When Nate Silver's Next Model Comes Out, What Odds Will It Give the Democratic Nominee?
➕
Plus
469
Ṁ200k
resolved Jul 30
100%98.6%
36-45%
0.1%
25% or less
0.1%
26-35%
0.1%
46-55%
1.1%
56-65%
0.0%
66-75%
0.0%
76% or more

What do we think Nate's model will say about the Democratic Candidate (likely Harris) compared to Trump? If multiple versions of the model are put out, this will resolve based on the outcome of the candidate most likely to be the Democratic Candidate based on the Manifold at the time that the model is published. In the unlikely event that Silver's model is not published again this year, all these options will resolve to NO.

Get
Ṁ1,000
and
S3.00
Sort by:
bought Ṁ1,000 36-45% YES

What is the link to this?

https://www.natesilver.net/p/nate-silver-2024-president-election-polls-model

Though I should mention that the actual probabilities are behind a paywall. The polling average is public, though

Pretty insane that she may win the popular vote by 8% and not get the presidency…

You're reading that wrong @JaimeSantaCruz.

Whoops, I misunderstood the misunderstanding

Please, explain to this uncultured swine how it should be read 🙏🏼

The 53.5 to 46.5 bit in the middle column is the probability that she will win the popular vote, not what her projected vote share will be.

If you imagine there are 1000 worlds, the 53.5 means that 535 of them are worlds where Harris wins the popular vote, and 465 of them are worlds where Trump wins the popular vote. But a bunch of the worlds where Harris wins the popular vote, she does so by a very slim margin, say 47% to 46.5%. In other worlds, she wins by 5 or 6 points. But both of those worlds are covered in the 535 worlds. The second column in this screenshot is purely talking about the number of worlds in which one or the other candidate wins the popular vote, not how much they win by.

Additionally, the Electoral College does diminish Harris's chances, but not to the degree your original comment suggested. The Electoral College favors Trump this election by perhaps two points or so, meaning that Harris could win the popular vote by a point or two, but still lose the election, since the Electoral College is what actually determines the winner. There are about 140 worlds in which this happens, according to the forecast.

But none of those worlds are ones in which Harris wins the popular vote by 8 points and still loses the election.

(You probably know this already, but the reason Trump is favored by the Electoral College is because of how the two different parties have different demographics and different representation in different states, it happens to be the case that Trump is very unlikely to lose the election if he wins the popular vote, while Harris has to win by a couple points in order to be similarly likely to win.)

One other thing to note is that Nate Silver published an updated model today in which Harris's chances improved slightly, from 38% to about 42%. So basically 40 worlds of the thousand shifted from being Trump victories to being Harris victories.

Thanks a lot. Probabilities hard, me dumb.

"Oops" is the sound we make when we improve ourselves.

opened a Ṁ2,000 46-55% YES at 25% order

Adding a 2k limit Yes on 25%. I suspect Nate’s “Kamala mode” has some bullish adjustments that will give Harris an initial boost in odds.

registering my actual probabilities before model release

26-35: 20%

36-45: 75%

>45: 5%

Biden's absolute peak in the Silver model was 36%, and Harris' horserace polling isn't hugely different from Biden's pre-debate. The reasons to be bullish on Kamala relative to Biden (in my view) will likely not be captured in this model.

I thought it might be higher because Nate Silver said there would be more variance on Harris for different reasons (ie fewer polls).

i think he meant that more qualitatively

@SemioticRivalry The model would capture the fewer polls too though.

@traders Nate's Kamala v Trump model is allegedly coming out today, so make your final bets now!

bought Ṁ50 56-65% YES

I'll guess 43% for Harris from my own modeling (simpler than Silver's model). However I'd estimate we only have the equivalent of ~1-3 top-quality polls for Harris in swing states. A top pollster releasing swing state polls tomorrow could swing things some.

While the 36-45% bucket is likely, the 46-55% bucket is close and at a ~24%-price is probably underpriced a bit.

26-35% seems underpriced. When Nate had Biden-Trump tied in popular vote, he gave Biden 33% chance of victory. Had Kamala even tied Trump, or is she still 1 or 2 points behind? Does the fundamental part allow her to make up the deficit?

https://www.natesilver.net/p/silver-bulletin-2024-biden-trump

@Dgfold love your writeup btw, very detailed and intuitive

@PlainBG 🎉 thanks!

Yeah, agree 26-35% is probably underpriced at 6%. It has also been several weeks since that time when Biden-Trump were tied (chance gets worse for the person behind as gets closer. However, in my estimate the majority of the variance is from poll miss, not movement). In national polls Harris is about 1 margin-point behind. At least in Biden times, Silver didn't really seem to give much fundamentals advantage or disadvantage.

opened a Ṁ300 26-35% YES at 12% order

Anyone know what programming language Nate uses?

nice

The Harris-Trump model will be released tomorrow:

https://www.natesilver.net/p/kamala_mode-will-be-activated-tomorrow

Saw this earlier today. Seemed like the first real good news I’ve seen for Kamala.

© Manifold Markets, Inc.Terms + Mana-only TermsPrivacyRules