or a participant in the BCS title game in the event of a tie.
This method tries to give more credit for the "good" wins, while at the same time penalizing a team for a "bad" loss.
The method is described in full detail after the jump.
This method gives positive points for a win and negative points for a loss. It attempts to factor in the quality of opponent (the better the opponent in a win gets more positive points, and the worse an opponent in a loss gets more negative points). You could even have margin of loss factored in (I think margin of loss (personal opinion) means more than margin of victory) if you want to.
Basically an overly simplistic way would be to take a ranking system (Sagarin for example) sorted in the usual fashion from from best to worst (regular sagarin) for the losses and the same list sorted worst to best (reverse sagarin) for the wins. Add up the rank of all the teams for wins and subtract the rank of all the teams for losses.
Here are some teams of interest that I have already done this for (Bama, OkieSt, KSU, and OU):
Bama wins (reverse sagarin):
Kent State +122, Penn St +222, N Texas +118, Ark +240, Florida +214, Vandy +208, Miss + 139, Tenn + 196, Miss St +216, Ga Sou + 174, Auburn +213
Bama loss (regular sagarin):
Bama’s "Jeremy’s end of regular season" ranking: 2061
OSU wins: (assuming a win over OU)
UL Laf +160, Ariz +184, Tulsa +220, aTm +235, KU +159, UT +233, Mizzou +229, BU +231, KSU +236, TTU +203, OU +243
OSU’s "Jeremy’s end of regular season" ranking: 2304 (interestingly, OSU’s score NOT including OU at all is 2061, so losing to OU would drop the score to 2057)
KSU wins: (assuming a win over ISU)
EKU +92, Kent St +122, Miami +200, BU +231, MU +229, TTU +203, KU + 159, aTm +235, UT +233, ISU +218
OU -4, OSU -3
KSU’s "Jeremy’s end of regular season" ranking: 1915
OU wins: (based on the above assumption that OSU beats OU)
Tulsa +220, FSU +221, MU +229, Ball St +145, UT +233, KU +159, KSU +236, aTm +235, ISU +218, OSU +244
TTU -44, BU -16, OSU -3
OU’s "Jeremy’s end of regular season" ranking: 1833
I developed this the other night, and I haven't made any changes to the mathematics of how this works, but I do now see one small problem with this method. It really only works to compare teams that have played an equal number of games. For instance, if UGA beats LSU in the SEX title game, then using this method to determine how LSU stands up to Bama and OSU is not really fair, because LSU would have one more game with which positive points were gained to help their cause (similarly how winning 20 games in a basketball season in 2011 is "easier" than winning 20 games in a season in 1960 because more total games are played these days).
That being said, if you are trying to determine whether Bama or OSU is the more deserving 1 loss team to play in the BCS title game, or who is the "true champion" in a 3-way tie for the Big 12 regular season among 7-2 OU, 7-2 OSU, and 7-2 KSU, I think this method does a pretty good job.
Maybe since this is a fanpost it will attract more comments than my original comment did in the 11-29 slate.
I know this isn't perfect (or somebody else would have already implemented something like this), and if this was done for all 246 FBS and FCS teams, there might be a glaring reason why this is NOT used. It is a rather slow process for me to do for a single team (takes me about 5 minutes per team to look up everything and do the math), or I would have done the remaining teams in the Big 12 to get a slightly larger sample.
Update Dec 2, 2011: As BlackCats pointed out in another thread where I posted something about this, if this method is going to be used to determine the "true" champion of a conference that has no championship game (Big 12), then you would not add in the points earned (or subtracted) from the non-conference games. You would only perform this exercise for the conference games. You would also only include the conference games to determine the team that should represent a division in the championship game (for example to settle the 3 way tie in the Big 12 South that occurred last season).
I welcome all comments and criticisms.