Dec. 9, 2007

Finding the Best: The College Football Championship in America

By John V. Lombardi

This is the time of endless speculation about which division I-A college football team is the best in America. We have polls, computer rankings, conference championships, and the high profile BCS (Bowl Championship Series) program. Our experts (which include just about everyone who follows college sports) argue with great passion about which scheme is the appropriate method for anointing the “best college football team in America.” This controversy lives and repeats itself because we can never get it right.

The notion of “best” in college football represents an unattainable goal. If football were organized like track or swimming with absolute standardized measurements of performance for everyone, we might be able to measure the best. But football is not organized this way. Football is a regionalized team sport (through the mechanism of conferences), and it is constructed in a format that guarantees the impossibility of defining the best.

No college football team in America can play every other college football team. Some football teams play some other football teams. Moreover, not every football team or athletic conference is equal to every other. Some are very rich (the SEC for example), some have institutions with huge athletic programs (like the Big Ten with Ohio State), and some struggle with financial difficulties or low attendance (like the University at Buffalo). Some are in conferences that have 12 teams divided into east and west divisions, play an extra conference championship game, but no team plays every other team in the conference. The won-loss record of one college football team almost always has a different value from the same won-loss record of another team. If a mediocre team plays other mediocre teams, the won-loss record means something different than if the mediocre team had played very good teams. This illustrates why we have no standardized measure of a football team’s performance.

Absent a clear measurement of football performance, we end up with subjective measures. For many years, we identified the national champion of college football by consulting expert opinions, expressed in various polls, from sports writers, coaches, and others who, in theory, had seen most of the contenders play during the year and who we hoped had a reasonably objective basis for identifying the best. This, however, violated the fundamental premise of sports which is defined by a competition within a strict set of rules that produces an unambiguous winner. People who play and watch sports keep score. If we only care about the beauty of the game, then we do not need to keep score. Not keeping score is inconceivable to most sports enthusiasts, and so everything in sports turns on determining winners and losers.

This context helps explain why the college football season’s end is so fraught with controversy. We can’t figure out how to get a clear, single, winner. Some argue that a tournament would make it work better, but that’s not any clearer because we have to seed a tournament by using a formula based on season performances of teams that did not play each other. The tournament format gives the illusion of effectively determining a winner, but it only creates a second season of football for a group of teams chosen by a complex ranking process that relies on judgment, not direct competition. At the end of a tournament, a team is crowned champion, but the team is only champion of the group invited to the tournament. Given the inconsistency of collegiate football performance, the wide range of resources and capabilities of the various college athletic programs, and the large number of potential participants in a tournament, one can easily imagine a team not included in the tournament that might well have ended up a winner.

The BCS, the mechanism used to identify a football champion, takes a different approach. It recognizes that football is organized into conferences, that the conferences are of differing quality and significance, and that the teams from these conferences will accumulate records in ways complicated and difficult to compare. The BCS identifies the best football teams by combining first, the opinions of two groups of experts (a panel of observers representative of the experienced public and a poll of coaches), and second, rankings based on six different computer models (that offer an illusion of impartiality and fairness) to identify the best football team. These computer models are of course just as arbitrary as the opinions of the human experts, but at least they eliminate the possibility of individual bias against particular teams by substituting systematic bias in favor of certain characteristics of measurable team performance. The resulting rankings produce a playoff between number one and number two in the BCS formula, and then provides a set of secondary playoffs between other ranked teams paired up for their presumed performance quality and television draw. (For a primer on the BCS see http://www.bcsfootball.org/bcsfb/about )

The owners of the BCS, the football conferences and the various bowl organizations, have adjusted the methodology for constructing the ranking to persuade the football-expert public of its validity. However, no ranking system based on subjective or computer analysis can produce a true competition between the best teams for an objectively identified BEST football team in America. We have intentionally constructed a college football system to produce conference champions, not national champions.

Should we worry about all this? No, not at all. While some might yearn for the certainty of knowing which college football team is the best in the country this year, not knowing exactly which team is best has many advantages. The argument over the schemes employed to designate the best two teams to compete for the championship gives us endless conversational opportunity, it gives our alumni and friends something to focus on, it allows newspapers and sports shows to fill their pages and air time with speculation about the relative performances of our teams, and it delivers a discussion that requires no closure because it presents a question that has no answer.

