In my previous articles, I presented “newer” basketball statistics that provide clearer insight into basketball games and the performance of players than do some of the box score statistics. We are entering an era in which some NBA teams are beginning to rely heavily on statistical approaches to analyzing games and players.  This is especially true of the Houston Rockets, as Daryl Morey is their new general manager. Morey has degrees from Northwestern as well as M.I.T. and is viewed as a non-basketball person with great analytical and quantitative skills (see the article about him at
http://www.houstonpress.com/2007-11-01/news/rocket-science). Certainly the performance of the Houston Rockets will be closely watched and analyzed this year since Morey has helped put the team together.

Other NBA teams have hired people with recognized quantitative skills and a strong sports background who have become important team members.  In particular, Dean Oliver, who has a Ph.D. in engineering, played basketball in college, has worked as a scout, and is the author of Basketball on Paper, was hired by the Denver Nuggets last year and named their Director of Quantitative Analysis.  Dan Rosenbaum, a college professor at UNC-Charlotte who has a Ph.D. in economics and was also a college athlete, is a part-time consultant to the Cleveland Cavaliers.

It is well-known that the Boston Red Sox’s operations have been run in the past few years with heavy reliance on quantitative approaches and they have just won their second World Series title in the past four seasons. Whether or not the Rockets, in particular, enjoy similar success remains to be seen, but certainly many basketball people will be scrutinizing their performance very closely, especially those who prefer to go by what they see on the court rather than what the numbers say (as is the case with Kenny Smith, who was quoted on page four in the article on Morey).   

This season the NBA will be keeping some of these newer statistics on an unofficial basis and perhaps make them official statistics in a year or two.

Unfortunately, there is not a parallel movement in college basketball. When the Board of Directors of the National Association of Basketball Coaches (NABC) met about a year ago to discuss possible adoption of the composite field goal percentage (called the “effective field goal percentage” by some basketball people), I was told that the prevailing question was “How will this make me a better coach?”. Of course such an understanding will probably not come without firsthand experience with these statistics.

What I had in mind with this article was to use it to initiate discussion regarding the role of statistics in basketball, as the title states, at all levels. If new quantitative approaches to assess individual and team performances are being deemed important at the NBA level, then clearly they should also be useful at the college and high school levels.

The acronym KISS is often used for “Keep It Simple Statistically”.  If we believe that newer approaches are necessary, then it seems reasonable to assume that the starting point should be the introduction of statistics that are both effective and intuitive.   

Kubatko, Oliver, Pelton, and Rosenbaum (2007) published a recent paper that sets forth their views on the use of basketball statistics in what they term “A Starting Point for Analyzing Basketball Statistics”, which is the title of their article.  Although I am acknowledged as a contributor to their paper, I don’t necessarily agree with all of their views that were espoused, nor do I agree with all of their terminology.  (For people who don’t subscribe to the journal, as I don’t, the article can be downloaded after first going to the website for the article, http://www.bepress.com/jqas/vol3/iss3/1, and registering as a guest.  I caution that the early parts of the article require knowledge of regression analysis, a statistical technique, in order to understand the presentation.)  

I will add that Rosenbaum (speaker) and Oliver (panelist) participated in the New England Symposium on Statistics in Sports that was held at Harvard University on 9/29/07.  The symposium may be viewed online by selecting the individual videos at http://www.amstat.org/chapters/boston/nessis07/videos.html.

It is obvious that the heavy use of statistical methodology is “catching on” in the NBA, at least with certain teams, and undoubtedly more teams would jump on board if the Rockets were to win the NBA championship this season. 

With that season and the college season now underway, the following questions seem relevant.  Should other NBA teams create positions for scholars with strong reputations in the field of statistics and other quantitative disciplines who have played basketball at the high school level and beyond and are students of the game?  Should college teams attempt to emulate the Rockets, Nuggets, and Cavaliers and, say, appoint one of their graduate assistants to become knowledgeable in the new statistical approaches and keep such statistics as part of the team statistics?   What are your views on this?

To respond, please click on the “Ask a Contributor” icon on the right side of the Online Magazine Home Page.  Enter Dr. Ryan’s name into the “Contributor” box and fill the “Your Specific Question” box with your response.  Thoughtful responses will be compiled into second article.  Dr. Ryan will then respond and sum up in a third article.
      
REFERENCE
Kubatko, J., D. Oliver, K. Pelton, and D. T. Rosenbaum (2007).  A starting point for 
     analyzing basketball statistics.  Journal of Quantitative Analysis in Sports, 3(3), 1-22.