To be added in tonight But, first, now the first-annual Meph rankings. There are four sorts of rankings to put out. The first, of which, are the Conference Ratings, for me. Note: I've been using these all season, just never wanted to post them until now. Note: For everything, 50.0 is average and the metric is measured in points. Conference Ratings: 1. 60.01 Big 12 2. 59.69 Atlantic Coast Conference 3. 56.47 Southerneastern Conference 4. 55.81 Big Ten 5. 53.30 Big East 6. 52.97 Pac 10 7. 51.13 Mountain West Conference 8. 42.56 Western Athletic Conference 9. 41.92 Mid-American Conference 10. 41.60 Independents 11. 40.70 Conference USA 12. 34.13 Sun Belt Thoughts: As easily seen, the Big 12 and ACC were considerably better than the other nine conferences in D1A. The SEC was number two going into the last two weeks, before the ACC pretty much handed their ass to them outside of Florida vs Miami and FSU. Note: The conference ratings themselves aren't a large part of the final rankings. I'm just averaging the teams in the conference to get an average rating for the conference. Now. If I had a vote, for top 25 including everything, not just how good a team is nor how much they accomplished, but a combination of the two, with the current AP ranks in parentheses. 1. 83.81 Oklahoma (2) 2. 82.50 Florida (1) 3. 79.58 Texas (3) 4. 77.80 Southern Cal (5) 5. 75.35 Penn State (6) 6. 74.84 Texas Tech (8) 7. 71.49 Utah (7) 8. 69.68 Ohio State (10) 9. 69.51 Alabama (4) 10. 68.77 TCU (11) 11. 68.43 Boise St (9) 12. 67.75 Missouri (25) 13. 67.17 Oklahoma St (13) 14. 66.42 Georgia Tech (14) 15. 65.72 Florida St (28) 16. 65.13 Georgia (16) 17. 64.97 Oregon (15) 18. 64.29 North Carolina (No Votes) 19. 63.44 California (31) 20. 63.08 Oregon St (24) 21. 62.68 Iowa (26) 22. 62.56 Boston College (27) 23. 62.50 Virginia Tech (21) 24. 62.21 Mississippi (20) 25. 62.10 Pittsburgh (18) Notable Absences: #28 BYU (#17), #29 Cincinnati (#12) These particular rankings were derived using only three things: WL record, Point differential, schedule. Nothing else was used. It's akin to the overall rankings that Sagarin posts on his site. Now, if we wanted to rank accomplishments on the field, meaning ignoring point differential and looking just at WL record and schedule. We have the accomplishment, or the best resume, stuff, argument. These rankings, are what, I would have to send to the BCS. 1. 80.40 Oklahoma 2. 78.37 Florida 3. 77.14 Texas 4. 76.23 Texas Tech 5. 75.00 Utah 6. 72.24 Southern Cal 7. 71.67 Penn State 8. 70.94 Georgia Tech 9. 70.26 Alabama 10. 69.64 Boise St 11. 69.60 Ohio State 12. 68.92 Florida St 13. 67.96 Georgia 14. 67.70 Cincinnati 15. 67.24 TCU 16. 67.22 Virginia Tech 17. 66.88 Boston College 18. 66.61 North Carolina 19. 66.13 Pittsburgh 20. 66.02 Oklahoma St 21. 64.38 Michigan St 22. 64.30 Missouri 23. 64.08 Brigham Young 24. 62.83 Nebraska 25. 62.50 Oregon Thoughts: Georgia Tech in the top ten? Yeah, that's what I thought. I was surprised to see Oklahoma State so low, myself. Weird. Alright, now for the pure prediction type rankings. These ones were derived only using point differential and schedule, nothing else. These would be akin to the Sagarin predictor. 1. 87.22 Oklahoma 2. 86.63 Florida 3. 83.36 Southern Cal 4. 82.03 Texas 5. 79.02 Penn State 6. 73.45 Texas Tech 7. 71.19 Missouri 8. 70.30 TCU 9. 69.76 Ohio State 10. 68.76 Alabama 11. 68.32 Oklahoma St 12. 67.97 Utah 13. 67.44 Oregon 14. 67.34 Iowa 15. 67.23 Boise St 16. 66.62 California 17. 65.45 Mississippi 18. 64.81 Arizona 19. 64.24 Oregon St 20. 62.52 Florida St 21. 62.30 Georgia 22. 61.98 North Carolina 23. 61.90 Georgia Tech 24. 61.19 Nebraska 25. 59.69 Clemson There is another incarnation of the rankings that's are a bit better, but i will go into these a little later. Anyways, OU is #1 in both and UF is #2 in both. Ought to be a great game. Now defending the method. Of the BCS computers, only one has been released with this weeks games included. So I wanted to check the accuracy of said methods. Alright, the average rankings (the ones that use WL and Point Diff) at predicting all 683 division 1 football games. Winner Accuracy 565 82.72% - Meph 559 81.84% - Sagarin Score Differential RMSE 13.61 - Meph 14.31 - Sagarin Now let's look at points only methods, sagarins predictor and mine. note these WILL be less accurate in predicting wins because they dont use wins, but are more accurate at predicting future matchups. this is all expected Winner Accuracy 81.11% - Meph 80.09% - Sagarin Score Differential RMSE 13.18% - Meph 13.60% - Sagarin Finally, the one the BCS requires. Only using WLs and who and where they played. Winner Accuracy 84.04% - Meph 83.46% - Sagarin Score Differential RMSE 14.87 - Meph 15.38 - Sagarin