The conference structure of football ensures that we will have many champions, many contenders for the best in the nation, whatever the results of the BCS contests. If our team is very good, but didn’t get into the top game, we can feel confident that, but for the vagaries of a peculiar polling system and ranking calculation, our favorite would have been able to demonstrate its supremacy. If we are among the 64 teams playing in one of the 32 bowl games, we have another chance for a celebration.

This is the genius of the BCS system. It invents an imprecise standard based on subjective measurement (the polls and computer models). It applies this standard of measurement to one of the most rigidly structured games in the nation. It then builds a second season of bowl championships of various degrees of importance around this ranking system and the entrepreneurial enthusiasm of bowl promoters. America’s sports fans, who spend endless hours debating the merits of the imprecise measurement, consume this manufactured championship process and its subsidiary bowl games with ever-growing enthusiasm.

Whatever else we can say about the Bowl Championship Series, it is a one of the most visible demonstrations of America’s unique ability to combine serious academic enterprise with intense competitive spirit, remarkable entrepreneurial ingenuity, and spectacular entertainment value.

[Disclaimer: the author’s institution, LSU, will play in the BCS championship game.]

Comments

The purpose of competition is not to identify the best team or teams participating in a competition. It is simply to determine a winner. That is the role objective rules (computers) play. Objective rules do not offer an illiusion of impartiality and fairness. They define impartiality and fairness. That different rules favor different statistical criteria or value criteria differently does not make them biased or unfair. Furthermore, the results produced by objective rules are not just as arbitrary as those produced by subjective means. The objective rules I favor can only produce one answer given the exact same results. However, pollsters can produce any number of answers given the exact same results. This year’s coaches’ ballots included votes for nine different title game matchups. That is arbitrary. The bottom line is that competition is about determining a winner and only objective rules can produce a undisputed winner.

Chad Hansen, at 2:05 pm EST on December 10, 2007

Genius?

“This is the genius of the BCS system.”

I don’t think there is much genius in the BCS system. I’ll stick with Joe Theismann who once said, “Nobody in football should be called a genius. A genius is a guy like Norman Einstein.”

justaguy, parent & taxpayer, at 5:45 pm EST on December 11, 2007

BCS

Mr Lombardi — your analysis and reasonings are incorrect. College basketball teams do not play all other basketball teams during the year but somehow a real national champion is crowned. Same with the Championship Subdivision (formerly IAA) where your previous institution UMass was a national champion — undisputed. There is no beauty to the BCS — it is corrupt, driven by money, meaningless and thus ugly.

David Proulx, at 7:55 am EST on December 14, 2007

BCS

With each line of this article, I kept expecting the author to say. . . “Just kidding!” Fans celebrating the values of the BCS? Seriously? Seriously??? That’s like a spouse in a bad marriage reasoning, “Well, I’m stuck in this marriage, but if I lower my expectations, surely I can rationalize. . . I mean, my marriage isn’t as bad as some peoples’ marriages. . .”

I work at a college, so I love healthy debate, but imagine if conversations about the NFL were reduced to “scholars are divided about the worthiness of the Baltimore Colts or the Dallas Cowboys to face off against the Patriots in the Super Bowl.” I would boycott it.

Sports fans are most satisfied by on-field action where the players determine the outcome. We don’t like it when a referee’s call decides the outcome. We react to the term “Olympic judge” with nausea or disinterest. We want it sportsmanlike, but we want it Darwinian and Decided.

D-I football has evolved to what it is, an admittedly flawed system. If we were to create it all from scratch today, there’s not a chance in million that we would organize its post-season the in its current configuration. It evolved this way and we are stuck with it.

I understand it, just as I understand how bad marriages have similarly evolved. But to celebrate it? To recognize the beauty of it? Are you sure this article wasn’t tongue in cheek?

reader, at 9:55 am EST on December 31, 2007

Who Cares?

If these men are truely student athletes, the NCAA should get them back into the classroom sooner rather than later.

