Every season is different, so the postseason structure should be different, too.
(Originally written on 3/22/06)
The NCAA tournament is in full swing, and what fun it is. And it really is fun. I'm sincere when I say that. Woo-hoo!
But I am, first and foremost, a college football fan. And it's this time of year, in our media-saturated hyper-advertised culture, that college football is near forgotten in favor of "the best time of year," the NCAAs. (Until the next "best time of year" comes along, that is...) My purpose here is manifold. First, to talk about what really is wrong with college football's post-season format. Second, to also demonstrate that there are genuine ways in which college football's set-up is very good, in fact better than what basketball does. Third, I want to discuss, yet again, how to make college football better. My style of argument will be fairly dialectical, talking about all these points at once (especially the first two), rather than going through them topically, one-by-one.
Let's start with this: there are both good things and bad things about having a team "prove it on the field."
Folks who want a playoff for college football usually put their argument in terms similar to this: that the national champion needs to be determined through some sort of "objective" process, and that in sports the only real objective process that makes any sense is to have the best teams play each other and "settle it on the field."
There is something to this, obviously. The whole idea of a sport is to allow two teams to compete against one another for the glory of victory by actually playing the sport. This is hard to deny. All things being equal, the very best teams in a given year should meet up to settle things, to whatever extent this is possible.
There is some sort of desire for "fairness" lurking in all these sentiments. What we want is to be able, at the end of the year, to take all and only the teams that have performed best throughout the year (this is an "objective" determination, it seems, based on how the various teams actually play during the year, how many games they win, etc.) and give them all a "fair" shot at winning it all. If you are in the group of teams that are clearly the best, clearly separated from the rest of the pack, then we want you to get to have a shot at winning it all.
For instance, college football has moved more towards this ideal, it seems. Under the old Bowl system, there were often two (or more) teams that seemed equally deserving of a national championship but they would not even play one another. Even when there were two clearly superior teams, they often would not play one another. #1 v. #2 matchups were very rare in the bowl games, because of the pre-set rules and traditions that governed how the bowl match-ups were determined. #1 would play #5 and win by 7, while #2 would play #3 and win by 3. The end. Season over. Nothing left to do but take some polls and see who has been "elected" to the national championship. And whoever it is, how likely were we to feel comfortable with the result?
But now college football has a system where we know every year that the two "best" teams (as determined by a controversial forumla, to be sure) will play each other. This is a move towards the kind of "objectivity" that people like. But still many are left unsatisfied. Part of their disatisfaction, obviously, lies in how "subjective" the process of picking the "best" two teams is.
So, it might seem that the real 'problem' with college football is that, while it has made some strides toward picking its national champion 'objectively', it still isn't 'objective' enough. We still have no guarantee that the very best teams will be able to settle it on the field at the end of the year. We leave it up to polls and formulae and human persuaders and persuadees, and something "weird" almost always comes out of this process. A good team that should get a shot at the national championship does not (USC '03, Auburn '04, etc.), or a team that isn't thought to deserve a shot ends up getting one (Nebraska '01, Oklahoma '03, etc.), or both. The problem, in short, is that the current post-season format in college football does not give any guarantee that the teams competing "on the field" at the end of the year will really be the best teams (an underserving team might get a shot), or will be all of the best teams (an equally-deserving team might get left out).
Hence, the argument might (and often does) go, we need to make things even more "objective", and settle things on the field with a playoff. Such a playoff will force the two teams in the final game to first prove themselves by earning their way there by defeating other very good teams. The problem with college football is it is too "subjective" in the way the final two teams are determined. Human error, computer nerd polling (which, depending on which critic you ask, are either too prone to huamn error, programmed by computer nerds and all; or are two rigid and standardized, which is objective but the "wrong kind" of objective), etc. More teams need to be included, to be given their "fair" shot at winning the national championship. Then we will be more confident that the winner is really The Best (peace be upon her).
