The Category Mistake at the Heart of College Football

By Sean Carroll | December 16, 2008 10:44 am

Too many things I would blog about, if only I could slip into an extra timelike dimension and experience several weeks in just a few of your Earth minutes. Between now and New Year’s I’m going to clean out my collection of blog-worthy things; if you’ve read enough of Cosmic Variance in the past, you should be able to extrapolate to a full post.

Today: “Top Ten Stupidest Arguments in College Football,” which is itself so full of stupid arguments one suspects one is being punk’d. College football is the only major sport that decides who plays in the championship game on the basis of a vote, rather than by a playoff. One can debate the merits vis-a-vis excitement and revenues, but the whole operation is based on an epistemological blunder: the idea that there is something called the “best” team. The point of sports is not that there are better teams and worse teams, it’s that some teams win and some teams lose. Winning and losing is not some approximation to the true measure of excellence that we are forced to put up with; it’s what the games are all about. A sensible world would have a playoff, and let the teams play. (I’ve actually heard people argue that a playoff would be bad idea because the “best” team might not win.)

(If I could just train myself to make posts that are that short all the time, I’d blog twice as often. Maybe five times as often. Are more/shorter posts better?)

  • Mike


    Not only does the quality of play for a particular team change drastically with time, but — especially in a game like football — team A might be able to consistently beat B, B consistently beat C, and C consistently beat A.

    Of course, in a tournament there is still luck-of-draw in specific match-ups and weaker brackets — but I still think there is a greater sense of real, meaningful accomplishment in winning a playoff tournament, as opposed to being selected to play a single game and winning that.

  • Rien

    f I could just train myself to make posts that are that short all the time, I’d blog twice as often. Maybe five times as often. Are more/shorter posts better?)

    You would run the risk of turning into Andrew Sullivan. But he does have a great beard.

  • Dan

    “Are more/shorter posts better?”

    I don’t know… let’s vote! :-)

  • Stephen

    I disagree with the notion that there is an indisputably best team in college football, and any attempt to rank the teams is subject to some pretty limiting statistical factors, and you’d have to rank teams with a poll even before setting up a playoff. Here’s why:

    In Division I-A college football, there’s something around 120 teams, and each team plays 12, perhaps 13 games in the regular season. So each team, ostensibly, samples 10% of college football. However, the overlap in schedule between two top teams in two separate BCS conferences is limited, so of that 10%, most of it is disconnected between, say, Oklahoma and Florida, so their record is not necessarily a reflection of their ability against the league as a whole, but only the small part of the graph they are connected to. Utah and Boise State are great examples: nobody would argue that these teams are as good as Florida or Oklahoma, but both have better records.

    In contrast, the NFL takes twelve teams for the playoffs, which is more than 33% of the league. Also, roughly calculated, any two arbitrary NFL teams will have at least one opponent in common, I believe (this is what happens when you crudely try to apply the pigeonhole principle after working on a presentation for hours), so there is some level of scheduling overlap guaranteed. For college football to pull this off, the season would have to be six months long, and the playoff would have to incorporate around 32 teams.

    The comparison breaks down, of course, because there is such a disparity between the top ten teams in the country and, say, the thirtieth best. But the point of my line of thought isn’t to work out the efficacy of a playoff system in college football, but to point out that there is no meaningful way to determine the best team in college football. A single elimination tournament implies that the best team should be able to beat every other team in the country — see Utah and Boise State?

    So why not have split titles like they did fifteen years ago? If there are, say, six undefeated teams at the end of the season, why not have two national champions? Two or three teams get the satisfaction of being dubbed first among peers, and the history reflects the absence of a clear national champion. Having a single national championship game implies incorrectly that there is a well ordering principle in college football.

  • Elliot Tarabour

    The problem with football in relation to other sports is that you cannot play more than 1 game a week. (at least that’s what they say) This is due to the tremendous physical damage done to the players in each game.

    I think we would be better off with a switch to touch football and then they could play more games and a playoff would make sense.

    But one thing I am grateful for is that we won’t need to have a thread with “National Championship” and “Notre Dame” any time soon.


  • Sili


  • Ginger Yellow

    A sensible world would have a league

  • Ginger Yellow

    Gah. It removed my comedy tags. That post was meant in jest.

  • Fabian Ledvina

    There is such a thing as a best team. The best team doesn’t necessarily win every game though. A moral relativist would not understand that objective value does in fact exist, but it does. The point of athletic competition is to perform at one’s highest physical ability and to match up against others to determine who stands above the rest. The point of sports is not that some teams win and some teams lose, this epistemological blunder is the attitude that has ruined athletics. This is the type of attitude that has parents forcing a pitcher in little league in Connecticut out of the game because he is too good. He actually is the best player in the league and because their kids can’t handle it and the parents have misguided morals, they tell their kids it isn’t about being the best, it is about playing the games, and proceed to have the best player removed from the game to make things “fair”.

