Residents Demand Answers at Council Meeting on Police Killing of Sayed Faisal


Bob Odenkirk Named Hasty Pudding Man of the Year


Harvard Kennedy School Dean Reverses Course, Will Name Ken Roth Fellow


Ex-Provost, Harvard Corporation Member Will Investigate Stanford President’s Scientific Misconduct Allegations


Harvard Medical School Drops Out of U.S. News Rankings

Markets, By Way of The BCS

By Prateek Kumar, Crimson Staff Writer

Football is not just for jocks, as Harvard Business School Professor Alvin E. Roth showed by using college football bowl games to analyze inefficient matching in markets.

Working with economists M. Utku Ünver and Guillaume R. Fréchette of Boston College and New York University, respectively, the researchers Roth compared the selection of teams for college bowls prior to and after the creation of a team matching system in 1992.

The three professors developed the idea to study bowl games while Ünver and Fréchette were research fellows together at the Business School between 2002 and 2003.

“Guillaume asked if there was a way to find a market that didn’t work and whose evidence we could estimate accurately in real life,” Ünver said. “We came up with college bowls because there is a matching system in place, even though we wouldn’t consider this to be a normal market.”

According to Roth, prior to 1992, teams would get bowl bids before the season was over, thus resulting in inefficient matchups.

“Lots of markets unravel over time, with participants making moves too early,” Roth said.

“For example, the Sugar Bowl would offer a bid to a top-ranked team that has four games left,” Roth continued, “but if that team loses one of its remaining games, then the team drops precipitously in the rankings. Thus, the rankings of the two teams are no longer comparatively equal.”

The inefficiency in getting equitable bowl matchups had a number of negative effects, including lower viewership ratings and lost advertising revenue, according to Ünver.

“Private entities want to get the best matches to earn money through ads and ticket sales while schools want to go to good bowls so that they get better recruiting for their athletic programs,” Ünver said. “This is a very important market, but the problem is that it was very unregulated.”

To mediate between the various participants in the market and correct the inefficiency, the NCAA had to act.

“This problem began to lessen between 1991 and 1992 when the NCAA pushed back ‘Pick’em Day,’ and said that [colleges] can’t get a bid before that day,” Roth said. “This increased the flexibility in the market.”

After a reorganization of the market under a centralized authority, the researchers found that, controlling for other parameters such as seasonal effects and varying yearly interest in sports, the switch was a success, as measured by higher Nielsen viewer ratings and increased revenue for market participants.

Despite focusing on the college football market, the researchers noted that their work had implications for a number of different markets.

In 2004 and 2006, Roth helped design the high school matching system in for students in New York and the matching system for Boston Public Schools.

“Our research applies to markets where price doesn’t clear the market,” Ünver said. “One example is when medical practitioners are matched with hospitals for residency. Another is when students are assigned to public schools of choice.”

“Price doesn’t play much of a role,” Ünver added, “and as a result, centralized systems are proposed to efficiently clear the market.”

—Staff writer Prateek Kumar can be reached at

Want to keep up with breaking news? Subscribe to our email newsletter.