We can use the preseason rankings to see how the new structure would be implemented using data for the current season, and compare the sample schedules with the actual games being played this year. The list of games looks much more exciting—and eliminates matchups no one really wants to see.

~~~

In an earlier proposal I recommended that we restructure college football to improve the quality of the matchups and eliminate meaningless games between powerhouses and second-tier opponents. I recently read an article that raised this issue again. But how would things look if we implemented the proposed change?

We can use the rankings and schedules for the current 2015-16 season to see quite clearly the impact made by these changes.

Preseason rankings for 128 teams were created by a number of sources. I used this one as a starting point. I also compared the top 25 with the AP rankings, and while the exact rankings differed, the top 21 teams were the same. So there’s quite a bit of consistency.

We can use these rankings to generate the following four leagues of 32 teams:

This would be the starting point for implementing my proposal.

The next step would be to divide each of the four leagues into divisions of eight, and to allocate the teams. I did this in two steps. First, I allocated them based on ranking. I took the first four teams and put one in each division, then took the next four and put one in each division, and so on. In assigning teams to divisions, I tried to put teams with others in their current conferences, or at least in their region. But for the first attempt I didn’t deviate from the approach of taking four teams at a time and allocating each of them to a division and then moving on to the next group.

This yielded a structure in which teams played several teams not in their current conference or region.

In reviewing the resulting divisions, it was apparent that I could re-allocate teams among the divisions and generate groups that align relatively well with the current conference structure. While I don’t think this is critical, I think it’s helpful to demonstrate that these aspects of college football could be preserved in part. Matching teams by regions could also reduce travel time and distance, which is important for colleges and their students.
Using these divisions, we could create schedules for each team based on the structure in the initial proposal: each team plays the other teams in its division, and also the team from each of the other divisions that has the same ranking within its division—the top team plays the top teams from the other three divisions, the second team plays the second team in the other divisions, etc. I compared these resulting schedules with the current schedules for the four teams with the highest preseason rankings: Ohio State, Alabama, Oregon, and Baylor.

   Ohio State and Oregon

Alabama and Baylor

Don’t these new sample schedules look exciting? Lots of meaty matchups all season long. It’s also interesting to note the caliber of competition for these top teams in their current schedules. Each of them plays at least one team ranked worse than 97, and TCU plays four of them! Alabama has the toughest current schedule, with six games against teams in the top 32. But that still means that half of their games are against significantly weaker teams, including three ranked worse than 65. Oregon plays only two teams ranked in the top 32–imagine if they played 10, and everyone else in “League A” did as well. That would be an exciting season!

If we made this change, it would be a much more exciting and compelling experience for players and fans. If everyone also really wants to include a “rivalry” game towards the end of the season like many schools do today, this would certainly be possible—my proposed schedule has only 10 games and many teams already play 12. The new structure would also help address another minor but rather perplexing aspect of the current scheduling structure: teams don’t all play the same number of games, yet the number of losses is still a critical factor in final rankings. How is it fair for a team that plays an extra game and therefore risks an extra loss to be compared in rankings with others that don’t? In this new proposed structure, all teams would play the same number of games until the top eight teams in each league started the playoffs.

I think this proposal would dramatically improve college football, and it was helpful for me to see exactly how it might work using real rankings and comparing with real schedules. I hope it’s helpful for you, too!