Same Player, Different Roles
Posted by Neil Paine on December 29, 2010
I was browsing the stats this morning when I noticed that Ron Artest is currently using 14.7% of the Lakers' possessions when he's on the court, the 19th-lowest possession-usage rate of any qualified player in the NBA. Before joining L.A., Artest was accustomed to usage rates well over the league average of 20%, which had me wondering how Artest's decline in usage compares to other players who changed roles at varying times in the their careers.
It turns out that Artest is currently on pace to be one of only 5 players in NBA history (since 1952, at least) to have one qualified season with a possession rate of at least 25% and another with a rate of 15% of less:
| Player | Max Usg | Year | Min Usg | Year |
|---|---|---|---|---|
| Ron Artest | 25.7 | 2004 | 14.7 | 2011 |
| Wilt Chamberlain | 32.7 | 1962 | 12.8 | 1973 |
| Gary Payton | 28.0 | 2002 | 14.5 | 2006 |
| Guy Rodgers | 25.0 | 1966 | 14.4 | 1962 |
| Sidney Wicks | 27.4 | 1972 | 14.7 | 1980 |
Here are the players with the biggest difference between their largest and smallest-usage qualified seasons:
| Player | Max Usg | Year | Min Usg | Year | Diff |
|---|---|---|---|---|---|
| Wilt Chamberlain | 32.7 | 1962 | 12.8 | 1973 | 19.9 |
| Tony Parker | 32.1 | 2009 | 18.4 | 2002 | 13.7 |
| Gary Payton | 28.0 | 2002 | 14.5 | 2006 | 13.6 |
| Grant Hill | 30.1 | 1999 | 16.6 | 2010 | 13.5 |
| Tiny Archibald | 31.6 | 1973 | 18.7 | 1980 | 12.9 |
| Sidney Wicks | 27.4 | 1972 | 14.7 | 1980 | 12.7 |
| Michael Adams | 29.2 | 1991 | 16.7 | 1988 | 12.5 |
| David Robinson | 31.4 | 1994 | 19.3 | 2002 | 12.2 |
| Kevin Garnett | 29.5 | 2004 | 17.5 | 1996 | 12.0 |
| Kobe Bryant | 36.5 | 2006 | 24.7 | 1999 | 11.8 |
| Ron Harper | 28.4 | 1987 | 16.6 | 2000 | 11.8 |
| Jason Kidd | 27.4 | 2003 | 16.1 | 2009 | 11.3 |
| Ron Artest | 25.7 | 2004 | 14.7 | 2011 | 11.0 |
| Dwyane Wade | 36.5 | 2009 | 25.6 | 2004 | 10.9 |
| Antonio McDyess | 26.5 | 1999 | 15.7 | 2008 | 10.8 |
| Charlie Scott | 29.0 | 1975 | 18.2 | 1979 | 10.8 |
| Rick Barry | 30.5 | 1975 | 19.8 | 1979 | 10.8 |
| Dwight Howard | 28.1 | 2011 | 17.4 | 2005 | 10.7 |
| Guy Rodgers | 25.0 | 1966 | 14.4 | 1962 | 10.6 |
| Michael Finley | 25.8 | 1998 | 15.2 | 2009 | 10.6 |
| Paul Pierce | 32.6 | 2003 | 22.0 | 2011 | 10.6 |
| Steve Nash | 27.0 | 2011 | 16.4 | 1999 | 10.5 |
| Bernard King | 33.1 | 1985 | 22.7 | 1981 | 10.4 |
| Danny Granger | 27.8 | 2009 | 17.4 | 2007 | 10.4 |
| Carlos Arroyo | 24.1 | 2004 | 13.8 | 2011 | 10.3 |
| Phil Ford | 24.4 | 1981 | 14.1 | 1984 | 10.3 |
| Dennis Rodman | 20.0 | 1988 | 9.7 | 1994 | 10.3 |
| Andre Iguodala | 23.6 | 2008 | 13.4 | 2005 | 10.2 |
| Tracy McGrady | 34.4 | 2007 | 24.3 | 2000 | 10.2 |
| Allen Iverson | 35.8 | 2002 | 25.6 | 2009 | 10.2 |
Of course, not all of these players changed roles per se -- we wouldn't necessarily say Allen Iverson dropping to 25.6% in 2009 represented a completely different role, just a reduced one compared to the near-record mark he posted in 2002. To capture players who were both low- and high-usage at different points in their careers, let's limit the list to include only players whose max usage season was 23% or more and whose minimum usage was 18% or less (the same bins I put players in here):
| Player | Max Usg | Year | Min Usg | Year | Diff |
|---|---|---|---|---|---|
| Wilt Chamberlain | 32.7 | 1962 | 12.8 | 1973 | 19.9 |
| Gary Payton | 28.0 | 2002 | 14.5 | 2006 | 13.6 |
| Grant Hill | 30.1 | 1999 | 16.6 | 2010 | 13.5 |
| Sidney Wicks | 27.4 | 1972 | 14.7 | 1980 | 12.7 |
| Michael Adams | 29.2 | 1991 | 16.7 | 1988 | 12.5 |
| Kevin Garnett | 29.5 | 2004 | 17.5 | 1996 | 12.0 |
| Ron Harper | 28.4 | 1987 | 16.6 | 2000 | 11.8 |
| Jason Kidd | 27.4 | 2003 | 16.1 | 2009 | 11.3 |
| Ron Artest | 25.7 | 2004 | 14.7 | 2011 | 11.0 |