At this point defenders of the current system, traditionalists of a certain sort, will come in and argue (usually persuasively) that every team was given its "fair" shot at winning it all. We call this fair shot "the regular season." Every week in college football is like a playoff, because one loss is often all it takes to knock you out of the national championship picture (since only two teams get to play for the national championship at the end of the season, and those teams are often but not always both undefeated, one single loss can be devastating to a team's hopes of making it in the end). This argument has strengths, the biggest of which is that the thing it is trying to preserve is absolutely, without-a-doubt worth preserving. The regular season in college football is, hands down, the best thing American sports has going. The importance invested in rivalry games, conference games, and high profile inter-conference matchups is unrivaled in any other major American sport. This is indeed, at least in part, because of how important every individual game really is to a college football team. You do not want to lose, even once. There is likely going to be no revenge game, no chance to beat 'em later that year when they come to your place. No chance to beat 'em in a conference tournament at the end of year (and thus, no chance to sneak into the national championship picture by winning such a conference tournament, either). There are other teams playing on television when your game is over, teams that look about as good as you on paper and which will get their own chance to impress a national audience when you are finished. You must perform, week after week.
But this argument also overlooks some things that are important. As important as the regular season is, it simply does not, if we look at it with an unmuddled mind, produce the same sort of result every year. The regular season gives us a pyramid of who is really good and who ain't as good, but the top of the pyramid is configured differently year after year. Sometimes there are two undefeated teams, and only two undefeated teams. Both are from major conferences, and played respectably-difficult schedules. For instance, Texas and USC this year. When all the dust settled in early December, everyone was satisfied that the right two teams were playing for the national championship. These were the two teams that deserved a "fair" shot at the national championship, and there was no feeling that anyone else was being left out. Based on the regular season, nobody had any right to complain that Texas and USC were going to the Rose Bowl instead of them. But this doesn't always happen. Sometimes, after the regular games are played, you have three undefeated teams ('04, four if we count mid-major Utah), or no undefeated teams but three one-loss teams ('03). Or some other configuration which cannot be satisfactorily resolved by only letting two very good teams play each other and calling the winner "national champion." The problem here is not with the regular season, per se, but with the fact that the regular season cannot be counted on to always produce two and only two qualified national championship contenders. It seems this just illustrates the problem anew: We want to be able, at the end of the year, to take all and only the teams that have performed best throughout the year and give them all a "fair" shot at winning it all. But college football right now is pre-locked every year into only allowing two teams to play for the championship, and this seems to produce an unfair result when there are more than two teams who deserve a shot.
What I'm getting at is this: whenever anyone talks about fixing college football, the conversation is always about what one playoff structure should be put in place. But this way of thinking about the problem will never lead to its solution. What we need to be asking is why we feel committed to setting up any particular structure before the regular season has played itself out in the first place. Flexibility is the key, and flexibility really is doable.
It should now be obvious, though, that this is a problem common to all sports! No sport currently in existence (that has a post-season of some kind) has any guarantee that the very best, and only the very best, teams in a given year will be given a "fair" shot at the championship. Why not? Because every sport has its own pre-locked post-season structure that must be conformed to, no matter what happens during the regular season. In the NBA, 16 teams get to enter a playoff tournament, but are there really 16 teams every year that deserve a chance to win the NBA title? Was it really "fair" to make the 1995-6 Bulls, for example, play through three rounds of playoffs against Eastern Conference opponents they had already clearly dominated during the regular season, before they could play the Western Conference champion for all the marbles? If the Atlanta Hawks had by some fluke beaten the Bulls in a seven game series, would that have made the Bulls less than the best team in the Eastern Conference that year? Hardly, it would be a fluke, and everyone would have seen it as such (except perhaps for the two over-enthusiastic Hawks fans in the world).
