Race in basketball

May 2, 2007 10:55 AM

Are NBA referees racially biased when calling fouls? In a paper [PDF] released yesterday, economists Wolfers and Price claim that an all-white team would win two extra games over an 82-game season.
posted by Aloysius Bear (99 comments total)

This study means nothing without the inclusion of statistics related to Bill Lambeer.
posted by billysumday at 11:04 AM on May 2, 2007

Except, an all white team would be inferior talentwise, almost surely. I mean, I don't follow the NBA, but I don't think it's my imagination that black players are, by and large, the better players.

The most likely explanation here seems to be that better players are more aggressive players and thus more likely to draw more fouls.
posted by TypographicalError at 11:06 AM on May 2, 2007

This is an example of why economists get no respect.
posted by Kirth Gerson at 11:07 AM on May 2, 2007

It says they controlled for white players drawing more fouls due to position, though I don't understand how, so I'll give them the benefit of the incompetence-on-my-part and say they controlled for the two colors actually fouling at different rates.

That said, 2.5% racist sounds pretty good to me in the context of broader society.
posted by DU at 11:07 AM on May 2, 2007

Also being discussed on SportsFilter
posted by terrapin at 11:09 AM on May 2, 2007

Given that your league is mostly African-American, maybe you should have more African-American referees — for the same reason that you don’t want mostly white police forces in primarily black neighborhoods

I get it. Separate but Equal... now where have I heard this before.
posted by French Fry at 11:10 AM on May 2, 2007

I'm all for pogroms against economists.
posted by Burhanistan at 11:11 AM on May 2, 2007

I honestly can't name a single white basketball player other than Larry Bird.
posted by Faint of Butt at 11:14 AM on May 2, 2007

Except, an all white team would be inferior talentwise, almost surely.

Not actually. There has actually been a huge increase in white players coming into the league as a result of basketball’s wild popularity in Eastern Europe and other non American locals. I think a lot of people’s assumptions about race are changing as a result of that.
posted by French Fry at 11:14 AM on May 2, 2007

Also, I've seen the Harlem Globetrotters play and those guys get away with a TON of fouls.
posted by billysumday at 11:14 AM on May 2, 2007 [7 favorites]

An all-white team would be inferior talentwise not because "black players are, by and large, the better players" but because any group of people selected on the basis of race will not be the best at what they are selected to do, unless you are selecting a group of people to be a certain race. If you are selecting a team of individuals to be white, and you select an all-white team, then you are fine. Other than that, only a purely merit-based (and therefore not race-based) selection process will produce the best group.
posted by ND¢ at 11:15 AM on May 2, 2007 [2 favorites]

I am skeptical.
posted by caddis at 11:16 AM on May 2, 2007

Did they not include Shaq in their calculations? He would skew things pretty heavily in the black-player-not-getting-called-for-fouls direction.

Oh, and this:

MetaFilter:2.5% racist.
posted by Mister_A at 11:17 AM on May 2, 2007

More work by Justin Wolfers (scary picture included)
posted by phaedon at 11:18 AM on May 2, 2007

TypographicalError: The most likely explanation here seems to be that better players are more aggressive players and thus more likely to draw more fouls.

They control for "player fixed effects", which means that they've already taken this explanation into account. If the referee is unbiased, a particularly aggressive player should draw as many more fouls under a white referee as under a black referee. It's relatively easy to remove this effect from the data, so you are (hopefully) just looking at the effect of the race of the referee and the race of the players.
posted by Aloysius Bear at 11:20 AM on May 2, 2007

Also, I've seen the Harlem Globetrotters play and those guys get away with a TON of fouls.

What kind of foul should be called for deflated basketballs and throwing buckets of confetti on the audience?
posted by NationalKato at 11:20 AM on May 2, 2007

They ought to get all-Latino referees. That would clear this right up.
posted by Bookhouse at 11:21 AM on May 2, 2007

What kind of foul should be called for deflated basketballs and throwing buckets of confetti on the audience?

I'm pretty sure that constitutes a technical.
posted by billysumday at 11:22 AM on May 2, 2007

I guess they're also sexist, since 100% of the players they call fouls on are men.
posted by triolus at 11:23 AM on May 2, 2007

Bookhouse: sure, fine, but then Eduardo Najera is gonna become a huge star.
posted by taliaferro at 11:24 AM on May 2, 2007

Assuming these guys aren't complete statistics idiots, they've controlled for (meaning, ruled out) all of the faults you folks have pointed out in this thread.

The only way to say this is wrong, and prove it, is to point out how their data or statistics are flawed.

Their study also showed that black referees were more likely to call fouls on white players, as well, btw, which also shoots holes through some of the explanations for why this study is "crap."
posted by teece at 11:24 AM on May 2, 2007 [1 favorite]

Typographical error, that has really nothing to do with the study -- the study didn't claim that an all-white team would do better than an all-black team and so teams should pick more white players. What it said was if you could change the race of a player from black to white without changing any other characteristic of that player (skill, position, team, etc), that player would on average be called on 2.5% fewer fouls. Clearly, a player's race cannot be "changed" in that manner. The study accounts for all those variables you mentioned -- aggression, skill, age, position, etc.

