FanPost

My computer didn't hate the Cards' bye week

Most of the computer rankings used by the BCS aren't published this early in the season, and with good reason. There isn't enough interlinking of schedules via common opponents for a prejudice-free system to reach reasonable conclusions.

My rankings last week, for example, had the top 5 as Baylor, Florida State, Oregon, Oklahoma State, and Washington in that order. The PC (score-independent) version had a top 5 of UW-Whitewater, Clemson, Alabama, Florida State, and Washington. (The previous week it had LSU, OK St, Southern Utah, Troy, and Washington). But things are beginning to settle in.

Here's my points-based top 10 after yesterday's games:

 Rank  Rating     SOS Team                         PF   PA    W    L
---- ------ --- ---- -- -- -- --
1 90.37 28.66 Baylor 209 23 3 0
2 78.92 30.22 Oregon 239 43 4 0
3 71.18 42.60 Washington 159 43 4 0
4 70.98 35.08 Florida St 205 60 4 0
5 68.05 27.07 Louisville 192 27 4 0
6 65.64 45.60 Alabama 140 58 4 0
7 63.77 38.40 Clemson 172 69 4 0
8 63.45 32.57 Ohio State 241 85 5 0
9 63.37 28.99 UCLA 158 54 3 0
10 63.17 54.97 Georgia 165 130 3 1
Louisville is up from 7th last week.
And in my score-blind ratings, where we moved up from 35th to 19th, the top 10 looks like this:
 Rank  Rating     SOS Team                          W    L
---- ------ --- ---- -- --
1 2.42 1.44 Alabama 4 0
2 2.26 1.78 Georgia 3 1
3 2.18 1.20 Florida St 4 0
4 2.14 1.15 Clemson 4 0
5 2.02 1.03 Washington 4 0
6 2.01 1.02 Stanford 4 0
7 1.88 0.89 Texas Tech 4 0
8 1.87 0.88 Fresno St 4 0
9 1.83 0.84 Oklahoma 4 0
10 1.83 1.34 South Carolina 3 1
Like I said, things are beginning to settle in. The rankings look more reasonable every week.

I don't make any claims to accuracy as far as who's best, especially at this time of year, but my method is nothing if not fair. I feel it accurately reflects the rating a team has earned. The method is similar to the one I use in basketball. A team is rated by its points scored and points allowed or by its won-loss record, added to 99% of its strength of schedule (which is the average of its opponents' ratings), computed iteratively until the results converge on a solution.

Anyway, I just thought it was interesting that our close game with Bye Week didn't drop us in either of my rankings. We rose in both - significantly in the score-independent one, although still not high enough in my opinion to have any shot at the NCG. For that we need a lot of things outside our control to fall into place; but if the BCS computers track anywhere close to mine, we still have a chance.