There has been a lot of debate surrounding the BCS since its inception in 1998. You have no doubt heard the exhausted rhetoric from both sides by now. College Football Cafeteria is unabashedly against the BCS and in favor of a playoff. I’ve decided to use the most logical argument I can make against the BCS and it’s inherent flaw, so here goes.
First, what is the BCS? What was it designed to do? This is straight from their website:
The Bowl Championship Series (BCS) is a five-game showcase of college football; designed to ensure that the two top-rated teams in the country meet in the national championship game, while creating exciting and competitive matchups among eight other highly regarded teams in four other games.
The BCS has done a great job at what it was designed to do. It always puts the two top-rated teams against each other in a championship and then puts other highly regarded teams in bowl games against each other. It has been successful in growing both interest and revenue for the sport. It is a much better bowl system with the BCS than it was without.
The case against the BCS can be made in the logic it employs. The BCS says “the two top-rated teams in the country meet in the national championship game.” This raises questions and makes assumptions that I think are the inherent flaws of this system.
First of all, how are the teams rated? We have 120 teams in the FBS, so of all 120 teams, the BCS claims to know the two best. It doesn’t say this, but if the BCS rates two teams higher than the other 118, then allows them to play for the championship without including anyone else, then the assumption is that the BCS knows who the best two teams are.
The BCS employs a three part method in determining the two best teams. To two highest rated teams in the country are the teams whose average of the three parts are higher than everyone else. The three parts are a poll of about 60 coaches, a poll of about 100 something former players and coaches and 6 computer rankings.
The two polls are straightforward. You get points based on the votes you get. So if there are 60 coaches and a first place vote counts for 25 points, the most points a team can get in the poll is 1,625. If a team gets 1,625 points in the poll, their BCS score is 1.000. If they get 1,500 points in the poll, their BCS score is 0.923. Same methodology applies to the other poll.
The computer rankings are a bit tricky. There are six computer rankings. Your highest and lowest rankings are dropped and only four are used. Like the polls, a 1st place ranking from a computer is 25 points, so the most you can get is 100 points. They add up the points from your four computers and then just divide by 100. So if your six rankings were 4, 3, 4, 9, 6, 6, they would drop the 3 and 9 and you would get 4, 4, 6, 6. Your BCS score would be 0.840.
Your three BCS scores are then averaged and that’s how they rate the top two teams. They then feel confident enough in these ratings that they can simply proceed to the championship game since they are sure they have the two best teams in the country.
Here is why this approach is logically flawed. One third of the ranking is determined by the vote of 60 or so coaches. Most of these coaches have never seen more than five of the teams in their top 25 actually play a game. They obviously cannot evaluate each team based on performance, so what metric do they use to rank their teams? And what about coaches whose teams are in the top 25 that vote in the poll? Is their opinion an objective measure of a team’s actual rating?
A computer ranking is possibly even worse. The computer rankings are completely beholden to who you play. If two teams go undefeated, then the team who played the harder schedule is rated better. Why? Why was Auburn the unanimous computer #1 last year? Because they beat Alabama, LSU, Arkansas and other good teams. So what? If TCU had played these teams, who’s to say they wouldn’t have beaten them worse? Suppose TCU crushes everyone and goes undefeated, why are they discredited because of who they played? If Auburn had played the exact same schedule, wouldn’t they have crushed everyone too? Everything is determined by who you play with the computers. Heck, one computer ranked TCU 7th last year and put two-loss Oklahoma, Arkansas and LSU ahead of them. Fail.
So what is a fair rating system you ask? Well, we’ve clearly shown that the BCS uses an unfair rating system. They always get two top rated teams each year, but the way they get there is incredibly flawed. Now the ideal ranking system doesn’t exist nor can it. Ideally you would have all 120 teams play each other. We all know a 119 game schedule is impossible. The next best thing statistically would then be to put teams into small groupings and then have each team play the others in their groups and play a couple teams from every other group. Then you have cross linking data. This would still be a 35 game schedule. Not happening.
So there really is no way to stack up all the teams against each other and see how they compare. Which means there really is no way of knowing which team is better than another if those two teams never played each other or a common opponent. Which means the BCS is inherently flawed for thinking it can use an unfair rating system to find which two out of 120 teams are the best. What they think they are doing simply cannot be done. They can always get 2 very good teams, but we will never know if they are the best teams.
So there are two main flaws with the BCS. First is their ratings system which is in no way an accurate measure of how good any team is compared to another. Second, they operate under the false assumption that any rating system in a 12 game schedule can determine the two best teams out of a group of 120. It is unequivocally impossible.
Case in point: the BCS thought Michigan State was better than Alabama after 12 games last year. The BCS thought Missouri was better than Iowa after 12 games last year. The BCS thought Cincinnati was better than Florida after the regular season in 2009. They thought Miami was better than Wisconsin after the regular season in 2009. I can do this all day long. The BCS has a very, very long history of rating one team better than another only to have that lower rated team prove they are better than the higher rated team. The BCS has been proven wrong time and time again, yet we trust it to tell us which two teams are the best in the country? It cannot be done.
Logic and common sense dictate that the BCS is flawed. It cannot do what it implies it is doing.