FiveThirtyEight, Nate Silver’s famous brainchild, ostensibly exists in order to aggregate the polls in any given election cycle, and then make accurate predictions based upon those poll aggregations.
In other words, Nate Silver is a stats geek. And that’s exactly what we want him to be.
There’s only one problem.
Nate Silver no longer has any faith in polls.
Here’s what he’s been saying recently in regards to the fact that FiveThirtyEight is consistently reporting Donald Trump to have a better chance at winning the presidency than any other polling aggregator, any betting markets, or other prediction markets.
Why Nate Silver thinks Trump more of a chance - Nate Silver: “Trump remains an underdog, but no longer really a longshot: His Electoral College chances are 29 percent in our polls-only model — his highest probability since Oct. 2 — and 30 percent in polls-plus…. From a set of simulations the polls-only model ran earlier this evening, I pulled the cases where Clinton won the national popular vote by 3 to 5 percentage points. In other words, we’re positing that the national polling average is about right, and seeing how the results shake out in the states… Trump’s chances are slim-to-none in this scenario. His odds are 10 percent or below in all of the Clinton firewall states except for Maine and New Hampshire… The question is how robust Clinton’s lead would be to a modest error in the polling, or a further tightening of the race. So here’s a second set of simulations, drawn from cases in which Trump or Clinton win the national popular vote by less than 2 percentage points… This isn’t a secure map for Clinton at all. In a race where the popular vote is roughly tied nationally, Colorado and New Hampshire are toss-ups, and Clinton’s chances are only 60 to 65 percent in Wisconsin, Michigan and Pennsylvania. She has quite a gauntlet to run through to hold her firewall, and she doesn’t have a lot of good backup options.”
So look at what Nate says in these two paragraphs, and look closely. Nate says that if the national polling average is about right and those results came to fruition in the states, Trump’s odds would be 10 percent or below in nearly all Clinton states.
Huh.
Kind of like how Sam Wang at Princeton Election Consortium has Clinton with a better than 97% chance to win, Drew Linzer at Daily Kos has her with 92%, and NYT Upshot has her at 86% currently to win?
So why exactly is it that out of all of the very experienced and accurate statisticians who are seeing this race as relatively stable and relatively clear cut win for Clinton, Nate and FiveThirtyEight are seeing a very different race?
Look to Nate’s second paragraph: The question is how robust Clinton’s lead would be to a modest error in the polling…
He is essentially hedging his model against the idea that the polls are going to be off by a fairly wide margin this election season—at least 2%, perhaps more.
Silver is also spooked by this notion: “My view of this general election — both how the FiveThirtyEight forecast models see the race and how I see it personally — is that it’s characterized by high volatility and high uncertainty. Those two things go together, although they’re not quite the same.”
And yet, Sam Wang, amongst others, has pointed out that in reality, voter polarization has made it so that volatility is actually quite low amongst the electorate. Says Wang:
Measured in terms of polling margins, this year’s campaign is the most stable of any race in the era of modern polling, going back 65 years. I estimate that 2016 has been slightly more stable than 2012, though not by much. Really, those two elections are basically the record-setters when it comes to stability.
So what is really going on with Nate Silver and his doom and gloom prognostications?
Many have noted that this election cycle, his model has been incredibly sensitive, moving up and down with every single poll and changing in ways that are hard to comprehend.
During the primaries, Silver notoriously made a prediction that Trump had less than 2% chance to become the republican nominee. Said Silver at the time:
If you want absurd specificity, I recently estimated Trump’s chance of becoming the GOP nominee at 2 percent. How did I get there? By considering the gantlet he’ll face over the next 11 months — Donald Trump’s Six Stages of Doom:
After Trump proved Nate wrong, he took stock of his error and claimed that he’d learned from his mistake. What he’d learned was that he needed to trust the data, trust the polling and his models and not resort to punditry.
But I would argue--as Silver has now appeared to reach peak Trump insanity with his recent estimation that Trump has a 35% chance or better to win the Presidency—that in fact Silver learned nothing from his primary blunder.
The reason his model differs so fundamentally from everyone else is because he is once again trusting his gut and intuition over what the data is saying. Nate is basically saying that he “feels” there’s a good chance the polling will have missed the true nature of the race on election day.
He once again does not trust the polls.
A highly ironic turn for a website that prides itself on using data and cold hard facts to come to conclusions about the state of the election.
Comments are closed on this story.