Show Us Your Model

It might be easy to believe we're approaching Peak Trutherism, what with good old-fashioned birthers now being supplemented by BLS truthers and poll truthers. But just you wait—should Barack Obama win this election, we'll see an explosion of election trutherism that will be truly unprecedented in scope. In the meantime, we can content ourselves with the newest variant, Nate Silver trutherism, which isn't coming just from conservatives.

In case you don't know, Silver runs the blog FiveThirtyEight, which after producing a series of highly accurate predictions during the 2008 campaign got swallowed up by The New York Times. Silver makes electoral projections by taking as many different polls as he can find and running them through an algorithm. Rather than just averaging the polls' results, the algorithm uses a series of variables, including state polls and each pollster's prior record, to produce a number of different estimates. As of today he gives Obama a 77.4 percent chance of winning, higher than it has been at some points but not too far off from where he has estimated for most of the campaign.

In the last few days, we've seen a couple of different Silver narratives emerge as attention to him has increased. First, you have stories about how liberals are obsessing over Silver, "clinging" to him like a raft in a roiling sea of ambiguous poll data. Then you have the backlash, with conservatives criticizing him not because they have a specific critique of the techniques he uses, but basically because they disagree with his conclusions (No way is Obama going to win!) and because he's a liberal so therefore he must be intentionally warping his data to produce more Obama-friendly results. For instance, Joe Scarborough recently declared that Silver has to be wrong, because while Silver says Obama has an advantage, Scarborough knows the race is a toss-up, which I guess he feels in his gut. The most hilarious criticism came in this column in The Examiner, which noted, "Nate Silver is a man of very small stature, a thin and effeminate man with a soft-sounding voice." So there.

Then you've got the reporter backlash. At Politico, Dylan Byers raised the possibility that Silver would be completely discredited if Mitt Romney won, because "it's difficult to see how people can continue to put faith in the predictions of someone who has never given that candidate anything higher than a 41 percent chance of winning." But of course, if you say there's a 41 percent chance of something happening and then it happens, you wouldn't actually be discredited, because 41 is not zero. And that's not even mentioning the huge number of pollsters who make absurdly wrong predictions all the time and continue to ply their trade. Furthermore, as Ezra Klein points out, a lot of these criticisms come from writers at Politico, the most strategy-obsessed, who-won-the-day publication there is. Looking at the election's eventual outcome systematically is almost an affront to their business model.

But Silver is only a threat to reporters if they see explaining who is going to win as their primary job. And if that's how they see their job, they really ought to apologize to the public and find another career. Because there are few things as useless to the citizenry as a reporter giving his or her opinion on who is going to win.

After all, there are lots of interesting and revealing things going on in campaigns. There's a debate about where the country is now, where it has been, and where it ought to go. There are interesting characters we can learn about. There are policy issues aplenty. Even the dramatic moments of the campaign can be reported on and examined without asking, "Will this change the race?" I'm not saying you have to banish any of the who's-up-who's down stuff completely (I certainly talk about it here), but if you actually feel threatened by someone coming along with a persuasive answer to the question of who's going to win that wasn't derived from listening to campaign spinners, then you really ought to re-examine what you're doing.

Finally, let me address the question of Silver's liberal fans. Do they love the fact that he gives them reason to feel optimistic? Sure. But that's only half the reason they love him. If he were not as rigorous as he is but was producing the same results, liberals wouldn't be as taken with him as they are. There are Democratic polling outfits out there, and while liberal blogs might cite them fairly often, none of them have produced the same devotion Nate has. On the other hand, would liberals be as interested in Silver if his analysis consistently predicted a Romney win? Probably not. In other words, it took both to make Silver a liberal hero. His projections had to make liberals feel optimistic, and he had to be going about things in the kind of nuanced, detailed way he is. Liberals are pleased as punch that the thing they want to be true can be supported by a highly complex and sophisticated analysis. They want to feel good, but they also want to feel smart.


The "Nate Silver truthers" should be aware that Silver very accurately predicted the Republican congressional sweep in the 2010 midterm elections.

Exactly. Dems believed in Nate's models in 2008 when he showed an Obama win. We believed in his models and cried in our beer (microbrewed, natch) in the 2010 elections when he said we would get shellacked. We believe in his models now with him showing an Obama win again. That is because we are the party of science, math and facts. We believe Nate because he has the best ideas on how to evaluate polls and he has a track record of success that shows he is right.
Trutherism is when you believe something is true because you FEEL like it is true. An example of trutherism would be that it FEELS like Obama is from Kenya because real Americans (especially Presidents) should be white. You can show a truther the birth certificate, newspapers that report his birth and the fact that his family has always said he was born in Hawaii long before anyone thought he would ever be president, but they won't believe them because Trutherism doesn't accept reality. In trutherism you ignore, distort, obfuscate, twist and suppress facts that disagree with what you feel.

In my model the electorate is evenly split between Dems and Reps, and Romney has a double digit lead amongst indies. In 2008 Dems had a +7 advantage over Reps and won indies by 8 points. SO my model says Romney has a 63.24% chance of winning, which means if he loses I still win because I said ther was a 36.66% Obama would win.

Look how accurate I am! /sarc

Saying Silver's model is "rigorous" doesn't make it so. I could make a 10,000 line computer model that is still as subject to GIGO as anything else. Its not "sophistication" its sophistry. And, yes, there is a difference.

Riiiight... The model is sophistry and every single one of the polls is systematically biased! GIGO!

It's hard for me to imagine someone who wants to feel bad, and dumb. Or someone who wants to feel good, but dumb. Of course liberals want to feel that they are right, and that they are right for the right reasons.

And yes, if Silver was predicting a Romney win, he would not be as popular or as well regarded, and there would be more efforts to pick apart his methodology

Well this is just silly. So Silver can't be called wrong if Romney wins because he said there was a chance he would? Please. Wrong is wrong. If Silver's model predicts the wrong winner, his model does not work--by definition--. I appreciate that he attempts to give odds, but the question we have asked him to answer is "who will win" not "who is most likely to win".

I saw you the other day - you were walking in the rain soaking wet without an umbrella, because the weatherman didn't predict '100% chance of rain'!

You need to be logged in to comment.
(If there's one thing we know about comment trolls, it's that they're lazy)