With less than 24 hours to go before actual votes replace prognostication, Nate Silver’s election blog fivethirtyeight.com has become comfort food for Obamaphiles. At a time when public opinion polls find the Presidential race to be within-the-margin-of-error close, Silver says it’s a near-sure thing for the incumbent.
Silver— a number-cruncher who first gained fame analyzing baseball stats and coming up with a now widely-accepted talent evaluation system called PECOTA— says on Election Eve there’s an 92.2% chance Obama will win at least the 270 electoral votes that would give him a second term (though Silver pegs Obama’s chances of winning the popular vote at only 50.9%).
He’s liked Obama’s chances all along. All fall, Silver’s never pegged the prospects of an Obama electoral win at anything less than 61%, even after the first debate that so many saw as a game-changer for Romney.
Just within the last 24 hours, I’ve talked to three friends who admit to consulting Silver’s blog several times a day, to see if his predictive needle has moved at all (in fact, it’s gone up several days in a row). The fact that all three back Obama suggests that checking “538” provides a shot of confidence when electoral uncertainty becomes unsettling.
Obamaphiles love citing Silver’s numbers on blogs. “Nate Silver's 538.com blog gives me hope”, writes blogger Joe Cook. Jack Dawson summarizes his coping mechanism as “Rely on Nate and ignore everything else (which helped get me through the post-Denver debacle).“ Holly Cara Price has put her mantra in a widely-circulated graphic on Facebook— “Keep Calm and Trust Nate Silver.”
Silver— whose blog is found on the New York Times website— is not alone in projecting an Obama victory. University of Virginia political scientist Larry Sabato figures Obama will win 290 electoral votes in his “Crystal Ball” website, while electionprojection.com predicts the President will take 303 electors.
But Silver’s the highest-profile projector, having gotten 49 of 50 states right in 2008 (even calling some surprises like Obama’s victories in Indiana and North Carolina).
He’s been attacked for skewing the numbers, been the subject of snarky headlines (“Nate Silver Could be a One-Term Celebrity”, wrote Politico), and his methodology has been picked apart by stat freaks.
Plenty of conservative pundits opine that undecided voters will break heavily against the President, and predict Silver will have electoral egg on his face, come Wednesday morning.
Karl Rove forecasts a Romney win, 285-253. Conservative columnist Michael Barone of the Washington Examiner, sees Romney getting 315 electoral votes. These forecasts, unlike Silver’s, rely on political gut instinct rather than statistical modeling.
Romney backers can also take heart in a forecast model with only mystical statistical relevance— the Redskin Rule. The Washington Redskins lost on Sunday. Since 1937, Politico reports, a pre-election Redskins loss has meant a loss for the party that won the popular vote in the prior election— in this case, Obama and the Democrats.