Here's The Real Reason Nate Silver's Perfect Election Call Was Such An Awesome Breakthrough

nate silver on stephen colbert

Photo: Comedy Central

So for the second election in a row, Nate Silver’s electoral forecast was basically perfect. Like, ridiculously so.Of the 50 states, Nate Silver only had one a tossup, and that was Florida… the only state that needed an extra day to have someone declared a winner.

Here’s the awesome thing about Nate Silver being so spot on: We can go ahead and dismiss almost all of the ex-post facto excuses/spinning about why the race went the way it did.

On the Sunday before the election, Tim Murphy at Mother Jones did a hilarious piece called Every Single Political Game Changer Of The 2012 Election. Basically, it’s a big list of every event that some pundit somewhere called a “Game Changer” during the race.

Here’s a chunk, listing some of the game changers:

The Benghazi cover-up. A new book from WorldNetDaily. A hypothetical question about Simpson–Bowles. First-time Puerto Rican voters. Latinos. Asian-Americans. The 47 per cent tape. Paul Ryan’s “makers and takers” riff. Suburban women. October jobs report. September jobs report. Democrats’ ground game. Obama’s tech operation. James O’Keefe’s voter fraud sting. Gallup early voting poll. October 27 Gallup tracking poll. Gloria Allred. The Walker recall. Paul Ryan’s interview with WJRT-TV, Flint. January jobs report. The Des Moines Register endorsement.

The big one that people are talking about now is Hurricane Sandy. Dick Morris, for example, said he totally whiffed his election call because Hurricane Sandy changed the game. Others have said something similar, that Romney had some momentum going into the final week, and that Sandy (and Chris Christie) killed it, not letting Romney get his message out in the final days of the race.

But this is bunk. Here’s a chart of Nate Silver’s Forecast going back to the summer:


Obama was handily favoured every single day of the race. The closest thing to a game changer (by far) was the early disastrous debate in early October. But at no point, did Romney’s odds of winning hit 50 per cent. And by the middle of October, Obama was back on the re-ascent. Sandy didn’t change anything. The 47 per cent tape didn’t change anything. Benghazi didn’t change anything. Paul Ryan’s selection didn’t change anything, and so forth. Romney didn’t lose because of a glitch in his election day Get Out The Vote app, which is something a lot of folks are talking about right now.

Political punditry like market analysis is lousy with ex-post facto explanations of events. Market coverage might be worse (“Stocks Rise On Fiscal Cliff Deal Hopes!” “Stocks Fall On Fears The Fed May Tighten!”). Robust models that get outcomes right really let us cut through all that noise and figure out what matters.

Business Insider Emails & Alerts

Site highlights each day to your inbox.

Follow Business Insider Australia on Facebook, Twitter, LinkedIn, and Instagram.