BCS Odds for Notre Dame, Alabama, Georgia, Florida, Oregon, Kansas State, and FSU

Two weeks of football are left in the 2012 season, and after last weekend's carnage 7 teams still have a measureable chance of getting into the BCS championship game.

What are those odds? We looked at each team's remaining schedule and their current BCS ranking and computed rough odds for their ability to finish in the top two. For the game-winning odds we used our Strength power rating and its comparisons of offense and defensive strength, and we looked at all the relevant scenarios that could affect the team's placement. Note that there is still a lot of art rather than science in determining exactly what the pollsters will do, but we gave it our best shot.

Here is the current (11/17/2012) BCS top ten:

1. Notre Dame (11-0)
2. Alabama (10-1)
3. Georgia (10-1)
4. Florida (10-1)
5. Oregon (10-1)
6. Kansas State (10-1)
7. LSU (9-2)
8. Stanford (9-2)
9. Texas A&M (9-2)
10. Florida State (10-1)

And here is our take on the teams that have the best shot at finishing in the top two, in order:

Alabama	66%	The Crimson Tide have to do one thing: win out. If they beat Auburn and win the SEC title game, they're in. Beating Auburn is almost a foregone conclusion; we give the tide a 99% chance to beat the Tigers. The SEC championship game is a bit harder. We give them a 2/3 chance of beating Georgia in Atlanta. Dock them 1% for the chance to lose to Auburn and Alabama has a 66% chance to be in the BCS title game.
Notre Dame	55%	Notre Dame has the clearest path to the to BCS championship: Beat USC. If they fail that, they have perhaps the most convoluted path to getting back to #2. The game at USC is pretty close to a tossup by our measures; we give the Irish just over a 50% chance to win and finish #1 in the BCS standings. If they lose, they aren't quite out of it, as they would be another of the 1-loss teams and still strong in the computers. But since "backing in" to the title game is frowned upon it would make the pollsters loathe to keep them too close to the top. It's hard to say how far they'd fall—and that would be affected by the score of the USC loss—but they'd probably need losses by Florida, Oregon, and Kansas State, which adds up to under a 10% chance to happen, so that adds just under 5% to their odds, which finish at 55%
Georgia	25%	Georgia controls their own destiny, too, one of only three teams to do so. To get to the championship game the Bulldogs have to handle in-state rival Georgia Tech, then beat Alabama to win the SEC. Beat Georgia Tech is the easier part of the job, and our numbers give them almost exactly 75% winning odds against the Yellow Jackets. Against Alabama they're aided by partial home field advantage by playing in-state in Atlanta, but still have just a 1 in 3 shot to win that game. 75% x 1/3 = 25%, or a 1 in 4 chance that Georgia will be playing for the title.
Oregon	22%	Now it gets complicated, as we are past the teams that control their destiny, and since they need help from other teams the calculations get more difficult (and the outcomes less exact). Oregon is at #5 currently and has one game left against Oregon State—unless UCLA beats Stanford, in which case the Ducks would be back in the Pac-12 title game. Beating Oregon State on the road is about a 2/3 proposition. 50% of the time—the odds we give UCLA at home against the Cardinal—they'll play UCLA, and beat the Bruins 80% of the time. This gives them a 60% chance of emerging unscathed. Note that it would benefit Oregon to have the extra game, despite the fact that it gives them another chance to lose, they're going to need all the help they can get with the computers and pollsters. Winning out alone doesn't help if Notre Dame and Alabama win out. If the Irish lose they still need to pass Florida, and the best way is for the Gators to lose to FSU. If the Irish win, they need all 3 SEC teams to finish with 2 losses, which means FSU still has to win, Georgia Tech needs to beat Georgia, and Georgia has to upset Alabama. By far the more likely scenario is the Irish losing/FSU winning, which is about a 1 in 3 chance. Overall there's about a 36% chance that enough things go their way; x the 60% chance of winning out = roughly 22% chance of Oregon making it back to the title game, most likely against the SEC winner. Note that there's no guarantee Oregon will pass a 1-loss Irish team that narrowly lost to USC, or stay ahead of FSU if they beat Florida. That's why they should root for UCLA to beat Stanford.
