The New Mo Knows Ranking System

During the course of this week I developed an objective ranking system. While it still has flaws in it, it remains objective as it is entirely rule-based. The only subjective part is that I decided which rules to use. It will be a bit complicated to explain how the rules work, but it is simple to apply them, trust me, I wouldn't develop something too complicated to keep tabs on.

First, I only use teams that have better than .500 records. If a loss drops you to .500, you drop from the rankings
Second, the preseason rankings were determined as follows: I took all of the teams from last season that finished 7-6 or better and sorted them by winning percentage. Kansas and Hawaii were both on top, by virtue of their 12-1 records. Then I sorted within the winning percentage groups by teams that won their bowl games and teams that lost, placing teams that won ahead of teams that lost their bowl games. Kansas won their bowl game, getting the nod over Hawaii who lost. Then I checked for any head-to-head matchups to distinguish teams with identical records and similar bowl results. If Kansas and Hawaii had both won their bowl games, but had played during the season, the winner would be placed ahead of the loser. At that point I placed them in order of strength of conference, as determined by the final Sagarin Conference rankings from last season. Back to Hawaii and Kansas, say they were both 12-1, won their bowl games, and did not play each other. Kansas would have been ranked over Hawaii because the Big 12 was rated higher, as a conference, than the WAC. I hope that made sense.
Third, if a team wins, they can only maintain their ranking or move up, there is no moving down because Georgia's win over Georgia Southern wasn't as impressive as people wanted it to be or because Texas Tech's win over Kansas State was more impressive than BYU's win over Utah State (which it was, but that doesn't play a role here, a win is a win). You can jump a team based on rules to be explained later, but you cannot jump them if it moves them down in the rankings.
Fourth, if a team loses, they will be segregated from the rankings, all of the teams that won will move up and the loser will be placed at the bottom of the section containing teams with the same number of losses as them. For instance, South Florida lost, to go to 5-1. They are placed at the bottom of the 1-loss teams that won or had a bye that weekend. Teams that lose retain their ordering from their previous spot in the polls, i.e. if number 5, 10, and 15 all lose for the first time, they move to the bottom of the 1-loss teams, with number 5 staying ahead of 10 and 15, unless number 10 or 15 have more wins than number 5 (i.e. if number 5 loses to become 4-1 but number 10 and 15 both lost to become 5-1, they would be ranked ahead of the 4-1 loser).
Fifth, if a team wins to push them over .500, they are re-inserted into the rankings at the bottom of the rankings, placed ahead of teams that lost the same weekend, giving them the same number of losses. So West Virginia won to go to 3-2. They took their place at the bottom of the current rankings and Air Force, who lost to go to 3-2, was placed below West Virginia.
Sixth, if you win to go ahead of .500 and it is your first time to be ranked in the season, you will be placed below teams that win to go ahead of .500 who have previously been ranked during the season. If there is a tie-breaker, the previous season's record will be used to determine order of teams entering the rankings for the first time.
Seventh, starting after week 6 (so this upcoming weekend this will apply), if a team beats someone ranked in the top 10, they can jump teams ranked ahead of them, provided they have the same number of losses or only 1 less, until moving up would require rule 3 to be broken. For instance, if Kentucky (4-1 currently) were to beat undefeated Vanderbilt this weekend, they would move ahead of all 1-loss teams, and ahead of all undefeated teams until they arrived at a point where someone would have to move down. If a 2-loss team beat a top 10 team, they would move ahead of all 2-loss teams, and all 1-loss teams until moving them up would cause someone to move down in the rankings.
Eighth, starting after week 6, if a team beats a team ranked between 10 and 20, they can move ahead of all teams with the same number of losses as them. So a 1-loss team beats the 13th-ranked team, they would move ahead of all other 1-loss teams until rule 3 would have to be broken to move them up any higher. The rationale behind these last two rules is: by the 7th week teams in the top 10 and 20 have established themselves as teams capable of winning ball games. Teams are already in conference play, by a game or two or three, so if a team has managed to continue winning and stay in the top 20, they actually are a pretty good team. If this rule applied starting in week 1, Florida would have vaulted to the front of the back because they beat a very bad Hawaii team. These are the only two rules that have a strength of schedule element to them. Losing to a team in the top 20, no matter how close the game is, do not improve your ranking in any way. Anybody can lose to a top 20 team. Also, losing to an unranked team does not hurt any more than losing to a ranked team.