And let's be clear about the NCAA basketball tournament. The idea that letting 64 (excuse me, 65) teams play it out in a free-for-all is any way to decide who is the "best" team is ludicrous on its face. This is at least as bad of an "objective" way to measure a national champion as having seven computer nerds put together some number-crunching formula before the season begins that is then used to guage teams at season's end. No, it really is worse, because the computer nerds' formulae are being used to fix upon only two teams who are deserving a shot at the national championship. Going into a season (in any sport) sight unseen, if you had to guess how many teams would clearly separate themselves during the regular season as being better than everyone else and as deserving a chance to play each other for the national championship, would you guess that that number of teams was closer to 2 or 65? Exactly.
Whatever college football's credibility problem in picking a convincing two every year who are "truly" deserving of a shot at the national championship, it represents a more sincere effort to actually be "objective" in determining such a champion than college basketball. College basketball has thrown all sincerity in picking a genuinely deserving national champion to the wind, and has opted instead for a de facto lottery in which the very best teams are rewarded by...being forced to run a gauntlet against other good (but clearly not as good) teams for six straight games. If you can win six in a row at the very end of the season, then college basketball will call you "national championship." This is a devil's bargain for anyone with eyes to see and ears to hear. (But it really is a lot of fun. I love it; just let basketball keep it for itself.)
So people feel like college football is inadequate because it predetermines that only two teams will get to play for a national championship at the end of the year. But then other sports are also inadequate, for they also predetermine a certain number of teams that will get to play it out for the championship, and the odds that this forced number will always line up with what reality shows to actually be the case are long indeed. Pick any number of teams you want--2, 4, 8, 64--is it really "fair" to always guarantee that this will be the number of teams who get a shot at all the glory? Of course not! So now we see that the problem college football has is actually the same problem every other sport has. As it stands right now, college football often allows too few of the deserving teams to have a shot at the national champtionship. But by the same token, college basketball (and other sports) frankly allow too many teams (many of which are not deserving) to have a shot. In both cases the problem is of the same kind--the list of teams that are allowed a shot at the championship and the list of teams that deserve a shot at the championship are not the same lists.
Remember what our goal supposedly is: We want to be able, at the end of the year, to take all and only the teams that have performed at the very best level throughout the year and give them all a "fair" shot at winning it all. But this number of teams will vary from year to year, and so any pre-set post-season format will always be inadequate to this goal.
What is the solution? There is one. Basically, we need to build flexibility of format into the college football postseason.
How should we do this? My first suggestion is this: let a committee of some sort determine how the post-season will work, on a year-by-year basis.
The exact composition of the committee (fans, coaches, commentators, regional affiliations, etc.) is actually not all that important. What does matter is that there is some sort of oversight. Its decisions as a body, as well as all deliberations and individual votes, must be publically-accessible and reviewable. Some sort of reviewing body might need to be put in place. Committee membership should rotate regularly. Finally, the committee membership must not be predisposed towards or away from any particular kind of post-season format. (For example, nobody who simply opposes all playoffs, no matter what, should be allowed on the committee.)
In this case, the devil really isn't in the details. Think back on the last several seasons of college football when there was some sort of major controversy at the end. The vast majority of people are always clamoring for the same thing. Auburn should have been given a shot in '04. LSU in '03. There are some questions about how to given them that shot, and there will always be some room for debate here. But the general kinds of post-season formats that make sense in any given year will be limited to a few. Let's consider some possible formats this committee might choose.
1. Two and only two major conference teams are undefeated or have only one loss (or only two losses, etc.; the point is that two and only two teams are tied for the best record, whatever the best record is in that particular year). The committee will simply let the two teams play each other for the national championship. This is what we get now, and it works just fine in this sort of scenario.
2. There are more than two "best record" major conference teams. All of these teams will be "in". The question is what to do next. Let's look at 2004. USC, Oklahoma, and Auburn finished 1-3, and all were undefeated. Clearly all three of them should have been given a shot. But 3 contenders is an awkward number for a post-season. There seemed to be two possibilities for a fair post-season set up in 2004.