Florida	18%	Though they're at #4 right now, Florida has a tough road ahead, and even if they do their best by beating Florida State, they won't pass the SEC winner if both win their final regular-season games. That makes a Notre Dame loss their best shot. The Gators play at Florida State, and it's going to be tough to get a win—about a 1 in 3 chance as we see it. That already cuts their odds down quite a bit. Their scenarios of hope basically require either 1) Notre Dame loses and the Gators face the SEC winner, or 2) Georgia Tech upsets Georgia and Georgia beats Alabama, leaving Florida to face Notre Dame. Scenario #1 has about a 50% chance of happening, but Florida would probably have to dominate FSU to make the pollsters want to have another all-SEC title game. Scenario #2 requires a Georgia loss (1 in 4) followed by a Georgia upset of Alabama. And this is after a required win over Florida State. It's rough for Florida that their best odds are for another all-SEC title game, which the pollsters might rebel against. But Florida is #2 in 5 of 6 computers, and they'll go to #1 if Notre Dame loses. That gives them a good chance as long as the pollsters aren't so sick of the SEC that they move someone else ahead.
Kansas State	10%	Kansas State fell from #1 to #6 after their loss to Baylor, and it's going to take some doing to get back in the top 2. But it's not impossible. First, of course, they have to beat Texas, which we see as a 2/3 chance (a lot of 2/3 chances in these game-odds, it seems). Then, they need to clear out all but one of the teams ahead of them. They'll either need Notre Dame, Oregon, and Florida to lose, hopefully putting the Wildcats in the championship game against the SEC winner. Or Georgia must lose to Georgia Tech, and upset Alabama, while two of the other three lose. That doesn't guarantee them the #2 spot, but it's close enough if you add in the slim possibility that all of the above end with 2 losses. Oregon losing another game is a 40% proposition; Notre Dame losing is about 50/50. Already down to a 20% chance—and 13.33% including their own game. They also need EITHER Florida to lose to FSU (67%) or Georgia to lose to Georgia Tech (25%). The odds of one of those happening is 75%, which we multiply by the 13.33% to get 10%. And that's as far as we're going to figure it.
Florida State	4%	#10 Florida State is in a terrible position with the computers, ranking 17th in the composite. They aren't poorly positioned in the polls, though, and a win over Florida would boost their standing quite a bit with most computers. That win alone would allow them to pass the 2-loss teams at #7, 8, and 9, which is good because the ACC title game isn't going to help them much at all. We're giving FSU a generous 2/3 chance to beat Florida in Tallahassee. Then the question is, who is their ACC opponent? It could be Miami, if the Hurricanes beat Duke and decide to play in the post-season. Otherwise UPDATE: Miami has banned itself from the post-season, so it will be Georgia Tech. Both are about 85% wins for the Seminoles, so overall FSU has about a 57% chance of finishing 12-1. Then they need help, and lots of it, to have a chance to pass Kansas State, Oregon, and either Notre Dame or the SEC winner. Florida would be out of the way, and either Georgia or Alabama will lose, too. So if Kansas State loses to Texas (1 in 3), AND Oregon loses a game (40%), AND Georgia Tech upsets Georgia AND Georgia upsets Alabama, FSU would be the only viable 1-loss team and hold a win over Florida! That could get them a date with Notre Dame. Otherwise, K-State, Oregon, and Notre Dame could lose, and FSU would have to hope to jump the Irish to face Alabama/Georgia. All of these scenarios are so inexact—and FSU has such a long way to go—that the 'Noles need every bit of help. Clemson beating South Carolina (55%) would help FSU with the computers, and it wouldn't hurt for Stanford to lose again even.

That all adds up to 200%, leaving no room for the real long shots such as LSU, Stanford, Texas A&M, and maybe even #11 Clemson, to have a crazy run to the top if the remaining weekends are both like last weekend or we have a repeat of 2007. Numbers are rounded off so there's still hope for those not mentioned, but not a lot.

I haven't calculated the exact odds for every matchup, but using the numbers above it's pretty easy to figure out the 5 most likely pairings, noting that an Alabama-Georgia pairing isn't possible:

1. Alabama vs. Notre Dame
2. Alabama vs. Oregon
3. Notre Dame vs. Georgia
4. Alabama vs. Florida
5. Alabama vs. Kansas State

Comments

[Crayton]

Florida can not win the SEC East. Georgia and Florida already finished conference play at 7-1, with Georgia winning the head-to-head tie-breaker.

[SportsRatings]

Thank you for mentioning that...I've fixed things. I blame making this post at 2am for my brainfart. Now the numbers are actually a better fit, too.