There are several shortcomings in this system, I realize. For instance, there is not an element to rank teams based on head-to-head matchups. If Florida and Ole Miss were to have the same record later in the season and Florida was ranked ahead of Ole Miss, there is nothing to place Ole Miss ahead of Florida. I tried to come up with a way to include this but there were too many complications with having a head-to-head element. Another issue is that an early season loss is less damaging than a late season loss. I recognize this but I also thing underlying issues might actually favor this system: a loss later in the season is probably due to playing a better team, a rival, or at least a conference opponent and so you lost. A loss earlier in the season is because your team was breaking in new starters and trying to gain experience. Now after your team has gotten experience, losing, even to a better opponent, should be more damaging to your ranking because it means you weren't as good later in the season once all of the early season mistakes were ironed out. If anyone has any suggestions for improving the system, let me know.

There are a lot of plusses to it though. For one, preseason hype can't carry you through the season. If you lose, you go to the back of the line. But doing well in consecutive seasons can help you, since being ranked higher earlier in the season means you get to stay ahead of teams with similar records to you that lose the same week you do. But if you didn't do well last season, it doesn't matter too much. If you win this season, you are rewarded for doing so, even if your program isn't viewed as a higher echelon program. It is decided on the field of play, a novel concept for college football, I know. If you win, you move up the ladder.

Clear as mud yet? Good. Without further ado:
1. LSU (4-0)
2. Missouri (5-0)
3. BYU (5-0)
4. Oklahoma (5-0)
5. Texas (5-0)
6. Boise State (4-0)
7. Tulsa (5-0)
8. Utah (6-0)
9. Penn State (6-0)
10. Texas Tech (5-0)
11. Alabama (6-0)
12. Oklahoma State (5-0)
13. Ball State (6-0)
14. Northwestern (5-0)
15. Vanderbilt (5-0)
16. Michigan State (5-1)
17. Virginia Tech (5-1)
18. Western Michigan (5-1)
19. Kansas (4-1)
20. Ohio State (5-1)
21. California (4-1)
22. Georgia Tech (4-1)
23. Arizona (4-1)
24. Boston College (4-1)
25. Cincinnati (4-1)
26. Pittsburgh (4-1)
27. Florida State (4-1)
28. North Carolina (4-1)
29. Notre Dame (4-1)
30. Georgia (4-1)
31. Minnesota (5-1)
32. Wake Forest (3-1)
33. Florida (5-1)
34. TCU (5-1)
35. USC (3-1)
36. South Florida (5-1)
37. Connecticut (5-1)
38. Kentucky (4-1)
39. Central Michigan (3-2)
40. San Jose State (3-2)
41. South Carolina (4-2)
42. Navy (4-2)
43. Clemson (3-2)
44. Arkansas State (3-2)
45. East Carolina (3-2)
46. Nevada (3-2)*
47. Illinois (3-2)*
48. West Virginia (3-2)*
49. Colorado State (3-2)*
50. Maryland (4-2)
51. Auburn (4-2)
52. Oregon (4-2)
53. Duke (3-2)
54. Fresno State (3-2)
55. Air Force (3-2)
56. Kansas State (3-2)
57. Wisconsin (3-2)
58. Colorado (3-2)
59. Nebraska (3-2)
*new entry this week
Dropped from the rankings:
32. Louisiana Tech (3-3)
51. Ole Miss (3-3)
53. Rice (3-3)
56. Stanford (3-3)
58. Marshall (3-3)
59. UNLV (3-3)
62. Iowa (3-3)