The math behind what I’m doing with this is pretty simple – try to estimate the chances a team wins out, and thus the chances it plays for the “national” title. The hard part is deciding which teams fall where.
Oregon and Auburn are easy. The two teams are a strong 1-2 in the BCS, and if they win out they will play each other. It’s deciding who’s next in line if one of them should lose that’s tricky.
For this week’s model, I’ve come to a new conclusion that sharply increases Auburn’s chances of playing for the BCS title – I believe that if Auburn loses “respectably” to Alabama, and wins the SEC title, they will still play for the BCS title ahead of any other team. Here’s why –
-Auburn is a close second in the human polls, and a lock number 1 in the computers.
- Auburn, after playing Alabama, will have the nation’s toughest schedule. This is before playing for the SEC title game.
The hard part was including one loss Auburn in the spreadsheet I’ve created. My solution was to “create” a second team I called “Auburn II”. Auburn II is a team that loses a respectable game to Alabama then wins the SECC. (More on this in a moment)
Also, I did not model a title game between TCU and Boise State. In my opinion, the human voters simply won’t let it happen.
Next, I’ve virtually eliminated the Big 10 teams from any contention. Wisconsin, OSU and Michigan State are 7th, 9th and 12th in the BCS, and their computer rankings are even worse (12th, 13th and 10th, respectively). There’s simply not enough games left to be played for them to jump up.
Lastly, I still think some Big 12 team has a shot. Nebraska is 8th overall (and 8th in the computers), and Oklahoma State, though 10th overall, is 6th in the computers, tied with Boise State. I think the Cowboys are a long shot, but if they win out the humans should come around, and with a couple of key losses they could make it.
With Auburn & Auburn II combined, here are the estimated chances of a team playing for the BCS title -
74.62% | |
55.86% | |
TCU | 43.88% |
18.00% | |
LSU | 3.73% |
1.87% | |
"Other" | 1.28% |
0.77% |
“Other” in this case is mostly 6th ranked Stanford, who would need 2 Oregon losses, so I didn’t model them specifically.
And here are the individual matchup choices. AUII is a 1-loss Auburn team –
TCU-ORE | 24.90% |
AUII-ORE | 24.18% |
BSU-ORE | 13.37% |
TCU-AUII | 11.43% |
AU-ORE | 10.20% |
TCU-AU | 5.59% |
BSU-AUII | 2.34% |
LSU-TCU | 1.16% |
LSU-ORE | 1.16% |
BSU-AU | 1.10% |
LSU-BSU | 0.70% |
0.58% | |
NEB-TCU | 0.57% |
LSU-AUII | 0.42% |
NEB-BSU | 0.34% |
0.24% | |
OK ST- ORE | 0.24% |
NEB-AU II | 0.21% |
LSU-AU | 0.18% |
OK ST-BSU | 0.14% |
0.09% | |
OK ST-AU II | 0.09% |
NEB-LSU | 0.08% |
0.04% | |
OK ST-LSU | 0.03% |
As for Oklahoma State – LSU, “So you’re saying there’s a chance?”
3 comments:
'As for Oklahoma State – LSU, “So you’re saying there’s a chance?” '
Does your spreadsheet account for the "Les Miles Effect", by which LSU's probablity of victory increases exponentially in proportion to the number of dumbass decisions he makes?
I stand corrected.
LSU 100%
Beware mocking the powers of the hat, my friend.
Post a Comment