January 21, 2009

Should the BCS be made stronger?

As most of you know I have been obsessing about a concept to improve the championship system used by college football's premier division. The research and thought lead me to an unusual conclusion.

The problem with the BCS is that it is not strong enough. The BCS has performed admirably under the restrictions placed on it.

The BCS is claimed to have been formed to allow the #1 and #2 teams to meet in the bowl games. One could argue it was really designed to pit the top 2 teams against each other when the polls differed. Either way, it's mandate is to create a single game with the top two teams in the nation.

Since 1998 the BCS did not fail to pair the top two teams in the nation. College football failed to consistently provide two teams who clearly were above the rest. If there are not a pair of unanimously top team the BCS's task of pairing the top two teams is intractable.

In computer science talk: "Garbage in, garbage out."

Most cases where the BCS formula is claimed to have failed involved teams with close scores in the rankings or an undefeated team that was untested and got a boost in schedule strength with a quality win in a BCS bowl.

In each case the BCS formula was telling us something. In the former it said there are no clear cut top two teams the BCS was designed to find. In the latter the formula was telling us that the team in question has been untested. One BCS bowl win has proven it can close the gap significantly. An undefeated team beating two top 5 teams with the championship on the line would push anyone over the top in most polls.

The formula is working, we are using it badly. Instead of looking at a specific number of teams we should be looking for where the natural gaps in the standings are. Also, undefeated teams should be tested.

What should the BCS's role be?

The conferences are contractually tied to the bowls. The NCAA has let the bowls be virtually ungoverned, only playing a role in certifying them and establishing bounds for the bowl season. This leaves the college post season virtually ungoverned.

The result is the bowls have lost their identity, history and prestige. A system is needed to manage and govern the bowls. The BCS is the right entity to take on this role. This is a far greater position than the paper entity the BCS currently is.

The BCS should have three functions:
1) Apply criteria agreed upon by all involved to determine the prestige levels for the conferences. (Elite, automatic qualifier or at large)
2) Determine the prestige of the bowls based on their conference tie-ins, thereby setting the window they would be played. (BCS on New Year's Day, premier between December 26th and December 31st or regional between December 19th and Decemebr 25th)
3) Organize, with its bowl partners, the national championship game plus wild card games and play-in games as needed.

It would then be possible to create a system that:
- Builds the bowls and their purpose into its core philosophy
- Restores the excitement of New Year's Day as a celebration of elite college football
- Restores the bowl timing to represent each bowls level of prestige

- Allows all teams a chance to earn the national title by going undefeated
- Increases consensus at the eligibility cutoff

- Remains within the current post season bounds established by the NCAA
- Keeps the total number of games possible consistent with other NCAA football championships
- Does not require a shortening of the regular season

- Keeps the field limited enough to preserve the proven marketing value of the BCS ranking system and polls to add to the excitement of the regular season
- Builds on the existing structures, using the same governing bodies that the current system uses

What constituency of college football would be against a championship system that met all these goals?

Comments:

Should the BCS be made stronger?
As most of you know I have been obsessing about a concept to improve the championship system used by college football's premier division. The research and thought lead me to an unusual conclusion. The problem with the BCS is that it is not strong enough. The BCS has performed admirably under the restrictions placed on it. The BCS is claimed to have been formed to allow the #1 and #2 teams to meet in the bowl games. One could argue it was really designed to pit the top 2 teams against each other when the polls differed. Either way, it's mandate is to create a single game with the top two teams in the nation. Since 1998 the BCS did not fail to pair the top two teams in the nation. College football failed to consistently provide two teams who clearly were above the rest. If there are not a pair of unanimously top team the BCS's task of pairing the top two teams is intractable. In computer science talk: "Garbage in, garbage out." Most cases where the BCS formula is claimed to have failed involved teams with close scores in the rankings or an undefeated team that was untested and got a boost in schedule strength with a quality win in a BCS bowl. In each case the BCS formula was telling us something. In the former it said there are no clear cut top two teams the BCS was designed to find. In the latter the formula was telling us that the team in question has been untested. One BCS bowl win has proven it can close the gap significantly. An undefeated team beating two top 5 teams with the championship on the line would push anyone over the top in most polls. The formula is working, we are using it badly. Instead of looking at a specific number of teams we should be looking for where the natural gaps in the standings are. Also, undefeated teams should be tested. What should the BCS's role be? The conferences are contractually tied to the bowls. The NCAA has let the bowls be virtually ungoverned, only playing a role in certifying them and establishing bounds for the bowl season. This leaves the college post season virtually ungoverned. The result is the bowls have lost their identity, history and prestige. A system is needed to manage and govern the bowls. The BCS is the right entity to take on this role. This is a far greater position than the paper entity the BCS currently is. The BCS should have three functions: 1) Apply criteria agreed upon by all involved to determine the prestige levels for the conferences. (Elite, automatic qualifier or at large) 2) Determine the prestige of the bowls based on their conference tie-ins, thereby setting the window they would be played. (BCS on New Year's Day, premier between December 26th and December 31st or regional between December 19th and Decemebr 25th) 3) Organize, with its bowl partners, the national championship game plus wild card games and play-in games as needed. It would then be possible to create a system that: - Builds the bowls and their purpose into its core philosophy - Restores the excitement of New Year's Day as a celebration of elite college football - Restores the bowl timing to represent each bowls level of prestige - Allows all teams a chance to earn the national title by going undefeated - Increases consensus at the eligibility cutoff - Remains within the current post season bounds established by the NCAA - Keeps the total number of games possible consistent with other NCAA football championships - Does not require a shortening of the regular season - Keeps the field limited enough to preserve the proven marketing value of the BCS ranking system and polls to add to the excitement of the regular season - Builds on the existing structures, using the same governing bodies that the current system uses What constituency of college football would be against a championship system that met all these goals?