Stern must be pleased: Study of N.B.A. Sees Racial Bias in Calling Fouls

I don't know if this has been posted yet, but it's John Hollinger's commentary (ESPN INSIDER) on the study. He makes some interesting points here that I thought might contribute to this discussion.

http://insider.espn.go.com/espn/blog...hollinger_john


Closer look at ref study
by: John Hollinger
posted: Wednesday, May 2, 2007 | Feedback | Print Entry

The basketball world is buzzing today over an academic study on racial discrimination by NBA referees, which states that the racial composition of a three-man officiating crew can have an impact on a game's outcome.

In particular, the study states that the difference between how white referees treat black players and how they treat white players is "large enough that the probability of a team winning is noticeably affected by the racial composition of the refereeing crew."

Predictably, the NBA has launched a PR offensive defending its officials' work and saying its own study, which looked at individual refs who made a call rather than the results of a three-man crew, showed no bias by the officials.

But the bigger point that everyone is missing is that, in fact, this study showed remarkably little bias as well. Maybe I'm a cynic here, but I had expected there would be some level of bias by both black and white officials -- refs are human too, after all, and when they step on the court they unwittingly bring their life experiences and values with them.

Yet the affect is almost totally insignificant. The study reports that a black player will rack up an added 0.16 fouls per 48 minutes with an all-white officiating crew, as compared to an all-black one.

In other words, if he plays 3,190 minutes in a season -- the league-leading total posted by LeBron James this year -- he would pick up 11 extra fouls. Eleven.

Even that scenario depends on the difference between all-black and all-white crews, which isn't realistic -- in reality most games will be officiated by a mixed crew (32 percent of the league's officials are black), so the effect will be much smaller. Thus, the difference between a black player and white player of similar skills and abilities would be something like six or seven fouls all year, out of the 200 or more that most players accumulate in a season. That's if you lead the league in minutes, mind you -- it would be much less for anyone else.

So when the authors talk about a noticeable impact on results, I guess it depends on what they mean by "noticeable." The authors chose to play up the fact that a bias was found, but to me it's even more of a story that it was found to be so small.

That said, I'm sure the league is unhappy about the study and the implications, and obviously we'd like to see the difference become zero. But let's not lose sight of the microscopic impact we're talking about. Thus, don't expect Gregg Popovich to walk on the court tonight, see three white refs, and decide Beno Udrih and Matt Bonner will get the bulk of the minutes.

While we're talking about this study, one other item in it drew my attention: the finding that during the 13-year study period, teams with the greater share of playing time by black players won 48.6 percent of games. The authors seemed to imply some kind of mild institutional racism against black players by this result.

In fact, there's a much more obvious explanation -- the league imported a bunch of talent from Europe during the study period, almost all of it white, and the poorly-run teams were the last ones to figure out there were good players on other continents. Thus, by default they ended up with more black players on their roster.

Look back on the drafts of the mid-to-late-'90s and you'll see what I mean. Players like Peja Stojakovic, Arvydas Sabonis, Zydrunas Ilgauskas, Manu Ginobili, Dirk Nowitzki, Andrei Kirilenko and Mehmet Okur were all basically stolen in the draft by smart, forward-thinking teams. That their teams won more games than average is an effect of their superior front offices, not the officiating.

Finally, let's get to the real juicy stuff -- I was a little disappointed they didn't look at technical fouls and ejections. If there were a pattern of bias, wouldn't we expect it to come out most often when emotions already were running high? Based on some recent high-profile events involving white referees and black players (Crawford vs. Duncan, Salvatore vs. Davis, etc.) I can't be the only one who wishes this had been part of the study.

ESPN Conversation

Here a few essential twists.

Firstable, it would be EXTREMELY interesting to separate charging fouls from all other fouls and correlate with race.
Those are usually the most questionable calls.

57 players on the active rosters of playoff teams are non-american citizens (plus Bell and Duncan).
The vast majority are whites (38).
Overall, this is 25/28% of active rosters of playoff teams.
How does that very impressive 4.5% of additional fouls called on black players apply to the "international" players?
Is there a bias in favor of international players?
Or against?
Is it fair to have international players BUT american-only referees?

