Introduction
This is a review of the literature concerning the BCS and the question surrounding the controversy about the existing system of championship playoffs. Currently there is a conference division system for choosing the champion team of college football. However, it is flawed, because too often, very deserving teams with excellent win-loss records wind up with no chance at all for the championship. “The BCS includes the Atlantic Coast Conference, Big East, Big 10, Pac-10, and Southeastern Conference. The NCAA has no role in the BCS. As a cartel, the BCS seeks to arrange a national championship game and other bowl games among leading contenders. Controversy has swirled around its choices, however, and traditional bowl rivalries have been compromised.” (Staudohar, and Zepel, p. 39).
About the Methodology of the BCS System
“A number of theoretical models of sports leagues have been developed in attempts to describe the incentives and methods of operation of a league and its member franchises’” (Demmert, 1973; Noll, 1994; El-Hordiri and Quirk, 1971; Scully 1989, 1995; Quirk and Fort, 1994; Vrooman, 1995; and Rascher, 1997). (Depken, and Wilson 198) The main governing body is the National Collegiate Athletic Association and over the past ten to twenty years, many changes have been implemented for college sports, and not all for the good. “Our findings run counter to the stated purpose of changes in NCAA regulations, to wit improved competitive balance, a more “level playing field,” and a protection of the spirit of amateur sport.” (Depken, and Wilson 198) “The BCS would enhance competitive balance if more teams were able to reach higher rankings through wins and strength of schedule. On the other hand, the BCS might perpetuate the dominance of a small number of teams, thereby reducing competitive balance.” (Depken, and Wilson, p. 205).
In order to understand further how the new rules have impacted college football, we should look at some statistics in the following two tables:
“Column 1 of Table 2 reports the regression results using the variance of performance points. While the variance in performance points in the NCAA has been declining over time, the various regime changes instituted by the NCAA have had a significantly different impact on the variance of performance. The creation of the NCAA and the splitting of NCAA football into Division I and Division I-AA are found to improve the competitiveness of NCAA football. However, the implementation of minimum GPA of 1.6 is found to increase the concentration of playing talent at the expense of competitiveness. The effect of the variables SANITY, ENFORCE, and BCS is not found to be statistically different from zero. These results indicate that NCAA decisions have little impact on the competitiveness of NCAA football” (Depken, and Wilson, p. 206).
Interpretation is offered by Depkin and Wilson following the charts found in their book (2004).
“The intertemporal comparison of HHI indices does not account for differences in the number of teams included in the calculation. An increase (decrease) in the number of teams included in Division I-A football will naturally decrease (increase) the HHI. As can be seen from Table 1, the average number of teams classified as Division I-A has changed rather dramatically over time. While the average number of teams included in the concentration measures is approximately 98, the minimum number of teams was 15 and the maximum number of teams was 145. The relatively large variance of teams with Division I-A status has the potential of distorting the inter-temporal comparisons of the HHI.” (Depken, and Wilson, p. 207).
“From column 3 of Table 2, it is apparent that the various regulatory changes included on the right-hand side of the regression have had an impact on the distribution of performance points within the NCAA. However, unlike the common justifications that such regime changes are in the best interest of competitiveness, the Sanity Code, credible enforcement policy, the minimum GPA, and the dividing of the NCAA into multiple divisions are found to have a significant negative influence on the competitiveness of NCAA football. The implementation of the BCS is also found to be detrimental to competitiveness, but is only significant when using a one-tailed test.” (Depken, and Wilson, p. 207).
According to these charts, the competitiveness of the current system increases with the addition of more teams and decreases with the removal o teams. This means that the system can easily be manipulated, as has been charged by detractors. The number of teams is changing dramatically and so the system is not balanced (SEC, 1999).
Problems with the BCS System
“This article investigates two conflicting hypotheses regarding the role of NCAA rule enforcement within Division IA football conferences.” (Depken, and Wilson) Either of their choices look better than the current system, since they add an additional dimension to bare numbers.
However, not everyone sees the current system as a problem. Billingsly mentioned reasons that make the current system the best choice. “Another point that I would like to expand on is about the lack of a college football playoff. This is an incredibly complex issue, but the core of the problem is that college football is a unique sport, totally unlike any other. There are currently 117 teams in Division I-A, far greater than the playoff-manageable NFL.” (Billingsley, p. 190).
