Saturday, January 1, 2011

Behind the FAIR Rankings

According to Drew Pasteur, “the selection and seeding process for the NCHSAA football playoffs is greatly improved from the days when pre-determined brackets matched up designated conferences. However, the current system still relies too much on a team's place within its conference (even though conference sizes vary from five to nine teams). The current system also ignores strength-of-schedule, discouraging teams from playing tough non-conference foes. A power rating system could fix this, and would help ensure that the most deserving teams receive the highest seeds.”

A team earns points for each team it defeats (and more points for beating higher-classification teams). The point total is divided by the number of games played, so that playing ten games versus eleven is not a factor. Teams earn additional points (for strength-of-schedule) based on wins by teams they played. Beating good teams will result in the most points, but there is some credit for playing good teams, even if those games are lost. Bonus points are given for winning a conference championship or for being the top team from the lower class in a split conference. The ratings can be used to select and seed the playoff teams.

The FAIR rankings are based on the following criteria.

Fresh start - All teams in a particular class start the season with identical ratings, or none at all.
All games weighted equally - There is no distinction made between conference and non-conference games, and a week one endowment game counts just as much as a week twelve contest.
Ignore non-NCHSAA games - Because the quality of opponents outside the NCHSAA can very widely and can be difficult to judge, only NCHSAA games will be counted. Any contests against NCISAA or out-of-state opponents will be treated the same as an open date.
Reproducible - The formulas used should be transparent and simple enough that coaches and members of the media can verify the calculations for a particular team.