In 1905, Caspar Whitney, a co-founder of Outing magazine and one of the creators of collegiate football’s All-American Team, named Yale as the season’s “national champion” in college football. Four years previously, The (New York) Sun had named Harvard as the 1901 collegiate football champions. Although the National Collegiate Athletic Association (NCAA) was established in 1906, it first named national champions only in 1921, starting with track and field. Despite naming national champions in literally dozens of intercollegiate sports since then, the NCAA has not and still does not name a national champion in what is now the Football Bowl Subdivision, formerly known as Division I or Division I-A Football.
This, of course, has not stopped those in the sports world from naming national champions in collegiate football. Beginning with The Sun and Caspar Whitney at the turn of the 20th Century, roughly three dozen systems have endeavored to select the national champion in collegiate football. Several of these have used statistical data to retroactively select national champions all the way back to the very first season for collegiate football in the United States, 1869. In 1926, professor Frank Dickinson, of the economics department at the University of Illinois, created what seems to be the first mathematical system to determine the national championship, with the nod going to Stanford. The system devised by Dickinson attracted the interest of the coach at Notre Dame, Knute Rockne, who persuade the professor to apply the model to several previous seasons, with the result of Notre Dame becoming the retroactive 1924 national champion and Dartmouth the 1925 champion.
According to the listing of national championships provided by the NCAA, there are five colleges with claims to the national title for the 1926 season. The first is Alabama, with nine systems ranking it as the national champion: the Berryman (QPRS) System, which began in 1990; the Billingsley Report, 1970; the College Football Researchers Association, 1982 and 2009; Helms Athletic Foundation, 1941; National Championship Foundation, 1980; and, the Poling System, 1935. The school with the second high number of rankings, four, as the national champion was Stanford: the Dickinson System, 1926; Helms Athletic Foundation, 1941; National Championship Foundation, 1980; and, the Sagarin Ratings, 1978. Navy was selected by two systems as the 1926 national champion: the Houlgate System, 1927; and, the Boand System or Azzi Ratem System, 1930. The other two schools with selections as the national champion for 1926 are: Lafayette, Parke Davis, 1933; Michigan, Sagarin, 1978.
While it quite probably simply a coincidence that after naming Stanford as the 1926 national champion that professor Dickinson then used his formula to create retroactive champions for the 1924 and 1925 systems, that the Contest Board of the American Automobile Association (AAA) seems to have done something similar in 1927 is quite striking. One could suggest, however, that it might not be quite as coincidental that several automotive journals and newspapers named champion drivers from 1909 to 1915.
While it is entirely possible that we may never know exactly what prompted members of the AAA Contest Board to create retroactive champion drivers in 1927, nor why they were apparently so readily accepted, the many efforts to create retroactive collegiate football national champions might suggest that this inclination in the sports world is not unheard of. If professor Dickinson’s model of 1926 was used to determine the possible national champions of 1924 and 1925, this was also the case with several of the other systems used to select a national championship team.
A year after the Dickinson System was introduced, 1927, Deke Houlgate, created another mathematical model to determine the national championship team, which was used to determine championship teams beginning with the 1885 season. First used in 1930, the system devised by William Boand, the Boand or Azzi Ratem System, created national champions for the 1919 to 1929 seasons. In 1933, Parke Davis, a former player for Princeton and coach at Wisconsin and several other colleges, created a listing of national champions that began with the 1869 season until the 1933 season. Another former player, Richard Poling, devised yet another rating system based upon a mathematical model beginning with the 1935 season, creating championship teams back to 1924.
I would suggest from this consideration of historical revisionism in American collegiate football that the revisionism undertaken by the AAA Contest Board and its national champions is not necessarily unique in the field of American sports history. It might also help in understanding why the retroactively-created champion drivers seems to have been accepted with few qualms, only the Chevrolet-Milton and Dingley-Robertson issues apparently drawing any attention over the decades.