Colley (2004) suggests that we “allow the computer rankers to use the system of their choice, rather than restricting us to win-loss only (that would affect three of the seven, I think). Statistical and intellectual diversity in the systems is important in establishing some sense of the parameter space.” He further complains that the team making adjustments are kept away from the public. This “sequestering”–when layered on top of the perceived arcanity of the BCS system–creates an “Ivory Tower” problem. … (We) most would like is for the BCS to adopt a simpler “non-tweaked” system (which apparently is likely), allow each ranker to use the system of his choice, and adopt a policy of expressing faith in the computer systems, and making their authors more available to the media and fans (Colley, 2004).
In other words, the current system works only with a stable number of teams, preferably including more teams, and it allows tweaking the variables and how they are defined. Colley suggests (2004) that a system without the possibility of tweaking would be more fair. An article from Massey includes his list of problems, “Reactionary patches have been successful only in hindsight, and recent debacles have many fans justifiably upset with the BCS process.” (Massey) “I, along with David Rothman, Jeff Sagarin, and Peter Wolfe, have suggested a more fluid poll measure based on the Borda count.” (Massey)
How it Works
Chosen boards members meet and use their computer formula to inform their decision making in ranking the various teams eligible to play in the bowl games. The formula used can be “tweaked” on a number of variables, so the group has a highly controversial fudge factor.
“A computer-driven formula that accounted for a team’s win-loss record, the difficulty of its regular season schedule, and its rankings in various polls would choose the teams that would participate in the “championship” game. 83 Renamed the “Bowl Championship Series, ” this arrangement is a compromise that aims to crown a credible “national champion” without instituting a playoff system that would render the bowl games irrelevant, causing them to fold, and would likely extend the college football season well into January” (Porto, p. 65).
Mease (2004) examined the various years since the new system was implemented and compared them to the selections using the old system, and then he considered two new models which are modification of the current one. “From this comparison it can be seen that the model proposed in this article in fact had the smallest average difference from the AP and Coaches average over the five year period among all BCS models.” (Mease)
Economic Inequality
This is seen as the primary reason in most of the literature that people are not totally happy with the current selection system. It seems that the more money colleges spend on football, the more likely they will play in a bowl, all of which are lucrative.
“Football is a moneymaker for some of the BCS colleges, but it is also the most expensive sport that they offer. Consequently, it is an enormous drain on the financial resources of Division I-A colleges that either belong to conferences that are not a part of the BCS or are athletically marginal members of BCS conferences. In 1999-2000 the average football budget for members of BCS conferences was $6.4 million, whereas the comparable figure for Division I-A colleges that did not belong to a BCS conference was $2.9 million” (Porto, p. 55).
Membership in the conferences of the “Alliance Bowl is a factor in who has a chance to participate, since these games are highly profitable.
“Alliance” bowl (Rose, Sugar, Orange, or Fiesta) plus five other bowls, the SEC and the Big Ten are guaranteed an appearance in an Alliance bowl plus four other bowls, and the ACC, Big East, and Pac-10 are each guaranteed an appearance in an Alliance bowl plus three other bowls. In contrast, the Mountain West, Western Athletic, Big West, Sun Belt, and Mid-American Conferences, along with Conference USA, all of which belong to Division I-A, have no right to participate in any of the Alliance bowls and, hence, are usually represented only in the so called minor bowls, where the proceeds are substantially less than they are in an Alliance bowl.” (Porto 66)
Specific Problems and Objections
Possible Bias of Human Element
Some of the literature points out that the human element can easily be swayed by their personal preferences. In addition, humans may also be more impressed by factors not relevant to a team’s actual ranking, such as winning by large margins, or previous record of even very different teams at the same college.
“First, the human pollsters are not objective observers and may have biases toward certain schools based on regional loyalty, historical perception, and so on. Second, it is impossible for a human pollster to recall all outcomes of all games involving the 117 Division 1-A teams over the course of an 11- to 14-week season, even if he or she had witnessed or read about every game…..Some people, including the creators of some of these computer models, would argue the human pollsters themselves can be highly influenced by large victory margins, citing examples in which a team that wins by a large margin climbs higher in the polls than a team that wins by a small margin.” (Mease)
Conclusions
Many of the objections to the current system arise from ignorance in how it functions. There have been numerous tweaks and improvements over the years since its inception that stem from identification of problems with using only win-loss statistics. “In the five years of existence up to the time of the writing of this article, ten different models have been used by the BCS for ranking the Division 1-A college football teams.” (Mease) In all of the literature dealing directly with the methodology of making the BCS choices, the consensus is that it is better than a playoff would be.
