In summary, according to Rasmussen’s turnout model, Romney has no “leaning” states that he needs to defend. He has to take at least one of the Obama-leaning states plus all of the tossups.
Leaning Obama:
Ohio - 18
Wisconsin - 10
Toss ups:
Colorado - 9
Florida - 29
Iowa - 6
Virginia - 13
The campaign will change a few positions, but the big issue is turnout. Particularly since turnout will decide several competitive Senate races, we need to get every possible conservative voter to the polls. Freepers can make that happen, and we need to do it this year.
To be precise, Romney has two +1% leaners in Florida and Iowa, which equate to roughly a 60%-40% probability of the leader winning. Obama has Virginia at +1%, too. I'm putting these states into the Toss-up category, although their actual probabilities are used for the expected value calculations.
For comparison, a +2% lead equates to roughly a 70%-30% probability of the leader winning, which is a more likely lean. These also depend on the poll's margin of error for each state.
For my Electoral College Balance chart, if the probability of the leader winning is between 50%-60%, I'm calling it a Toss-up. If it is roughly between 60% and 80%, it's Lean. If it's between 80% and 99%, it's Strong. If it's 100%, it's Safe.
Colorado is the only true Toss-up at a 45% tie.
-PJ