To make academics more of a focus, perhaps team size should be governed by some measure of academic performance (e.g. average graduate rate, average GPA, etc.). In such a system, teams with higher graduation rates would be allowed to suit up larger teams.

Adam, at 12:35 pm EST on January 9, 2008

Um...check your facts there

No team plays every other team in its conference? That’s news to the PAC-10, Mountain West, WAC, and Big East.

Matt, at 6:40 am EST on January 13, 2008

Adam—that would just lead to more academic fraud, like Auburn’s sociology classes, or basketweaving degrees. It would also reward schools that have weaker academic standards in general over schools where getting a degree is harder.

Matt, at 6:40 am EST on January 13, 2008

This is the stupidest argument that I have heard yet: “No college football team in America can play every other college football team.” You could say that about virtually every sport in existence.

Sam, at 5:45 pm EST on January 14, 2008

Finding the Best: The College Football Championship in America

Mr. Lombardi says of the BCS, “The resulting rankings produce a playoff between number one and number two in the BCS formula, and then provides a set of secondary playoffs between other ranked teams paired up for their presumed performance quality and television draw.” The latter half of that sentence is misleading, and that fact is responsible for much of the complaints against the BCS. For example, the University of Missouri was ranked #6 in the BCS final regular season standings, before the bowl season, but was not chosen to play in a BCS bowl based on the BCS’ own ranking formula. Though finishing #6 in the BCS, Missouri was not given the opportunity to play in a BCS game, while teams ranked below it in the BCS, such as Kansas (which Missouri had defeated two weeks previously) were selected for BCS games.If the BCS system wants to maintain a semblance of objectivity based on statistical computer models, then the BCS bowls should be required to select all BCS bowl participants based on the final BCS rankings.

David Bundrick, VP for Student Development at Evangel University, at 10:30 am EST on January 21, 2008

Who cares?

Why not let the NFL run its own damn farm system and keep higher education out of it entirely? The alternative is to watch once-reputable institutions (like my erstwhile employer, Rutgers) prostitute themselves to the money and the cynical commercial ethos of the big-time sports racket, which is pure poison, so far as educational values are concerned. Let’s worry a lot more about whose undergrads win the Putnam competition!

Norman Levitt, Prof. of Math, Emeritus at Rutgers, at 8:00 pm EDT on March 31, 2008

BCS and big-time Intercollegiate Athletics

Dear Sirs: This is a modified comment I made to a previous John Lombardi column. As an East Coast-to-Midwest transplant it was amazing to see the emphasis in Michigan placed on intercollegiate athletics. I should add that I have always loved competitive sports. It did not take me long to come to the belief that there is only one viable long-term solution to the problem of the professionalization of college sports: Big-time Intercollegiate sports programs should have a university affiliation in name, but they should be completely separated, financially and academically, from the universities’ core (intellectual) mission. So, for example, Michigan State University (MSU) teams would still carry the MSU Spartan logo, but the so-called students would be registered at the MSU Sports Institute, where it is understood that the explicit goal is to produce future professional athletes, not budding young scholars. The pretense and dishonesty of labeling collegiate sports as something done by students and young scholars would be erased, and, due to the financial separation the “big donor issue” (and many other issues besides) would be moot. Donors interested in academics would contribute to academic programs, and donors interested in the MSU Sports Institute could make their contribution there. What about our young scholars? They may, if they wish, take courses at reduced rate at the host university, but they are under no obligation to do so (i.e., there would be no need to graduate with a degree). All such athletics institutes should be at least partly supported by the professional organizations that draft their future star athletes from these programs.

MSU Prof, at 2:50 pm EDT on April 3, 2008

big time sports

Every time someone in education complains about sports I sense one of the deadly sins. I feel most of the problem is that our education system while having too many high school dropouts also grants way too many doctorates. The PhD glut is real. There is too much supply. Everyone with a PhD thinks they deserve a good job. They’ve been lied to and lied to themselves. The NCAA basketball tournament is cheap entertainment even for people in nursing homes. If a college sports program brings more money to college in donations, grants, student interest, etc. who deserves more money the college president or the coach? Education is supposed to make you think like an adult not a petulant child.

wayne bailey, at 9:05 pm EDT on April 10, 2008