Let’s Check In On How Nate Silver’s Weird Hexagon Analysis Has Held Up

We’re about two months out from the first contest in the Democratic primary, so I think it’s a good time to look at the current state of the race and compare how they compare with peoples’ expectations.

.

Now, I myself shared some thoughts on that back in march. Essentially my argument at the time boiled down to this: current polling may not predict what’s going to happen in the race, but you should more or less accept what people are telling you because however sophisticated you think your model is not actually that smart. If Biden and Bernie are polling at first and second, and have for a while, then it’s reasonable to assume they have something going for them and shouldn’t be treated lightly. Don’t contrive a lot of reasons to dismiss them just because you can’t reconcile their success with some preconceived notion of how politics works.

.

On the other hand you also had a lot of people at the time who did the opposite and contrived a lot of reasons to dismiss Biden and Bernie based on some preconceived notion of how politics works. The main example I pointed to was Nate Silver and his weird demographic hexagon analysis of who should hypothetically have the most appeal to the Democratic base.

.

So let’s take a look at how things are right now:

Okay, some candidates had some ups and downs, but overall things have been surprisingly consistent. Biden’s still up, Bernie is still in second. Looks like I was more or less right to be conservative and assuming that there was a reason those two candidates were doing well and not trying to outsmart the polls.

.

Now let’s see how Nate’s model is holding up…

.

… Oh dear God. Three of the top five candidates aren’t even in the race anymore, and the other two didn’t qualify for the next debate. All the candidates in the current 4-way race are ranked at 6th or lower (though, to be fair, Buttigieg wasn’t in the race yet), and the highest of them is arguably doing the worst.

.

I mean, did it just not occur to Nate that Kamala’s record as a prosecutor might undermine her appeal? And Julian Castro? What the hell?

.

Alright, what about betting markets. Surely the smartest people in the room were savvy enough to anticipate how the race would break down…

… oh, no, wait, I forgot, political betting markets are stupid. They were betting on Harris well into March, and pretty much over valued her since. Then they kept chasing more or less every ephemeral boom that happened in the polls, all the while consistently underestimating and criminally underrating Bernie. And they’re still doing it, quite frankly.

.

Now my point here’s isn’t (entirely) to rub too clever by half political analysts’ noses in their failures. Nor is it to say that there aren’t things we can reasonably expect to upend current polls (like Biden’s strategically dubious decision to ignore Iowa and New Hampshire until fairly recently). But if you’re going to play inside baseball on politics, at least try to be a little humble and not get so wrapped up in what you think is supposed to happen that you miss what actually is happening.

.

Or you can do what Nate does, deny you made a prediction and act like your model worked all along.