Musings on Connectivity and Scheduling

by LaFortune Teller , South Bend, Tuesday, January 26, 2010, 20:20 (5977 days ago)
edited by LaFortune Teller, Tuesday, January 26, 2010, 20:27

College football was better in 1989 than in 2009. More specifically, college football connectivity, especially among its elite teams, was stronger.

In 1989, there were 106 Division 1-A teams. They played an average of only 10.6 scheduled* games per team against one another. They played a total of only 50 1-A vs. 1-AA games. There were 25 independents (24% of Divison 1-A).

In 2009, there were 120 FBS (formerly 1-A) teams. They played an average of 11.3 scheduled* games per team against one another. They played a total of 94 FBS vs. FCS games. There were only 3 independents.

*"Scheduled games" here means non-bowl and non-conference championship games.

Connectivity in college football has been exhaustively studied by Paul Kislanko for the last several years: http://football.kislanko.com/. He measures the "connectedness" of the field of teams similar to the Six Degrees of Kevin Bacon game. You are connected by 1 link to all of the teams you play head-to-head. You are connected by 2 links to all of the additional teams with which you share an opponent; by 3 links to all shared opponent-opponents, etc.

In 1989, despite fewer teams playing fewer games, connectivity on average was better. Before the bowls, 18 percent of all D1A teams in 1989 had a 1- or 2-link connection to at least 60 percent of all other teams. 13 of those teams ranked in the final pre-bowl AP top-25, including Nos. 1 through 6. In 2009, only 11 percent of all FBS teams had a 1- or 2-link connection to at least 60 percent of all other teams. Only four of those teams ranked in the final pre-bowl AP top-25.

The independent teams led the way in 1989, both for connectivity and in the rankings. Nine of the "60-percent connected" teams in 1989 were independents. Only two independents were "60-percent connected" teams in 2009. That said, it isn't as though 1989 and 2009 were dramatically different in terms of direct connectivity among the very best teams:

The final post-bowl rankings in 1989:
#1 Miami had played #2, #3, #9, #16, and #17.
#2 Notre Dame had played #1, #4, #7, #8, #15, #16, #17, and #18.
#3 Florida State had played #1, #6, #11, and #12.
(All three were independents.)

The final post-bowl AP rankings in 2009:
#1 Alabama had played #2, #3, #10, #17, #20
#2 Texas had played #1, #14, #21
#3 Florida had played #1, #8, #17

But the meat of the schedules of elite teams was much stronger in 1989. None of the top three teams played a 1-AA opponent in 1989. In 2009, instead of another game or two against a ranked team, Florida and Alabama each played an FCS opponent and two Sun Belt opponents; Texas' out-of-conference schedule last year consisted of UTEP, Wyoming, New Mexico, and Louisiana-Monroe (a team that was 1-AA in 1989).

1989 might not be fondly remembered by Notre Dame fans, but let it not be forgotten that we are two decades removed from what might have been the best overall college football season ever. 1990 saw a 34% increase in FBS vs. FCS games. Miami joined the Big East in 1991. Florida State joined the ACC in 1992, the same year the SEC expanded to 12 teams. By 1993, only 11 total independents remained. Since 2004, there have been four independents or fewer. Meanwhile, the number of FBS vs. FCS games has nearly doubled since 1989: 94 games this past season.

Where were all the great games in 2009? The same place as all the independents. The landscape hasn't just changed; it would be virutually unrecognizable to elite teams from 20 years ago.