Perhaps the BEST B1G Forum anywhere, here at College Football Fan Site, CFB51!!!

Baylor, huh?

It's hard to mesh eye test with resume evenly, if that's even a goal for someone.  There are sold and/or impressive wins, but there are also close wins vs weak teams.  There are close losses to strong teams, even multiple games like that and more blowouts vs cupcakes than is comfortable.



I'm not criticizing anything, no one should without posting their own rankings.  But for Baylor:
3 pretty solid wins
vs
close wins vs Rice, ISU, TxTech, WV



Back to the abstract - what's "better" - the 3 point win vs WV or the 3 point loss to an Oregon?  The goal is to win the game, but with all of the variables and lack of cross-section data means we have to value this stuff on the same plane.  Meh, I type all this and we all know it already.

I like computer rankings because they don't have to feel good about it, lol.  Many, many times, the .500 team is better than the 10-1 team and would beat them 6+ times in 10 games.  But resume cannot be ignored.  


The worst sin is being a slave to W-L record (cue my annual post about lazy voters).

The BCS computer fiasco was what happens when ignorant people dabble in statistics.  The BCS folks took the computer poll people's best efforts and forced them to change in order to avoid hurt feelings (of running up the score).  Brutal stupidity.



I'm not a fan, however, of rankings that reward you for who you play more than the outcome of the game - a la moving up in the rankings because you lost to a good team.  I understand not punishing a team in that event but to reward them....nuh uh.

i'll take a stab at it

my version:

• osu
• lsu
• bama
• clemson
• ou
• psu
• oregon
• utah
• uga
• baylor
• florida
• auburn
• wisconsin
• minnesota
• michigan
• notre dame
• iowa
• cincinnati
• memphis
• kansas st
• wake
• smu
• navy
• mizzou
• ok state

what i think cfp committee will do:

• lsu
• osu
• bama
• clemson
• psu
• baylor
• ou
• uga
• oregon
• utah

why not?

Why the square root of points differential?

The BCS eliminated MoV for two reasons:

• So as not to reward teams for piling points on hapless opponents, and
• So as to prevent a team from compensating for a ridiculously weak schedule by blowing out their opponents

IMHO, the BCS's solution of completely eliminating MoV was a step too far.  MoV DOES matter.  However, I do agree with both of the motivations listed above so what I advocated was a modified MoV.  Here is how I would calculate MoV if I ran a computer poll:

I would take the sum of the point differential at halftime, the point differential at the end of the third quarter, and the final point differential and award that as the MoV to the winning team (and MoD or Margin of Defeat) to the losing team with the following caveats:

• The point differential at each check (halftime, end of third quarter, and end of game) would be capped at 21
• In any game in which the winning team has a negative MoV, they are awarded a 1 point MoV and the losing team gets a 1 point MoD
• In any game that goes to OT, the winning team is awarded a 1 point MoV and the losing team gets a 1 point MoD

I think this takes what the committee has typically referred to as "Game Control" into account and I think that matters.  If you lead a team by at least three TD's at halftime, the end of the third quarter, and the end of the game, you dominated and you get a 63 point MoV but winning by more than that doesn't make it any better.  The maximum MoV/MoD is 63/-63.