Nate Silver vs. the world

Matt K. Lewis Senior Contributor
Font Size:

In the past few years, you’d be hard-pressed to find a star that has risen higher or faster in the political prediction business than New York Times blogger Nate Silver.

This is for good reason. As On The Media notes, “In 2008, his blog FiveThirtyEight correctly predicted the outcome of the presidential race in 49 out of 50 states. (In that same election, he was also right about all 35 senate races.)”

Not too shabby. But one wonders if his reputation might be about to take a hit if Romney actually pulls off a victory.

So far, Silver isn’t hedging any bets. In fact, he’s all in. Using his prediction model, on October 4, he gave Obama a 87.1 percent chance of winning.

Today, he gives Obama a 70.4 percent chance of winning. (70.4!)

Don’t get me wrong, I think it’s entirely possible that Obama will win re-election. (Before the first debate in Denver, even I would have agreed with Silver’s analysis.)

But the race changed dramatically, and my guess is that, right now, it’s probably a 50-50 proposition. (Silver would likely dismiss this by arguing that political commentators always think every election is a coin toss. But empirical evidence suggests the race is actually close. At the time of this writing, Real Clear Politics poll average has Romney up .1 percent.)

So why hasn’t Silver adjusted accordingly?

It could be that he is just frankly smarter than all the other pollsters, pundits and predictors.

Maybe he just got lucky last time?

Or maybe it’s wishful thinking? — Silver was up front about being an Obama supporter in 2008, and it’s hard to blame conservatives for wondering if he might be working the refs.

Silver has become wildly influential among media elites — and some could argue that his predictions are even a sort of self-fulfilling prophesy (good numbers beget good fundraising and good press coverage, which, in turn, begets good poll numbers…)

And keep in mind, when it comes to face-saving, Silver can always just “adjust” his numbers in, say, early November — if he senses an Obama defeat. Who knows, maybe he will take credit for correctly predicting Romney would win?

For now, though, Silver’s predictions do have a psychological impact beyond merely providing information.

As BuzzFeed’s Ben Smith recently tweeted:

Another theory is that, unlike a real gambler, Silver doesn’t have much to lose (will his credibility really take a hit with his friends in the media if Romney wins?)

As one reader emailed me:

If you had a book and you only started taking election bets on October 1st, your line would be Romney favored by 8 to 1. All those early bets were sucker bets. It would be the same as betting a football game on Tuesday only to find out your star quarterback was hurt 10 minutes after you laid down your bet. Nate Silver’s numbers are shit because since he doesn’t take bets, he just leaves his numbers the same. He still has Obama at over 2.5 to 1. That is a huge difference. Nate Silver gives odds, but does not take wagers, that is not possible. It’s like playing poker with play money

Unlike Mitt Romney, I don’t casually make $10,000 bets. Still, I do wonder if Silver is confident enough in his models that he would be willing to put some real skin in the game.

Matt K. Lewis