How To Make College Football Better (Well, Maybe)

Aside from pro tennis, college football is my favorite sport (and luckily their two seasons rarely come into conflict). Yes, I think college football is more fun to watch than the NFL. Why? I can't really put my finger on it, but I think it has something to do with the fact that college football fans (for the big schools anyway) seem to be more passionate, and the games themselves are more exciting because there is a certain amount of variance in talent and consistency. NFL teams, for the most part, have roughly equal levels of talent, so the games usually become chess matches between coaches. While this usually results in close games, the games just seem to lack any sort of excitement.

Also, college football games are on Saturday, which is a much better day to watch any sporting event. On a side note, I know "Super Bowl Sunday" is like a big tradition and all, but the Super Bowl always takes place after the college season is over, which means that we could totally have the Super Bowl on a Saturday...and it would be so much better. Who the hell wants to rage on a friggin' Sunday night and show up to work hungover on Monday? Not me.

In any case, a lot of people don't like college football because of its inherent lack of fairness, and these people are called "everyone but Bama fans."

"What's it feel like to not make the playoff?" -Bama fans
Image Credit

This is, however, a valid criticism. The determination of teams in the college football playoff (which is now two rounds, but used to be just one) is based on the rankings of a committee of thirteen people, who get together and vote on the top four teams in the country to make the playoff. What's kind of interesting about this is that several of these people are current athletic directors (or are associated in some way) with universities that they may indeed be ranking...sort of a conflict of interest if you ask me. Granted, the fact that there are 13 people and only 4 teams in the final should help even this out. Still though, it irks me a little bit that a fairly small (however knowledgeable) team of human people is left to make such a critical decision about which teams are better. Some may prefer this over the old computer ranking systems, but I still say that humans are fallible, biased, and tend to think with emotions rather than reason.

This 2017 season (much like last year's) left an interesting decision on the table: there were two teams (Southern California and Ohio State) that each had two losses and were champions of their respective conferences; then there was undefeated Central Florida, who was a conference champion (but not in a Power Five conference); and finally there was Alabama, who had one loss but no conference championship. The decision was made to go with Alabama, because their strength of schedule and quality of victories/losses was deemed better than Ohio State's, with USC and UCF getting the boot due to poor strength of schedule.

I'm sure there are many Ohio State fans who really hate the committee right about now, and they have some right to be. They were, after all, a conference champion and Alabama wasn't even a division champion. And to this, Alabama fans could say, "we didn't get blown out by Iowa." Nevertheless, I still can't help but wonder if there is a better way to do things, a way that is both completely fair and completely objective, no old guys stroking their beards and consulting their crystal balls about who to invite to their new year's eve party.

"Okay it's turning a darkish red color...could be Bama or OSU...we should get a better ball"
Image Credit

As of now, there are 129 FBS teams in the league, which makes scheduling difficult. Since football is a very physically intensive sport, having more than one game per week would be ill-advised, and if we want the season to be a tolerable length, then it really only leaves a season of 12-15 games possible. In the current system, a team in a Power Five conference has to play 12 regular season games, a conference championship game, and two playoff games in order to win the National Championship, for a total of 15 games.

Because of the 129 team pool, a pure round robin would be impossible, since that would involve each team playing a 128 game season...which would last about two and a half years. We could employ a more efficient playoff system, which could work nicely with 7 rounds (if we kicked one of the teams out of the FBS). The problem with this is that half the teams in the league would only have a 1 game season, and the season would only last 7 weeks for the two teams in the championship. The other problem is determining the seeding of the playoff, as this plays a dramatic impact on a team's chances. Since college football teams change from season to season, there isn't really a way to seed teams based on the results of the last season, because it's a different team.

Clearly, a 7 round playoff is not a great option. This leaves us with a hybrid approach, which is what currently exists. The teams play a 12 game regular season which consists of a pseudo round robin with other teams in their conference, plus a few games with teams in other conferences. The results of these games determine the seeding for the final 2 round playoff (which I suspect will become a 3 round playoff...and then a 4 round playoff...and then eventually probably a 7 round playoff).

