Since Stephon Marbury agreed to a buyout with the New York Knicks on Tuesday, the Internet has been abuzz about the possibility--even probability--of Marbury joining the defending champion Boston Celtics. There are numerous ways to take a look at what adding Marbury would mean for the Celtics, but because of the soap opera that has surrounded Marbury the last couple of years, one that has tended to get overlooked is this: What kind of player is Marbury at this point?
It's been more than 13 months since Marbury last played in a regular-season NBA game, so the people with the best chance of answering that question are Marbury's Knicks teammates and the New York coaching staff, who determined that what Marbury could bring to the team failed to outweigh the downside of how his dominant persona would have affected his teammates. The best the rest of us can do is guess at what Marbury has left, but I think the SCHOENE projection system sheds some light on what the Celtics can reasonably expect from Marbury at this point.
Based on Marbury's abbreviated 2007-08 season, here are Marbury's ten most similar players at the same age, per SCHOENE:
Player Sim Year Tm WARP
Dennis Johnson 98.1 1986 bos 2.3
Rory Sparrow 97.9 1989 mia -2.6
Travis Best 97.5 2003 mia -0.6
Randy Smith 97.4 1980 cle 4.3
Bimbo Coles 97.3 1999 gsw -0.4
Cuttino Mobley 97.2 2006 lac -0.5
Antonio Daniels 96.9 2006 was 2.7
Derek Fisher 96.6 2005 gsw 4.3
Anthony Johnson 96.6 2006 ind -1.2
Michael Adams 96.2 1994 was 1.5
Gone are the superstar point guards who once dominated Marbury's comp lists. They've been replaced by a variety of role players. The presence of Dennis Johnson and Derek Fisher, in particular, show that Marbury's skill set can still be valuable to a championship contender--though Marbury will certainly never be confused for Johnson or even Fisher at the defensive end.
The overall perspective is less encouraging. At the same age, several players (most notably Sparrow) fell off a cliff. When a larger group of Marbury's most similar players (57 in all) are used to project his 2008-09 performance, as I did for all players before the season, the results are unimpressive: 13.5 points, 4.9 assists and 3.4 rebounds per 40 minutes, with an individual winning percentage of .433--not far above replacement level. By contrast, Eddie House has averaged 18.9 points per 40 minutes for the Celtics this season with a .475 winning percentage.
Alas, this method might be somewhat unfair to Marbury. Not only had he clearly checked out on the Knicks last season (in terms of predicting his on-court performance for Boston, I suppose this is actually something of a good thing), he was bothered to some extent by bone spurs in his left ankle which eventually led to season-ending surgery. So, to give Marbury the benefit of the doubt, let's also consider an alternate career path. We can use SCHOENE to project what Marbury's 2007-08 stats "should" have been, based on similar players, and then run those numbers through the system again to provide comparable players and a different projection.
Lo and behold, while this naturally predicts a better season than Marbury actually had in 2007-08, the impact on comparable players is relatively minimal, aside from replacing some of the backup point guard types with shooting guards.
Player Sim Year Tm WARP
Dennis Johnson 98.4 1986 bos 2.3
Derek Fisher 98.4 2005 gsw 4.3
Allen Leavell 97.7 1988 hou -1.0
Randy Smith 97.6 1980 cle 4.3
Cuttino Mobley 97.5 2006 lac -0.5
Rory Sparrow 97.4 1989 mia -2.6
Jeff Hornacek 97.2 1994 uta 8.2
Rafer Alston 96.8 2007 hou 3.0
Michael Adams 96.7 1994 was 1.5
Byron Scott 96.6 1992 lal 0.6
All told, this list is somewhat more favorable to Marbury. The downside is that SCHOENE takes a bigger bite out of Marbury's 2008-09 projection this time around. The results aren't a ton better than we got using Marbury's actual 2007-08 stats: 13.8 points and 5.0 assists per 40 minutes, with a .450 winning percentage.
Those projections, and the similar players, leave a lot open to interpretation. There is a reason the other elements of Marbury going to the Celtics have been so heavily discussed. Whether he is motivated to play a limited role while trying to help Boston win another championship is crucial to whether Marbury is worth adding. Still, I think there's something important to be taken from this exercise. Anyone expecting a return to Marbury's All-Star heyday or anything of the sort is dreaming. Even before his off-court behavior turned tabloid fodder, Marbury's game was on a steady decline.
At best, Marbury is an above-average backup point guard who might work well with House in a backcourt where Marbury handles ballhandling duties and defends bigger guards, letting House do what he does best--shoot. At worst...well, the last year-plus in New York has pretty much told that story. There's a chance that Marbury might be done. A year is a long time to go without meaningful basketball competition, and Marbury looked disturbingly gaunt in the public appearances designed to prepare the world for his return to the NBA.
Is adding a player whose upside is limited and whose downside is potentially enormous a worthwhile risk for a team that is already in position to compete for a second straight championship? My answer is an easy "no," but the Celtics appear ready to cast a dissenting vote. We'll know in a few months whether it works.
Kevin Pelton is an author of Basketball Prospectus.
You can contact Kevin by clicking here or click here to see Kevin's other articles.