March 14, 2002

Keeping the Game Close
NIU researcher says national TV spotlight
linked to how refs call college hoops

DeKalb, Ill.- In a new study that would seem to confirm the suspicions of every college basketball fan who ever griped over officiating, a Northern Illinois University researcher is whistling a foul on the referees.

Kendall Thu, a faculty member in NIU's Department of Anthropology, says his study detects an unusual foul-call pattern in NCAA Division I officiating. However, it isn't the screaming of home-court fans that influences referees, but rather the presence of a national television spotlight.

"Referees tend to keep nationally televised games close by calling a significantly higher number of fouls against teams that are ahead in the score," Thu said. "This results in more competitive games that maintain an edge of suspense for viewers. Interestingly, we don't find the same pattern in games televised on a regional basis."

In the games televised on national networks and analyzed by Thu, referees called 58 percent of the fouls against the leading (footnote) team (excluding calls during the last two minutes of each half and tie-game situations). Thu's study results are published as the lead article in the spring edition of "Human Organization," the flagship journal of the Society for Applied Anthropology (on the Web at http://www.sfaa.net/).

The study suggests that the evolution of college athletics into big business might alter the way in which rules are enforced. "The rules of fair play may actually be the rules of keeping the game close in order to create a product for spectatorship in a very commercialized setting," Thu said.

It's not unusual for cultural anthropologists to apply scientific methods of study to sports, which is one of about 70 traits found in all known cultures.

Anthropologists have a slightly different perspective that allows us to see this sort of 'exotic reality' in the midst of the mundane," Thu added. "Anthropologists do this all the time, it's just that the topic of basketball might happen to capture greater public interest than other topics."

Thu got the idea for the foul-call analysis more than a decade ago while pursuing graduate studies at the University of Iowa. "I've always been a basketball fan," he says. "Fellow faculty and students would always grumble about foul calls. I wanted to know if their perceptions had any merit."

Thu's initial game observations indicated the topic was worth pursuing. But when he contacted the NCAA and Big 10 Conference officials, he found little long-term data on officiating. "I was amazed because there's so much in the way of statistics kept on teams and players," he said.

After arriving at NIU in 1999, Thu enlisted help for his project from anthropology graduate students Kelly Hattman, Vance Hutchinson and Scott Lueken and undergrads Nathan Davis and Elmer Linboom. All are co-authors on the study, titled "Keeping the Game Close: 'Fair Play' Among Men's College Basketball Referees."

From Jan. 5 to April 3, 2000, Thu and his students spent hundreds of hours recording 67 televised games and logging 2,441 foul calls. At first blush, there appeared to be little difference in the proportion of foul calls against leading versus trailing teams. However, it became apparent that the last two minutes of each 20-minute half accounted for special circumstances-fully two-thirds of fouls went against the more aggressive trailing team.

"The last two minutes of each half, particularly the last two minutes of the game, reflect a distinct period because trailing teams are more likely to commit blatant fouls that can't be ignored by referees," Thu said.

When the researchers excluded the last two minutes of each half, they found a different pattern entirely. The average number of foul calls against the leading team was 16.1 per game, compared to 13.3 against the trailing team.

Next, the games were grouped into three types of television coverage-national TV networks, the primary ESPN network and regional telecasts. The researchers worked with the expectation that foul calls would be evenly distributed between leading and trailing teams. Excluding the final two minutes of each half and tie-game situations, the results were as follows:

• During 15 games televised on national networks, the researchers observed 478 foul calls. Leading teams were whistled for 58 percent of the fouls or on average 5.3 more fouls per game than the trailing team. (footnote)
• During 25 games televised on ESPN, the researchers observed 709 fouls. Leading teams were whistled for 55 percent of the fouls or on average 2.8 more fouls per game.
• During 27 games televised regionally, the researchers observed 782 fouls. Leading teams were whistled for 53 percent of the fouls or on average 1.5 more fouls per game.

"This last instance was the only case where our original statistical expectation that fouls would be about evenly split was correct," Thu said.

"There's something funny going on," he added. "It's improbable that our findings would be attributable to random chance. For the nationally televised games we examined, the likelihood of the proportion of fouls being skewed to this extent by chance is three out of 1,000."

The findings appear counterintuitive to what fans might expect. "Trailing teams would seem more likely to take risks and be more aggressive. So you would think they would be whistled for fouls more often," Thu said. "Also, teams are coached to avoid silly fouls when they're leading."

Thu said he hopes the study will be a catalyst for further scrutiny.

"I must underscore that we are not suggesting impropriety on the part of referees," he said. "I don't think referees are being paid off, and I don't think they are in the pockets of corporate television. Our study did not attempt to identify the mechanism by which this foul-call pattern occurs. The links might be very subtle, or even subconscious."

Thu also examined foul calls against both home and visiting teams. He found that referees called 51.8 percent of the fouls against the visiting team, versus 48.2 percent on the home team. Thu said the slightly higher percentage is not a statistically significant departure from the assumption that fouls should be evenly distributed between the two teams.

-30-