But how can we make this better, and remove (most) of the ambiguity related to different strengths of schedule?

Split The Leagues!

In case there was any doubt about this before, if you aren't in a Power Five conference, then it clearly doesn't matter if you go undefeated (see: UCF). Granted, they didn't have to play my Alma Mater (the mighty Georgia Tech Yellow Jackets) due to Hurricane Irma, but based on how our season went I think they probably would have beaten us. Even if they did, that might not have been enough to get them in.

Which begs the question, if you aren't in a Power Five conference, then what's really the point of playing? If you know on game 1 that even if you go undefeated, you won't get invited to the playoff, then why even try?

College football currently has several leagues, the main two being the FBS and FCS, with the FBS being the top league. FCS teams keep wanting to make the jump up to the FBS, with a few being added each season for the past several years. This, in my mind, is the wrong thing for the FBS to do. By adding more teams, they are just increasing the amount of teams that have next to zero chance of ever making the playoff. My solution: eliminate teams.

The FBS needs to be broken up, and a decent way to do this would be to just take the current Power Five conferences and make them the top league, and then take the other conferences and make them their own league. The "Group of Five" league could have their own playoff and championship game, which might give some meaning to their seasons and let them taste the thrill of victory for once.

As of now, there are 65 schools that are considered part of "Power Five" conferences, so if we made them their own league then that leaves us with some options.

My Utopian college football league setup is as follows:

  • The league is divided into conferences of equal sizes
  • Each team plays every other team in its conference once (a pure round robin), and plays no teams outside the conference during the regular season
  • The winner of each conference is put into the final playoff bracket

This system, while similar to the current system, has a dramatic departure in that there is no play outside of the conference during the regular season, and no conference championship game. It does, however, have some inherent unfairness that we will get to later, but first let's talk about the mechanics.

For this setup to work, the numbers have to be "nice", i.e. the number of teams must be evenly divisible by a power of 2. This ensures that no conference is left without a playoff spot.

Unfortunately, with the current 65 teams in the "Power Five", it wouldn't work. 65 is divisible by 5 (5 conferences, 13 teams each), but 5 is not a power of 2, therefore you'd have either a 2-round playoff where one team doesn't get to play, or a 3 round playoff where 3 teams have first-round byes. Neither option is particularly fair.

In order to see some examples of "nice" numbers for the total number of teams, consult the table below:

65 teams doesn't work, but there are other options for 11 and 12 game regular seasons

In my mind, the best option of the ones above is to drop 13 teams from the Power Five to get a 52 team league, which could be broken down into 4 conferences of 13 teams, giving each team a 12 game regular season (same as now) with a 2 round playoff (same as now). This comes out to a max total of 14 games (one less than now since each Power Five conference currently has a championship game).

Problems With This Model

While I think it's better than the current system, it does have some issues. One is that it's very inflexible when it comes to adding/subtracting teams. If you started with 52 teams and had some that thought they were deserving of being in the league, then you'd have to either kick some teams out, or add 44 more teams to get you to 96 (3 round playoff, 11 game regular season).

The second, and probably biggest, problem with this model is that it would be hard to make the conferences even in terms of overall skill level. Unfortunately, because the teams don't play out of conference games until the playoff, it wouldn't be known until after the playoff which conferences are really better. Even then, that conclusion would be made based on the outcome of a single game, which is hardly statistically significant.

In the current system, the conferences definitely aren't even, but it is remedied somewhat by the pretty ambiguous "strength of schedule" metric. This is also very flawed, because teams can vary in skill level throughout the season, which is hard to track, meaning your team may have beaten teams that ended up with bad records, but they all may have been good when they played you.

