WTF happened to improve Trump’s odds of re-election?

Single biggest? Probably not.

However, the election was close enough that it probably was enough that without it, she would have won.

But, that can be said for a dozen or so events and strategies that did not work to her favor, each just being enough to move a few people here and there away from supporting her.

That is what Nate silver said.

Next was poor strategy, then Russian bot attacks.

Yep. Silver compared the polling from before the Comey letter to afterwards and saw a dramatic swing, while noting a higher number of late deciding voters. Exit polling showed these late deciders went far more heavily for Trump.

You persist in conflating all protesters with rioters and you persist in ascribing all rioting, looting, violence and damage to “leftists” despite being repeatedly given evidence that both are not accurate descriptions of what is happening. I suppose I should be grateful that you’ve at least accepted the existence of “some right wing nuts” making “sporadic appearances”, one of which just made an “appearance” in the form of shooting a couple protesters.

The actual anarchist fringe aside, the “far left” don’t want “no law and order”. They want “better law and order” implemented in a different, less draconian fashion.

I mean, what in the world do you think happens when the individuals and institutions responsible for enforcing law and order are not held accountable to the same standard they enforce? What do you think happens when they are able to torture and murder people in the streets and in their homes with little to no consequence? What do you think happens when the victims and their families are denied justice time and time again by the same system persecuting them? And what do you think happens when all peaceful attempts to protest this simply result in half the country - and the government - engaging in coordinated vicious smear campaigns against the victims and concerted defense of the perpetrators?

You get this. Because all other courses of action have not worked. And they have not worked because the right have actively ensured that they didn’t. So, as I said - what did you think would happen?

I know we have a ways to go, and a week doesn’t necessarily equate with a long-term trend… BUT, today’s 538 update (model based mainly, but not entirely, on aggregate of recent polls, weighted toward high-quality ones) shows a real uptick in Trump’s chances. He’s gained about a point, maybe two, in four or five battleground states, and that’s enough to make him the winner in about 30% of the model’s iterations (versus closer to 25% a week ago).

Darn, just when my fingernails had grown back in…

Don’t worry about day-to-day stuff. The polls will jiggle about. We’ll know if there’s a real shift over the next several weeks, but even then, what really matters is the last few weeks before the election. I’m worried (and I will be until the election is over), but it’s too early to be concerned about daily poll changes.

Thumbs up.

It definitely hurt, but Clinton also failed to reach out to Black voters in key states.

Trump has never been within 4 points on the RealClearPolitics poll of polls, going back to last summer. I doubt Trump can get within striking distance nationally, but again, he doesn’t have to win the popular vote; it’s about where he wins and loses. Biden needs to win the national vote by at least 5% before we can assume a widespread impact on individual key state races.

I do wonder what the limit on this is. This article from 538 got me playing around with electoral maps and wondering how far from the national result we could make the tipping point state: Is The Electoral Map Changing? | FiveThirtyEight

5% may actually be too low. The most common tipping point states are (in order): FL, PA, WI, MN, MI, AZ. Based on the swing state numbers, all of those except PA and MI could end up more than 5% away from the national vote total (the error bars on PA put it between +5 and -1 for the GOP, and MI is +2 to -4 or so). All of the other ones have “pro-GOP” maximums above 5%.

If MN and WI move towards the GOP due to the social unrest, a situation in which Biden wins by 6 or 7% nationally but still squeaks by on the Electoral College is not impossible.

And after that, I have no idea what to predict. There has to be a limit at which the Constitutional arrangement is seen to be fundamentally undemocratic.

Rick Wilson, former Republican strategist, on his ‘The New Abnormal’ podcast noted that the Republicans have tipped their hand as to whom they are targeting this fall:

  1. White suburban women
    2-4. WI, FL, and ME-02

… based on the speakers and the night 2 & 3 messaging.

And to make it perfectly clear, to those that don’t have the math at their fingertips, holding those three from the 2016 map makes the EC 270-268 in favor of Trump. If WI is like 8% more Republican than the national average (not impossible, IMO) and Florida continues it’s rightward lurch (less likely, again IMO, but you also can’t discount shenanigans in FL), we could see a very large electoral vote/popular vote disconnect.

I believe that would lead to extremely large direct popular action of all types, and could lead to very unpleasant results.

I was cool as a cucumber a couple months ago. I’m scared that with:

  1. Cheating
  2. Low-turnout from people who want change
  3. High-turnout from the cult-of-Trump

Trump is going to win. Oh, he’ll lose popular vote by 3 million or more like Clinton did, but I am afraid he’ll pull off electoral wins in Florida and a couple other key states(Wisconsin?).

The Cohen book is coming out soon. I’m hoping that noticeably dings Trump’s chances.

Here are plots of those states’ PredictIt markets for the last six weeks.* I would agree that there is a worrying trend going on, but there is time for things to change.

Google Photos

Google Photos

Google Photos

Google Photos

Google Photos

Google Photos

The electoral college market doesn’t quite have the same shape, but still, things are not as good as they were six weeks ago.

  • Debiased for favorite-longshot bias using p’ = ϕ(1.64 ϕ^(-1)(p)) where ϕ is the CDF of the standard normal.

The worrying trend suggests that this race will inevitably tighten. The real danger is not that Trump will eventually overtake Biden in terms of overall popularity, and I am increasingly optimistic that Biden will carry the state races in terms of votes counted.

The real danger is, that might not be enough to guarantee a democratic result.

Of course, the RNC is going on right now. In 2016, it was in mid-July.

Maybe. But I think those numbers were true last week, after the DNC convention (from which I don’t believe Biden got any bounce) and the Republican convention.

The numbers that will tell the tale will come out next week, after both conventions are in the rear-view mirror. The next important ones will be the week after all the debates are finished.

I suspect the events that will determine the race haven’t happened yet. We have very short attention spans now. If there are any crises in September or October (and this is 2020, so there probably will be), they may swing the election.

Then there are the debates. In the past, debates haven’t moved the needle much, but this year is very different. Both sides have been having trouble getting their messages out, so for many people the debates will be the first time hearing what exactly each side has planned. And a good performance by Biden could eliminate the ‘senile’ criticism. On the other hand, if Joe goes off script and starts rambling about shooting people in the legs or brandishing shotguns or makes some obvious blunder, it will hurt him harder than most, because it will feed the narrative.

Nancy Pelosi thinks Biden should avoid the debates. That seems like a terrible strategy, unless she truly thinks that Joe will embarass himself.

Also important is whether there are still riots going on, and whether they are getting better or worse. If the riots are still going in October and nothing else comes along to upset things, I’d put my money on Trump winning. The unrest going on now is polling very badly. And if Biden comes out strongly for stopping it he risks alienating his progressive base. So he’s got a bit of a highwire act to perform unless the violence stops soon.

The Biden team has already addressed this. Biden wants to debate.

“We certainly agree with Speaker Pelosi on her views of the President’s behavior. But just as she has powerfully confronted that behavior in the Oval Office and the Cabinet Room, Joe Biden looks forward to doing the same on the debate stage.” - Andrew Bates, director of rapid response for Joe Biden

Source: Johnny Verhovek (ABC news) Tweet

That’s a smart response. Pelosi should know better - just floating the idea that Joe shouldn’t debate plays into the narrative that he can’t debate Trump and win because of his supposed cognitive defects. She didn’t help him at all.