BCS drops margin of victory from rankings formula

Dent: No computers in old school

Mel Kiper Archive

Tuesday, March 8

BCS needs margin of victory

The BCS is overreacting with its decision to eliminate margin of victory as one of the variables in the computer rankings.

No one can expect third-teamers to give anything less than 100 percent.
Margin of victory must be included because it helps in comparing and contrasting teams. The BCS may feel teams are running up scores to influence the voters and computers.

However, just because an outcome looks too one-sided on paper doesn't mean the winner purposely ran up the score. A team can win 60-10 and not run up the score if it merely ran its plays and the other team was unable to compete.

Victory Margin Reflects a Team's Depth
Whether a score has been run up is one of the great water-cooler debates each week. But do we really care if a team was pouring it on at the end? Is a team more humiliated losing 45-7 or 60-7? I don't know. In the players' eyes, a loss is a loss -- no matter what the score. If a team can't compete for four quarters, don't play. Forfeit the game.

Outlandish scores are mainly a result of one team being unable to compete physically. Teams must be prepared to play four quarters. They can't expect to put third-teamers in the game and pull back the throttle. These players are trying to win jobs, earn respect and impress coaching staffs. No one can expect them to give anything less than 100 percent. If one team can't compete against another, it shouldn't cry about how the other team ran up the score.

To evaluate the strongest football team, one has to look at its depth, not just its starters. The second- and third-teamers are part of the evaluation both when they enter the game as subs and when they have to fill in for injured starters.

Without the margin-of-victory variable, four of the eight computer rankings may no longer be part of the BCS process. But I have always believed the system is better off with fewer computer rankings. It's more important to have voters watching the games and seeing how the scores came about. A final score isn't always indicative of how the game went. If a team won by 30 points, how did it happen?

If a team is winning 35-7 late in the third quarter but won only 35-24 because the backups gave up some inconsequential points in the fourth quarter, will that team be downgraded because the score looked closer on paper? Or will the voter say, "One team was dominating until the other team scored a few cheap touchdowns at the end"?

Polls Come Out Too Soon
This leads to what I feel is the single biggest flaw in the current system: The votes in the polls are cast too prematurely each week for voters to accurately evaluate the teams, making the polling process less credible than it should be.

The voters -- whether they're writers, coaches or others -- cannot be expected to judge every team by Sunday. It's physically impossible. Nobody can study four quarters of each game that quickly and feel good about the votes they submit each Sunday.

The voters participating in the voting system need to be evaluated. Based on their job requirements, how many games are they watching each week? One or two games? If they are at a game on Saturday afternoon, they get reports and highlights of the others, including printouts of the stats for each game. But how many did they see?

By the end of the day Saturday, Dave Revsine and I watch about 30 games for College GameDay on ESPN Radio. We looked at them superficially, but how many did we study? Extra time is needed to evaluate the teams.

Instead of forcing voters to make rash decisions each week, I propose that there should be a board whose job requirement is to watch and study the games. With games now being played on Thursday nights, give the board until Thursday to come out with the rankings.

Why do the polls have to come out so early? Let the debate rage through the week, with anticipation building until the rankings are released. Knowing that the board carefully studied each game and team, we'll look forward to seeing what it came up with.

College football is a serious business with millions of dollars at stake. Every vote has a dramatic impact on what happens in the top 25, which team ends up in the BCS and which teams are playing for the national championship. Special time and consideration should be taken to ensure the process is fair, accurate and credible. And that goes far beyond merely eliminating margin of victory as a computer number.

Send this story to a friend | Most sent stories

Copyright ©2002 ESPN Internet Ventures.
Click here for Terms of Use and Privacy Policy applicable to this site.