As to the idea that black players "are by and large the better players", there are many excellent white players in today's NBA -- last year's (or was it 2 years ago? -- I'm not actually a basketball fan) MVP was the very, very white Steve Nash, for a start.
posted by modernnomad at 11:24 AM on May 2, 2007

Frankly, that NY Times article was retarded, and it's not like I'm about to sit down and read somebody's graduate thesis. All I can say is that this was most likely purposely released early (before peer review) so that it could be picked up by the media in the wake of the last week's Tim Duncan incident.
posted by phaedon at 11:26 AM on May 2, 2007

Throw Rasheed Wallace out of the study and the numbers even right out.
posted by mr_crash_davis at 11:27 AM on May 2, 2007

Assuming these guys aren't complete statistics idiots, they've controlled for (meaning, ruled out) all of the faults you folks have pointed out in this thread.

This comment applies to many, many, many MeFi threads, sadly. People, if you in 10 seconds can come up with alternative explanations for the claim made in the headline of a post that links to a news report of a simplification of some research, then rest assured that the scientists themselves could have thought of the same explanation and have almost certainly ruled it out.
posted by DU at 11:28 AM on May 2, 2007 [1 favorite]

I honestly can't name a single white basketball player other than Larry Bird

I'm not a huge fan of the NBA myself, but I do have to give a lil' shout out to fellow Canadian, Steve Nash (a white guy) of the Phoenix Suns, who, for the past two seasons has been voted the league's most valuable player. Steve's main competition for that title last year, and his quest this year for the accolade is the Dallas Maverick's Dirk Nowitzki, also a white guy. So you see...white men may not be able to jump, but they can play a little round ball...you know wha' um sayin?
posted by Shfishp at 11:29 AM on May 2, 2007

Whatever, dude. Prove to me causation.

This is not a statement in defense of old little white guys with whistles in their mouths. (So you're saying Gravel calling out Obama is racist?)
posted by phaedon at 11:30 AM on May 2, 2007

I'll tell you what, lets fix the problem of racial bias in the justice system (pdf), hiring, and everywhere else we can think of in the real world, then lets get to tackling it where grown men play a game for a living.

Also Shfishp, I'm pretty sure that your mother is an astronaut.
posted by ND¢ at 11:34 AM on May 2, 2007

What kind of foul should be called for deflated basketballs and throwing buckets of confetti on the audience?

The Chuck Taylor Rip and the Rip Taylor Chuck.
posted by jimmythefish at 11:34 AM on May 2, 2007 [6 favorites]

Also, white people have it tough, too.
posted by phaedon at 11:34 AM on May 2, 2007

What kind of foul should be called for deflated basketballs and throwing buckets of confetti on the audience?

Those are technical fouls; I'm not sure if they're included in the analysis (though they are shooting fouls, they don't go to a player's foul total for foul-outular purposes).
posted by Mister_A at 11:35 AM on May 2, 2007

Prove to me causation.

I don't believe a claim of causation is being made.
posted by DU at 11:35 AM on May 2, 2007

Frankly, that NY Times article was retarded, and it's not like I'm about to sit down and read somebody's graduate thesis.

So, because the original research involves too many big words for you you're pretty sure its all a waste of time? Well done.

Whatever, dude. Prove to me causation.

Just because you don't "get" statistics doesn't invalidate them.

They're not trying to say that the NBA is "rigged" against black players. What they are saying is that this discrepency exists, and it might be worth looking into to see if it can be resolved. Given that the NBA itself has done a (less rigorous) study into this phenonmena, it is clearly on the minds of some people high up in the NBA too.

This whole "head in the sand" thing about any race-related issue I find bewildering.
posted by modernnomad at 11:39 AM on May 2, 2007

Like I said, I did not, and I'm not going to, read the paper. But the title of this linked article reads, "Study of N.B.A. Sees Racial Bias in Calling Fouls". The causation that I'm referring to is just that - prove to me that racial bias is causing nba referees to call more fouls on black players than on white ones.
posted by phaedon at 11:39 AM on May 2, 2007

I'm not burying my head in sand. Frankly, I don't give a fuck about racism in the NBA, I deal with it every day. Me not reading a paper doesn't invalidate anything, you're right. But it damn sure doesn't make it right, either.
posted by phaedon at 11:41 AM on May 2, 2007

...prove to me that racial bias is causing nba referees to call more fouls on black players than on white ones.