    Also, the correct statement is not that college football does not have a playoff, Division 1A (now called the Bowl subdivision) does not have a playoff, the other 4 divisions of college football do have playoffs.

    For those of us that have played football, we understand that the nature of the game precludes a lengthy playoff process. It does bring about the possibility that the best team actually loses because of the single elimination tournament that would be required. The best we could hope for is a +1 system. We can use the computers and the voters to accurately determine the best four teams in college football (this year would be hard to get 4 out of the top 5, but not really as it is USC’s fault for having a weak schedule). You then play 1 vs 4 and 2 vs 3, and then have only one extra game to determine the true champion. A lot of people have wanted this system for a long time, but other issues have gotten in the way.

  • Joseph Brant

    The top four thing is something I’ve wanted. It wouldn’t be perfect, but it’d be a step up from where we are now.

  • SI

    Sean, not to be elitist, but unlike the professional physicists you spend a lot of time with, college football fans have an average IQ that is (gasp) average.

    Fabian, I think you mistook Sean’s words there. He’s not saying that every team has (or should have) an equal chance of winning and losing. His point is that, ultimately, all the arguments should be done on the field. I think you’ll agree with that one.

    Patriots fans may want to claim that they had the ‘best’ team in the NFL last year because they had the best record or the most talent*. But having the best record or the most talent is not what the NFL is about. It’s about winning the games you need to win in order to claim the championship. Yes, it’s kind of a tautology.

    This goes a little off topic, and I think I’m closer now to getting on the posted link, but USC didn’t get to choose 9 of their 12 opponents. Only opponents that they can be ‘faulted’ for are Virginia, Ohio State, and Notre Dame. I wouldn’t call that weak.

    *I just saw Gregg Easterbrook write in his column that the talent differential between the best team and the worst team in the NFL is ‘maybe ten percent’. How do you quantify ‘talent’? Now THAT is a category mistake. He’s also very anti-physics for some reason.

  • Elliot


    that needs some type of explanation. I’m sure he’d be pretty unhappy if physics were to simply go away….


  • SI

    OK, anti-physicist.

    He puts in a lot of non-sports news into his column (understandable since he’s not a sports journalist), and there’s about one piece of physics/astro news along with his commentary per column. His comments about academic physicists/astronomers are often not very kind. I’ll let you read it and decide. (This is now completely off topic, so I won’t discuss this here further)

  • Elliot

    whoa… he should be the poster person for ADD….

    does somebody actually pay him to write the column or does he pay ESPN to run it.


  • astromcnaught

    As far as posting policy goes: I like to see the quality physics stuff.
    I fear that more and shorter posts may end up revealing how your office plants are faring.

  • Peter Morgan

    “Are more/shorter posts better?”

    You don’t post about a dog, therefore you don’t have a dog? Get a dog. Get the best of a hundred dogs at the pound. Maybe get an office plant, but I fear there’s no best and worst between plants and dogs.

    But quality Physics stuff for preference.

  • Serge

    I don’t think I completely agree. Genetic algorithms are all about getting best, or fittest “solution”, and tournaments or duels are the way to estimate fitness. Sure duels impose only partial order, but who say that the “best” should mean “globally bes”t ? There could be several locally “best”. There is even the expression “one of the best” …

  • mr paul

    I don’t see how shorter posts leads to more posts. Do you assume short posts take less time? I don’t think that is necessarily true. See

    Perhaps you meant to say “If I could just train myself to make posts that are that devoid of much thought all the time, I’d blog twice as often. Maybe five times as often. Are more/less thoughtful posts better?”

    1. Short, thoughtful posts are good.
    2. Long, thoughtful posts are good.
    3. Off-topic posts that still relate to things many (but not all) of us have in common (eg, an interest in sports) are good, in moderation.
    4. More != good. (Note that this is not the same a “more is not good”. Subtle but important diffference.)

  • Brian Mingus

    Sean, I look forward to your physics-related posts because they make me feel like I understand physics, whether or not I actually do. I prefer your medium-long posts. I hate football.

    Cheers :)


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Cosmic Variance

Random samplings from a universe of ideas.

About Sean Carroll

Sean Carroll is a Senior Research Associate in the Department of Physics at the California Institute of Technology. His research interests include theoretical aspects of cosmology, field theory, and gravitation. His most recent book is The Particle at the End of the Universe, about the Large Hadron Collider and the search for the Higgs boson. Here are some of his favorite blog posts, home page, and email: carroll [at] .


See More

Collapse bottom bar