Column » Sports

Wichita State's flawed perfection: Why the NCAA should ditch the RPI rating system

Every year in college basketball, it seems like there is that one mid-major team that rips through its weak conference schedule and ends up ranked higher than many people would like. Last year it was Gonzaga, which somehow ended up first in both polls going into the tournament, despite its 1-2 record against ranked opponents. The Bulldogs pretty much ended up there by default since they started out with a preseason rank of 21 and consistently moved up as teams ahead of them lost. Yet how many fans, coaches or analysts actually considered Gonzaga the best team in the country? I boldly predicted that it would become the first one-seed to lose in the first round, and it ended up barely surviving Southern University before losing in the second round.

For some reason, the polls are unwilling to penalize a team for a “bad win” or jump a team past another inside the top 10 when both teams have won. To humans, the two-point difference between a one-point win and a one-point loss is far greater than any other two-point margin, whereas in most computer rankings, there is no premium placed on winning — a close loss to a good team can be better than a close win over a bad team.

If you are a college basketball fan, you might know where I am going with this. Wichita State is this year’s powerhouse mid-major, but it is not your garden-variety shocker. These Shockers (yes, this is its mascot) happen to be ranked second in both polls, a feat rarely achieved by a non-power conference school. But what has really gotten everyone talking is their undefeated record. At 31-0, Wichita State is the first team in 10 years to survive the regular season without a loss. If it wins the Missouri Valley Conference tournament, it will set the record for consecutive wins to start a season, currently held by North Carolina and Indiana, two teams that went 32-0 and won the NCAA tournament in 1957 and 1976, respectively.

All this is great, but what every college basketball season comes down to is the NCAA tournament. The Shockers are a projected number-one seed, based on their high ranking and RPI rating, however, I feel they should not be placed this highly due to the relatively poor predictive power of poll position and RPI rating. The goal of seeding is supposed to be to reward teams for their level of performance throughout the entire season. RPI, invented by the NCAA in 1981 for the specific purpose of seeding the tournament, does not take into account the actual scores of any games; it instead ranks teams based solely on their win-loss record, the win-loss record of their opponents and the win-loss record of their opponents’ opponents. It has been shown to predict fewer correct matchups in the NCAA tournament than any of the major computer rankings, which do take into account scoring margin. The BCS system, used in college football, has acknowledged the improved accuracy of the computer ratings and uses a number of them in its formula. The college basketball committee ought to stop clinging to its flawed traditions and use the computer rankings in the seeding process. Ken Pomeroy, the inventor of tempo-free basketball statistics, publishes his own computer ratings, which rank Wichita State sixth. Jeff Sagarin of USA Today, who creates computer ratings for a variety of different sports, has them just 16th. I have come up with my own simple rating system that also uses tempo-free stats and places the Shockers in 20th. Computers aside, the fact that they have not played a team in the top 20 in any kind of rankings make me loath to call them a one-seed. You have got to beat the best to be called the best, and Wichita State has not done this.

I probably would not be writing this unless I had some sort of Princeton tie-in. Princeton has made the NCAA tournament 24 times and has often encountered the injustice of the seeding system, albeit in ways opposite to Wichita State or Gonzaga. Take the 1997-98 team, for example. That season’s Tigers were 27-1 entering the NCAA tournament, with the lone loss coming in a near upset of No. 1 North Carolina in Chapel Hill. Teamrankings.com is the only website I could find with computer rankings that go back that far, and it had the Tigers at number nine headed into the tournament. My own rankings put them 10th. RPI, however, had them 24th, and they ended up as a five-seed instead of a three-seed as the computers suggested. They ended up losing a heartbreaker in the second round to fourth-seeded Michigan State, a young team with all the pieces in place that earned them a national championship two years later. No one can say how far the Tigers would have made it as a three- or four-seed, and it is a shame they were not given a chance.

comments powered by Disqus