BCS Follow Up

The computer rankings once again have Oregon ranked 8th in the country, once again pulling the Ducks down to the number 2 spot in the most recent BCS poll. Last week it was a Big 12 team; this time they find themselves second to an SEC team, Auburn.

Before I look too closely at Auburn, I want to talk about the six computer rankings that make up 1/3 of the weight in the BCS poll. None of them takes margin of victory into account, and only one–the Massey ratings–separate out comfortable wins from the nail-biting sort. Here’s what that means. Hypothetically, say UCLA marches into Austin, Texas and beats a ranked opponent on the road by 22. Then, pretend that UCLA waddles into Autzen Stadium, and loses by 47. The computers don’t give a shit what the score was. UCLA could have won an ugly game by 1 point against the Longhorns, then lost a heart-breaker on a last-second field goal to Oregon, and we would have the exact same rankings today. The first of those scenarios did happen, but Oregon is not reaping the benefits of dominating performances.

In fact, according to Rivals.com, the BCS chose this set of computer rankings specifically because each one didn’t use margin of victory. Says Peter Wolfe, creator of the Wolfe Rankings, “Running up the score is generally looked on as evidence of bad sportsmanship, behavior which should not be encouraged or rewarded.” SERIOUSLY!?!? We’re going to take out probably the most important component of assessing team strength in such a short season because we don’t want coaches running up the score to get higher computer rankings. Let them run up the frickin’ score. If a team wants to show its dominance, it should be allowed to do so. This would give us a good representation of a team’s true ability.

If you feel that winning teams will be running up the score in the fourth quarter against inferior teams’ third units–which is a definite possibility–then I suggest you take a look at the square root function. The square root function (among other functions) can be used to model diminishing returns. In other words, extending a lead from 40 points to 50 points would not be nearly as important as extending a lead from 0 to 10. There would be very little payout for running up the score, but 30-point wins would still get more weight than 10-point wins. This is an Econ 200 problem, and an organization which decides the fate of millions of dollars and about 12 major universities can’t figure it out.

Jeff Sagarin, one of the more well-known computer/football analysts, is in charge of one of the other six computer polls. “[His] rankings that include margin of victory are featured in USA Today, and he considers them his ‘best effort,’ ” says Rivals.com, yet the BCS won’t take his best effort, instead demanding a poll that doesn’t include margin of victory. So Sagarin trots out his non-margin of victory poll for the dumbshits at BCS headquarters. This is truly sad.

But back to Auburn. Out of conference, the Tigers beat Arkansas State, Clemson, and Louisiana-Monroe, all at home. They beat a Clemson team that’s down this year by only 3 in overtime. This is not an impressive resume. In conference, they have beaten LSU (3), Arkansas (4), South Carolina (5), Mississippi State (6) and Kentucky (9) by a combined 41 points, 8.2 per game (approx. conference rank in parentheses). Oregon beat the second best team in the Pac-10 by 21,  and the three worst teams by 78 points combined, an average margin of victory of 24.75 points.  While Oregon has yet to play some of the tougher teams in the conference like Auburn, a margin of victory three times that of the Tigers’ has to be taken into account.

When Tennessee went down to the Bayou to play at LSU, they lost by just 2. But only the Massey poll cared that it was an impressive two-point loss for the Volunteers. Oregon’s stock simply went down that week because Tennessee lost. There’s no way, in my opinion to even remotely rank teams correctly if margin of victory is not a factor.

Advertisements

2 Responses to BCS Follow Up

  1. Drew says:

    The irony of this whole thing is that the BCS used to incorporate margin of victory, but it was eliminated after the 2001 season because of the Oregon Ducks.

    From wikipedia: “The season was riddled with close games, 6 of which ended with a spread of one score or less, coining the nickname “Captain Comeback” for Harrington. The only loss of the season came to the Stanford Cardinal at home. At the end of the regular season, the Colorado Buffaloes routed the number two ranked Nebraska Cornhuskers[37] and then edged the number three ranked Longhorns in the Big-12 title game,[38] sending the BCS rankings into a tizzy. The Ducks ended up with a number two ranking on both the AP and Coaches polls, but the computers dropped them down to a number four ranking, eliminating them from the national title game. Instead, the Huskers went on to play the Miami Hurricanes for the national title at the Rose Bowl. This caused the BCS committee to alter the ranking system for subsequent years to a formula which, if applied in 2001, would have placed Oregon in the national championship game.”

  2. uoduckfan33 says:

    There are definitely drawbacks to margin of victory, but those drawbacks can be reduced to a certain extent. In baseball, teams with a good bullpen have a better-than-expected chance of winning 1-run games, and margin of victory may miss this. In basketball, that that play situationally well (manage the clock, the shot clock, and make freethrows) can win more than their margin of victory would suggest. I would bet that football is the same way, in that teams with good kickers and clock management have advantages late in the game that margin of victory won’t specifically pick up. That being said, adjusting MOV using something like a log or square root function can give 2 point wins more “relative strenght” to 30-point wins versus taking MOV at face value. I think there are many ways to work around the MOV issues, but to leave it out is not one of those.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: