For the Times, it is always racism

For the third time in a week, the New York Times has devoted space on the front page of its main news section to a sports article. The first story covered the baseball steroid hearings in Congress. The second was a feel—good puff piece on how victories by low—seeded teams make the NCAA Men's Basketball Tournament so exciting. Some Times staffer or editor who graduated from Vermont or Bucknell probably had this 'major' story placed on Page one.

Yesterday's Times' front page sports story  was different.

In the tradition of the earlier campaign by the Times (40 stories!) and the National Organization for Women to get the Augusta National Country Club to accept a female member, yesterday (March 22) the Times said that racism must be the reason why black NBA coaches have shorter tenures than white coaches. The authors of the article are David Leonhardt and Ford Fessenden. I know neither of them, and do not know whether they have any background in statistics. Their article fails to meet the most basic standards for serious statistical research, and their conclusions are belied by the small print in a box describing the methodology for their 'study'.

The Times editorial and op ed pages are not known for subtlety these days (think Maureen Dowd, Paul Krugman and almost every editorial).  Some have accused the news pages of similar over the top offenses (consider the phony missing weapons story pounded home every day for a week before the presidential election in November). This latest sports article turns out to be an opinion piece more than a news story. It is the kind of thing Bob Herbert would write, a man who sees racism in his soup.

It is no accident that this article appeared Tueday and not some other day.The day before, the Cleveland Cavaliers fired Paul Silas, their head coach, who conveniently enough is African American. The headline on the Silas story is: 'Cavaliers Join Trend and Fire Silas'.  One might think the trend that is being talked about with the Silas firing is the firing of black coaches.  In fact, in this case, the trend that is discussed in the Silas article is that NBA coaches, whatever their race, are getting fired very quickly these days. Silas' firing was the 16th NBA coaching change in 16 months. That is more than half the league's 30 coaching positions that have changed hands in this period.  The article on Silas does not make any mention of the race of the 16 replaced coaches. In fact, the great majority of them were white. Some white coaches were replaced by black coaches, Some black coaches were replaced by white coaches.

Silas was probably fired because his team was going in the wrong direction. It has lost 9 of 12, and if the playoffs began tomorrow, it would not have the home court edge. Several weeks back, the Denver Nuggets fired their coach as the team continued to play under .500 ball, after an impressive season last year, including a rare playoff run. The Nuggets hired George Karl as their new coach, and the team has since won 17 of 22, and now holds the final playoff spot in the West. They are the hottest team in the League as the playoffs near. Sounds to me like Denver management pulled the trigger at the right time.  Maybe Cleveland wanted a similar shot in the arm.

The conclusion one would reach by carefully reading both the Silas article and the front page article is that if you are interested in job security, do not coach an NBA team. For white coaches, average job tenure of those who started and either retired or were fired in the 15 year period beginning in 1989 was 2.4 years.  For black coaches, the average tenure was 1.6 years. The Silas article suggests the half life for NBA coaches, whatever their race, is shrinking.

Was race a factor in the difference between the two averages? There are ways to determine this statistically. I do not have the full data base, so I can't test this. But the box called 'How the Study Was Conducted' is quite revealing. The third paragraph reads as follows:

'Team performance turned out to be the most important factor in a coach's job security: winning games  and making the playoffs strongly correlated with longer coach tenures. The quality of the team before the coach took it over was also a factor. But white coaches tended to keep their jobs longer than black coaches even when these factors were taken into account.'

This language is very cute. The statistical tests in other words identified two factors as correlated with coaching tenure — success in winning games, and the team's prior success (presumably teams that were used to success had a shorter trigger if the new coach did not perform). It is not clear from the paragraph above that race is correlated with job tenure. The explanation only notes that even when the other factors were taken into account, white coaches had longer tenure.

What does this mean exactly? Does it mean that race is a factor that is positively correlated with job tenure? To what degree? At best, if it is a third ranking explanatory factor, probably not much (technically, r squared closer to zero than to one).  If there is a significant statistical correlation between race and job tenure in the NBA coaching ranks, wouldn't the authors have included that fact in the article (which took up almost a full page in the sports section in addition to the front page lead)?  After all, this paper is read mainly by blue staters, many with advanced degrees. They can handle the statistical stuff.  Rather, the authors provide some averages that seem to show that white coaches last longer on average under each separately considered criterion.

Sorry, but averages, with one factor considered at a time, do not equal causality or relationship in statistics. In fact, all the other factors that are considered separately, one at a time, in this 'Shortcomings Remain' table, as well as other factors not tested or listed, could explain the difference in average job tenure between the white and black coach job tenure average in every one of the four comparisons listed: victories and losses, playoff success, experience, and former NBA players. 

