Tuesday, December 05, 2006


BCS Gets It Right

For the wrong reasons

Unbeaten Ohio State will play 12-1 Florida for BCS national title. Michigan, 11-1, had been #3 with Florida #4 in the BCS rankings. USC, which lost to UCLA on Saturday was #2. Florida passed Michigan in the final rankings after an unimpressive victory over Arkansas in the SEC title game. Having played the tougher schedule between the two remaining major conference, one-loss schools, Florida deserved the nod. The computerized third of the BCS ranking formula rated both schools even. So it was up to the two polls to determine the outcome. ESPN's Pat Forde comments:
On Nov. 26, the Wolverines led the Gators by 86 points in the Harris Poll and 30 points in the USA Today poll. By Sunday morning there had been a 154-point reversal in the Harris poll and a 56-point swing in the USA Today poll.

That was shocking. If you were already predisposed to voting Michigan ahead of Florida, I didn't see enough in that game to merit that kind of turnaround. We certainly didn't see anything from Michigan to merit a demotion, given the fact that the Wolverines weren't playing.
Teddy Greenstein further explains the controversy:
Even Ohio State coach Jim Tressel got swept into the controversy after he decided not to cast a vote in the USA Today poll.

"We felt it was somewhat of a conflict of interest," he said.

You think? Had Tressel voted Florida No. 2, he would have dissed the Big Ten and given his top rival's fans fresh ammunition. Had he voted Michigan No. 2, it would have gone beyond bulletin-board material. He would have supplied new wallpaper for the Gators' locker room.

The remaining 62 coaches gave an emphatic endorsement to Florida for the No. 2 spot, with 44 apparently voting the Gators ahead of the Wolverines.

The Harris Poll, made up of 114 former players, coaches, administrators and media members, also gave a solid edge to Florida. One of them, former Washington State and Iowa State coach Jim Walden, actually voted the Gators No. 1.
Many, including Forde and Captain Ed, maintain that this whole sordid affair is a compelling argument for tournament style Division I college football playoff. It is not.

A playoff would destroy the the one aspect of Division I college football which makes it truly unique in the world of sports: No team is ever guaranteed a second chance; to guarantee at least a chance to play for a national championship, you must win each and every game. The best illustration of this is the fate of Michigan this year. Under a tournament format, the loser of the Ohio State/Michigan game would simply receive a lower tournament cede. What was at stake in the game would have been akin to a winning a conference title in college basketball. Instead, both teams knew that the loser could still, with luck, get a second chance; but more likely than not, the loser was out of national championship contention. The game was, in essence, a playoff. While Michigan is arguably a better team than Florida, they have only themselves to blame for their fate.

The regular season frequently produces games games like this, though its not always this obvious at the time. Many games that appear this meaningful at the time turnout not to be. And sometimes the stakes are apparent only in retrospect. That is what makes the regular season in college football so compelling. Its something very special and should be preserved.

Though I think Michigan is a better team than Florida, and a Michigan/Ohio State rematch would make for a more compelling and competitive game, Florida deserves the spot due to its strength of schedule. They are the more accomplished team.

But the way the Gators were selected is scandalous. The 'human element' should be removed from the business of ranking teams altogether as bias and conflict of interest are inevitable in any polling system. Instead, teams should simply be ranked by fewest losses, with ties broken by a strength of schedule formula which uses both the winning percentages of a team's opponents and the winning percentages of their opponent's opponents, with 2/3 weight given to the former. To qualify for the national title game, a team's strength of schedule must fall within, say, the top two quintiles of all Division I schedules. Not a perfect system, perhaps, but at least the only 'human element' involved in the process is the one actually on the field.

Diego corrects: If you want to add a rule that you have to win your conference to make the BCS title game I can agree with that. But your strength of schedule basis for picking Florida is wrong. In fact it is ridiculous. They both played tough schedules in tough conferences. Your reasoning might hold up for Boise St. but not for Michigan. Florida played two non conference games: Central Florida and Western Carolina. I'm okay with a local game but when given a choice they went soft. Michgan played Central Michigan and then Vanderbilt but they also played Notre Dame. Give them credit for that.

