Let's quickly review what the college football model is measuring. The model looks a the on-field statistics that significantly affect the ability of a team to score points and also the ability of a team to stop their opponent from scoring points. To do that I use on-field statistics measuring the teams (and their opponents) ability to: acquire possession of the ball; move the ball down the field; maintain possession of the ball; scoring efficiency; and adjust for the number of games played and the conference the team plays in. Once I have collected theses statistics for each team and similar statistics for the team's opponent, I use a statistical model to determine whether the variables I am using are significant, and if they are, to estimate the weights of all the variables I am using. Once I have the weights of the statistically significant variables I calculate the overall production of the team to score points (which I refer to as offense) and the overall production of the team to stop their opponent from scoring points (which I refer to as defense). Taking these two calculations I subtract the defense from the offense and arrive at the final team production. Then all I do is sort from highest to lowest all the team's total production to determine each team's rank.
When a team performs better than other teams in terms of either offense or defense, they will increase in their relative rank to other teams in the football bowl subdivision and when team perform poorer than the others they will fall in the overall rank.
Here is Northern Illinois University through week #11 (although they did not play this week).
| Offense | Defense | Total | Week | Date | Opponent | PF | PA | Result | Place |
| Rank | Rank | Rank | |||||||
| --- | --- | --- | 1 | 9/1/2012 | Iowa | 17 | 18 | Loss | Neutral Site |
| 85 | 4 | 42 | 2 | 9/8/2012 | Tenn.-Martin | 35 | 7 | Win | Home |
| 54 | 58 | 50 | 3 | 9/15/2012 | Army | 41 | 40 | Win | Away |
| 48 | 54 | 56 | 4 | 9/22/2012 | Kansas | 30 | 23 | Win | Home |
| 26 | 69 | 39 | 5 | 9/29/2012 | Central Mich. | 55 | 24 | Win | Home |
| 16 | 65 | 31 | 6 | 10/6/2012 | Ball St. | 35 | 23 | Win | Away |
| 6 | 22 | 3 | 7 | 10/13/2012 | Buffalo | 45 | 3 | Win | Home |
| 10 | 29 | 9 | 8 | 10/20/2012 | Akron | 37 | 7 | Win | Away |
| 37 | 35 | 28 | 9 | 10/27/2012 | Western Mich. | 48 | 34 | Win | Away |
| 2 | 26 | 4 | 10 | 11/3/2012 | Massachusetts | 63 | 0 | Win | Home |
| 9 | 16 | 4 | 11 | 11/10/2012 | --- |
Notice that other than the first game against Iowa (a narrow loss), Northern Illinois University has been above average in terms of their overall rank. As the season has progressed, Northern Illinois University's on-field production on offense has increased and their opponents on-field production has decreased. In weeks they have been very productive, they have moved up in the rankings and in weeks they have been less productive, they have moved back. Take for example the Western Michigan game on Oct. 27th. Western Michigan gained more yards, had more first downs and had greater scoring efficiency than Northern Illinois. Given this, NIU's overall ranking dropped from #9 in week #8 to an overall ranking of #28 in week #9. As NIU's overall on-field performance has relatively increased, so has their overall ranking.