Two other tables in the article are also revealing. One table shows the coaches with the longest tenures in recent years who were hired in 1989 or thereafter.  All but one of them (Lenny Wilkins) is white. Interestingly, this coaching list also includes all the coaches of NBA champions since 1989, except for Chuck Daly, who coached the Pistons to titles in 1989 and 1990, but was hired before 1989.  All the coaches of champions since 1986, Pat Riley, Rudy Tomjanovich, Phil Jackson, Gregg Popovich, and Chuck Daly are white.  Pat Riley won no titles with the Miami Heat, but four with the Lakers in the prior decade. Might that record of prior success have gotten him a few more years to win one in Miami? You might suspect so.

The recent experience suggests that many of the longer tenured coaches — Popovich, Tomjanovich, Jackson — were winners.  Their longevity skews the averages towards the higher white coach tenure average. K.C. Jones was the Celtics' coach for five years in the 80s and won two titles (the last in 1986). Jones is black, but his experience was not included in the Times survey, though it is consistent with the longer tenure for coaches who win. Winning means longer careers, losing means early exits. Not a real shocker. This might be your conclusion about this data set if you were an honest researcher, rather than somebody digging to find racism, whether or not it exists.

The other table lists all the black coaches who have ever coached either in the NBA (49), Major League Baseball (22), or the NFL (10). No list is provided of black coaches in the NHL, another case of racism at work, I must assume. The NBA has clearly had a much better record than the other major sports in hiring black coaches. As the authors point, out, a third of the league's current coaches, even with Silas gone, are black. If anything, the NBA's effort should be applauded rather than faulted. But obviously, their record is not good enough for the Times. Rather than the NBA owing an apology to the fired black coaches, or worse, having to meet with shakedown artist Jesse Jackson to explore ways of improving coaching diversity, or improving general manager sensitivity on the firing issue, the Times owes an apology to the NBA for besmirching the league's decent record in integrating its coaching ranks. The history is not perfect of course, but like much in American society, things have been getting better in this regard.

This year, it was a big story when Notre Dame's football coach, Ty Willingham, was fired. It was a big story because Notre Dame does not fire coaches before their contract term is up, and because Willingham was one of the very few black head coaches in Division 1 college football.  Willingham had a solid first season, and disappointing seasons the next two years at Notre Dame.  He was soon offered a job as the new coach at the University of Washington, which won only one game last year. Willingham will probably improve on that record next year, but won't win national championships with the Huskies so long as Southern California continues to play like a pro team. Willingham probably does not have to win a national championship to have secure job tenure at Washington. Winning seasons is probably enough.  But at Notre Dame, national championships, not winning seasons, are the only way to wake up the echoes. And that is very tough to accomplish when you are at a school that actually expects even the football players to go to class and take calculus.  New coach Charlie Weis will be feeling the pressure soon enough.

I would argue that the Willingham firing was a good thing for the future of black coaches in NCAA football. So long as a school is afraid that they won't be able to fire a black coach, they won't hire one.  This is also true in the NBA. Coaches come and go, and race should not be an issue in the decision. In the NBA black coaches can be both hired and fired. That is a good thing.  I do not think the evidence proves that race is a factor in coaching longevity in the NBA, unlike the Times writers' poorly reasoned conclusion.

Winning is the test in the NBA. The NBA is not a sport in which empty arenas can still produce profits for the owner (some NFL teams could make money with little or no attendance due to the very large annual TV revenues distributed equally to every team). Attendance matters in the NBA. Winning is positively correlated with attendance, not just a coach's job tenure. In Chicago, attendance at Bulls' games fell 30% in the horrible years after Michael Jordan retired the second time, and the team proceeded to lose over two thirds of its games each year.

The NBA had had its problems the last few years. The league's once great role models and fan favorites are now all gone — Larry Bird, Magic Johnson, and finally for the third time, Michael Jordan. The heir apparent role models have not filled their shoes — Vince Carter, Grant Hill, Kobe Bryant, Tracy McGrady. Now LeBron James, Nike's $90 million dollar man is on the stage. He scored 56 the other night, but his team, the Cavaliers, lost. If the Cavs kept losing, they might have fallen out of the playoffs altogether. That would be a hit to the NBA's TV ratings.  It is also only three months since Ron Artest starred in NBA Players Gone Wild.

The NBA did not need the Times' attack piece yesterday, and did not deserve it. But the paper seems to have found its next major sports cause. And as the Masters controversy demonstrated, good common sense is not enough reason to get the paper off its high horse.