Here is Nate Silver’s map from his FiveThirtyEight blog on the eve of yesterday’s election:
Compare it to CNN‘s results map as of noon PST:
He nailed it. Every state he predicted would go to Obama went to Obama, and every one he predicted would go to Romney went to Romney. Florida is the only one undecided, and it’s leaning ever so slightly to the President. (It’s also the lightest shade of blue in Silver’s map and was his only real “tossup” left).
At worst, Silver will match his accuracy from 2008 (49 of 50). If Obama holds his lead in Florida, he’ll be perfect.
He got a pretty good read on the electoral vote count too:
Note the two highest spikes: 332 and 303. If Obama wins Florida, he’ll get 332, if not he’ll stay at 303. (The third spike at 347 includes North Carolina, which was Romney’s weakest “lean.”)
And he knew which state would decide the race:
It was around 8:10 p.m. PST when the networks called Ohio and the election, though in this case, I think we all saw Ohio coming:
So what was Silver’s secret. Why did he get it right when so many pundits either said it was too close to call or that Romney would win? One word: Math.
So score the election a win for Obama and a loss for Romney, but also score it a loss for punditry, spin, and propaganda, and a win for math.