Next, it is absolutely mandatory to calculate the travelling and three seconds violations and correlate both with international/white, international/black, american/black and american/white. Unless the latter cohort is too skinny to have robust data on.

Last, but definitely not the least, is there a correlation between laughing on the bench, technicals and race. I guess this can be limited to international players that have a -valid- american passport.

There is work for AT LEAST two PhDs, 3 years fellowships at $50000/year funded by the generous donations of the MCNCAFF (Mark Cuban Numbers Come Always First Foundation).

Anybody interested?
It would be fascinating.

Interesting take by Hollinger - what do you bet that Mr. Stats is figuring a way to do some of this analysis himself?

of course it does.

Now i see why Manu doesnt get that many calls and don´t go to the line like other guards in the league.

They should study on fouls and ejections vs square feet of tatoos and corn-rows. :hungry:

Quote:

---

Originally Posted by ambchang

The study did insinuate that, but to a lesser degree.
Again, I am not saying that the study is true and the conclusion correct, but I am still struggling to find the reason for the statistical difference.

---

Well, since the difference in fouls amounts to significantly fewer than one per game, let's just think about when the most fouls occur in a game, and that's at the end of a game when one team is trying to come back. Since the statistical majority of players on the floor at that time are going to be black and the statistical majority of officials are going to be white then it stands that most of the legitimate, intentional fouls whilstled will be by white officials on black players. That is WAY more than enough to tip the results of the study; do you think Lebron James has more than 11 intentional fouls over the course of the regular season?

Quote:

---

Originally Posted by Obstructed_View

Well, since the difference in fouls amounts to significantly fewer than one per game, let's just think about when the most fouls occur in a game, and that's at the end of a game when one team is trying to come back. Since the statistical majority of players on the floor at that time are going to be black and the statistical majority of officials are going to be white then it stands that most of the legitimate, intentional fouls whilstled will be by white officials on black players. That is WAY more than enough to tip the results of the study; do you think Lebron James has more than 11 intentional fouls over the course of the regular season?

---

I believe it is 0.2 fouls per player per 48 minutes, so that amounts to 0.2 fouls x 5 players = 1 foul per game.
The statistical majority of the player on the floor at any time is mostly black, and refs mostly black, and I still fail to see how your scenario will affect how the black and white refs will play out. The thing is, if Lebron James commits 11 intentional fouls per game over the course of the season, so will Nash in the same situation, and those should be 11 intentional fouls regardless of whether a black or a white ref is calling it. Refs do not substitute, and calls throughout the 48 minutes.
Correlation != Causation, but somebody better find a cause.

Quote:

---

Originally Posted by ambchang

I believe it is 0.2 fouls per player per 48 minutes, so that amounts to 0.2 fouls x 5 players = 1 foul per game.
The statistical majority of the player on the floor at any time is mostly black, and refs mostly black, and I still fail to see how your scenario will affect how the black and white refs will play out. The thing is, if Lebron James commits 11 intentional fouls per game over the course of the season, so will Nash in the same situation, and those should be 11 intentional fouls regardless of whether a black or a white ref is calling it. Refs do not substitute, and calls throughout the 48 minutes.
Correlation != Causation, but somebody better find a cause.

---

No. Don't get sidetracked. Lebron leads the league in minutes, and he's just used as an example of how small the numbers are; it has nothing to do with what color his skin is.

It is a statistical certainty that there are going to be numerically more intentional fouls called by white refs on black players at the end of games due to the fact that that's who is on the floor at the end of the games. That is enough to skew the stats by 0.8 fouls per game. 1.6 does not equal 2.

This study is GARBAGE.

They don't know WHICH ref called the fouls, so in about 2/3 of the games (those in which there are black and white refs together) we don't know if a black ref or white ref called the foul...

GARBAGE

Quote:

---

Originally Posted by LEONARD

This study is GARBAGE.

They don't know WHICH ref called the fouls, so in about 2/3 of the games (those in which there are black and white refs together) we don't know if a black ref or white ref called the foul...

GARBAGE

---

And that's the real killer. Since they had to get all their information from box scores, which don't indicate who blew the whistle, they had to just guess how many whistles went to each ref. That sufficiently muddies the numbers in and of itself.

Most everyone in this thread has made great points and rendered the study completely meritless.