Ingale's Recursive Power Rankings

Ingale's Recursive Power Ranking System (IRPRS) is an iterative method that uses only three pieces of data about each game/match to rank all of the teams within a particular league:
  • The two teams playing
  • The location of the game
  • The winner
Further, this system avoids a blanket homefield advantage by iteratively adjusting the value of being at home for each team. A team only gets extra credit for a road win if the opposing team has proven to be a tough opponent on their home turf.

I specifically don't use the score because there are far too many variables that play into the final score of a game, and many of these things do not reveal any real information about the teams' strength. For example, in 2006, Ohio State won at Texas by 17 points, while winning at Illinois by 7. Is Illinois really better at Texas? Do they deserve less of a loss than Texas does? Perhaps Ohio State got to an early lead against Illinois and felt so unthreatened by Illinois' offense that they ran the ball to protect it. Meanwhile against Texas, they felt the need to keep scoring to protect their lead. Did a a freshman key player get some time against Illinois to help him get some game experience? The final score of the game doesn't tell us any of this. What we do know is that Ohio State won the game.

Another important goal I had in developing the method was that I didn't want to punish teams for playing really bad teams. At the very worst, they should get no benefit from staging the game. At the same time, teams shouldn't get rewarded simply for playing a good team. If the number 1 team in the country wins a 66-0 nail-biter over the worst team in the country, it doesn't really tell us anything new about either team. So the number one team gets no benefit from playing the game. And the worst team won't suffer for playing in the game either.

The IRPRS is a recursive method. The teams are given a rating based on their wins and losses. Then they are reranked based on the rankings of their opponents. Teams are essentially treated as completely different entities for their home and road games. As the process continues, the rankings settle out and a composite ranking is made of the home and away rankings.

Neutral Site games are treated as half a home game for both teams. Ties are treated as half a win for both teams. Based on this, a tie game at a neutral site would be treated as 1/4 of a win and 1/4 of a loss both home and away for both teams. Confused yet?

The goal of the IRPRS is to rank a group of teams based on their body of work for a particular season. This is particularly important in sports such as NCAA division 1-A football, where there are roughly 120 teams but each only plays 12 games, and there is no playoff system to crown a champion. For this reason, I've included my rankings for each NCAA season since 1869, except for 1871. (There were no documented football games played in 1871).

In other cases, I'm giving my rankings as a point of reference. The best way to crown a champion is on the field, and I do not dispute that in any way. In fact, one should be wary of any system that claims anything else.

NCAA football

1869 1870
1871 1872 1873 1874 1875 1876 1877 1878 1879 1880
1881 1882 1883 1884 1885 1886 1887 1888 1889 1890
1891 1892 1893 1894 1895 1896 1897 1898 1899 1900
1901 1902 1903 1904 1905 1906 1907 1908 1909 1910
1911 1912 1913 1914 1915 1916 1917 1918 1919 1920
1921 1922 1923 1924 1925 1926 1927 1928 1929 1930
1931 1932 1933 1934 1935 1936 1937 1938 1939 1940
1941 1942 1943 1944 1945 1946 1947 1948 1949 1950
1951 1952 1953 1954 1955 1956 1957 1958 1959 1960
1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
1971 1972 1973 1974 1975 1976 1977 1978 1979 1980
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
2001 2002 2003 2004 2005 2006


National Football League

1941 1942 1943 1944 1945 1946 1947 1948 1949 1950
1951 1952 1953 1954 1955 1956 1957 1958 1959 1960
1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
1971 1972 1973 1974 1975 1976 1977 1978 1979 1980
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
2001 2002 2003 2004 2005 2006


NCAA Men's Basketball

2006-07 NCAA Men's basketball rankings (as of 3/10/2007)
2005-06 NCAA Men's basketball rankings


Credits

I did *NOT* compile the game data myself and I want to give credit where credit is due.