Keep in mind that they are fighting for #2 here. If your only loss is a 3pt loss to the #1 team on their home field then you don't really have a blemish. You would think that a #2 team would lose to #1 on the road, no? But you would not expect a #2 to lose to a #9. There is the difference. Michigan played like a #1 or #2 team all season and finished that way. Florida did not. They did not play like a #1 or #2 team against Auburn. Same goes for USC against Oregon St. But I guess some games just don't count anymore. That is the situation that college football should try and avoid. The fact that every game counts is "something very special and should be preserved."

USC had no business leaping over Michigan after their ND win. And let's be honest. If the UofM/OSU game was played just last Saturday Michigan would be ranked #2. Florida is only in because of timing and some politics. Such reasoning behind the selection of Florida over Michigan places in effect a playoff but without the inclusiveness and excitement that should come with it. You don't have the pro system but you don't have the college system anymore either. This is a shame.

John O responds: Alas, I miscalculated Florida's strength of schedule in my head. (I guess I shouldn't have been so lazy.) In doing the math, Florida's opponents had a combined winning percentage of .568, Michigan's opponents, .587; the difference being .019 in Michigan's favor. So my conclusion was in error. Considering the schedules of the two schools were substantially similar in strength, however, it wasn't by any measure ridiculous.

(I must correct Diego on something. He overlooks the fact that Michigan's non-conference schedule included Ball State whereas Florida's included Southern Mississippi and perennial powerhouse Florida State, a local game where they hardly 'went soft.'

Also, if Diego believes it reasonable to expect a #2 team to defeat #9 team, would he be willing to make book on such a basis? If so, please let me know.)

Diego argues that Michigan's loss to OSU was, in the parlance of the NCAA selection committee, a 'quality loss,' which indeed it was. But this is a poor basis for distinguishing between teams; football games are binary. Whether or not Florida or Michigan had a quality loss, or for that matter played like a #1 or a #2 all season long, or finished that way, should ultimately be irrelevant. It matters only because the ranking process is so arbitrary, open to blatant manipulation. The ranking system I have proposed allows matters to be resolved where they belong -- on the field.

Diego responds: I missed the non-conference games played in the 'conference' portion of the schedule. Florida State is not a soft opponent.

The next time a #2 plays a #9 I will take the #2 and give you 1 point. Or if you like I will give you odds of +101 for #9 to win outright.

If you are going to rank reams based on their accomplishments then just have a playoff. That way the most accomplished team will be the one that accomplishes the task of making it through the tournament. All accomplished on the field.

If you don't recognize the difference between 'quality losses' then why bother with quality wins? Both Michigan and Florida played tough shcedules. The losses are the only real way to compare the two.

John O responds: A few points:

I wasn't asking for a line on the next #2 vs #9 matchup. I simply wanted to point out that no bookie makes lines based on the assumption that the higher ranked team ought to win. A bookie who does this won't last very long.

Obviously a playoff would produce the most accomplished team as champion. Its just that I adamantly oppose a playoff for the reasons I stated in my post. Besides, I never said I want to rank teams based on accomplishment. My idea is to rank teams by lack of failure, with ties to be broken by an objectively defined measure of accomplishment, primarily for the purpose of producing a national title matchup. (Though I do think its a good way to rank teams in that it removes the 'human element' from the equation.) If you have any ideas about how to measure the quality of failure objectively, or eliminate the subjectivity inherent in any poll, I'm all ears.

Diego notes: Had Wisconsin beat Michigan they would be undefeated (they did not play OSU this year). I wonder if voters would be reluctant to have what would effectively be the Big Ten championship game played out in the BCS Championship game. That situation also would have made for some complaints.


<< Home

This page is powered by Blogger. Isn't yours?