Sunday night, I referenced the possibility that Florida might win Saturday and still finish behind Texas in next week's BCS standings because of the Longhorns' huge lead in the computer polls, setting up an Oklahoma-Texas mythical championship rematch if the Sooners beat Missouri. This fell into the category of "unlikely but interesting" then, and after reading an assessment in my inbox today from Zach Rosenfield of the stat site AccuScore, it seems even more unlikely -- but no less interesting, by Zach's account:

I actually have access to the BCS spread sheet and have been able to plug in multiple scenarios. Texas will not be able to hold off Florida on the computers alone. But their margin in the computer section of the polls means that this is going to get interesting and will come right down to it.

It breaks down as follows. Oklahoma and Alabama win and they are in. That much we know. But here is how Florida can get left out. If OU wins big and Florida struggles, then it gets murky. A big win by OU coupled with Florida struggling could give OU a overwhelming majority of No. 1 votes and could move Texas up to No. 2, or in some cases, No. 1. It is only a matter of time until Mack Brown shifts the discussion to quality losses: UT in the last minute on the road and UF losing at home to Ole Miss. Whatever it takes to creep into the voters minds.

If Texas stabilizes in the human polls, then it gets interesting. Particularly because the only way UT can stabilize is if Florida struggles. If Florida wins big, then the UF will get a bunch of first place votes and be in. But if the Gators struggle, the votes get divided and it gets murky. Very murky.

A Florida win will in all probability increase its computer score from .82 to .88
The magic number is .944 for the Gators. Florida needs to average .944 between the Harris and Coaches Poll to make up the computer defecit. That means they need to go from 2619 points in the Harris to 2669. In the Coaches, they need to go from 1385 to 1439.

To put those numbers in perspective and roughly estimate how many points Florida can expect to pick up for beating Alabama, we can compare some recent historical precedents for other teams making major one-week moves.

There weren't many car-crash moments from 2002-2005, when the teams at the top of the standings tended to get there and stay there, but both Florida and LSU in 2006 and 2007 made the jump Florida is trying to make again Saturday by winning in the SEC Championship, and just to add more data points and get a sense of how the poll has seized and shook at times this year, I've added the point totals for each of the three Big 12 South frontrunners for the weeks before and after their big wins over a previously undefeated rival (as you know, that's over No. 1 Oklahoma for Texas, over No. 1 Texas for Texas Tech and over No. 2 Texas Tech for Oklahoma). According to each of the human polls and the average of the BCS computers (not the final BCS score itself, but the computer score), this is what a big jump in the polls looks like, by the numbers:

Going back to Zach's estimates, Florida needs to pick up 50 points in the Harris and 54 points from the Coaches, both well, well below not only the average gains from big wins the last three years, but well below even the lowest gains on the chart. If all the Gators are looking to do is to pick up roughly 50 points per poll, recent history suggests they have no reason to worried about style or running it up. Unless voters launch a full-fledged, conscientious campaign in Texas' favor -- the odds of either poll coordinating anything like that are probably somewhere near zero -- all Florida has to do to pass Texas is win.