Sunday, October 17, 2010

My Ballot

As you may have heard, the first BCS standings come out tonight. As you may have guessed, no one involved with the BCS in any capacity has asked me to submit a ballot. But why not be prepared?

The Ballot I Would Submit, If Anybody Asked:
1. Boise State
2. Oregon
3. Oklahoma
4. TCU
5. Auburn
6. Utah
7. Michigan State
8. LSU
9. Missouri
10. Oklahoma State
11. Arizona
12. Alabama
13. Iowa
14. Stanford
15. Florida State
16. Wisconsin
17. Ohio State
18. Nebraska
19. West Virginia
20. Nevada
21. K-State
22. Northwestern
23. South Carolina
24. Virginia Tech
25. East Carolina

My Method:
First Principle: Ability to get into the championship game, as determined by number of losses.
Exactly one two-loss team has made it to the BCS championship game, and that was a weird, weird year (LSU, 2007). And I'm still kind of annoyed about it. So I decided to include every unbeaten and one-loss team, regardless of conference. There are only 22 of them, so I did go ahead and try to pick the best three two-loss teams to round the ballot out to 25. (That was possibly the hardest part of the whole thing; more on that later.) I realized as I was ranking the teams that, as a corollary to this principle, all unbeaten teams should be ranked ahead of all one-loss teams. This makes some parts of the poll look a little wonky (do I believe that Oke State would beat Alabama on a neutral field? Not particularly). However, all that an unbeaten team has to do at this point is win all the rest of their games; one-loss teams have to do that and root for unbeatens to lose.

Second Principle: Head-to-head matchups.
If a team has beaten another team on the ballot, they are above the team they beat, end of story. That's the one nice thing about a ballot that comes out this early--there are not yet any "circles of death" where within a group of three or four, each team has beaten one of the others and lost to another one. I know that Wisconsin might not beat Ohio State six times if they played ten times on a neutral field or whatever, but they did win the one time they did play, and that's that. (Again, later in the season, this would get more muddied due to larger bodies of work, but it's still pretty straightforward right now.)

The answer to any complicated research problem: 3X5 notecards!

Third Principle: Body of work, so far
This is where it gets really muddy and subjective. Being a fake pollster isn't easy, y'all. I'll just say this: people like to kill Boise for their weak schedule, but it hasn't gotten weak yet. The combined record of Boise State's opponents is 16-22; Oregon's is 13-23 . . . and they've played an FCS opponent (2-3 Portland State), and Boise hasn't (and won't, ever). Yes, by the end of the season, Oregon will have played higher-quality opponents than Boise gets to, but they haven't yet.

This principle would have been a lot easier if I factored in margin of victory (particularly when comparing the non-AQs), but since that's not kosher, I tried to avoid it.

Fourth Principle: so, um, who would win on a neutral field?
AKA the eyeball test. I've seen West Virginia play, and I don't feel like they'd beat most the other teams on the list. Even though this principle sounds the sketchiest, it's theoretically my favorite--if only I could watch every game. My kingdom for a DVR and a ruinously expensive cable package!

Until then, I'll take comfort in my 'cards.

Fifth Principle: Wait, all these two-loss teams suck . . .
Principles 1-4 got me through the top 22, but then I had to come up with someone, anyone, to round out the 25. And none of them look that great. I was tempted to slap all three two-loss SEC teams in there, but I worried that was the media's pro-SEC bias talking. So I took what I considered the best of the SEC teams, the best of all the others (Va Tech has actually acquitted itself very nicely since the James Madison disaster, have you noticed?), and then East Carolina because they make me smile. I know that's a terrible reason, but that's why nobody asked me to submit a ballot.

No comments:

Post a Comment