden’s slow but steady rise in the 538 model would have continued (because the closer we get to Election Day, the less inherent uncertainty), and he’d be at about 90% now, maybe 91%. Instead, he plateaued a week and a half ago, at 87%. So the plateau really does reflect how the polling shows a slight, but real, tightening in a bunch of key states.
(At about 94% or 95%, he would plateau no matter what, because of the unavoidable uncertainty built into the model.)
I had not thought of that. Overall is it easier to vote early and/or by mail than in 2016?
Trump and his people were hoping to turn the race around using Biden’s not-really-a-gaffe about phasing out the oil industry. So much for that idea.
Yes. I’ve heard it said in a couple of 538 podcasts that, all things considered, there are more ways for more people to vote this year. They said about half of these are temporary measures due to COVID (e.g., allowing fear of COVID as a valid reason for an absentee ballot), and half are permanent changes (well, except where Republicans will get them rescinded for no good reason).
The latter category would include things like states where some ex-felons had their voting rights restored, plus more states with Election Day registration, some with ballot drop-boxes a week or two before Election Day that didn’t have them before — that sort of thing.
The GOP and Trumpist establishment seems to have been REALLY banking on the Hunter Biden laptop thing.
They were. It was the lead story on their evening shows for days (Hannity spent almost an hour raving about how Trump won the debate because of it) and on the website. Now the story is nowhere to be found and they are concentrating on oil and fracking
I don’t think that’s a “fact” at all. I’ve shown that the swing states polling leads are essentially the same (even slightly above) where we were 7-8 weeks ago. Biden’s lead in the national average is as high as it’s been this entire campaign. And Nate Silver’s forecast has inched up for Biden since the beginning of September. Forget the day-to-day bops up and down. The polling lines showing Biden’s lead is rock-steady, and the trend line on every forecast projection I’ve seen is moving up for Biden and down for Trump.
I’ve checked the trend lines and overall numbers on 538 and the Economist, and I just don’t see this tightening of the race that you see. Biden’s numbers are better and Trump’s are worse at the end of October than they were at the beginning. Same thing goes for the numbers from the beginning of September til now. There was a two-week bubble in the middle of October where polling was reflective of Trump’s shitshow of a week early in the month, but that bubble seems to have come to an end-- and Biden’s numbers are higher now than they were before that bubble began.
Happy, agree to disagree, then. It’s not that important, and mainly a matter of definitions — time scales, probabilistic predictive models vs. just polls, etc. We agree on the big picture, and the important stuff. FWIW, Nate Cohn agrees with you more (see todays’s NY Times piece), and Nate Silver with me more (see a couple of comments in the mist recent podcasts). The important, thank-goodness thing is that, whatever tightening there is, is VERY slight — well within the margin of error (I.e., could also be explained as mere statistical noise, but also might not be).
The one swing state where it’s unarguably real st this time is Ohio — but you prefer to emphasize that: 1. as of a month ago, no one was expecting Ohio to be close to 50-50 anyway, so at bigger time scales, seeing it still close to that counts as a wonderful surprise for Biden — I agree!; 2. how truly anomalous were the lows for Trump following the first debate — I agree with this, too, but 538’s model was set up to dampen post-debate movement a little (in either direction), so again, Joe’s plateauing over the past 12 days represents a tiny, but real, change in trajectory. Not a fatal one, and probably not even a worrisome one.
ETA: Checking, I see you actually have to go back almost two months (not one month) to the last time Ohio seemed out of reach for Biden. It started to become a real possibility for Biden aroundj the first week of September.
Again, I agree that the fact we’re even taking about Ohio this late in the game is GREAT! It’s very unlikely to br the crucial state. If Biden wins Ohio, he’s almost certainly won the presidency for OTHER reasons already.
But, along with a few good-but-not-great recent polls for Biden in states like Wisconsin (again, he’s now up around 5 there, rather than the steady 8 of one week to one month ago), or Florida (now up around 2 there, rather than the steady 5 of one week to one month ago), this does contribute to a slowing down (really, a halt) to Biden’s otherwise nearly constant march (over the past two-plus months, at least) toward ever-better chances of winning.
On the Economist model, Trump’s win percentage has dipped sharply to an all-time low of 5%. Not quite sure which polls have caused that dip.
Anyway we will have to wait for the high-quality pollsters to release their final state polls to get a proper final picture of the race. And since they probably prefer to release their final polls nearer the election, there may be a bit of a fog till later in the week.
Looking more closely at the bigger picture (the 538 state-by-state trend lines for actual polls, not any model), I see that Florida’s tightening is real, but it happened back in mid-to-late August. Basically, Biden had enjoyed a 4-to-6-pint lead there through most of the summer, but since then it’s been more like a 1-to-3 point lead.
As for Wisconsin, I concede to Happy that the slight tightening I perceived there in the past couple of weeks may be illusory. The smoothed-out Biden lead there really has held remarkably steady for months now, I’m relieved to note.