CLAASSEN RANKINGS RULES

The Claassen Rankings were started in 1999 to rank Division I-A college football teams and rankings were computed for that season mostly to establish a starting point for the system for the year 2000. All of the calculations were run for 1999 and I experimented with several different point values for each factor in my equations and determined what gave me, in my opinion, the most accurate results. In 2000 I added all of the I-AA teams to make the rankings for the I-A teams more accurate. The rankings for the I-AA teams are not as important as the I-A teams but I list the I-AA rankings because I have produced them. Below is an overview of how my system works. It is quite unpredictable in the first few weeks but be assured that it does settle down about half-way through the season.

Here is my background. I played football for 11 years starting in 7th grade and going through 5 years of college including a red-shirt season. I played at a small high school and at Kansas State University beginning with Bill Snyder's first season. I was a four-year letterman and I have a bowl ring to my credit. Since then I have been officiating high school football since 1995.

Here is a summary of my system based on the definitions utilized by David Wilson's web site. My system recalculates all of it's equations up to 25 times until the system's values quit changing. This system ranks teams based on wins, margin-of-victory, and strength-of-schedule. The system is also a mixture of predictive and retrodictive but is mostly a retrodictive one.

My system assumes that a team is as good at the start of the season as it was at the end of the previous system. Each year a team starts out with the same score as they ended the previous season. I see no unbiased way to change the rankings based on players or coaches leaving or entering each program. With 117 head coaches and over 2500 starting players just in Division I-A, I don’t have enough time to analyze all of the changes.

This system assigns points based on the outcome of games and tries to project the winner of upcoming games. Points are awarded based on performance relative to the level of competition.

A win is awarded points. Based on 2002 numbers the points earned from a Win make up approximately 22% of a team's total.

Margin-of-victory is calculated by comparing the predicted outcome of the game to the actual outcome. This difference in points is then assigned a point value and then multiplied by the Ave Pt percentage. This factor is limited to a 21-point difference between the predicted and actual outcomes. Thus if one team is predicted to win the game by 28 points but only wins by 7 then they will be penalized for playing worse than predicted and the loser will be rewarded for the game being closer than predicted. Based on 2002 numbers the points earned from Margin-of-Victory make up approximately 8% of a team's total.

The Ave Pt percentage of the opponent is calculated by taking the Ave Pt of the opponent and dividing it by the Ave Pt of the #1 ranked team that week and is multiplied by a constant to provide a value for strength-of-schedule. This Ave Pt percentage is recalculated each week so the point value of each game can change throughout the season. Based on 2002 numbers the points earned from Strength-of-Schedule make up approximately 70% of a team's total.

Finally, all of the point values are totaled. This equals the total value for each game. All games played to that point are averaged. All teams are then ranked based on that value.

Until a team has played three games in the current season, the Ave Pt from the previous season will be factored in. For the first game of the season all rankings will be the Ave Pt which each team earned from the previous season. Before a team has played three games in the current season, their Ave Pt will be the average of the completed games and the previous season’s Ave Pt. From the third game and on in each season, last season’s Ave Pt is discarded and the rankings are based on this season only.