| Antonio McDyess | 26.5 | 1999 | 15.7 | 2008 | 10.8 |
| Dwight Howard | 28.1 | 2011 | 17.4 | 2005 | 10.7 |
| Guy Rodgers | 25.0 | 1966 | 14.4 | 1962 | 10.6 |
| Michael Finley | 25.8 | 1998 | 15.2 | 2009 | 10.6 |
| Steve Nash | 27.0 | 2011 | 16.4 | 1999 | 10.5 |
| Danny Granger | 27.8 | 2009 | 17.4 | 2007 | 10.4 |
| Carlos Arroyo | 24.1 | 2004 | 13.8 | 2011 | 10.3 |
| Phil Ford | 24.4 | 1981 | 14.1 | 1984 | 10.3 |
| Andre Iguodala | 23.6 | 2008 | 13.4 | 2005 | 10.2 |
| Christian Laettner | 25.0 | 1994 | 15.1 | 2003 | 9.9 |
| Jim Jackson | 26.7 | 1994 | 16.9 | 2004 | 9.8 |
| Artis Gilmore | 25.2 | 1978 | 15.6 | 1987 | 9.6 |
| Anthony Peeler | 23.9 | 1997 | 14.3 | 2003 | 9.6 |
| Clarence Weatherspoon | 23.2 | 1995 | 13.9 | 2003 | 9.3 |
| Reggie Miller | 24.4 | 1996 | 15.1 | 2004 | 9.3 |
| Johnny Newman | 23.5 | 1991 | 14.3 | 1996 | 9.2 |
| Quentin Richardson | 23.4 | 2004 | 14.1 | 2010 | 9.2 |
| John Starks | 26.2 | 1994 | 17.1 | 2001 | 9.2 |
| Dale Ellis | 24.8 | 1988 | 15.7 | 1999 | 9.1 |
| Thurl Bailey | 23.6 | 1989 | 14.6 | 1984 | 9.0 |
| Mike Bibby | 24.9 | 2006 | 16.0 | 2011 | 8.9 |
That's more like it. And here you see Artest with the 9th-biggest change of any player since the NBA started tracking minutes, a testament to the statistical sacrifice he's made in order to play for a championship-caliber team.
December 29th, 2010 at 12:28 pm
Shouldn't we be looking only at his usage from 2008-2009 versus his average usage from 2009-2011? His usage in 2004 is somewhat irrelevant when considering what he gave up to be on the Lakers.
Might also be more interesting to limit the comparison to seasons where the player switched teams, or added a significant player. These are just ideas, and represent a significant amount of work - which I am clearly too lazy to do myself.
December 29th, 2010 at 1:22 pm
Sure, that's something I can look at too. The main point here was just to capture guys whose careers have seen seasons as both "the man" and a role player.
December 29th, 2010 at 3:03 pm
How confident are you about Wilt's usage rate? Not that a 32.7 rate means he's Marcus Camby, but I was surprised to see his number that low compared to some current players. Then I looked at his player page, and saw that there is no usage rate posted for him, which I would have guessed to be the case. Any brief description of how you came to the number and whether you think it adequately compares to usage rates that are calculated for today's players would be welcomed.
December 29th, 2010 at 5:12 pm
I described the process here:
http://www.basketball-reference.com/blog/?p=6269
I estimated turnovers according to Justin's formula and estimated that 30% of TRB were offensive, which allows you to get a possessions estimate.
If it looks weird to see Wilt that low, consider that in 1973 he took 6 shots per 36 minutes, which would rank among the lowest rates in 2010-11:
http://www.basketball-reference.com/play-index/tiny.cgi?id=QP758
Add the fact that the average 1973 game had roughly 16 more possessions than the avg. game in 2011, and it's easy to see why his usage estimate was so microscopic.
Obviously these are just estimates anyway, so going back to the 1960s & 70s really tests their accuracy. But I don't think they're very far from what we'd estimate if the NBA officially tracked the necessary stats in that era.
December 29th, 2010 at 5:24 pm
Re-reading your comment, I see you're actually saying his 1962 rate is too low. That can also be explained by the fact that Wilt played almost every minute of every game that season, in a pace environment with 30-40 more possessions per 48 minutes than the typical 2010-11 game. Kobe's 2006 usage record came with 27.2 FGA/G, playing 82.5% of LA's minutes, on a team with 80.6 FGA/G. Wilt in 1962 had 39.5 FGA/G, playing 100% of PHW's minutes, on a team with 111.6 FGA/G.