Key excerpt (But it's worth reading the full thing):
But the real value-add of the model is not just in calculating who’s ahead in the polling average. Rather, it’s in understanding the uncertainties in the data: how accurate polls are in practice, and how these errors are correlated between the states. The final margins on Tuesday were actually quite close to the polling averages in the swing states, though less so in blue states, as I’ll discuss in a moment. But this was more or less a textbook illustration of the normal-sized polling error that we frequently wrote about [paid only; basically says that the polling errors could be correlated be correlated between states]. When polls miss low on Trump in one key state, they probably also will in most or all of the others.
In fact, because polling errors are highly correlated between states — and because Trump was ahead in 5 of the 7 swing states anyway — a Trump sweep of the swing states was actually our most common scenario, occurring in 20 percent of simulations. Following the same logic, the second most common outcome, happening 14 percent of the time, was a Harris swing state sweep.6
[Interactive table]
Relatedly, the final Electoral College tally will be 312 electoral votes for Trump and 226 for Harris. And Trump @ 312 was by far the most common outcome in our simulations, occurring 6 percent of the time. In fact, Trump 312/Harris 226 is the huge spike you see in our electoral vote distribution chart:
[Interactive graph]
The difference between 20 percent (the share of times Trump won all 7 swing states) and 6 percent (his getting exactly 312 electoral votes) is because sometimes, Trump winning all the swing states was part of a complete landslide where he penetrated further into blue territory. Conditional on winning all 7 swing states, for instance, Trump had a 22 percent chance of also winning New Mexico, a 21 percent chance at Minnesota, 19 percent in New Hampshire, 16 percent in Maine, 11 percent in Nebraska’s 2nd Congressional District, and 10 percent in Virginia. Trump won more than 312 electoral votes in 16 percent of our simulations.
But on Tuesday, there weren’t any upsets in the other states. So not only did Trump win with exactly 312 electoral votes, he also won with the exact map that occurred most often in our simulations, counting all 50 states, the District of Columbia and the congressional districts in Nebraska and Maine.
I don't know of an intuitive test for whether a forecast of a non-repeating event was well-reasoned (see, also, the lively debate over the performance of prediction markets), but this is Silver's initial defense of his 50-50 forecast. I'm unconvinced - if the modal outcome of the model was the actual result of the election, does that vindicate its internal correlations, indict its confidence in its output, both, neither... ? But I don't think it's irreconcilable that the model's modal outcome being real vindicates its internal correlations AND that its certainty was limited by the quality of the available data, so this hasn't lowered my opinion of Silver, either.
Jump in the discussion.
No email address required.
Notes -
This is tangential to my main point. Nate's beliefs about the world, if expressed, would've cost him a lot of money. There are probably large numbers of people who trusted Nate's modelling and lost money thinking 'oh well if Nate says it's 50/50 then I can profit from Polymarket being 70/30'.
I think Nate is trying to claw back lost prestige. It reeks of cope to say 'my model exactly predicted the most likely election map' when his model gave it a 6% chance of happening. He struggles to even say what he wanted to happen on election day from the perspective of 'what makes my model look good'. If you're not producing a useful, falsifiable prediction then what is the point?
The important thing is getting it right. I want models that give alpha. If I'm going to pay for a substack, I'd rather pay for someone who got it right - Dominic Cummings who said confidently that Kamala would lose, based on his special polling methods. He actually predicted a Trump victory vs Kamala 311-227 in 2023. He foresaw that Biden would likely become obviously senile, that the Dems needed a better candidate.
https://x.com/Dominic2306/status/1854275006064476410
https://x.com/Dominic2306/status/1854505847440715938
Let's say he had bet $100,000 at 50-50 odds that he wouldn't roll a six on a die. Then he rolls a six. Does that prove something about his beliefs? It's only profitable in expectation. There is no guarantee of making money.
To take the election example, 50-50 means losing half the time. It's only profitable because, when you do win, you win more than you would have otherwise lost.
That is just not possible to get from a single sample. You need to look at his track record over many elections.
More options
Context Copy link
If that is so, I accept this correction in good faith, and I do believe this elaboration of the main argument is substantially stronger. I am not attempting to change your opinion on Nate Silver's accuracy.
I am still curious of if there was ever evidence that Silver's bet was accepted, both for it's own sake in addressing the question and to updating priors, but an argument of 'he would have lost money if he made the bet' is a substantially different argument than 'he refused to respect his own challenge,' and if you did not mean the later interpretation I am grateful for the clarification.
More options
Context Copy link
how does it work?
More options
Context Copy link
More options
Context Copy link