Um....if race of ref vs race of player is correlated to number of fouls called, that is racial bias by definition. Do you even know what the word "bias" means?
posted by DU at 11:45 AM on May 2, 2007

Finally, my background is not in economics. Reading the paper isn't necessarily going to make me any smarter either way. This is similar to my stance on behavioral psychology. I'm not going to read some fucking 50-page paper on how the fact that boy apes like to play with toy cars, and girl apes like to play with barbies, and thereby conclude that this is some awesome insight into the genetic history of mankind. No. I was taught that BF Skinner was categorically wrong, and I believe that argument. I'm sorry, I'm throwing out the baby with the bath water here. But I have broader methodologies to defend.
posted by phaedon at 11:46 AM on May 2, 2007

I think the 2.5% is easily within the margin that could be explained by the 'fact' that black players are 'more athletic' than white players, playing the game at a slightly faster pace and with stronger moves. More black players (inner city) play to dunk and attack the hoop, whereas many white players might be jump shooters (Dirk, Stojakovic, Gasol). Obviously, this line of thought cannot be proven by me whatsoever. I'm not even sure I buy it, but I think the delta is the right general magnitude. Don't think its neccesarily race either from the calling of fouls standpoint, (maybe they actually create contact at a slightly more violent level) or the style of game that they are used to playing (inner city US versus the international game).

Also (didn't read the paper - I am a statistical moron) but was any thought given to the incidence of fouls amongst same race versus different race.
posted by sfts2 at 11:47 AM on May 2, 2007

Like I said, I did not, and I'm not going to, read the paper... prove to me that racial bias is causing nba referees to call more fouls on black players than on white ones.

Well if you read the paper that tries to demonstrate exactly that, perhaps we could have a constructive discussion. Until you do, we can't really talk because you're sticking your fingers in your ears and saying "it's retarded, it can't be true."

Frankly, I don't give a fuck about racism in the NBA, I deal with it every day.

You deal with racism in the NBA every day? And you don't give a fuck? Impressive.
posted by modernnomad at 11:48 AM on May 2, 2007

I guess that would explain the Bulls' dominance in the 90s, and the Jazz' success during that same time period.
posted by The World Famous at 11:49 AM on May 2, 2007

I was taught that BF Skinner was categorically wrong, and I believe that argument. I'm sorry, I'm throwing out the baby with the bath water here. But I have broader methodologies to defend.

Just for future reference, are the broader methodologies you are defending the dogmatic reliance on what you were categorically taught or the throwing out of babies with bathwater?
posted by DU at 11:50 AM on May 2, 2007

Mostly dogmatic reliance (but don't worry, my blood is rich with post-modern anti-bodies) - but the occasional baby will be jettisoned out of the airlock.

Sorry, I realize my contributions here qualify as trolling. Continue. Inspire me.
posted by phaedon at 11:54 AM on May 2, 2007

Also Shfishp, I'm pretty sure that your mother is an astronaut.

Hey ND¢, out of good taste, in this throwdown, I'm gonna leave your mother out....

...so what time do ya want the bitch back?

OH!!!!!! SNAP!!!!!!
posted by Shfishp at 11:58 AM on May 2, 2007

Except, an all white team would be inferior talentwise, almost surely. I mean, I don't follow the NBA...

...but don't let that get ahead of your stereotypes, clearly. Some pointers for your re-education. The last two most valuable player awards? To Steve Nash, a white guy. Dirk Nowitski, also a honky is favored for this year's award. While the majority of NBA players are black, this doesn't mean anything with regards to relative talent, just relative melanin.

The most likely explanation here seems to be that better players are more aggressive players and thus more likely to draw more fouls.

Again, not necessarily. Nash (not that I'm a fan of his or anything) isn't on that list. You can be good and play clean. Most of the good players are also going to be in more minutes, therefore will have more opportunities for fouls.

Its also important to note that fouls are often times not "drawn" but "given". Its confusing to many casual basketball watchers but there's a fair amount of strategy to giving fouls: in most cases a foul will be given when its clear that the shooter has an easy shot. Its better to take the foul and worst case have them make those two points at the line.

I'd be itchy as the author of this report by the number of radical factors involved in calls. Style of play, nature of the matchup, thug factor all greatly affect the calls on a game. The methodology used in this study bases it solely on refereeing crew and foul count -- the best correlation they have ignores who is calling what on whom...
posted by Ogre Lawless at 11:59 AM on May 2, 2007

White referees are 2.5% more likely to throw out the baby with the bathwater? That's crazy talk. I want to see some numbers I don't have to read to understand.
posted by goatdog at 11:59 AM on May 2, 2007

sfts2 - the study shows that as the number of white referees in a refereeing crew increases, the number of fouls called on black players increases while the number of fouls called on white players decreases. The reverse is true for black referees. The issue is not that blacks have more fouls called on them - it's that black referees caller fewer fouls on blacks and more on whites, and white referees call more fouls on blacks and fewer on whites.

Which, as DU noted, is a racial bias. (yeah, phaedon, I think you're misreading the title of the post here.)
posted by taliaferro at 11:59 AM on May 2, 2007

Lets get off mothers Shfishp . . . cause I just got off yours!
posted by ND¢ at 12:02 PM on May 2, 2007

(yeah, phaedon, I think you're misreading the title of the post here.)

by title, I mean post.
posted by taliaferro at 12:02 PM on May 2, 2007

This comment applies to many, many, many MeFi threads, sadly. People, if you in 10 seconds can come up with alternative explanations for the claim made in the headline of a post that links to a news report of a simplification of some research, then rest assured that the scientists themselves could have thought of the same explanation and have almost certainly ruled it out.