"Nah, the ACC is totally as good as other conferences" -Clemson fans
Image Credit

The third problem with this model is related to the second, and it is that because the teams have no out of conference games before the playoff, how is the playoff seeding determined? Seeding is important in any playoff bracket, as it prevents the two best teams from playing each other in an early round. My solution to this is to keep the playoff to two rounds, which basically makes the seeding less important, since the top 4 teams are likely to be competitive regardless. For a three round playoff, the seeding becomes more important, and the only real way to determine it in this model would be to give conferences "scores" based on previous performance, and try to seed accordingly. Again, not exactly fair.

Another problem has to do with the lack of a conference championship, and how a pure round robin with a large pool of teams can sometimes lead to ugly multiple-way tie situations. My solution to this: objective metrics for quality of victory, explained below.

Quality of Victory

The results of most sports are binary, i.e. you either win or lose, there is no in-between. But what if there was?

In the current college football system, losses are incredibly detrimental. One is manageable, but two is pretty much a death wish. Even worse, it doesn't matter how these losses occur. A team with one loss by a 30 point margin could get into the playoff, whereas a team with 2 losses by a combined 2 points might not. That really irks me, and I think it needs to change. Quality of victory has probably played a role into certain ranking decisions, but it is currently a nebulous measure that lacks any sort of quantification or consistency.

The concept of quality of victory is controversial for a few reasons: 1) it encourages teams to "run up the score" on sub-par opponents, which is considered "unsportsmanlike", and 2) it favors teams that have a few blowout victories, rather than those that have consistent play.

In my mind, quality of victory can't be adequately dealt with under the current college football schedule, because it necessitates that each team plays the same opponents in order to be fair. Currently, teams must schedule 4 non-conference opponents, which are usually a mix of quality opponents and "easy" opponents. Teams do this because the committee takes strength of schedule into consideration, so having all "patsy" opponents makes your schedule look worse. If, however, quality of victory was taken into account, it would be advantageous to schedule only weak opponents in order to capitalize on blowout wins.

In the pure round robin system I outlined in the previous section, I think quality of victory could be implemented, because each team within a conference plays the same schedule. How to implement it is another question, and will probably warrant another article. The simple way to do it is this: instead of each team having an overall record of wins and losses, they have a single "season score". If they win a game by 7 points, then 7 points are added to the season score. If they lose a game by 4 points, then 4 points are subtracted from the season score. My suggestion is to cap the amount at 14 points per game (maybe 16, it's open for debate), because any win by more margin than this is fairly definitive, i.e. a 21 point victory isn't really that much different than a 35 point victory, both are comfortable wins. Capping it at a certain point would also not encourage teams to run up the score.

The problem with this simple method is that not all margins of victory in football are created equal. A 4, 5, or 6 point victory are, for all intents an purposes, the same, because a touchdown and extra point would win the game for the other team in all 3 cases. With this in mind, different scores could be given for different margins of victory, based on the probability that the other team could have won. This could get more complicated, so stayed tuned for some mathematical analysis in a future article.

Would this change the way football is watched and enjoyed? Yes, and that does have potential benefits and drawbacks. It would definitely take some of the sting away from a close loss, but would also take some of the joy out of a close victory. It may, however, make the end of regular season potentially more exciting, because certain teams that would be mathematically eliminated under the old system could still have a chance in this new system by getting a big win and other teams getting bad losses, so something to keep in mind. Toward the end of the season, though, the games would retain the same level of excitement, just in a different way. Instead of "if we win this game, we get to the playoff", it could be something like "if we win by 5 points or more, we get into the playoff", or "as long as we don't lose by more than 10, we're in." It would simply shift the definitions of "winning" and "losing".


There's no correct answer for how college football can be fairly decided, there are simply too many teams that can't play enough games to significantly determine the true champion. I think my way would be pretty good, but it also has its flaws. At the end of the day, college football is exciting due to its inherent randomness, and while it may not be completely fair, it is usually pretty exciting.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.