As we count down the last few days of this interminable campaign (and I see Erick Erickson is just wishing it was over with, I'll take that as a good sign), the Blue Team has been buoyed by math, while the Red Team has "the math". It is actually a fascinating test in a few days between the quants and the pundits. If Obama wins 300 EV or more, and the states and senate races go as the quants predicted, they will have huge credibility going forward, but if Romney actually wins, I think we will all be scratching our heads.
One interesting split is between Sam Wang at Princeton Election Consortium and Nate Silver. PEC has Obama's chance of election at 99%, while Nate is around 80% (using their "nowcast" what would happen if the election were held today). If Nate is right, Romney still has a 1 in 5 chance, if Sam is right, the election is over. I'm not absolutely sure of why the two differ, Sam does use state polls only while Nate adds his secret sauce. Nate's model also implies a bit more likelihood of Romney winning a state like Wisconsin or Pennsylvania. If I apply a simple math model and assume the only states in play are FL NC VA OH NH IA and CO I tend to think Sam is right. I don't see Wisconsin or Nevada as up for grabs at all. Looking at the 7 swing states, Romney has to win all four of FL NC VA OH, then one of NH IA or CO to become President. Let's assume his chances are 60% FL, 70% NC, 50% VA, and 40% OH. I would think these are clearly favorable numbers to Romney, but the chance of him winning all four is only 8.4%. If Romney's odds are 50% NH, 50% IA, and 50% CO, again generous, he has an 87% chance of winning at least one. His odds of winning the election are then about 7% with very favorable assumptions. Put more realistic numbers up, and it gets down to 3% or less, more in line with Sam's odds than Nate's. If I was a betting man, I'd buy me some Obama shares.