If you haven't visited FiveThirtyEight.com , I strongly suggest a quick stroll over there. It's a pretty clever site where the various polls aren't just aggregated and averaged, they're weighted according to fairly comprehensive factors to provide probabilities of a winner, not just a raw percentage difference. Nothing's guaranteed, of course, particularly with the possibility of dirty tricks at the polls. 
Among the more interesting aspects of this site is the way it takes into account the irregular topology of the electoral college math to try to figure out what the most important states are, ultimately. I won't try to explain it here, read the FAQ at the web site. But I will highlight their idea of "Tipping Point" states. These are states where if they fall to one candidate or another, in their simulations using polling data ranges, the state causes a sort of cascade of scenarios that throws the election one way or another. So, for instance, Ohio is a tipping point state, but in the sense that if McCain loses it, there's almost no way for him to win the election. Pennsylvania's another, maybe the most important tipping point state this time around, because while not as hard for Obama to win without it as McCain to win without Ohio, it becomes quite difficult.
Click on the image to see a full-resolution version of the tipping point diagram as of today.
So this is yet another reason to be so dedicated to keeping Pennsylvania blue. As far as importance to an Obama victory, it's the center of the nation, much as it was dating back to the very first federal elections.
Recent Comments