Except, you know, when they haven't. Especially when there's an agenda to push. When there's an agenda, I can often read the whole report with no mention of the flaw I thought of in ten seconds.
posted by TheOnlyCoolTim at 12:07 PM on May 2, 2007

One thing to point out: a lot of the conversation here has talked about the aggresion of offensive players relating to their race. But this only has a small impact on the study, since the vast majority of fouls are called on defenders.
posted by taliaferro at 12:13 PM on May 2, 2007

I can often read the whole report with no mention of the flaw I thought of in ten seconds.

As long as you reading the whole report before denouncing it, I have no problem with that.
posted by DU at 12:16 PM on May 2, 2007

The most likely explanation here seems to be that better players are more aggressive players and thus more likely to draw more fouls.

Nope. The "better" players are called for less fouls than the scrubs b/c the NBA needs to keep its stars on the floor.

It's not as bad as it once was (ahem, Michael Jordan), but it's still pervasive. Officials will also shift fouls around when possible (two guys in the same area) to eliminate fouling out.

That said, the subconscious racism angle is interesting, as is the study. I'm not too surprised, or concerned, actually. I think the NBA should continue to investigate, but as somebody else noted above, we have more pressing racial equality issues to work on. (I'm not even sure this one is "fixable," without major changes in the larger society.)
posted by mrgrimm at 12:19 PM on May 2, 2007

As a basketball referee, there are MANY facotrs that go into officiating a game.

I would like to know how many fouls are called by each referee on each racial demographic when the referee is in a certain position, and how many of each basketball position is visible to that referee or under their jurisdiction at each time. Many forwards will be on the side of teh court, in the eyesite of the trailing or middle referee, so if there are statistically more white referees trailing and middle, and statistically more black forwards, than it makes sense that statistically, there are more fouls called on black basketball players because the situation is created as such. Also, if one referee is lead and trailing most of the game and the centers for one team are mostly white and teh guards for the otehr team are mostly white, he is going to call more fouls on white people, because those are the players he is watching.

Basketball officiating is a team effort. Each ref watches a part of the floor and part of the play. If black referees happen to oversee the part of the floor that on a given night is occupied mostly with white players, they are going to call fouls or infractions on white players FAR more than on blacks, because whites are occupying their area of the floor more often than blacks.

Seriously, I want their data, so we can hack it up in several different ways to get to the real root cause. Im pretty sure that in the EXTREMELY FAST NBA game, there is a lot more to worry about than if the guy you just called the foul on is white or black.

Im not saying there isnt racisim in teh NBA, but I am saying that is a lot more than saying white referees call more fouls on blacks to the story.
posted by subaruwrx at 12:32 PM on May 2, 2007

I thought it was interesting that the difference in the rate at which white and black referees called fouls on white players was greater than the difference in the rate at which referees of each race called fouls on black players.

It occurs to me that one explanation could be that white referees are biased with respect to interracial fouling.

Since the NBA is predominately black, most fouls will be against black players. Consequently, even if white referees were particularly hard on black-on-white fouls, it wouldn't necessarily raise the overall rate of black fouling that much.

On the other hand, since most of a white player's fouls and potential fouls will be against black players, any leniency on the part of white referees toward white-on-black fouls would go a long way to reducing white foul rates.
posted by Mr. President Dr. Steve Elvis America at 12:34 PM on May 2, 2007

Im pretty sure that in the EXTREMELY FAST NBA game, there is a lot more to worry about than if the guy you just called the foul on is white or black.

Actually, this is support that racism *is* a factor. Subconscious biases show up more in snap decisions than they do in reasoned, considered ones.
posted by DU at 12:37 PM on May 2, 2007

The Stern and the League's pathological inability to deal with criticism directed at officiating (see Cuban, Mark) is really disappointing, especially in a case where it's not people talking about incompetence but issues of underlying racism.

“If that’s going on, then it’s something that needs to be dealt with,” James said. “But I’ve never seen it.”

The perspectives from Mike James and Alan Henderson are probably necessary from a journalistic standpoint, but their observations aren't at all valid in refuting the case being made. It would be superhuman for them to accurately perceive the 2.5 percent correlation that the study demonstrates.

The broader point, though, is that there's not an immediately obvious course for correcting a racial bias that's almost certainly unconscious on the part of the refs and imperceptible on the part of the players. Dealing with issues of this kind is what makes confronting these sorts of biases so difficult. (Not to mention the horde of people who would prefer to ignore it because a. It's just sports and b. It's too small a phenomenon to bother with.)

