Electoral math

The amazing thing is that Obama doesn't even need both Colorado and Virginia, just one of them.

That's right, Obama can lose Ohio, Florida and win only one of Colorado and Virginia, and he still wins.

How? By holding onto (from the 2004 results) Wisconsin, Michigan, Pennsylvania and New Hampshire, and adding Iowa, New Mexico (both look likely) and either Colorado or Virginia.

Here's a map you can play with.

Now, try adding Iowa and New Mexico to the 2004 states. Just doing that puts Obama at 264, just 6 points shy.

What will be the tipping point state? I would venture the following, in this rough order:

1) Colorado (9 points)
2) Virginia (13 points)
3) Ohio (20 points)
4) Florida (27 points)
5) Missouri (11 points)
6) Indiana (11 points)

If Obama wins Nevada, it would make it a 269-all split, in which case a stray electoral vote peeled away from Nebraska could affect the outcome.

But what if Obama loses both Ohio and Michigan (a much more likely scenario than Wisconsin or Minnesota moving)?

Losing Michigan would be a real disaster for the Dems. But not the end of the world.
Ohio & Michigan to the Repubs, Colorado and Virginia to the Dems would again result in a 269 split. The Dems would need a tiebreaker of some sort, most likely Nevada.

The key message here for the Dems is that they need to work Colorado and Virginia slightly more than Ohio and Florida, but they need to work Michigan most of all to keep it in the fold.