Note: Not long after the College Football Playoff's announcement, I decided to write a book explaining the new system. Well, I didn't actually budget time to write a full book and the playoff is old news, so it never came to be. I did write the first chapter, which was on why the BCS had to go away, and that's what's below. It contains some important points about the new playoff system that are still as valid today as when I wrote them years ago. —DW
The Bowl Championship Series was a bold step forward for college football, albeit a fairly conservative one by the standards of any other sport. It grew out of the Bowl Alliance (1996-97), which itself followed up the Bowl Coalition (1992-94). Those systems attempted to set up a postseason No. 1 versus No. 2 national championship game, but they couldn’t always succeed at their goal because the Big Ten, Pac-10, and Rose Bowl refused to join the party. Eventually the holdouts caved and the BCS came into being.
With no overarching scheduling governance and relatively few inter-conference games, the BCS architects had to come up with a method to rank more than 100 teams that had relatively few head-to-head links among them. They first picked a formula that included the Coaches Poll, the AP Poll, an average of computer-based polls, and a penalty for losses. The six major conferences–the ACC, Big 12, Big East, Big Ten, Pac-10, and SEC–would get a place for their champions automatically, Notre Dame would get its own special provisions, and all other conferences and independents would not be guaranteed anything. It was confusing and unwieldy, a byproduct of design by committee.
The formula changed almost annually in the early years as the sport’s leaders tried to solve whatever the previous season’s controversy was by tweaking the parameters. Eventually in 2004, they found a stable system. Two thirds of the formula would be the opinion polls, though the AP asked for its poll to be removed when the writers finally realized they were making and not just reporting on the news. It would be replaced by one conducted by the polling firm Harris Interactive. The final third would be an average of six computer polls with the best and worst rankings for each individual team tossed out.
While the formula ended up fairly simple in the end, nothing else about the system was all that straightforward. It had an increasingly large number of rules about eligibility and guaranteed spots in games as more controversies arose and pressure from the non-automatic qualifying conferences forced the leaders’ hands.
The fact that the BCS was constantly in a state of flux was the first big strike against it. Eventually all of the tweaking turned it into a Frankenstein’s monster with the core idea somewhat intact but with arcane rules bolted on all over. Every other postseason system is easy for casual fans to understand. Even the most devoted college football fans could not explain every last regulation the BCS had without having its Wikipedia page open for reference.
The next big strike against the BCS is the formula itself. Under close inspection, the logic of all three elements of it falls apart.
Including the Coaches’ Poll sounds like a good idea, because who knows the game better than those who teach it? However, coaches tend to focus on only their own teams and their upcoming opponents. They don’t take the time to really study the national landscape, so they can’t be trusted to make good calls on the truly difficult situations. It also is a conflict of interest, because bonuses for BCS game participation were common in coaches’ contracts. Nothing in the poll prevents them from boosting their own teams, their conference mates, and their opponents while dropping teams they’re fighting for position with as long as they’re not comically blatant about it.
The Harris Poll was on even shakier ground. The conferences and independent schools would send a list of potential voters to the Harris Interactive polling firm, which would then pick 115 of them at random to participate. Harris neither did anything to verify that the candidates were worthy voters nor made sure that they actually watched any games. Stories about how Harris Poll participants didn’t follow the game all that closely would come out from time to time, such as one in Yahoo Sports in 2008 when three voters admitted to not having seen undefeated Utah even play a game.
Finally, the computer polls had two big factors that made them sketchy. The first is that five of the six had proprietary formulas. That fact made it impossible to check whether they had any errors. Jerry Palm, then of CollegeBCS.com, actually found an error in the final 2010 regular season Colley Matrix system, the one algorithm that is open, showing that accuracy problems were possible. System creator Wes Colley accidentally omitted a single score between I-AA schools Appalachian State and Western Illinois in the season’s final regular season week, and that was enough to cause two pairs of teams to switch positions in the BCS top 25 rankings once he corrected the error. Fortunately neither party in the swap affected the BCS bowl participants that had already been decided before Palm caught the error, but there is no way to know that the other computer polls did not make similar errors that could have changed BCS matchups and therefore the millions of dollars that went along with them.
The other problem was that the people in charge of the BCS made a rule that the formulas could not include margin of victory. The hope was that excluding it from the algorithms would discourage teams from running up the score. That strategy did not work, however. Teams still ran up scores anyway because gaudy numbers impress the human voters that comprised two thirds of the rankings. The restriction made the computer polls much worse than they could be, and Bill James, one of the fathers of advanced statistical analysis in sports, advocated in 2009 that statisticians boycott the BCS for that reason.
