Last Sunday on the political talk shows, I noticed an interesting thing. An electoral math scenario for the outcome of the election in the Electoral College is being floated by some pundits: what happens if we get a 269-269 tie? I've noticed it more from right-wing commentators than left, but I think its true appeal is to the political wonk of either stripe -- a "what if" game to make this already exciting election even more so.

Here is Bill Kristol's version of the tie scenario, from last week's Fox News Sunday show:

"Here's an amazing fact. If you take the 2004 results, give McCain New Hampshire, which Kerry won, which I think is reasonable, give Obama New Mexico, Colorado and Iowa, which Bush won, in all of which now Obama is ahead, you end up -- leave everything else the same, which is quite possible, you end up with 269-269 result. That would be fun."

So what would happen if the Electoral College did tie? Well, it's a little complicated. Some of us might remember from our seventh grade civics classes that the election then goes to the House of Representatives. And you might think that because Democrats have an advantage right now in the House (and are expected to pick up seats this year) that this would automatically mean "President Obama." But you'd be wrong.

There are two quirks in the system that make it almost impossible to accurately predict what the outcome would be. The first is that each state's House delegation gets only a single vote in choosing the president. In this instance, the House works more like the Senate, in that states are considered of equal weight to each other. Each state's House delegation in such a vote first gets together and holds their own vote amongst themselves. Whoever wins this intra-state vote gets that state's vote for president. If a state has a tie in the intra-state vote, then it is not counted in the presidential vote.

The second problem is that it is the incoming House which votes, not the outgoing House. This means that you'd have to accurately predict almost every House race to come up with who would ultimately win.

Now, the advantage definitely rests with Obama, since the incoming House is likely to be more Democratic than the outgoing House. Even the Wall Street Journal is confirming this with headlines of late such as "Republicans Gird for Big Losses in Congress." But losing a few House races here and there to the Republicans could change the outcome of the vote. And there are a surprising number of states where one or two House seats changing hands either way could affect the outcome. [Note: I'm putting the lists of state data at the bottom, if you're interested.]

For a baseline, let's first consider the current makeup of the House. If everyone in the House kept their seat, and neither party gained a single seat, this is how the vote for president would play out: Democrats 28, Republicans 21, Tie 2. Obama wins, 28 votes to 21 (total is 51 because Washington, D.C. is counted as state for presidential elections).

Now a worst-case scenario for the Democrats. If every state that was separated by party by only one vote in the House went Republican, then the vote would be: Democrats 9, Republicans 34, Tie 8. McCain wins in a landslide, 34 to 9. But this is assuming wildly unlikely events will take place -- like DC's only House member switching to Republican hands -- so Democrats shouldn't get too worried about the lopsided numbers.

Conversely, if all of the one-vote states switched to Democratic hands (much more likely in the current political climate), then we'd have the best-case scenario for Obama. If all the one-vote states switched from Republican to either Democratic or a tie (it depends on whether the state delegation is odd or even in number), then the result would be: Democrats 40, Republicans 8, Tie 3. Obama wins in a landslide, 40-8.

Again, neither of these is likely to happen, at least not to this magnitude. The Democratic best case involves flipping such states as Utah and Wyoming, neither of which I'd bet the farm on. But you can see what a wide range of possibilities there are, just counting states that could flip parties by changing only one House seat. There are a whopping 34 of these states, 19 currently Democratic, 13 currently Republican, and two states that are tied. And there are a further 11 states that would flip by a change of two House seats, which I didn't even include in the best and worst case scenarios.

You can do the math, and make your own guesses as to which states are actually in danger of flipping. Factoring in Bush's approval ratings, the wave of new Democratic voters, and the fact that most of the country believes we're on the wrong track, and it looks like a pretty good year for Democrats. But we can't expect Democrats to hold every single one of their seats, as they astoundingly managed to do in 2006. Overall, the new Congress will be more solidly Democratic, but losing a key seat here and there might be decisive.

My armchair analysis of the state races, which I must admit I didn't do much research on (for each individual House race) shows that the House vote could be closer than expected. Call it an educated guess. In my off-the-cuff scenario, twelve states flip. The Democrats win an extra seat in Alaska, Missouri and New Mexico. The GOP picks up a seat in Mississippi, changing their vote to a tie. And the Republicans pick up a majority with a single seat switch in Arizona, Indiana, Louisiana, North Dakota, South Dakota, Tennessee, and West Virginia.

My choices are arguable, at best, I realize. Eyebrows might be raised about calling Alaska and Louisiana at-risk, but in both states there have been local scandals which could cause a change. I chose other states for other reasons, and no doubt you can come up with a believable alternative to what I've laid out. But the scary part is that this scenario winds up: Democratic 25, Republican 24, Tie 2. And that's counting on a win for Democrats in Alaska.

Meaning, in the unlikely event of a 269-269 Electoral College tie, the subsequent vote in the House could be a lot closer than Democrats may think. Once again, even the Electoral College tie scenario is a longshot. So be aware that we're speculating on top of speculation here. But if a tie does indeed happen, the President of the United States may be chosen by a single House district changing hands in some obscure state.

[Below is the data I used, getting delegation counts from the House of Representatives' website. "Solid" states are those which have more than two votes in the majority. For the one-vote and two-vote difference states, the states [in brackets] would change from one party to a tie. The states without brackets would change from one party to the other. Only states with an even number of House delegates have the possibility of a tie.]

Solidly Democratic House delegations:
CA, MA, NY

Solidly Republican House delegations:
FL, TX, VA

States currently tied:
AZ, KS

One-vote difference states:

Democratic:
[AR], CO, [HI], IN, IA, [ME], [MN], [MS], [NH], NJ, NC, ND, [RI], SD, TN, VT, WV, [WI], DC

Republican:
AK, DE, GA, [ID], [KY], LA, MO, MT, NV, NM, [SC], UT, WY

Two-vote difference states:

Democratic:
CT, IL, MD, OR, PA, WA

Republican:
AL, MI, NE, [OH], OK

Cross-posted at The Huffington Post

-- Chris Weigant