APBRmetrics

Not many people are going to read the 5,000 plus posts in the early Yahoo group files. Here is a brief recap of some of the bigger highlights from those first 4 years, 2001 to early 2005. This group is a spin-off from the original APBR board / group with a heightened focus on "analytics".

I am Dean Oliver and I am interested in basketball analysis. http://www.apbr.org/metrics/viewtopic.php?f=2&t=7863

The shot clock affects shooting efficiency. Raw plus minus is available now. Lets compare Tendex to the other boxscore metrics. Bob Chaikin shows up. Within plays there are touches which can be estimated. Amount rest may affect player performance. We need more stuff about defense. Baseball guys compute similarity scores. Rodman is interesting. Who won the trade? Mike Goodman and Mike Tamada show up. Iverson is someone to talk about. Kidd seems to help team. What is clutch?

Average of about 1 post per day.

2002

Dave Berri ran a regression. Would be interesting to look at assisted and unassisted fg%s. What should the ft multiplier estimate be? What do you think the linear weights should be subjectively? How do player stats in NCAA translate to NBA? Who are the ball hogs? Can we use point differential to predict team wins? Hey this guy John Hollinger put out a Basketball Prospectus with a new metric called PER. But is not that new.

In active months, up to 6 posts per day.

2003

Expansion kinda diluted the league. Igor Kupfer and Kevin Pelton show up. Dean Lavegne makes a stab at building a play by play database. Bob Chaikin has a lot of yearly data. Antoine Walker is polarizing. What recent players deserve the Hall of Fame? There is this high school guy named LeBron James. Wanna guess how many wins teams are going to get? Who is going to win title? Schtevie shows up.

2004 to early 2005

What is an assist really worth? How should we handle offensive rebounds? Hey this guy Roland Beech launched 82games.com. Here is how to order a copy of my new book Basketball on Paper. What's in it? A possession is defined as... Pace is... Win-loss records should be adjusted for strength of schedule. Anybody want to calculate Win Shares for basketball? Mark Cuban thinks we need some new, more specific stats. Gabe Farkas shows up. Dan Dickey shows up. Kevin Broom shows up. KG is a force. Guys, I got hired by the Sonics. Best wishes Dean. Justin Kubatko gives updates on Basketball-Reference developments. Dan Rosenbaum shows up. Is Shane Battier a great player? More people starting reading and some eventually start posting.

Eventually the forum moves to Kevin Pelton's site. And the saga continues for another ten years, til now... (for highlights of these years, see the recovered threads here)

Nice summary, reminiscent of "The Complete Works of Shakespeare in Sixty Minutes" whirlwind summaries. Before APBR had a board, it had an email list. IIRC the switch away from the email list happened at around the same time that APBRMetrics split off from APBR. The email list (and website) had been going on for years, I believe that both were started by Robert Bradley.

So in the early years it was just the APBR email list, covering both quantitative and non-quantitative topics. There was also a Usenet group, rec.sport.basketball (later rec.sport.basketball.pro) which AFAICT in the earlier years of the internet was the best and perhaps only place for these sorts of conversations until analytics became a more common topic on the APBR email list. E.g. there was an early attempt to emulate baseball's Project Scoresheet with a volunteer group NBAStats, a project to make detailed (by the standards of the day) boxscores permanently and publicly available. But Doug Steele's SteeleStats (sort of a predecessor to 82games.com, in terms of an energetic guy or group making all kinds of data available) made that project redundant, and then those basic stats became widely available on the web. And Usenet gradually became irrelevant.

Thanks, Crow, for the refreshing trip down memory lane. So inspired, I took a trip to the archives and was pleased to note that my chronic nattering about NBA shortcomings in three point shooting began (albeiit indirectly) in my inaugural post. Good times!

Thanks Mike T for adding to the early early history that I was only dimly aware of and forgot to mention. Thanks both for reading. Glad you found one of your main themes in the beginning Schtevie. As big as the analytics question list, knowlegebase and toolkit are now, they had to be built one by one. They may have been built separately by several groups redundantly but a lot was built by this group in its different incarnations. Perhaps someday Sloan will have origins panels for the different sports or one big one. Maybe not. There might be a fuller history coming out in print in future. This was intended to help with or tease for that.