May 17, 2009
NBA Teams Win With Ethnic Diversity
By Dan Peterson
Two new studies have shown that an NBA team's level of racial or ethnic diversity does not have any significant impact on its winning percentage or its players' split-second decision making on the court. These reassuring findings on player unity contrast with a 2007 report showing same-race bias among NBA referees when making foul calls.
The demographics of the NBA have changed dramatically over the last 40 years. African-Americans make up about 76 percent of the league's players, while Latinos and Asians account for three and one percent, respectively. According to the NBA, 77 international players from 32 countries contributed just over 17 percent to team rosters. There are not only potential ethnic and cultural barriers, but also language differences that may impact a team's chemistry.
For any organization, results matter. However, few groups of co-workers have their teamwork watched, measured and analyzed to the extent of an NBA team.
Paul Sommers and Jessica Weiss of Middlebury College wanted to see if the level of an NBA team's diversity affected its ability to win. For the last three complete NBA seasons (through 2007-08), players who had at least 800 minutes of court time were divided into one of five racial or demographic groups; African-Americans, Caucasians, East Europeans, Asians, and other foreign-born players who did not play either high school or college basketball in the United States. Using the Herfindahl-Hirschman index (HHI) to measure diversity, a number was assigned to each team for each season. An index of 1.0 would indicate a completely homogeneous team, while more diverse teams would score lower (between 0 and 1).
When the HHI was regressed against each team's regular season winning percentage, no significant correlation was found. In other words, a team's diversity did not help or hurt their success on the court. As supporting evidence, the last three NBA champions, the Boston Celtics (2007-08), the San Antonio Spurs (2006-07), and the Miami Heat (2005-06), had dramatically different HHIs of 1.0, .360, and .781, respectively.
What about that language barrier? If communications suffered, then there should be passing mixups and team turnovers should rise. To find out, Sommers and Weiss divided the teams into two groups, more diverse and less diverse at the median HHI for the league. Over the three seasons, there was no significant difference in total turnovers between the two groups.
The findings were detailed in last month's Atlantic Economic Journal.
Carrying that on-court cooperation theme even further, Brigham Young researchers searched for same-race bias in NBA players when passing to their teammates. To put it bluntly, would a white player subconsciously prefer to pass to another white player if given a choice and, conversely, a black player to a black player? In an exhaustive study, Joseph Price, Lars John Lefgren and Henry Tappen dug into six seasons of NBA data to look at every assisted basket and recorded the race (noted simply as "black" or "not black") of the passer and the scorer. They also noted the other three players on the floor when the basket was made. Of course, there were numerous decision variables that the researchers had to eliminate to isolate just racial preference.
The conclusion: No same-race bias was found in the passing patterns of NBA players. Study details are available from the Social Science Research Network as part of their working paper series.
Referees don't play fair
Joseph Price is known for his controversial paper in 2007 that concluded there is significant same-race bias shown by NBA referees. In that study, more than 600,000 officiating calls over 13 seasons were analyzed to see if white referees would call fewer fouls on white players than black players and vice versa (black referees whistling black players).
They concluded that the difference was "large enough that the probability of a team winning is noticeably affected by the racial composition of the refereeing crew assigned to the game.”
In fact, their data showed that players earned up to 4 percent fewer fouls and scored up to 2.5 percent more points on nights in which their race matches that of the refereeing crew. From a team perspective, the bias factor may change the outcome of two games out of an 82 game season. For some teams, that may be the difference that keeps them out of the playoffs.
Please visit my other sports science articles at Livescience.com