A weekend of sensitivity training isn't going to correct any behavior here, and surely these refs don't consider themselves racist. So, how does one deal with an issue like this, either from a person standpoint (a ref evaluating his own deeply rooted prejudices) or an organizational one (the League doing whatever it can to make sure that awareness trumps embarrassment when issues of race come up). These are difficult questions, but ones that are worth considering no matter what the context.
posted by camcgee at 12:44 PM on May 2, 2007

I think that this is a bunch of cow chips. If it was true, the Sacramento Kings would have beat the Lakers in the championship series back in 02. I think it was Mike Bibby that got knocked flat on his ass by Kobe Bryant, and no foul was called. Most blatent foul I'd seen in ages, and the white guy got pwned.

IIRC, the majority of the referee staff were black anyway.
posted by drstein at 1:01 PM on May 2, 2007

drstein, with your one anecdote, you really put their 266,384-strong dataset in its place!
posted by Aloysius Bear at 1:11 PM on May 2, 2007

Yeah, but how many African-American NASCAR drivers get back on the lead lap due to a debris caution compared to white NASCAR drivers, when the person calling the yellow flags is of either race?

Yeah, I don't do NBA, but I didn't want to be left out. And yes, there are African-American NASCAR drivers.
posted by davejay at 1:17 PM on May 2, 2007

Mike Bibby is black.
posted by otio at 1:18 PM on May 2, 2007

I think the problem is that the refs are human beings, and to a certain extent, all human beings are racially biased, no matter what race they are (even though race is just a construct and doesn't really exist). We are hardwired as human beings to 'side with' people who look like us because we can place ourselves in their shoes more easily, and therefore we empathize more with them. That is one reason that white defendants get lesser sentences from white judges than black defendants that commit the same crime. It is not because the judge is sitting on his bench thinking "Well, let me just harshly sentence this black person because black people are all crooks, and then head on to the Klan meeting." It is because he sees the white defendant and thinks, "Well, when I was that age I was not too different from this young man, and if things had been different..." but he just doesn't think the same things when he sees the black defendant. This does not make the judge (or the refs in this example) evil, but it does make them racially biased, the way all people are. It is part of being human. It is hard-wired into our brains. It may get better, and if the NBA has gotten it down to 2.5% and improves that after this study then good for them, but we as human beings will never look at people that look different from us and stop seeing "the other" more so than we see ourselves. Before that happens all "races" will have ceased to exist because of the fact that people are less bound by geography now, and so most people will look pretty much the same because of "interbreeding" (that term is completely nonsensical in this context but was the best that I could come up with), and, despite all the afterschool specials where the wise narrator asks "Wouldn't the world be boring if everyone looked the same Timmy?", that will probably be for the best.
posted by ND¢ at 1:19 PM on May 2, 2007 [1 favorite]

the Sacramento Kings would have beat the Lakers in the championship series back in 02

I think that's the first game Bill Walton used the classic "Thrown it DOWWWOUUN BIG MAN! Throw it DOWWWUN!"
posted by phaedon at 1:20 PM on May 2, 2007

I think that this is a bunch of cow chips. If it was true, the Sacramento Kings would have beat the Lakers in the championship series back in 02. I think it was Mike Bibby that got knocked flat on his ass by Kobe Bryant, and no foul was called. Most blatent foul I'd seen in ages, and the white guy got pwned.

Lol. Just cause someone falls doesn't mean they were fouled. Besides the Kings tried to give the Lakers food poisoning and Bibby is black. The Kings lost because the Lakers had Big Shot Rob and that is it. And because God Loves the Lakers (till they become rapists anyways). They also beat his dad like Bill Walton should be beaten with a microphone. I think that covers it.

Interesting link.
posted by dame at 1:33 PM on May 2, 2007

I think David Stern owes Al Sharpton an apology.
posted by dzot at 1:40 PM on May 2, 2007 [1 favorite]

I think that this is a bunch of cow chips. If it was true, the Sacramento Kings would have beat the Lakers in the championship series back in 02. I think it was Mike Bibby that got knocked flat on his ass by Kobe Bryant, and no foul was called. Most blatent foul I'd seen in ages, and the white guy got pwned.

You know, I don't remember the article saying "black guys always get called for fouls and white guys never do." I'll go back and check, but I'm pretty sure it didn't. But, hey if it did, then you sure blew that out the water.
posted by modernnomad at 1:45 PM on May 2, 2007

And here I thought the Lakers won in 02 cos no one had the stones to sucker-punch Horry to keep him from hitting at-the-buzzer threes.
posted by absalom at 1:57 PM on May 2, 2007

The NBA dude cites privacy for the players and refs in not releasing the ref-specific data to the researchers.

WTF?

Uh, aren't these games, like, open to the public? What is this, privacy by poor memory?
posted by Mental Wimp at 1:58 PM on May 2, 2007

Freakonomics weighs in.
posted by terrapin at 2:00 PM on May 2, 2007

Are there enough white players in the NBA to represent a statistically valid sample? (ahem)
posted by Steven C. Den Beste at 2:06 PM on May 2, 2007

You know who else showed racial bias when calling fouls, right?