2a. Pick up one 1-loss team to round it out into 4 final teams. There are problems with this, though. For one, there might be more than 1-loss team, and whichever one doesn't get picked will seem to be "screwed." (This was precisely the case with Cal and Texas in 2004, actually.)
2b. Stick with the 3 undefeateds, but choose one as the favorite and give them a bye. For instance, in 2004, there was little doubt in most people's minds that USC was a significant favorite to beat either Oklahoma or Auburn, so give USC a bye and let the Sooners and Tigers fight it out for the right to take on USC. All three teams are still getting a shot at the national championship this way, so there's really little to complain about. I would have preferred this option in 2004 over 2a, because it lets us avoid the Texas/Cal debacle, when it was clear that neither of those teams was a valid candidate for "the best" in any case.
2c. Consider an undefeated mid-major team. In 2004, Utah was undefeated and looked very impressive in their wins, so this would have (in my opinion) been the best outcome at all. Just make a 4 team playoff between USC, Oklahoma, Auburn, and Utah. Done.
3. Only one team is clearly better than the rest. For instance, Miami in 2001 was the only undefeated team. Frankly, the most "fair" thing to do here would probably have been to just declare Miami national champion without playing any post-season games at all (though of course Miami would still go to a bowl game as an exhibition, like in the old days when the final polls always came out before the bowls were played). Whether any committee would ever be content to do this or not, I doubt. Let's assume not. What to do, then? Well, do something similar to 2b above. Take the other teams that are even remotely close to Miami, and let them battle it out for the right to play Miami for the national championship. The point in such a scenario would be to not put the clear favorite through too much "stress"--it has earned the right to be the presumptive favorite and to benefit from that presumption. So, in 2001, the thing to do might have been to let Oregon and Nebraska (the only one-loss teams) play each other, and the winner plays Miami. Or perhaps to let all the 1- and 2- loss teams play it out in a mini-playoff, with the winner playing Miami.
There probably needs to be exceptions to these basic guidelines, and there are no doubt others that could be added. But the basic approach for determining how the post-season will work should be clear. Whichever precise appraoch is taken in any given year, it will be more likely to "fit" the current season than whatever formula was in place at the beginning of the year.
Let's be very clear here, though. It's not as though this plan would eliminate all controversy. Of course it wouldn't. But, no matter which precise solution is chosen in a particular year, the scenario where any team that deserves a shot is getting shut out would be much less likely to occur. There may be a few teams who are less-deserving that are given a shot, but this is less of a travesty than a deserving team getting screwed. (Though, I'm not even sure such a compromise would be necessary...see 2b and 3 above.)
But another thing about controversy. We like controversy. Part of the fun of the BCS is the debates it engenders. This system would allow this speculation-machine to continue (especially if BCS rankings were continued for purposes of major bowl games), and would add to it even more. What will The Committee do this year? Different pundits would offer there ideal post-season format for that year. In the end, the committee would emerge from a smoky room with its plan, and whatever that plan is, it would most likely be better fit to the particularities of the regular season that had just transpired than any pre-set formula would be. In a world of uncertainty, this is probably as good as we can do. But it is much better than what anyone is actually doing now.
It might seem like we should bring "objectivity" back into the process at this point. Some sort of formula or something should be set up which tells us, in advance, how the post-season will be figured out each year. It will vary, yes, but we should use some formula to tell us how to vary it from year to year.
This isn't the worst idea I've ever heard or anything. It would take us almost all the way to an ideal system, just not quite there. First of all, such a formula would have to be incredibly complex. And yet we know how we feel about letting complex formulae dictate unchangeable arrangements at the end of a season, no matter how carefully planned or researched the formula might be. When all is said and done, we want human beings who have experienced the ebb and flow of the particular season in question, who have watched the games as fans, coaches, commentators, or general experts, and who therefore have some sort of "existential" knowledge of how things should be arranged. It would be fine to lay out some sort of formula as an official guideline for how the decision should be made, but then the individuals on the committee should be allowed to deviate from the formula if it seems appropriate. The point is that the final decision-making power for arranging the post-season should be in the hands of people who have experienced the season, not formulae that were put in place before any games were even played.
