As its name implies, The College Football Data Warehouse, maintained by David DeLassus, keeps a myriad of information about college football.  Some years back, I found that among “national championships”, the list of “selectors” included my own Fleming System.  I certainly don’t offer a trophy to the team that ends up on top of my ratings at the end of the season—I’m just a schmoe with a computer.  And as a college football fan, I don’t even always believe that the team on top deserves the accolades of “national champion”.  But such are the underpinnings of the most interesting of discussions over a few beverages at the local establishment where such discussions take place.

It all begs the question: What is a “national champion” in a sport where no official champion is recognized?  Major college football is an odd egg in this regard.  The Football Bowl Subdivision of the NCAA’s Division I is the only NCAA-sanctioned sport that does not crown a champion (at least of the sports of which I am aware.)  As such, the distinction of being named a “national champion” has always come down to opinion.  Even in the years of the BCS, it has often been the case that a very deserving team was given no chance of winning due to exclusion from the BCS Championship game.  Other years, people questioned the worthiness of a participant in the game.

What computer rating systems bring to the table in these debates is a means of weighing who won their games versus how much did a team prove in winning their games.  In particular, how much does strength of schedule count?  When it comes to the Boise State and Northern Illinois-types of the college football world, it seems to be the most important criterion.  But when it comes to the teams in the more prominent conferences, the scheduling of out-of-conference patsies is generally overlooked.  It is all fodder for discussion.  And as college football fans, we crave such discussions all the time.

With all of that said, rankings matter.  A lot.  So, I went back to 1900* and listed the top rated teams.  The list of top-place finishers is given below, ranked by which schools have finished at the top most often.

[table id=9 /]

It should be noted that I have listed Harvard as the top team in 1901, despite the fact that they actually finished at No. 2.  The No. 1 team, Homestead Athletic Club, was actually a professional team that played a number of games against top college programs.  This was common back near the turn of the previous century since the game was pretty new and finding opponents was not a trivial task.  None the less, I made the arbitrary and capricious (well, maybe not) choice to exclude Homestead from consideration.  But it should be noted that games involving these professional athletic clubs are included in the ratings in many of the datasets from early years.  Another thing to note is that my ratings do not exclude games that have been “vacated” due to NCAA sanctions or similar after-the-fact alterations to the record.  But it is fun to see the war-time programs at various military bases getting their due.  And I especially like that (now NCAA Division III) Washington and Jefferson fare so well in the 1926 results. Go Presidents!

*Data for my ratings comes from Peter Wolfe’s page for the years 2005 and beyond.  Data for the ratings prior to 2005 comes from the excellent database provided by James Howell

The Fleming System: Week 9 Rankings

[table id=8 /]