Quotelog

Up to this point, the content on this site has come mostly in the form of confused one-liners. However, much more nonsensical talk goes mostly unnoticed, because it comes wrapped in convential wisdom or social biases and might even sound logical. As it gets batted around by media, it’s no wonder that many people remain clueless. So, from time to time, I will use this site as a means of exposing such nonsense.

The issue I would like to discuss is this study by business professor Justin Wolfers and economics grad student Joseph Price which studied the incidence of fouls in the NBA based on the skin color of the referee and the players. It concluded that there is a small but measurable bias; that is, white referees call a higher percentage of fouls on black players and black referees call a higher percentage of fouls on white players.

NBA administrators and players were quick to denounce the study. Players say there is no bias, because…there is no bias. The NBA also says there is no bias and has their own study to disprove it. More on that later. I was watching ESPN and they trotted out a couple people to supposedly make a counterpoint. The first was Charles Barkley. Now, all I could see was the closed captioning, but I remember seeing the words “most stupid thing I’ve seen in years” and “asinine” or words to that effect referring to the study. I’ll give Sir Charles a half point for making an effort to refute the study with a logical point instead of just ridiculing it. Mr. Barkley reasoned that because there are more black players in the NBA, of course more fouls will be called on them.

Um, sorry, but I’m pretty sure that if two academically gifted people would bother to study 14 years of data and write a paper on the subject, they would compare the rate of fouls, not raw numbers. And they did. After Mr. Barkley, I’m honestly not sure who the next representative was, but I think he was somehow affiliated with the NBA. He tried to attack the credibility of the study by making some claims about errors in the authors’ statistical methods. I’m just going with percentages here, but I’m guessing that whomever he was, he had no standing to critique statistical methods. I have a math degree and I wouldn’t even say I’m qualified to do that. But we do have this sentence from the Times article:

Three independent experts asked by The Times to examine the Wolfers-Price paper and materials released by the N.B.A. said they considered the Wolfers-Price argument far more sound.

Strike two. He basically concluded his argument by saying that players, coaches, and general managers all agree that referees get calls right. Huh? I’m sure the officials do a fine job, but how often have you seen a game in which nobody complains about the calls?

The problem with perception of statistical studies is that few people really understand the scope of what they suggest. Some studies are flawed, and occasionally a sound one will suggest an incorrect conclusion. But usually the act of actually counting up lots of discrete events and doing calculations on them is better than human perception alone for large data sets. The problem is that whenever some group doesn’t like the results of a study, they can find some expert to doubt its accuracy. The general public has no idea whom to believe.

So what exactly does Wolfers-Price say? It counts up thousands of fouls, controls for other influences such as position and skill of the player, and concludes that there is a correlation between skin color and foul rate. It does not say why this is. It does not claim that referees are racist. It does not claim that referees do a bad job. It appears to be a part of a larger trend that subconscious racial associations are present in society, in areas such as workplace hiring and police arrests. And studies tell us that the correlation is strong enough that it is probably not due to coincidence, usually to a certainty of 95% or more.

Numbers don’t lie. Interpretations of what they mean could be wrong, but ignoring them because we don’t like what they say isn’t the right approach. Stephen Dubner wrote in Freakonomics that economics tells us what the state of the world is, while morality tells us what we would like it to be. Economic studies are a form of statistics applied to a particular discipline, so the same could be said of all statistical studies. Mark Cuban has the right idea: “We’re all human. We all have our own prejudice. That’s the point of doing statistical analysis. It bears it out in this application, as in a thousand others.â€?

Denial won’t help solve anything. Seeking to understand the data with an open mind might.