The primary asset of the Guru's system is its predicative value. Instead of rating teams based on (immediate) past performances, the system seeks to predict how each team will perform from this point forward. A team's schedule thus has an overwhelming influence on a team's rankings.
And it makes sense. The Guru's system is designed to predict which two teams are most likely to end up in the BCS national championship game, not necessarily which are the best teams. It goes without saying that a BCS conference team that manages to go undefeated would be a prohibitive favorite to play for the BCS title.
A team playing a tough schedule and/or is in a tough conference will have less of a chance to go unbeaten - but that's balanced by the fact that such an arrangement may bolster its chances in case of a loss. Most of the BCS formula's idiosyncratic nature is taken into account for this system.
Testing the system for preseason 2008, it yielded a USC-Florida BCS title game. The Trojans didn't cooperate by losing at Oregon State; but the Gators, despite a home loss to Ole Miss, did make the title game - and went on to win it. Oklahoma, the other title-game participant, was fourth, behind No. 3 Georgia.
This year's preseason rankings forecast a rematch in the BCS title game, between Florida and Oklahoma. The Gators, in fact, nearly lapped the field. Florida has a talented team that returns quarterback Tim Tebow and the entire defensive starting lineup. It also doesn't have to play two of the SEC's better teams: Alabama and Ole Miss.
So without further ado, here's the Guru preseason Top 25. And this formula will be applied throughout the season, with the results released every Monday until the end of the regular season.
Rank | Team | | | | | | | | | | | | | | | RATING | Conf | Pvs* |
1 | Florida | 23.695 | SEC | 1 | ||||||||||||||
2 | Oklahoma | 11.484 | Big 12 | 5 | ||||||||||||||
3 | USC | 11.348 | Pac-10 | 3 | ||||||||||||||
4 | Texas | 10.134 | Big 12 | 4 | ||||||||||||||
5 | Ohio State | 6.322 | Big Ten | 9 | ||||||||||||||
6 | Penn State | 5.837 | Big Ten | 8 | ||||||||||||||
7 | Virginia Tech | 3.601 | ACC | 15 | ||||||||||||||
8 | Pittsburgh | 3.491 | Big East | 27 | ||||||||||||||
9 | Boise State | 3.244 | WAC | 11 | ||||||||||||||
10 | Notre Dame | 3.003 | Ind | NR | ||||||||||||||
11 | Alabama | 2.594 | SEC | 6 | ||||||||||||||
12 | Texas Christian | 2.172 | MWC | 7 | ||||||||||||||
13 | Gerogia Tech | 1.903 | ACC | 22 | ||||||||||||||
14 | Utah | 1.740 | MWC | 2 | ||||||||||||||
15 | North Carolina | 1.293 | ACC | NR | ||||||||||||||
16 | Oklahoma State | 1.239 | Big 12 | 16 | ||||||||||||||
17 | Oregon | 1.155 | Pac-10 | 10 | ||||||||||||||
18 | Ole Miss | 1.106 | SEC | 14 | ||||||||||||||
19 | California | 1.105 | Pac-10 | 26 | ||||||||||||||
20 | West Virginia | 0.814 | Big East | 23 | ||||||||||||||
21 | Brigham Young | 0.662 | MWC | 25 | ||||||||||||||
22 | Michigan State | 0.647 | Big Ten | 24 | ||||||||||||||
23 | Iowa | 0.501 | Big Ten | 20 | ||||||||||||||
24 | LSU | 0.196 | SEC | 28 | ||||||||||||||
25 | Oregon State | 0.193 | Pac-10 | 18 | ||||||||||||||
26 | Cincinnati | 0.147 | Big East | 17 | ||||||||||||||
27 | Georgia | 0.139 | SEC | 13 | ||||||||||||||
28 | Kansas | 0.071 | Big 12 | 37 | ||||||||||||||
29 | Illinois | 0.033 | Big Ten | NR | ||||||||||||||
30 | Florida State | 0.013 | ACC | 21 | ||||||||||||||
30 | Nebraska | 0.013 | Big 12 | 29 |
*Pvs- Final 2008 AP poll
3 comments:
How about posting more details of how this ranking system works? I for one would be interesting in checking things out under the hood, so to speak.
I notice that the ratings add up to 100. (Well, 99.895 for the 30 teams listed, provided I didn't hit a wrong button on my calculator.) Does this then imply that you give Florida a 23% chance at winning the BCS title game, and a 57% chance that it will be either Florida, OK, USC or TX? And there is a 4.5% chance that a MWC school wins it all???
You are not a guru if you think Notre Dame is the #10th best team in the country this year...
Clark- Thanks for your interest and it seems like you're onto something.
I can't let you get under the hood (it's bolted), but your assumptions are pretty accurate, with this one caveat: It's not so much who's going to win the BCS title, but who's going to get in.
This system is still somewhat in Beta mode, as I'll need an entire season to verify its accuracy/usefulness. So, yeah, I am kind of like playing Microsoft, but I sure hope this thing doesn't turn out to be like Vista.
Post a Comment