[Today's column is brought to us from KnickerBlogger Head West Coast Analyst Kevin Pelton, who serves as the Sonics and Storm beat writer for SUPERSONICS.COM and storm.wnba.com. He formerly wrote the APBRmetric "Page 23" column for Hoopsworld.com.]
When Steve Nash unofficially ascended to the role of NBA MVP candidate, if not favorite, early this season, the response of the statistical community that follows the league was more indifference than the harsh criticism heaped on Allen Iverson’s 2000-01 selection or Jason Kidd’s candidacy the following season.
Still, you could probably find more APBRmetricians who believe the moon landing wasn’t real than those who will be planting “Nash for MVP” signs in their front yards. About a month ago, my gracious host posted a poll at the APBRmetrics message board querying posters on their first-half MVPs. 20 people voted, and nary a one selected Nash.
Personally, I voted for Tim Duncan, with Kevin Garnett second. Over the last month, however, Nash for MVP has been knawing at me a little bit. Were we all too quick to judge because Nash’s PER isn’t up to snuff, or because we’re tired of the overly simplistic arguments about how Team X was bad before adding Player Y and is now outstanding? If Nash isn’t a serious, serious MVP candidate this year, can any point guard ever be?
One of the things I’ve been thinking about is how good Nash’s season really is in terms of assists. Nash’s 11.5 assists per game are the best in the NBA since Mark Jackson averaged 11.4 in the 1996-97 season, but he won’t even get the Suns record for assists per game without a remarkable run. (It’s owned by Kevin Johnson, who averaged 12.2 assists in 1988-89, setting up a pair of players who averaged 20 points per game, Tom Chambers and Eddie Johnson. KJ himself added 20 ppg as well.)
To level the playing field, I’m going to borrow from baseball analysts. In baseball, two things have to be corrected for: Park effects and league context. In basketball, pace factor replaces park effects. The first step, then, is calculating assists per possession. Nash is averaging assists on 16.2% of the possessions he’s played this season. Johnson, by comparison, had assists on just 13.6% of his possessions in 1988-89.
In 2004-05, the average NBA player gets an assist on 4.5% of his possessions; in 1988-89, that mark was 4.9%. So if we divide Johnson’s rate by the ’89 average and multiply it by the ’05 average, his 2004-05 assist rate is 12.9%. That’s a relatively meaningless stat, so what I’ve done is translate it to assists per game, assuming 96 possessions per game (two for each team every minute) and 32 minutes per game. By this standard, Nash averages 10.4 assists per game. (His average goes down because he plays more minutes and for a team playing at the league’s fastest pace.) Johnson’s average takes a much bigger hit, down to 8.2 assists per game.
If you had more free time than I have (and the length of my columns has in the past drawn suggestions I have too much time on my hands as is), you could with the help of Basketball-Reference.com calculate this equivalent assist average for NBA player dating back to 1973-74. (It would be impossible before that because team turnovers were not tracked, meaning no ability to determine possessions.) I didn’t do that, but I picked out the best 50 or so passing seasons of the past three decades and translated them. Here’s the leaderboard from amongst that group:
Player Year Team 05APG
John Stockton 1990 UTA 12.0
John Stockton 1988 UTA 11.5
John Stockton 1991 UTA 10.9
John Stockton 1992 UTA 10.8
John Stockton 1995 UTA 10.6
Steve Nash 2005 PHO 10.4
John Stockton 1994 UTA 10.1
John Stockton 1996 UTA 9.8
Magic Johnson 1991 LAL 9.8
John Stockton 1993 UTA 9.8
I believe the technical term for that list is “select company”. Steve Nash is, at the moment, having the best passing season of the last 30 years put together by a player not named “John Stockton”. When you consider what an efficient scorer Nash is as well, you could drop his line this season into the middle of Stockton’s career, adjust for league and pace, and I doubt anyone would be the wiser.
Using my similarity system, 13 of the top 15 comparable seasons for Nash 04-05 were posted by Stockton, including the top five. It’s worth pointing out that Stockton only topped Nash’s 04-05 PER (23.35) twice during his career.
Check back Wednesday as Part II of this column looks more carefully at Nash’s MVP credentials.