"Besides a few members from NBA teams and a brief exchange with [Houston Rockets GM] Daryl Morey, I was able to meet a few of the writers of other great basketball stats sites. I met Kevin Pelton from Basketball Prospectus, as well as Ryan Parker from Basketball Geek. They're all great guys who are eager to learn more about the field in which they excel. I didn't realize Justin Kubatko was there until it was too late. It would have been nice to meet him."
-- Jon Nichols, Basketball-Statistics
If you've been a reader of Third Quarter Collapse for a little while now, you'll know that 3QC makes reference to Jon Nichols' work at Basketball-Statistics periodically in analytical pieces and news stories, alike. Nichols is one of the more intelligent stat heads around the internet, today, and an active member of the APBRmetrics community.
I'm a big fan of Nichols' material and have chatted with him several times on the APBRmetrics forums regarding different advanced statistics he's created these past few months. Thus, it made logical sense to interview Nichols and gather his thoughts on a variety of topics (including the Orlando Magic, of course). In my conversation with him, Nichols divulges a bit on his background, provides a backstory on the various metrics he's come up with, talks about the 2009 MIT Sloan Sports Analytics Conference, and more.
This interview will be a two-part series (Part II will be revealed on Thursday).
Today is Part I of my Q/A with Nichols.
Click after the jump for the full transcript.
How long have you been running your website, Basketball-Statistics?
I don't remember the exact date, but the site was up and running some time last October.
What spurred you to create the site in the first place?
I had two main reasons for doing it. First, I had been doing a lot of work for different sites such as NBADraft.net, Courtsidetimes.net (which no longer exists), and 82games.com. Those sites were all great and did a lot to get me readers, but I also wanted a site of my own so that I would have total creative control. Now I still write for other sites (mostly HoopsDaily.com but occasionally RealGM.com), but I have all my articles running simultaneously at Basketball-Statistics.
My second reason was not as glamorous. I've been trying to find my way on to the basketball operations side of an NBA team, so I used the site as a resume builder. The site makes it easy to show off what I've done and shows that I'm dedicated to this stuff.
What are your current credentials at the moment? I know that NBA teams are like hawks, waiting to prey on available "stat geeks" with one felt swoop .. have you've been approached or collaborated with any NBA teams at this point in your career? If not, would you like to? If yes, why? If no, why not?
As I mentioned earlier, my goal is to work for a team eventually. I've spoken with guys from a few different teams, but so far I'm still a "free agent," so to speak. I have not done any consulting work, although I know teams have read what I've done. As for why I'd like to work for a team, I guess it's pretty simple. I love basketball, so why wouldn't I want my job to involve something I love?
I know you're aware that I've made reference and use of some of the various metrics you've created the past few months over at Third Quarter Collapse. I'm curious, how long did it take for you to create Composite Score, PAC, Value Rating, & the Box Score Prediction System? What was the initial basis and motivation to introduce some new statistics to the table?
It actually took me quite a while to develop Composite Score, my first original statistic. I was doing different research studies (many on the draft) for about a year using other people's numbers. Eventually, I decided to make one of my own, which is basically a combination of all the different numbers I was using. My motivation for making Composite Score was that I noticed how each stat had a flaw of its own. I figured maybe if I combined them all, those flaws would be minimized and the different numbers would make up for each other's limitations.
PAC, Value Rating, and BSPS are all relatively new. My motivation for making PAC was actually a book on coaching that I read. It resonated with me how important roles are in the NBA and play ers having different skills. I wanted to be able to develop a system that I could isolate who's good at what. Whether or not PAC does that is up for debate. I was inspired to make Value Rating after doing a lot of reading on the Collective Bargaining Agreement. I guess all that financial stuff got me into that mode. The Box Score Prediction System has been something that's been on my mind for quite a while. Ever since I started doing draft studies three years ago, I thought it would be useful to be able to predict NBA numbers based on college stats. Recently I figured out a nice way of actually doing it, so I got to it.
Do you believe any of your metrics need further refinement? If so, could you share what you believe could better fine tine the various number systems you've created?
I know for a fact that all of my metrics could use further refinement. They are certainly not perfect, and I try to make the limitations of each stat clear when I release them. I think Composite Score's biggest problem is that it is too team-dependent. So-so players on good teams look better than decent players on bad teams. If there was a way to take that away, it would be nice, but the problem is that some of the stats CS is based on are heavily team-dependent. PAC is an interesting case. In some ways, it may be too broad and doesn't break down players specifically enough. In other ways, it seems like there are too many categories and it's hard to keep the different ones straight. That's something I'm wrestling with right now.
As we've talked about before, Value Rating could do a better job of measuring the value of the high paid players. It really is useful with the medium-level guys, but guys with max contracts don't look great with VR and that may not be fair. The biggest weaknesses with BSPS are that it is not adjusted for pace or strength of schedule. Guys who play on fast-paced teams or guys who beat up weaker competition are going to project to be better than they actually are. The problem with that is the lack of historical advanced college stats.
You were able to attend the 2009 MIT Sloan Sports Analytics Conference this year in Boston. Could you briefly describe the experience of being there?
It was great. A lot of the great statistical minds were there, and there was a decent amount of representation from the teams. To be honest, the information provided in the panel discussions wasn't the highlight for me; it was meeting all the people.
Likewise, what are some things you took away from the Sloan Conference, with regards to further enhancing your capability and knowledge with advanced statistics?
Like I said before, there wasn't a ton of groundbreaking stuff revealed at the conference. I think it was a case of a lot of team personnel being afraid to give away any of their secrets. I follow what's going on in the basketball statistics world pretty close, so there weren't too many surprises. The presentation on whether or not basketball players get hot was interesting, though. Also, I came away with the impression that things such as game intelligence and team chemistry may be more important than many stats guys think.
Who are some of the individuals you were finally able to meet with person and what were some of them like?
Besides a few members from NBA teams and a brief exchange with [Houston Rockets GM] Daryl Morey, I was able to meet a few of the writers of other great basketball stats sites. I met Kevin Pelton from Basketball Prospectus, as well as Ryan Parker from Basketball Geek. They're all great guys who are eager to learn more about the field in which they excel. I didn't realize Justin Kubatko was there until it was too late. It would have been nice to meet him.
Part II of my interview with Jon will be unveiled on Thursday. Stay tuned.