Skip to Main content

He's Got Game ... Theory

Aug 11, 2010
by Priscilla Proctor

Associate dean’s research debunks idea of bias in NBA scheduling

Was there some kind of nefarious conspiracy at work in scheduling NBA basketball games? That was the question Associate Dean of Academic Affairs Yvan Kelly answered last fall in research published in the “Journal of Sports Economics.” His article was titled, “The Myth of Scheduling Bias with Back-to-Back Games in the NBA.”

Kelly, a former economics professor and scout for the Seattle Supersonics, employed game theory to analyze five seasons of data from NBA schedules to see if there had been any potential scheduling bias. His research was prompted by a noticeable pattern: certain teams were playing games two nights in a row, and most of the time they would lose the second game.

“Teams get tired, and they’ve been traveling,” he said. “They’re not left with much time to practice or rest in between games. So the conspiracy theorists wondered, ‘Are back-to-back games being rigged by the NBA?’ I found out that it was not.”

Using game theory, which uses mathematical models to predict outcomes, his data showed that the number of back-to-back games assigned to each team appeared to be random, and that there was no evidence of bias in scheduling.

Kelly said, “Although differences in the number of back-to-back games do result in a slight variation in competitive balance, one positive externality is that these types of games reduce travel costs by millions of dollars per year.”

The NBA article was actually Kelly’s second on game theory in the past year. He also had an article on the Austrian connection to early game theory published in “The Quarterly Journal of Austrian Economics.”

Kelly teaches a course at Flagler on game theory, as well as economics of sport.

Tagged As