Inspired by the Transitive Victories fanpost, I figured I'd post my pet project.  It's a computer point system that factors in margin of victory/defeat.

The teams are listed by average game score which is:  

Winning Team= {.6+(location)+(margin of victory}*{opponent's average margin of victory}*{opponent's opponent's average margin of victory}


Losing Team = {.4+(location)+(margin of defeat/100}*{opponent's average margin of victory}*{opponent's opponent's average margin of victory}

Location=+0.03 for away, -0.03 for home, 0 for neutral     margin of victory/defeat is capped at 40

In short, if you beat a good team by a lot, you get a lot of points.  If you lose to a good team by one, you still get a good amount of points.

I realize the problems involved;  mostly, the circular references and rewarding teams the run up the score. 

Below are the rankings:

Change is the number of spots the team moved up or down since last week.

Top 40:

Conferences:

Bottom 10: