Last week, the simulation map on 270 to Win was overwhelmingly favoring Obama based on the available polls. They warned that it would be a week before meaningful data would appear there, and they were right. I've been playing with the maps this evening (not wanting to run into any more stories about barnyard beasts or cosmetics) and have saved out six different maps showing narrow victories for each side, medium victories and large victories. The simulator itself currently reflects a 55% probability of an Obama win, with a median of 272 electoral votes. What I think the below maps illustrate is how little has to change for significant results.
(Click on images for larger versions)
Large Democratic win: 320 D/ 218 R. The key here is sweeping the upper Midwest plus picking up Florida. Colorado and New Mexico are not necessary to the scenario, but would likely be present if the win was this big. This was always the optimistic map for Dems. I think Iowa would go blue in a scenario like this and probably Nevada, too.
Medium Democratic win: 295 D/243 R. Colorado and Ohio go by the wayside, but Florida is salvaged. I actually think those combinations would be flipped in this scenario, but I also have doubts that Virginia would go Dem while Wisconsin went Republican. Flip those as well and you'll have roughly the same EC outcome.
Narrow Democratic win: 272 D/266 R. And I think the closest to a likely win this round, and is the result most likely in the simulator. Colorado and Florida go red, as does Ohio, while the party barely holds on to Michigan, Pennsylvania and Washington. Note that New Hampshire, Nevada and South Dakota could each be a deal breaker in this scenario. I think the party needs to be prepared for a solid red South.
Narrow Republican victory: 268 D/270 R. What is interesting in the simulator is that this scenario was far less likely than the bigger Republican wins. They lose by small margins but win by large ones. In this, the Dems retain all of the Midwest except Indiana, but the Republicans lose Ohio and Colorado, keep Florida, and pick up New Hampshire, Nevada, South Dakota, Minnesota and Washington. These are plausible pick ups. I think it more likely that they would keep Ohio and lose Minnesota, but they would need then either keep Washington or snag back Colorado. Note - compare this map and the narrow Dem victory one to the current map on electoral-vote.com. It has the same EC breakdown as this one, but has Washington and Minnesota in the Demo column, and Ohio, New Mexico and New Hampshire in the Republican one.
Medium Republican win: 250 D/288 R. I think this the most plausible winning scenario for Republicans. The West stays solidly blue, Colorado and New Mexico are blue but the south stays red and McCain squeezes out narrow victories in Wisconsin and Pennsylvania. This is why the "Appalachian" vote matters. I doubt South Dakota would stay blue.
Large Republican win: 219 D/319 R. The key to this is winning the upper Midwest. Minnesota and Washington could easily be flipped in this scenario for much the same result. Can McCain take all of the intermountain West? I think it possible, though they would not be easy pickups. The simulator shows Washington going red more easily than Colorado, for example.
I find all of these combinations well within plausible expectations. For me, simply looking at maps and shifts, a bare handful of states will determine the EC, but the particular combination that will do so varies greatly. The focus tends to be on the larger battleground states, but undecided small states may suddenly shift and provide the 1-3 EC margin.
Visit 270 to Win and use the simulator to see the changing possibilities based on probabilities.