AFTER 17 YEARS OF RESEARCH AND 8 YEARS ON THIS SITE, I landed myself a job in the NBA. I've been working with the Seattle Supersonics since before 2004 Training Camp. Given that we were predicted for last in the West and we are, as of this moment, best in the entire league, I think the science works. Yeah, this stuff really helps teams win. (Late note: In no way do I want the previous statement to imply that I am even a big part of the reason the Sonics are winning. I am apparently not a very good interview subject because it may have come out that way as I've read articles where I've been the subject. I've got to learn to manipulate the press the way politicians do. I will say flat out that I provide information to the team that they use, but I don't make a shot or make a substitution. I think my work helps, but the win improvement is due to a lot of people.)
It's about more than personnel decisions, which is what most other statistical analyses have addressed. Coaching does matter. My role with the Sonics has been to help on both management and coaching sides. I've assembled plans for improving the defense, development plans for our young players, and I've demonstrated what players should play and do more. This is all fine and good, but useless without the cooperation of the coaching staff there, which has overcome the traditional bias against numbers to actually use the work. They have taken what I do, blended it with their own unique skills, and turned it into what I always felt was possible -- winning through good chemistry. It's the best science I've done.
I will admit that I get hugely nervous before every game. I can't sit down. I pace incessantly. After games, I work for hours to better understand our strengths and weaknesses. I build the numbers into my programs and look for significant trends. I sleep a lot less than I used to. But every hour I'm awake, I smile because I've come to realize that the dream I've been chasing only becomes real if I do stay awake.
The reception I've gotten in the league has been friendly but still muted. I have introduced myself as a former coach and scout, which has gotten me credibility. But I have followed it up with Basketball on Paper and the phrase, "I'm a stat guy, bringing Moneyball to the NBA." That has intimidated people or turned some off. Some have asked me to explain a bit and it's been clear that I have a lot of educating to do. For some people, the education has been that stats can make a difference. For others, especially those who have read Michael Lewis' book on the Oakland A's (coincidentally my favorite team) use of stats, they have known that basketball is not like baseball. They have suggested that stats can't be used as broadly as in baseball because of the dynamics among basketball players. But what I have done through years of research is emphasize those dynamics. The most complex formula-filled chapter of Basketball on Paper is that way because I illustrate some of the things I account for in assessing player interaction or "chemistry." This hasn't convinced anyone thoroughly that I know what I'm doing. Mark Cuban said I was "half wrong" in dismissing my book (something that, ironically, convinced me that I was right). I had to take the Sonics job at a very low pay rate because it was "an experiment" for the team. But I'm confident that my rate will go up significantly with the use of all my ideas in the success of the team.
The tools I have used to get here are mostly in Basketball on Paper, some of which are here. There are a few tricks that I haven't spelled out much, including Roboscout, which has been hugely valuable for assessing tendencies of other teams quickly in such a way to see where they are vulnerable. Beyond this, Roland Beech of 82games.com has been great. His site provides data from play-by-plays that has helped us up in Seattle. I will vouch for the usefulness of his data and am gladly advertising it for what he supplies me. Dan Rosenbaum, an economics professor at UNC-Greensboro who is also an expert in the NBA luxury tax, has also built a statistical tool that estimates individual player contributions to team success by tracking the score while who is in the game. His tool has also been useful to me. When Roland, Dan, and I went to the Chicago Pre-Draft Camp with our set of tools, we got to meet numerous coaches, scouts, and people in management. Stan Van Gundy of Miami is a great guy who read anything I passed his way. Jeff Bower of New Orleans has been great to me, buying my book on the spot. Sam Presti in San Antonio filled me in on some of the rigors of doing the job. Dave Griffin in Phoenix asked really good questions. Greg Bittner in Denver pushed me in the right direction many times. Chris Mullin listened to me and put me in touch with his organization. Jordan Cohn, a scout now with the Nets, filled me in on so much of the way things work. Irving Thomas of the Lakers gave me what he might have thought wasn't a very inspirational little talk ("It took me several years to get into the business."), but was hugely inspiring. These guys (and others I haven't named) gave the group of us "outsiders" some credit for having something to offer. I'd like to think that Seattle's early success is proof of their initial faith.
I won't be doing a lot of public research. I can't. Too many secrets I've got to keep in order to really make a living. I miss the public aspect of doing good science and maybe I'll do it in a different forum, like football or economics. But, as much as I miss that, the NBA is a dream I'm living now.