March 2009 - Posts

 As I noted previously, I've been getting into analyzing NBA stats recently, which seems all the rage these days. Anywho, here's a random fun stat for the day:

Guards and forwards are both more than five times as likely to have a season with more offensive rebounds than defensive rebounds. The only centers to have more offensive rebounds than defensive rebounds in a season were (when considering only players who played 41 games or more):

  • Chris Dudley (1987)
  • Ed Horton (1989)
  • Stacey King (1992)
  • Scott Lloyd (1978)
  • Granville Waiters (1985)
(Keep in mind that rebound stats weren't divided into offensive and defensive stats until the 1973 season.)

 A few weeks ago I also added a new feature to the homepage - statcenter's Twitter feed. statcenter routinely posts various NBA stats via his (her?) Twitter feed, which then gets displayed on this site's homepage.

Posted by Scott Mitchell | 50 comment(s)
Filed under:

Over the past decade there's been quite the groundswell of interest in the gathering and analysis of basketball statistics. Purportedly, many NBA teams have at least one full-time statistician, if not several, trying to tease out trends and gain a deeper understanding of player and team efficiency based on the numbers. Perhaps one of the better known statistics is John Hollinger's Player Efficiency Rating (PER).

Being a numbers geek at heart, I've often thought it would be interesting to plough through some statistics and make observations or discover some insight from the sea of data. Of course, before you can analyze data, you need to get your hands on the data.

There are some websites that sell statistics marked up in the official SportsML markup language. However, being a hobbyist, I wasn't too keen on paying for stats and, fortunately, stumbled across free basketball stats in Microsoft Excel format at There are different spreadsheets for team stats by season, player stats by regular season, player stats in the playoffs, and so forth. Moreover, the statistical history dates back to the 50s and includes NBA and ABA teams, although certain stats, such as assits, weren't recorded until the 70s, and things like three pointers attempted and made were not relevant prior to the three point line. I spent some time getting the data from the Excel spreadsheets into a Microsoft SQL Server database, where I can more proficiently delve into the data. I'm still in the process of integrating all of the spreadsheets, but I currently have the team and regular season player data.

One area of interest for me has to with how the game changed from the 80s to the 90s, and now into the 00s. The 80s was typified by high scoring, fast break offense. Fun, free-flowing ball with high scoring and breakneck action. But in the late 80s the Pistons showed that defense could win championships, and the Knicks and Bulls exemplified a stronger defensive mindset. Scoring dropped precipitously. Over the past decade, the NBA has had several rule changes to encourage a faster paced game, which has brought more energy to the game and increased scoring, although not to the levels in the 1980s (or even to the levels of the early 1990s), but the offensive numbers are steadily increasing with each season this decade. I am curious as to how this change in strategy from the 80s to the 90s affected various facets of the game, and whether today's trend of more offense is going to return us to an 80s style of play or not.

I plan on writing up a more in-depth post as I continue to poke at this data, but for now consider this:

  • Average ppg in the 1980s: 108.6
  • Average ppg in the 1990s (Excluding 1998): 100.5
  • Average ppg in the 2000s: 96

The high-scoring 80s did not end abruptly with the start of the 1990 season, but were starting to wane from the high of 109 ppg in 1988 to 106 in 1990. From 1990 to the 1997 season scoring decreased every year, from 106 ppg down to 95 ppg. (I have exlucded the 1998 lockout year from my calculations, in which half a season was played; in this lockout year there was an abysmally low 91 ppg, but that was certainly due in part to players being out of shape when the season started.) The lowest ppg season since 1980 was 2000, at 94 ppg, but this number has been steadily increasing every year since, and last season (2007) was at 99 ppg. And this year currently stands at 99 ppg, on average. Point being, we're unlikely to see a decade ppg average eclipse 100.

The more interesting question is how did the improved defense of the 90s affect the offensive game. Sure, fewer points were scored, but was that because there was poorer shooting or fewer attempts at shots because of stiffling defense? Or were there fewer fast break attempts? Fewer fouls, and therefore fewer free throws? Or some combination of the above? Moreover, how have these factors changed as the scoring has increased over the past eight seasons? Are we seeing a return to the conditions in the 80s or is today's offense more the product of quicker foul calls and one-on-one plays? The answers lie in the data! Smile
Posted by Scott Mitchell | 108 comment(s)
Filed under:
More Posts