Referee Bias

On the last day of every golf tournament, Tiger Woods insists on wearing a bright red polo shirt. Woods says the habit is merely superstition, but new research suggests that his fashion sense might actually come with athletic benefits. A paper published this month in Psychological Science reports that referees and umpires subconsciously favor competitors in red uniforms. The experiment was clever: the scientists showed 42 experienced tae kwon do referees video clips of five different male competitors. Each clip featured one athlete in red and another athlete in blue. At first, the referees were shown the original videos, and asked to score the match. Then, they were shown the same clips but with the colors digitally swapped, so that the athlete originally wearing red was now wearing blue. This single alteration had a significant impact on the outcome, with competitors dressed in red scoring, on average, 13 percent more points than their opponents in blue.

This experiment builds on a 2005 paper demonstrating that athletes wearing red in the Athens Olympics consistently outperformed athletes clad in blue uniforms. While some scientists speculated that the effect was due to the color red causing athletes to perform with more aggression, this new data suggests that the effect is largely due to referee bias. Call it the Chris De Burgh effect.

Of course, this isn't the first time scientists have examined the fallibilities of referees. As I noted in a recent Globe article:

A 2002 experiment showed professional soccer referees a videotaped match and had them make officiating decisions. Half of the referees watched the game without sound, while the other half were exposed to simulated crowd noise. These cheers significantly biased the calls of the referees: On average, they called 2.3 fewer fouls against the home team when listening to the sound of the crowd.

Studies of actual soccer matches in the English Premier League support the experiment. (The English leagues are a popular subject for researchers because the matches feature animated audiences and take place within the same time zone, reducing the complicating factor of travel fatigue.) A 2007 analysis of more than 5,000 soccer matches found that, on average, home teams scored 1.5 goals while away teams scored 1.1 goals. This difference increased with crowd size, so that each additional 10,000 spectators increased the home-team advantage by 0.1 goals. The most surprising element of the research, however, was that the scoring disparity was largely the result of referees, with less experienced referees calling significantly more penalties against the visiting team. They seemed intimidated by the rowdy fans.

A study of the Winter Olympics found that, although host countries enjoyed a large advantage - they collected significantly more medals than normal - this edge was limited to "subjectively judged events," such as freestyle skiing and figure skating. (There was no host-country advantage in timed events.) This suggests that cheering crowds influenced the verdicts of the judges, not the performance of the athletes.

Some researchers, however, caution that it's difficult to tease apart referee bias from player performance. "It's possible that visiting teams actually commit more fouls or perform at a lower level," says Carron. If that's the case, then the "bias" is actually evidence of referee impartiality. When Carron investigated the number of penalties called in error during 42 televised NHL games, he found no difference between home and away teams. The cheering crowd had no effect on officiating mistakes.