/ob
posted by Devils Rancher at 2:14 PM on May 2, 2007

There's an easy fix to this. The NBA should just hire blind refs - it works for baseball!
posted by rtha at 2:14 PM on May 2, 2007

Those of you who are not going to bother to read the actual information, please take a look at the following paragraph, which was highlighted in the comments of the Freakonomics link. The article and paper do not say what you think they say, and this sums things up nicely:

"There are also two ways in which these own-race biases may emerge: they may reflect referees favoring players of their own race, or alternatively disfavoring those of the opposite race. The arbitrary assignment of referees to games means that we can test whether our estimates reflect an influence of referee race on black players, or on white players. Table 3 is instructive, showing that the rate at which fouls are earned by black players is largely invariant to the racial composition of the refereeing crew. By contrast the rate at which fouls are earned by white players responds quite strongly to referee race. Further regression-based tests yield a similar pattern (see in particular the coefficient on white referees in Table 4), suggesting that the impact of the biases we document is on white players, who are either favored by white referees, or disfavored by black referees."
posted by kyrademon at 2:37 PM on May 2, 2007

You all do realize that NBA refs intentionally call more fouls on black players in order make up for years of not calling fouls on Jordan right? It could take another decade to make up for that. That being said perhaps the NBA can make a rule that only black refs can call fouls on black players?
posted by MikeMc at 2:38 PM on May 2, 2007

And here I thought the Lakers won in 02 cos no one had the stones to sucker-punch Horry to keep him from hitting at-the-buzzer threes.

Of course they didn't: they were all white.*









*Except Bibby and Webber, the Greatest Choker of All Time.
posted by dame at 2:58 PM on May 2, 2007

phaedon: one thing I would caution you on. Doing a statistical study that proves a correlation between referee race and # of fouls on black/white players (which the claim here. Assuming the statistics is value, they have proven correlation), does NOT mean that white referees are KKK members out to get dirty n*ggers. OK? I get the feeling that is what you are reading when you see "racial bias."

Yes, the study (if valid), proves a racial bias, but that is not the same thing as "racism," as most of us use the term in day to day usage. People make decisions based upon race: whether those people are MLK or KKK cross-burning bastards. In that sense, we are all racist. In that sense, the statistical data here (purport) to prove racial bias in refereeing.

This more general kind of racism is what is being alleged here (the kind of racism where 'racial' is an adjective the describes making decisions based upon race) — not the KKK, cross-buring kind of bigoted haters that most folks think of when they hear "racism."

What's the causation? This study won't say, causation is much more complicated, and not generally the realm of (just) statistics. But a proven correlation is a very strong data point that must be considered, not tossed aside because we don't like what it says.

This is a real problem in discussing racism -- the term is so loaded that it short-circuits thought for most people. I am a racist. I am also as tolerant as I humanly can be, I'm married to a woman of a different race/ethnicity than my own, I detest "racism," etc. But there are very valid definitions of the word "racist" that I fit into: we all do. So it's important to try and differentiate between the multiple uses of the word.

This study does not prove, nor does it try to prove, that white referees are closeted KKK members (or that black referees are closeted haters of the blue-eyed devil).
posted by teece at 3:32 PM on May 2, 2007 [2 favorites]

Prove to me caucasian.
posted by jimmythefish at 3:39 PM on May 2, 2007

teece, yes, i think taliaferro did mention that i was probably misreading the use of the phrase "racial bias". i agree with him. being upfront about not reading the article didn't help me win any points either.

I was in Greece the year Dominique Wilkins decided to play for Panathinaikos, the year of the NBA lockout. Keep in mind greek basketball fans have about as much dignity as your english soccer hooligan, which means non-existent. Wilkins was hands down the best player in the league, took them to a championship i think. And when he shot free throws, opposing fans would should "K-K-K! K-K-K!". This is more or less still happening today in soccer, and acknowledged by FIFA to some extent as a league-wide problem. One guy was on HBO sports a few months ago during the world cip, talking about how fans would throw bananas at him every time he dribbled down to that side of the field.

So I'm giving the story of the "holy shit tell me when im supposed to freak out" treatment. But that is totally insensitive, and hinders other people who want to discuss this issue. i don't particularely like the way the nba is run - stern is what bloomberg is to new york - and i have no doubt "racism" is everywhere in one form or another. if its the phenomenon of preferentially treating one's own race over the other, as i tried to say earlier, shit i see that all the time! even if it means just not being friendly.

So I back out again, and ask this, what the hell does the 2.5% discrepancy mean, and does that sound like a fairly large or fairly small number?
posted by phaedon at 3:47 PM on May 2, 2007

You know who else showed racial bias when calling fouls, right?

No, who?
posted by grouse at 4:08 PM on May 2, 2007

This study does not at all demonstrate that both black and white refs were other-race biased.

It's possible that the black refs are entirely correct with their judgments and that the white refs are significantly biased. It is also possible that the white refs are entirely correct with their judgments and the black refs are significantly biased.

More importantly, it is also possible, since the white players were given significantly more fouls by both the white and black refs (4.970 vs 4.330 fouls per 48 minutes), that both the white and black refs are substantially biased against the white players, and biased in favor of black players, only that the white refs are somewhat less biased in their favortism towards blacks than the black refs.

