These ratings were devised by Joe Stanford, a 1990 graduate of Indiana University. They were developed as a hobby and inspired by systems such as The Dunkel Index and Jeff Sagarin’s Sports Ratings.
Special thanks to Fort Wayne (Indiana) News-Sentinel sportswriter Blake Sebring for his encouragement, and for posting the ECHL (East Coast Hockey League) Ratings in his blog “Tailing the Komets.” Thanks also to Kenneth Massey, who provided his computer rankings for the Bowl Championship Series, for posting the NCAA Football and NCAA Basketball Ratings in his Ranking Composite websites. (See the links in the sidebar to the right.)
The best measure of a computer rating system’s accuracy is how well it predicts future games (i.e., how often the higher-ranked team beats the lower-ranked team). For the 2013-14 NCAA Basketball regular season, The Stanford System predicted the outcome of the 5,362 games between Division I opponents with greater accuracy (71.1%) than all but two of well over 60 published rating systems, and better than any of the nationally-known systems. The Stanford System also correctly predicted more winners in the brackets for the 2014 NCAA Tournament (39) than any of the nationally-known systems.
Any newspaper or sports website operators who may be interested in adopting any of these ratings, or Stanford System ratings for any other sport, may contact The Stanford System by leaving a comment in the comment section of this site. Please include your contact information so that any such inquiries can be responded to directly rather than posted on this site.
EXPLANATION OF THE RATINGS
The Stanford System uses a series of algorithms to connect and compare sports teams within their leagues based solely on the results and scores of games. The ratings are therefore objective and unbiased.
The teams are ranked in order of their Power Rating, which is a measure of each team’s relative strength. Immediately to the right, each team’s Power Rating and Rank prior to the current week of games is shown.
The Schedule Rating and Rank is a measure of the strength of each team’s previous opponents, and takes into account whether the games were at home or on the road. Games scheduled but not yet played are not included in this rating.
These ratings include a Home edge. Since the factor is based on game results, it will change slightly each week. To predict the outcome of an upcoming game, compare the two teams’ Power Rating after adding the indicated Home edge to the home team. If the game is to be played at a neutral site, ignore the Home edge in the comparison.
The Retrodictive Rating does not consider margin of victory. It is intended to provide a retrodictive analysis of games already played, and therefore measures what each team has already accomplished. The Retrodictive Rating is not intended to predict the outcome of upcoming games.
The Power Rating, on the other hand, includes a factor for margin of victory, and is an excellent predictor of upcoming games.
These ratings are NOT intended to be used for gambling. They are for informational and entertainment purposes only, and intended to enhance sports fans’ enjoyment.