Certainty is disconcerting.

The true believer, in anything, causes you to reassess your convictions. Are you really sure? Is there no room for doubt?

Justin Wolfers is a true believer, as certain as he is of his own name that there is something wrong with NBA officials.

"Ask any expert to review the evidence," the Penn Wharton School business and public-policy professor said by phone Thursday. During a 20-minute conversation, he made the same point over and over: This isn't Wolfers' opinion. These are, he said, facts.

Wolfers and his partner, Cornell University doctoral candidate Joseph Price, are publishing a study that purports to prove a consistent pattern of racial bias on the part of NBA referees. Wolfers and Price studied box scores from all regular-season NBA games from the 1991-92 season through the 2003-04 season - a period in which there were roughly 600,000 calls made by officials.

Wolfers and Price maintain that their analysis proves that during this period, black players received between 0.12 and 0.21 more fouls per 48 minutes relative to white players when the number of white referees officiating a game increased from zero to three. (All-white crews constituted 30 percent of the crews in the study.)

They also claim that black players scored fewer points and committed more turnovers when playing in games officiated solely by white referees, and that the difference meant that, on average, teams with all-black lineups won around two fewer games per season.

White players also were discriminated against in games officiated by all-black referee crews, Wolfers and Price say, but not by as much, and not as often, since all-black crews constituted only 3 percent of officiating crews.

Their findings, they say, indicate unconscious bias on the part of referees; the white officials don't go into the game looking to make calls against black players. Wolfers and Price compare the behavior they claim NBA referees exhibit to that of companies that fail to promote or hire more members of minority groups, to judges deciding whom to sentence, and to police officers who decide whom to arrest.

"Theories of discrimination used to be pretty simple," Wolfers said. "It used to be one group hated the other group. There was malice. Thankfully, those types of bias have receded. [But] there's an emerging consensus in psychology and law that there may be unconscious forms of bias. We're trying to determine if that's empirically valid. . . . Our expectation was that we weren't going to find anything."

The head of the referees' union, Lamell McMorris, declined to comment Thursday. But the NBA, as you might expect, vehemently disputed the findings, which were first reported last week by the New York Times. The league pointed out that Wolfers and Price did not have access to individual referees' calls when perusing the box scores, so they didn't know which official made which call.

By contrast, the league's revised referee-evaluation system has charted each individual call for the last few years. But the league, when asked by Wolfers and Price to share that data, refused. Instead, it requested an outside analysis of about 148,000 calls made from the start of the 2004-05 season through January of this year.

You won't be surprised to find out that the league-requested study found no evidence of racial bias among NBA referees.

"Bias is not a fact," NBA general counsel Joel Litvin said by telephone Thursday. "Bias is a conclusion that [Wolfers] has drawn from analyses based on flawed data. We never said his data is false. [But] they can't figure out, they have no idea - but we do - as to each individual call, who's making the call, black on black, white on white. . . .

"To your question, 'Can you definitively say there's no bias in calling?' Yes. We proved it with the data, data that is superior to theirs."

Litvin said that the league provided the Times with the data from its study, but that the data and other explanatory information were ignored.

Players have complained about referees for decades. But almost none have accused them of being covert racists.

"I think I've gotten more techs from black refs than white refs," Kobe Bryant told reporters in Los Angeles. "That's reverse racism, probably."

There are other gray areas in this black-and-white story. For example, how did Wolfers and Price count players from mixed-race parentage, like Nets guard Jason Kidd? Wolfers says researchers looked at pictures of the players and made a determination (the researcher determined Kidd was black), and also used the recollections of an anonymous, retired referee. Players of non-black or non-white heritage, like Yao Ming, were counted as white for purposes of the study.

But how do you account for strategies like deliberate fouling at the end of games? And even if you accept the premise of bias, aren't the beneficiaries of that bias, most of the time, other black players - on the other team? So what would the point of the bias be?

Then again, why wouldn't the league release information that it claims definitively proves its referees are not biased?

For his part, Wolfers was unshaken.

"I'm so confident in the results that [you shouldn't] take my word for it," he said. "But certainly don't take their word for it."

Contact staff writer David Aldridge at 215-854-5516 or daldridge@phillynews.com.