This study does not provide any information that allows us to pick from these alternatives - and therefore the media may be entirely incorrect in presenting this as evidence of discrimination against blacks; sort of like reporting that the Armenians genocided the Turks, based only on the knowledge that someone was ethnic cleansing someone in that region.
posted by dgaicun at 4:28 PM on May 2, 2007

NB - I do not follow sports at all; I am not saying any of those interpretations are more or less plausible, only that that finding in and of itself says nothing about which fouls were incorrect.
posted by dgaicun at 4:36 PM on May 2, 2007

The paper by Mr. Wolfers and Mr. Price has yet to undergo formal peer review before publication in an economic journal

God damn, what low standards of reporting. The Times should be ashamed.
posted by dgaicun at 4:39 PM on May 2, 2007

Faint of Butt writes "I honestly can't name a single white basketball player other than Larry Bird."

There's that Steve Nash guy who I only know because he's Canadian and on TV all the time.

Mister_A writes "MetaFilter:2.5% racist."
MetaFilter:2.5% racist [NOT RACIST]

modernnomad writes "The study accounts for all those variables you mentioned -- aggression, skill, age, position, etc."

That's the crux isn't it. Put 20 scouts in a room with a list of a 100 players ask weach to rank them and I bet even their top ten don't match up. How do you equate the skill levels of several hundred players in the NBA? Statistics alone sure as heck aren't going going to do it.
posted by Mitheral at 5:04 PM on May 2, 2007

How do you equate the skill levels of several hundred players in the NBA?

I doubt it is as hard as you imagine. These guys have access to tons of metrics about players. Sports leagues keep crazy-OCD records on who does what.

Many of those metrics would be very reasonable proxies for player skill (points per game, blocks per game, years on the job, etc. Whatever. I'm not a basketball guy: the point is that the metrics are out there). You don't need to say "the rankings of the players are: " and then decide exactly what order all the players should be ranked in.

You simply create blocks of study based upon factors that might seem to effect number of fouls, for which you have metrics, and then you randomly assign players within those blocks, using the blocks to "control" for factors you think might have an effect on # fouls other than the two variables in question (independent: race of referee, dependent: # of fouls given, with one factor that has two levels: black and white).

This hasn't been peer-reviewed, so maybe these guys are real bozos and did a bad job on their stats. I don't know that, and I'm not inclined to read their paper and check their methodology (and nobody on Metafilter would care about my opinion if I did).

But controlling for confounding factors (such as age and skill) really can be done. It's not even that hard. There will always be some fuzziness, since skill is not a nice scale variable like "age" or "height," but it's still quite possible to generate a reasonable metric for it (and if I were going to design the study, the metric for skill would be a multitude of proxy metrics like points scored, etc., that the NBA will have readily on hand).
posted by teece at 6:56 PM on May 2, 2007

Kobe's black? That's crazy talk. He's a foreigner, like Pau Gasol or Tony Parker.
posted by breezeway at 7:47 PM on May 2, 2007

There isn't a commentator on ESPN tonight that has come out to support this study. Hm. Not exactly ground-breaking science, but thanks for the heavy hand, asswipes.
posted by phaedon at 10:48 PM on May 2, 2007

...an all-white team would win two extra games over an 82-game season.

In other news, economists discover that if pigs had wings...
posted by MarshallPoe at 4:40 AM on May 3, 2007

Oh, well, if commentators on ESPN disagree, then clearly everyone here is wrong. I mean, they are commentators! on ESPN! That is way more rigorous than being a scientist.

What exactly is your problem with this, phaedon? I don't think anyone here has said, OMG REFS IZ THE KKK. So why are you so invested in it?
posted by dame at 4:41 AM on May 3, 2007

The kid in me wants to remain totally indifferent, while the statistician in me wants to rip everybody's eyes out!
posted by GooseOnTheLoose at 5:11 AM on May 3, 2007

What exactly is your problem with this, phaedon? I don't think anyone here has said, OMG REFS IZ THE KKK. So why are you so invested in it?

My problem is that there is no problem. i get accused of sticking my fingers in my ears and calling this study retarded - and that's more or less what the top 4 black active-nba players are doing on national television.

Go figure.
posted by phaedon at 9:42 AM on May 3, 2007

You know, the problem is there's no way to know what the ideal ratio of fouls would be. It might be, for example (and I am free to wildly speculate here, since I don't watch or follow basketball at all), that white players are predominently European and black players are predominently from tough urban environments -- or some other factor that would mean that black players do commit a slightly higher percentage of fouls than white players. You can't use the calls of black referees as a control, since they could be biased in the opposite direction. So really, what is the point of all this?
posted by Dolukhanova at 10:25 AM on May 3, 2007

Dolukhanova, you're missing the point. If black players commit a larger number of fouls for some reason like "tough, urban upbringing" or similar, and referees aren't biased, surely if the black players commit 10% more fouls than white players with a white referee, they should also commit 10% more fouls under a black referee. Wolfers and Price show that this is not the case. The racial bias of referee decisions occurs even after player and referee 'fixed effects' are controlled for.
posted by Aloysius Bear at 10:43 AM on May 3, 2007

My problem is that there is no problem. i get accused of sticking my fingers in my ears and calling this study retarded - and that's more or less what the top 4 black active-nba players are doing on national television.

