Researchers' NBA officiating study detects biases, but not necessarily the ones fans suspect
by Rachel Bachman, The Oregonian
Wednesday June 03, 2009, 5:24 PM
But the same study found that NBA referees tend to favor home teams, teams trailing in a game and teams trailing in a playoff series.
The study, conducted by three economics researchers, fuels the perennial debate about the influence of NBA officials on games. It suggests that forces ranging from league executives to simple human psychology can influence calls in a measurable way -- though not always enough to affect a game's outcome.
The researchers looked at six seasons of turnover statistics. They used ones where referees wield relatively little influence, such as a bad pass or steal, as a "control" group, and compared them with ones where referees wield greater influence, such as traveling and offensive fouls.
The researchers found that each type of favoritism -- home, trailing in a game and trailing in a series -- resulted in a 5 to 10 percent advantage in "discretionary" turnovers, or ones over which referees have the most influence. The researchers do not attempt to explain what the percentages could mean in actual wins and losses.
Still, the study concludes that the detected referee biases, though probably unintentional, could increase the league's revenues through additional ticket sales and television appearances by reducing the number of blowout games and making televised games more compelling.
"We can say with fairly high confidence that the results are not just due to randomness or (statistical) noise, that even teams facing elimination have an additional advantage in these referee-based turnovers, discretionary turnovers," said Daniel F. Stone, assistant professor of economics at Oregon State University and one of the study's authors.
NBA executives condemned the study, which is dated March 2009 and has not been published in a scholarly journal, criticizing its methodology, tone and conclusions. The league, whose showcase event, the Finals, starts tonight with the Los Angeles Lakers hosting the Orlando Magic, rigorously monitors officiating and does not condone favoritism, said Joel Litvin, the NBA's president of league and basketball operations.
Litvin also dismissed the study's suggestion that fans influence referees' behavior, a finding that echoes those in research about European soccer. The basketball study finds that the home-team advantage in discretionary turnovers increases by 1 percent for every 1,000 people in attendance.
Based on that figure, when the Trail Blazers ranked third in the league in average home attendance last season at 20,524, they should have received more than a 5 percent advantage in discretionary-turnover calls over the 2005-06 season, when they were last in the league with 15,049.
Litvin responded that NBA officials are the best in the world, reaching the league through superior skill and imperviousness to criticism.
"I do believe, and I think it's the case," Litvin said, "that these people are, in fact, immune to the things that you and I would say are just human nature."
Study genesis, response
Stone and fellow graduate student Marc Remer decided while at Johns Hopkins University in Baltimore last spring to test the popular conspiracy theory that NBA officials favor teams facing elimination to extend a playoff series.
Stone and Remer recruited help from Joseph Price, an assistant professor of economics at Brigham Young University who co-authored a study publicized in 2007 that found that NBA officials tend to favor players of their own ethnic backgrounds.
After collecting data for the latest study, culled from play-by-play accounts on ESPN.com, the researchers stumbled into the "home" and "close-game" bias findings, Stone said. Price said the study's most interesting conclusion is that officials might exhibit biases -- conscious or unconscious -- that could benefit their employer.
The researchers also found foul advantages for home teams and teams trailing in a game, but did not emphasize them in the study.
"Our results on fouls are weaker since we can't draw the same distinction between whether they're referee or player-driven as we can for turnovers," Stone said.
NBA referees are paid salaries during the regular season and also for each playoff series they work, according to a league spokesman. So they have no apparent financial motivation to make calls that would extend a playoff series.
Referees must be selected to work the playoffs, however, and are chosen based on experience and regular-season and playoff performance, which is evaluated by league and team officials.
More playoff games typically produce more ticket and television revenue for the NBA. Consequently, officials could make calls to enhance those revenues in order to please their employer, the study says. It also says referees "may have made calls in attempts to please the crowd or players, without being cognizant of their effects on league profitability."
Litvin called the suggested link between referees' calls and league profitability "preposterous."
"But people still think we fixed the Ewing lottery, so I guess nothing surprises me," Litvin said. The 1985 draft lottery gave the large-market New York Knicks the No. 1 pick (and superstar college player Patrick Ewing) despite one-in-seven odds. Wild speculation surfaced that the league froze or bent the Knicks' envelope to make it identifiable by touch.
For five years, the NBA has used observers to log and scrutinize every call. It amplified its study of certain calls in the wake of referee Tim Donaghy's 2007 guilty plea to criminal charges in a gambling scandal involving games he worked and bet on.
"It's really a matter of whether or not the referees are correct in their calling," said Steven Angel, the NBA's senior vice president of league operations and officiating. "And we find that they are."
Although the executives said the league's vast data set is sortable by nearly any category, they declined to release all or a portion of that information, even with officials' names omitted. Litvin said the data are private, and that releasing them would not quell speculation anyway.
Study shortcomings, strengths
The study comes with caveats. Referees' tendencies as identified in the study do not determine what might happen in a particular game or playoff series. And even if bias is detected, it's nearly impossible to determine why it's happening.
The study also has not been accepted for publication, a process that includes a robust peer review. It was declined by the first journal to which it was submitted, though academic articles often are rejected by several journals before being accepted. The journal's editor said he does not comment on specific submissions.
Randy Bluffstone, professor of economics and department chair at Portland State, reviewed the study at The Oregonian's request and found it to be generally rigorous and relevant, especially given the frequent fan discussions of NBA officiating.
"I think the work is, in the main, quite well done," Bluffstone said. "The thing I just don't see is big effects. We can be really sure that they exist, but they're pretty small."
The home-team advantage, for instance, would yield about one extra turnover every three games, Bluffstone said.
He also called the study's detected advantage for a team trailing in a playoff series "statistically insignificant."
One of the study's findings most notable to fans, ironically, appeared in a footnote: that teams from large media markets received no strong favorable treatment during the playoffs.
"Our evidence for that is weak in either direction," Stone said. "It's not like we're proving that that bias doesn't exist, either."
The study notes that it did not examine another fan theory -- that referees favor star players -- because the researchers chose to focus on team-related biases.
Bruce Blonigen, Knight professor in social science in the University of Oregon's department of economics, also reviewed the study at The Oregonian's request. He praised its detailed data (taken from about 3,500 games, from 2002-2008) and the use of "non-discretionary" turnovers as a control group.
"All in all, I find their analysis very compelling and would not be surprised to see it land in a top economics journal," Blonigen said.
The study concludes with suggestions for the league, including increased monitoring of discretionary fouls, clarification of the rules and inclusion of traveling and offensive-foul violations in box scores.
It's unlikely the NBA will consider the suggestions any time soon. Said Litvin: "We don't plan to engage the authors."
Rachel Bachman: 503-221-4373; firstname.lastname@example.org
- INSIDE THE NBA
Browse by week: