Abstract:
|
The author gives a history of his involvement in college football, starting with a political economy seminar in 1963, adoption of the logistic model, realization that this pure model was unsatisfactory, adding a grading function based on margins in 1970 or 1971, and a failure to convince the NCAA to use such results as official standings. The author personally implemented the method for many years, and practical problems of acquiring a database, incorporating played and unplayed forfeits, and awarding cochampionships are explained in detail. The author's participation in the Bowl Championship Series from 1999 on is reviewed, including poll results in 2000 and 2001, which were controversial. New theory is presented on playoff nature and size and team selection, which gives an unexpected result. A simple set of requirements is given for any system which is a candidate for official college football standings, and the work is extended to basketball and ice hockey.
|