Go figure.

Yeah, I accused you of that because that's what you did -- you refused to read the report and just dismissed at as "retarded" with zero evidence to back up your points.

If the "top 4 black active-nba players" are doing the same, that doesn't make it right. And I love that you point out that they are "black", as though that somehow gives higher credence to what they are saying. If they had some training in statistical analysis and thought the methods used in the study, you might have a point.

But again, until you read the thing, your views on it are worth precisely nothing.
posted by modernnomad at 10:47 AM on May 3, 2007

See that's exactly my point. You concentrate too much on my views, my views, my views. What the fuck are your views? You've added nothing to this debate, other than fleshing out points about myself that I've alreaady conceded.

And if their experiences as black players on the court doesn't add credence to what they're saying, I don't know what else to fucking say. You act like you know something more about me, and about the NBA, and about race, than anybody else does. And you're using this data set as a way to exhibit that youthful arrogance.

Seacrest out.
posted by phaedon at 11:01 AM on May 3, 2007

See that's exactly my point. You concentrate too much on my views, my views, my views. What the fuck are your views? You've added nothing to this debate, other than fleshing out points about myself that I've alreaady conceded.

I think I made it clear that my view was that, having read the report and understanding its methodology, it's pretty compelling and needs to be addressed by the NBA.

You act like you know something more about me, and about the NBA, and about race, than anybody else does. And you're using this data set as a way to exhibit that youthful arrogance.

I know fuck all about you or the NBA. My beef is that whenever some sort of study comes out that suggests discrimination of some sort might still be at play in American society, a bunch of people immediately dismiss it out of hand as "retarded". If you had to bear the brunt of my frustrations with regard to this, my apologies. But your steadfast refusal to read the report while simultaneously dismissing it really got on my nerves. I have no fucking time for that Nancy Grace-esque "spidey sense" that lets people intuitively know when something is right or wrong without actually confronting it - or *gasp* READING it.

If you interpret my dismissal of your opinion as to the report being "retarded" as arrogance, then again, my apologies. But until you bring something more to the party, I stand by it. (And my arrogance, if anything, is not youthful -- thanks though.)
posted by modernnomad at 11:45 AM on May 3, 2007

And if their experiences as black players on the court doesn't add credence to what they're saying, I don't know what else to fucking say

These numbers are small enough that even individuals experienced in that environment would understandably not be able to personally notice them. Statistics don't always reveal intuitive trends.
posted by dgaicun at 1:18 PM on May 3, 2007

The only way to say this is wrong, and prove it, is to point out how their data or statistics are flawed.

Michael Stastny tries to do just that.
posted by Kwantsar at 1:50 PM on May 3, 2007

dgaicun: It's even worse. Intuition sometimes is diametrically opposed to underlying truth. If you take a bunch of underperformers on a test, let them listen to the Ramones, then retest them, their performance will improve. Most people's intuition would conclude that it's the Ramones that improved their performance. Regression to the mean is what explains the increase. Likewise with the gambler's fallacy: with a fair coin having yielded 10 heads in a row, the probability of a tail on the 11th toss is stilll 50%. And so on...
posted by Mental Wimp at 1:54 PM on May 3, 2007

Kwantsar

Stastny misreads the purpose of the model to some extent. The authors include a term in the model to account for a constant difference in rates between blacks and whites, but remain agnostic about whether it is bias or actually performance. That's smart, because the two effects are perfectly confounded in the study design.

The authors focus on the interaction between the racial composition of the referee team and the player's race. They assume that if there is a signficant difference in the foul differential for blacks depending upon the racial make up of the referees it has to be bias. I should note that I personally disagree. It could be that blacks play differently depending upon the racial composition of the referee team. I guess that would still boil down to a racial bias, but not necessarily in the referees. Of course, both could be true, as they are also perfectly confounded in the study design.
posted by Mental Wimp at 2:36 PM on May 3, 2007

teece writes "controlling for confounding factors (such as age and skill) really can be done. It's not even that hard. There will always be some fuzziness, since skill is not a nice scale variable like 'age' or 'height,' but it's still quite possible to generate a reasonable metric for it (and if I were going to design the study, the metric for skill would be a multitude of proxy metrics like points scored, etc., that the NBA will have readily on hand)"

Maybe you can create a metric that lets you rate the "skill" of any one player. Does the variability (over the season; away/home; phase of the moon; whatever the phenomenon that causes slumps and other extreme variability) of that ranking rise above the rate of bias?

This seems like an impossible metric to accomplish. The program would need to essentially act as a judge for every single foul called weighing whether the same call would have been made for a differently classed person.
posted by Mitheral at 12:23 PM on May 4, 2007