PDA

View Full Version : Dean Oliver on Quantitative Analysis



DenButsu
10-16-2010, 08:46 PM
This is the first page of a four-page interview.

Click this link to SLAM (http://www.slamonline.com/online/nba/2010/10/dean-oliver-on-quantitative-analysis/) to read the whole thing.


Dean Oliver on Quantitative Analysis
Oliver talks about the role his statistical analysis plays in the NBA.
by Kyle Stack

When most baseball fans think of statistical analysis in their sport, Bill James is usually the person that comes to mind. Yet with statistical analysis growing in basketball, many fans still might not know who Dean Oliver is. You might know him as the author of the well-known book Basketball on Paper, which was released in 2003. He serves now as the Director of Quantitative Analysis for the Denver Nuggets, which is a position heís held since 2006. Prior to that, he worked for six years as a statistical consultant to the then-Seattle SuperSonics.

Oliverís path to the NBA isnít like any you would imagine. Sure, he played college ball. But it was at Caltech, the academic home of NASAís Jet Propulsion Laboratory. He served as an assistant coach there for three years after graduating, then worked for a basketball scouting service from 1990-94. Once he earned his Ph.D in using statistics to make forecasts from the University of North Carolina in Ď94, he had laid down the path to eventually join the NBA.

Iíve wanted to speak with an expert in basketball quantitative analysis for some time. I felt like Oliver would be the ideal person to do that with, and I spoke with him recently about his role in statistical analysis, and how he uses his analysis to help him find the player he thinks the Nuggets needs. deano

SLAM: What were your expectations when you started with the Nuggets?

Dean Oliver: I actually had two seasons with the Sonics prior to that. That set my expectations with the Nuggets, a little bit. Between those two years [I was with the Sonics and Nuggets] I had time to take what I had known and done with the book on paper and apply it to more live situations. What does a coach have to answer? What does a GM have to answer? I had that. Coming in to the Nuggets, it was an opportunity to work with a little bit different people, because everyone has a different style. Working with the new group I was definitely hoping for a little bit more say-so. Thatís always what youíre looking for. I was able to do that.

SLAM: When you started, did you think that your value would be reporting to the GM and owner, or did you think it might be providing info to players and coaches?

DO: I never thought about providing information directly to players, except little bits and pieces here and there. I definitely felt like there was impact for the coaches. The coaches obviously know their own team very well, and I think you can only provide some information they donít know. Especially midway through the season. They know their guys pretty well. But they donít know some of their players, certainly ones who may be coming from other teams. It happens every year. There is some insight that can be gained. And then there is of course the other aspect of coaches dealing with opponents on a regular basis. Any information they can get on those guys, whether it comes from an advanced scout or from numbers can help. I felt like there was significant impact from a coaching side as well as the personnel side.

SLAM: Has the acceptance of quantitative analysis in the League changed much from 2006 to now?

DO: Yes. I would definitely say so. The immediate reflection of that is the rise of the MIT conference from the analytics. I think that was the first year it ran, if I recall correctly ó 2006. It was a handful of us. I think all the people who attended were able to go to the Boston Celtics game. Now, they canít fit people in. If I have to reflect the acceptance, I think some of it is what [the media] has done, just to publicize it a little bit.

SLAM: Is quantitative analysis beyond the stage of it being just a trend?

DO: I think itís established. Itís not established like it is in Major League Baseball. To my understanding, MLB has a pretty thorough infiltration across the league. Within the NBA, you may have half the teams that dabble a little bit, but I think itís probably only 7-to-10 teams that have it integrated into their decision-making process. Thereís an upward trend. I wouldnít say itís fully established.

SLAM: What are your responsibilities?

DO: My responsibilities are to help out with pretty much every aspect that numbers can. I respond to requests to coaches, I respond to requests from Masai and the other members of our management. But I also put things out there to go in certain directions based on my experience of what I think helps. I put things together with the advance scouts. I send things to them and send things to members of the coaching staff. Thereís a lot of tasks that are related to numbers. Itís a numbers game. Thereís so much that goes on thatís recorded. The fact that youíve got a lot of games and a lot of history of games over 30 or 40 years, you can make inferences based on a lot of history.

SLAM: Do you have people working under you?

DO: No. Denverís always been pretty small.

SLAM: Wow, thatís a lot of work.

DO: Yeah, there are definitely times during the season when I feel like Iím trying to do too much. Itís one of those thingsÖyou get more efficient every year but you also find more things to do. Some of those things that historically may have taken more time can now be done faster, but that just means you find new things to take more time. Thereís always a bit of a push-pull on that.

Patman
10-20-2010, 09:52 AM
It's all pretty general and sometimes he seems to evade the question. But overall an interesting read.

I finally received his book after 3 1/2 weeks. Started reading and it is very interesting, will reserve further judgment until i read the whole thing.