APBRmetrics

Hi all-

I'm looking to do some game tracking for the entire league- kind of like what the Football Outsiders guys do for the NFL, only for the NBA- based on the game tracking work of El Gee from the now defunct BackPicks.com as a foundation, with hopefully a few new wrinkles thrown in. If you're interested, please e-mail me at nbacouchside@gmail.com. I can't offer any money, obviously, but I think it'll make for a fun way to re-watch games for deeper understanding. I know 82games.com has done somewhat similar stuff, but I think things like Opportunities Created, Defensive Errors, and Missed FGA against are not things accurately tracked anywhere (publicly anyway) in a comprehensive manner. I want to change that. I have about 5 guys, including myself, currently interested in doing this. Ideally, I'd like to have at least 15 guys, but the more the better, because it would reduce pressure on everyone to do as many games.

I have 9 or 10 guys who seem to have expressed a good amount of interest (including myself), so if I could get a handful more, we could really get this thing off the ground.

Seems like an interesting idea... Just curious what the incentive is for somebody to track for you? I mean, I assume you are going to publish the results or something, correct?

mark kieffer wrote:Seems like an interesting idea... Just curious what the incentive is for somebody to track for you? I mean, I assume you are going to publish the results or something, correct?

Honestly, I don't have any money to pay people for their time, but should it be a successful endeavor, I'd like to eventually compensate people for their time to the extent possible. In the meantime, I guess it's just an opportunity to work on a cool project as a hobby, discuss hoops in depth with some other basketball addicts, and maybe see that work become spread wide over the internet. The model I was looking at was Football Outsiders, who have volunteer game trackers, but like I said, ideally, if I could, I would like to give people some sort of compensation for their time. As of right now, all I'd be able to provide would be credit for their work.

I don't have time to participate, but if I did I would assume that all participants would get full access to the data. Not to the individual analyses that each person does, but access to the raw -- but cleaned up and formatted -- full set of data.

The harder question is whether to make those same data available to the public. Pros and cons both ways; I would suggest that if the data are not made available to the public then the model that RolandB used at 82games.com is a good alternative: make some of the data available to researchers on a case-by-case basis with the understanding that the website could publish the results.

mtamada wrote:I don't have time to participate, but if I did I would assume that all participants would get full access to the data. Not to the individual analyses that each person does, but access to the raw -- but cleaned up and formatted -- full set of data.

The harder question is whether to make those same data available to the public. Pros and cons both ways; I would suggest that if the data are not made available to the public then the model that RolandB used at 82games.com is a good alternative: make some of the data available to researchers on a case-by-case basis with the understanding that the website could publish the results.

Yes, the full data is to be available to all of the trackers. Whether for their own personal amusement, for writing they do elsewhere, or what have you. As far as what will be shared publicly, I want to make it as widely available as possible, but I guess I wouldn't rule out having some things kept secret or available only with a subscription (set to a relatively small price) or maybe sell an e-book (ala the FO guys) with the full data. There's a lot of ways to do it, but yes, the trackers will definitely have access to the data.

I hope this gets off the ground. I'd suggest that publishing at least some of the early efforts might attract more assistance.

I am more likely to participate in a discussion of what to track and how to do it than the actual video work at this time. Others might fall in this group. If you want that kind of discussion here, give us more to chew on and react to.