"Present Tens"ions in PA

In his newly revived blog Present Tense, Cory talks about the McCain camp's focus on suburban Philadelphia. Despite the long-shot nature of this, it makes sense. According to today's analysis from FiveThirtyEight of the odds of various scenarios (shown left), Obama needs to win PA. If he loses OH & FL but not PA, he still has a 66% chance of winning (via states like CO, VA, etc.). But if he loses OH, FL, and PA, his odds of winning drop to 2%. Now, I'm not a statistician (Mike, want to comment?), but this is a dramatic difference in odds. And it's a fancy way of saying that if Barack loses PA, he has a 98% chance of losing.

So McCain has no other choice but to hope the polls are dead wrong in VA and CO (be it Bradley effect, bad polling, or last minute right-wing scare tactics), and then he needs to hope for the best in OH, FL, MO, MT, ND, NC and win all of those in addition to the states already in his column. In other words, the best McCain can hope for is to get all the states Bush got in '04 minus NM, and he can still win.

From the Obama point of view, the minimum to 270 is all the Kerry states (which, remember, includes PA) and then on top of that one of the following permutations:

• one of MO, OH, FL, or VA, of which VA looks likely, or
  
• NM + a state with at least 6 electoral votes, of which CO is the smallest likely shot
  
• NM + MT + ND (but none of the eastern battlegrounds, pretty unlikely)

You probably already know you can play around with electoral possibilities at 270towin.

The take-home message. If I were an Obama strategist, I'd fold operations in OH and FL and focus on PA, CO, and VA. At a more general level, I'd make sure we kept the Kerry states (incl. MN and WI) and then on try to win NM, CO, and VA with convincing margins.