Final chart:

https://x.com/NateSilver538/status/1837122995044651042

Looks like a big L for Nate. The wild swing because of the 'convention bump' in his model eventually reverted to the mean. Meanwhile prediction markets have been much more stable, with only a relatively small bump from the debate.

@cthor the swing persisted after the convention bump adjustment expired because the polls actually started looking worse for Kamala until the debate.

Current prediction is Harris 51.1%, Trump 48.6%. It is doubtful that there will be another update today, so this will likely be the final results. Obviously we should still wait until tomorrow to resolve though, since it’s not impossible that new polling will come out that Silver will want to run the model on.

This loss is going to hurt

What's the update for today 9/20?

A little bit more than a day left till closing

This will be extremely close.

Will this resolve based on probabilities at market close, or as soon as she re-takes the lead? Asking because she might overtake Trump then lose her lead again before the market closes

@dlin007

Quinnipiac today: Harris +6 PA, +5 MI, +1 WI. Wow!

Current numbers:

1500 limit order at 43% for no

Harris up +0.7% in Silver's polling average, looks like she needs at least another 0.7% or 0.8% to get to 50/50. Four days left. It might look alot different had it not been for the AtlasIntel poll.

@JakeLowery They would actually be tied at 49.8% in PA today had it not been for the convention bounce, so if that totally dissipates she should shoot back to 50%

Overall, Polymarket’s estimate of the race being pretty damned close to 50/50 seems reasonable. If I had to guess what the model will say in a week or two once we get enough post-debate polling, that’s the range I’d pick.

Nate Silver, from https://www.natesilver.net/p/a-guide-to-post-debate-poll-watching

Also from that blog post, Silver says that the model will fully take into account post-debate polling about a week after the debate, or around September 18.

Is the chart you posted under a paywall? If so, can someone post the updated chart currently post-debate?

@JaredAsh Harris is at 38.1% after today's update. Note that there hasn't been any post-debate polling yet.

When specifically will we snapshot the Nate Silver model to resolve based on that snapshot? Are we talking the exact moment the market closes is the moment that someone needs to snapshot the model and the market resolves based on that?

@Tripping yeah it's just whatever the chart shows at market close. The chart is only updated about once a day

2000 limit order at 50% for no, come and take it if you're confident that Kamala will overtake Trump

Isn’t this missing an option for “she’ll be worse”?

@jbca Worse than what?

Worse than 38.3%? Then bet No on "Harris will be at greater than 38.3% (her current prob)"

Worse than Trump? Then bet No on "Harris will be ahead of Trump"

@agabara I’m an idiot I misread what kind of market this was. Makes sense!