The various opinions on the current system range from “Leave it alone” to “We need to start over again.” However, most of the objections with the current system suggest fixes to add to it. “So what we, the computer “geeks,” most would like is for the BCS to adopt a simpler “non-tweaked” system (which apparently is likely), allow each ranker to use the system of his choice, and adopt a policy of expressing faith in the computer systems, and making their authors more available to the media and fans.” (Colley)
Some critics complain about the formula, while others object to the addition of human intervention. Billingsly believes there must be a human element,”I don’t believe a computer rating, regardless of how good it is (and I believe I have one of the best), is as good as a combination of computers and a knowledgeable, responsible, human element.” (Billingsley 190)
There are a number of reasons why there is so much controversy, but they center around misinformation and disinformation. Massey cites ignorance concerning the system on the part of those outside it. “One thing I’ve learned from the process is the importance of public relations. Comparable to the U.S. tax code, the current BCS formula is a strange concoction of polls, computers, strength of schedule, quality wins, and losses. A gray box clouds the inner workings, so that the average fan feels the whole thing is at the mercy of a few computer nerds stuck away in some top-secret laboratory.” (Massey 2003)
There has been a serious ongoing dialogue among statisticians in the American Statistician.
Mease (2004) would add more variables to make the system more relevant to the needs. “Whatever the fate of computer models in determining the college football national championship, if nothing else we have learned from the controversy to think more analytically about the problem and issues involved.” (Mease)
Harville sees need for improvement, but does not suggest abandoning it in favor of costly and very time consuming playoff system. “It seems clear that the current BCS ranking system has some serious deficiencies and that significant improvement requires something more than the kind of reactive ‘tweaking’ that has been the forte of the BCS decision makers.” (Harville)
Altogether, the various contributors to this discussion agree that a playoff would be cumbersome, expensive and not in the best interests of the colleges or the students. There are too many teams and football simply cannot be played at the same rate as other sports. Most of the objections found in the literature identified changes that would make the outcomes more equitable for all colleges in the system.
So it seems that the consensus is that more needs to be done to the formula used to make it more equitable and the number of teams in contention must be stabilized to provide stability. A true playoff system would be expensive and take too much time for college students, and not be in the best interests of the students. Currently, some teams never get a chance at a bowl game, even if they are deserving. It is my personal thought that an additional bowl or two might help matters. Finally, this new system is a work in progress, and can be criticized as such, but it will evolve, and it probably the best available system.
References
Billingsley, Richard. “Discussion.” The American Statistician 58.3 (2004): 190. Questia. Web.
Colley, Wes. “Discussion.” The American Statistician 58.3 (2004): 191+. Questia. Web.
Depken, Craig A., and Dennis P. Wilson. “12 Institutional Change in the Ncaa and Competitive Balance in Intercollegiate Football.” Economics of College Sports. Ed. John Fizel and Rodney Fort. Westport, CT: Praeger, 2004. 197-209. Questia. Web.
Depken, Craig A., and Dennis P. Wilson. “NCAA Enforcement and Competitive Balance in College Football.” Southern Economic Journal 72.4 (2006): 826+. Questia. Web.
Harville, David A. “Discussion.” The American Statistician 58.3 (2004): 187+. Questia. Web.
Massey, Kenneth. “Discussion.” The American Statistician 58.3 (2004): 185+. Questia. Web.
Mease, David. “Discussion.” The American Statistician 58.3 (2004): 192+. Questia. Web.
Mease, David. “A Penalized Maximum Likelihood Approach for the Ranking of College Football Teams Independent of Victory Margins.” The American Statistician 57.4 (2003): 241+. Web.
Porto, Brian L. A New Season: Using Title IX to Reform College Sports. Westport, CT: Praeger, 2003. Questia. Web.
Southeastern Conference (SEC). 1999. “History of the Bowl Championship Series.” Birmingham, AL: SEC, Total Sports and Host Communications.
Staudohar, Paul D., and Barry Zepel. “3 The Impact on Higher Education of Corruption in Big-Time College Sports.” Economics of College Sports. Ed. John Fizel and Rodney Fort. Westport, CT: Praeger, 2004. 35-49. Web.