Justin Sullivan/Getty Images
Mitt Romney speaks during a campaign rally at the Red Rocks Amphitheatre in Morrison, Colorado. Nate Silver says he probably won't win. Emphasis on the "probably."
In recent weeks, everyone — even casual political observers — has scrutinized the numbers at Nate Silver's FiveThirtyEight blog at the New York Times. Silver, if you've been living under a rock with no media for the past few years, is a 34-year-old rising star in the world of data journalism. He made a name for himself in baseball statistics and neatly transferred that fame over to politics. He basically called the 2008 election. When the New York Times absorbed his once-anonymous blog, he was a made man. (Watch him in uber-nerdy action on Comedy Central's "The Colbert Report" in the video below.)
The political right isn't happy about Nate Silver because, as of Tuesday morning, FiveThirtyEight indicates that Mitt Romney has a less than 10 percent possibility of winning the election. Indirectly, former Reagan speechwriter Peggy Noonan tried to discredit Silver in the Wall Street Journal on Monday with heartfelt rhetoric and a hodgepodge of theories, the most ridiculed of which was the idea that Romney is surging toward victory because across America he's winning the war of the yard signs.
The political left adores Nate Silver — or it does now, anyway, given that he's predicting a landslide re-election for Obama. Lefties have daily, and perhaps even hourly, studied FiveThirtyEight to see whether the President's prospects are improving, whether he's bounced back from that miserable first debate, etc. etc. etc.
So you look at Romney's less than one-in-ten probability of winning, based on Silver's take on the polls, and you say, "Well, the guy has no chance."
But of course he does have a chance. It's just an increasingly small one. Whereas Obama has an increasingly large chance of getting to the 270 electoral college votes he needs to win re-election.
It's not pure probability, because we don't have the luxury of playing out the actual election for a long period to test Silver's model. But it is a reasonable approximation based on what's evidently happening with the polling data, which is the basis for Silver's analysis (along with other pieces of information).
But, but, but! Romney could still win. Yard signs could trump math. And Silver's model, which hasn't been humming along for decades, could be flawed — although as he's pointed out, a lot of polls would have to be off for that to happen. The important thing to remember is that when you're talking about probabilities, there are good bets but absolutely no sure ones.