Some people inside the system knew the BCS wasn’t a long-term solution. In 2008, SEC commissioner Mike Slive proposed a four-team playoff with an additional sixth bowl game to the other BCS caretakers. Only the ACC’s John Swofford was willing to even consider the plan. The BCS survived its first real challenge, but pressure continued to mount from outside the system.
In 2010, some anti-BCS forces in the media put together a self-described "definitive case against the Bowl Championship Series". Subtly titled Death to the BCS, Yahoo Sports reporters Dan Wetzel, Josh Peter, and Jeff Passan compiled every argument they could against major college football’s postseason and combined them with finely detailed investigative reporting. The book is an impassioned broadside against the BCS, one that pulls the disparate arguments against the system from their various fields all into one place.
The noise from media opponents and legions of angry fans wore on the sports’ leaders. Notre Dame athletic director Jack Swarbrick admitted as much after the announcement of the new playoff system. All that the commissioners needed was a spark to set the BCS on fire once and for all.
That spark came on January 9, 2012. On that fateful night, Alabama defeated LSU 21-0 with five field goals and one garbage time touchdown in the BCS National Championship Game.
Fans of teams from outside the SEC were already unhappy with the national title game being a rematch of two teams from that conference. The first game between the Tigers and Crimson Tide was a rather boring (to casual observers' eyes) 9-6 overtime affair, and fans nationally were sick of hearing about how great the SEC was after a member school had won each of the previous five crystal footballs. But then to see LSU fail to make it to 100 net total yards and Alabama kick field goal after field goal was too much. Never mind the fact that the game validated the rematch in some ways, as LSU was the undefeated No. 1 team going into it. The system had lost its last semblance of legitimacy among the fans as a whole. It was time for change.
For all of those many factors that helped overthrow the BCS, the aforementioned lack of a centralized scheduling system is the root of the system’s demise.
Choosing non-conference opponents is an exercise entirely handled by each program, and getting an even mix of games for the sake of choosing postseason participants is not a consideration for any of them. Rare is the school among the power conferences that routinely schedules more than one big game outside of its league. Home games against weaker teams are essentially large fundraising events, and schools need them to make ends meet for their entire athletics departments. The incentive to line up a cupcake game for easy money is and will always be stronger than the incentive to contribute towards a more intellectually sound comparison of teams across the nation.
Think about a 12-game schedule for a typical national championship contender from a power conference. Three of its four non-conference games are probably going to be body bag games. At least two of its games will be against bottom feeders from its league that have scarcely a better chance than the cupcakes do. Right away, that is 41.6% of the team’s games having little to no comparative value. But that’s not all.
The contender’s sole marquee non-conference opponent may or may not be good, as the school likely lined up game many years prior. That choice also likely had as much to do with who out there was willing to agree to the game as it was the school’s attempt to find a worthy non-conference competitor. Another two to four games will be against foes from the middle of its conference, and a national championship contender would be expected to win those kinds of games by a couple of scores.
In the end, most title contenders only play two or three games in which they have more than about a 30% chance of losing. They also seldom have common opponents unless, like in 2008 or 2011, two come from the same conference. In perhaps the most contentious conundrum in BCS history, none of the three undefeated teams in 2004 had any common opponents. The only thorough way to compare most championship candidates is either by using statistical formulas, which most fans don’t trust, or through résumé comparison, which always leads to disputes and accusations of bias.
|Virginia Tech||Bowling Green||Louisiana-Monroe|
|Colorado State||Houston||Mississippi State|
|Arizona State||Kansas State||Louisiana Tech|
|Washington State||Oklahoma State||Kentucky|
|Oregon State||Texas A&M||Ole Miss|
Not a single common opponent.
Ultimately, a single game played between two teams does not necessarily reveal which team is better. The outcome of one game between fairly evenly matched teams is more likely to come down to who happens to be having a good day than anything. However, most fans agree that eliminating one of two teams from championship contention by having them play each other is the fairest way to go about it. As long as the referees don’t end up deciding the game on a borderline call, which is always a possibility as 2002 Miami (FL) can attest to, most people will accept a head-to-head outcome as a valid way to separate teams.
Even if the College Football Playoff manages to get teams to schedule more robust slates, this problem stemming from a lack of connectivity will still exist. With only three or four non-conference slots available for major conference teams, it will still be difficult to compare each season’s national championship contenders against each other at the regular season’s end.
That fact is why the any pressure on the new system will invariably be aimed towards a larger bracket rather than towards heading back to the old system.
The BCS had to die, and it is never coming back.