The irony of all this is that it turns out college football, for all its controversy, is already closer than any other sport to doing what needs to be done if the actual champion and the "true" best team are to be more likely to coincide. The irony is that the very thing that is often said to hold college football back is in many ways what it needs more of. Subjectivity. By introducing human subjectivity into the process in an even more fundamental way (i.e., by letting people decide how many teams to let in and how to structure the postseason each year, rather than just letting them decide which teams to plug into the pre-set number of available slots), college football could lead the way in doing what all sports really should do if they are serious about having their champion "on the field" line up as closely as possible to the "truly" best team. Every sport that takes this seriously (and maybe none of them really do) should stop trying to pre-set how many teams will be deserving of a champtionship shot every season, and should instead come up with a way to respond to the actual ebb and flow of each season after it happens.
This is a reactionary approach to finding a national champion, and so it is inherently more subjective, but this is not a weakness. It is (objectively) a better way to do things.
I've thought of several objections that might be raised to this, but I think they can be answered. I reckon there might be others I've not thought of, and I am looking forward to hearing them.
 USC, of course, ended up winning a "share" of the National Championship in 2003 anyway, as the AP defied the BCS system and voted the Trojans number one despite the fact that they didn't appear in the "national championship game."
 The fact that pro basketball, baseball, and hockey use series' of games, rather than a one-and-done format, for their post-seasons does make it less likely for a clearly inferior team to beat a more deserving team. But the possibility still remains of this happening, and it certainly has happened before. Putting all this aside, though, the fact remains that in the college games (of basketball and football, anyway) post-season series' are not really feasible. This is especially true of football, where even the professional version uses a one-and-done format.
 Perhaps an official appeals court of some kind, where questionable decisions are evaluated after the season, and anyone found to have acted unreasonably by such a court will be removed from the committee. Or something.
 No system can be perfect, of course. Neither can it do magic. I can't offer unicorns and leprecauns here. There will often be controversy over the "last team in" to any post-season format. In the NCAA tournament, we argue and speculate about 'bubble' teams. But bubble teams are never likely to actually win, anyway, and so the screw job is not nearly as bad as when (say) an undefeated major football program is simply locked out of a national championship because three's a crowd. Better to stick a really-good-but-clearly-not-the-best Texas or Cal team in with USC, Oklahoma, or Auburn than to tell one of the big 3 that they're just out of luck.
 2003 was also a type (2) scenario, so let's consider it as well. In 2003 we had three one-loss teams, and no undefeated teams. There was one two-loss team, Michigan. Giving any of the three 1-loss teams a bye seems less likely here, since none of them were clearly better than the others. So probably the better option here would have been to bring Michigan into the fold to make a 4 team playoff. But I don't know what my hypothetical committee would have decided, of course. Either way, it would have been better than what actually happened. (Unless you actually relish split championships, which some do.)
 I really don't like the idea of giving a 2-loss team a crack at an undefeated. If it were to happen that there was an undefeated team and a bunch of 2-loss teams, with no 1-loss teams in between, then I think it would most likely only be appropriate to award the national championship to that lone undefeated major conference team. But it is not such a travesty to give one-loss teams a chance at an undefeated, since if the underdog pulls off the upset the two teams would then have identical records (and the underdog would have the head-to-head win, so it could make sense to think of them as the better team). Hence my favored solution in 2001 would have been to let Nebraska and Oregon (two one-loss teams) fight it out for the right to play Miami, and leave the 2-loss teams (Colorado, Florida, Tennessee, etc.) out of it altogether. But again, I'm more interested here is sketching out the sorts of options that "The Committee" would have in structuring a post-season each year than I am in telling you what I personally would have done.
 Obviously I've given no consideration in this essay to how my suggested system could be best incorporated into the current bowl system. Perhaps I'll take that up in a later essay.