Critique of Kenneth Massey's Retro-Ratings

The Bowl Championship Series for the 1999 season will include, in addition to Herman Matthews, the computer ratings of Kenneth Massey. His ratings can be found at www.mratings.com/cf/years.txt and go back to the 1930 season. Here are some of Massey's more interesting rankings of teams through the years. All ratings are post-bowl.

Okay, here we go again. People who invent math formula rating systems like to promote the idea that ranking college football teams with rating systems is a superior way to rate teams as compared with the AP poll voters. But in the above example we have Mississippi number one in 1955 and Oklahoma number nine in 1956. What if a voter in the AP polls those years voted the teams that way? Are we supposed to accept teams rated this way because a math system put them that way instead of an AP poll voter? Not that ranking the teams this way is illegal. You can rank Kansas State number one in 1965 if you want. But why are these strange and unusual rankings of teams acceptable if a math system and not a poll voter produced them? Are off-the-wall rankings acceptable if a formula rating number is placed after them?

See also Historical Reality National College Football Champions.

<- Parent Directory

Bob Kirlin / cfrabob@hotmail.com