This is from a few days ago, but I didn't see it posted here.
A couple of weeks ago, as I looked to investigate the value of Kobe Bryant, I posted a chart that showed the five NBA players who had the most positive impact on their team's offense when on the court this season, a group topped by Bryant. Conspicuous in his absence was Phoenix Suns guard Steve Nash, who actually rates as surprisingly average by this measure this season, improving the Suns offense by just 4.3 points per 100 possessions.
Still, there is ample statistical evidence to show Nash's brilliance on offense. Again looking at difference in Offensive Rating with and without a player on the court, Nash led the league in 2004-05, with the Suns scoring 17.6 more points per 100 possessions with Nash. Nash has also led from his point-guard position the three greatest offenses since the NBA began tracking turnovers (2003-04 Dallas, 2004-05 Phoenix and 2002-03 Dallas), as measured by differential between the team's Offensive Rating and league average. Here is how I have them, defining possessions as .96*(FGA + (.44*FTA) - OR + TO):
Team Year ORtg League Diff
--------------------------------------
Dallas 2003-04 114.1 104.2 9.9
Phoenix 2004-05 116.6 107.4 9.2
Dallas 2001-02 114.0 105.8 8.2
Denver 1981-82 116.3 108.5 7.8
Chicago 1996-97 115.8 108.0 7.8
Dallas 2002-03 112.7 104.9 7.8
Of course, Nash has had some very talented teammates on those fine offensive squads, including All-Stars Dirk Nowitzki, Shawn Marion and Amaré Stoudemire. The question of the "real MVP" of the Suns has been hotly contested in APBRmetrics circles over the last year.
A year ago, Stoudemire, who finished fourth in the NBA in John Hollinger's PER rating, was the popular pick amongst the statistically-inclined, and it was widely speculated in the stats community that Stoudemire's microfracture knee surgery during last fall's training camp would demonstrate his true value. With the Suns on pace to win 54 games and the Pacific Division without Stoudemire, that argument has fallen by the wayside.
Camps are already beginning to form for this year's MVP debate, with a group of media members, led by ESPN.com's Marc Stein, touting Nash for a second consecutive honor. Meanwhile, many APBRmetricians are pointing to All-Star Shawn Marion as Phoenix's new "real MVP."
The case for Nash rests largely on his ability to make his teammates better. And while this ability is largely regarded as an intangible, the analysis of Bryant's teammates performance demonstrated that it can in fact be a very real statistical effect. As a result, it seems appropriate to subject Nash to the same scrutiny.
What I've done is break down each player who played extensively with and without Nash in Phoenix the last two seasons (minimum 250 minutes with and without last year, 150 this year); we'll look at them player-by-player:
Leandro Barbosa
0506 Rating TS% TORate FTA/FGA
-----------------------------------
with 111.0 .610 9.0 .267
w/o 101.8 .574 11.4 .300
-----------------------------------
Diff +9.2 +.036 -2.4 -.033
Quickly, "Rating" is points scored divided by possessions used times 100, with possessions defined here as FGA + (.44*FTA) + TO, ignoring offensive rebounds. True Shooting Percentage (TS%) is points divided by scoring possessions times two, the best measure of scoring efficiency. Turnover Rate (TORate) is turnovers divided by possessions (multiplied by 100), and FTA/FGA is, naturally, free-throw attempts per field-goal attempt.
Barbosa has been on fire this season with or without Nash, but has been slightly more efficient with fewer ballhandling responsibilities. (If you're wondering why no 2004-05 stats, the reason is that Nash and Barbosa rarely played together -- just 130 minutes all season.)
Raja Bell
0506 Rating TS% TORate FTA/FGA
-----------------------------------
with 104.0 .559 7.0 .163
w/o 98.2 .606 18.9 .123
-----------------------------------
Diff +5.8 -.046 -11.9 +.040
In a bit of a surprise, Bell has actually shot the ball worse when Nash is on the floor this season. He's been strictly a catch-and-shoot player with Nash, rarely turning the ball over, which is why his overall Offensive Rating is still higher with Nash.
Boris Diaw
0506 Rating TS% TORate FTA/FGA
-----------------------------------
with 90.1 .528 14.7 .243
w/o 87.3 .565 22.7 .376
-----------------------------------
Diff +2.8 -.037 -8.0 -.134
Because Diaw has taken such a dramatic step forward this season and is a leading candidate for Most Improved Player, there has been a school of thought that Nash has made him better. Not really so, however; Diaw is taking the ball more to the rack when Nash is on the bench, leading to more free-throw attempts and a better True Shooting Percentage. It has also meant more turnovers, however. The credit for Diaw's breakout season has to go to him, Bryan Colangelo and Mike D'Antoni and his coaching staff, which moved Diaw up front after he floundered as a guard in Atlanta.
Eddie House
0506 Rating TS% TORate FTA/FGA
-----------------------------------
with 104.6 .565 7.5 .059
w/o 92.6 .514 9.9 .058
-----------------------------------
Diff +12.0 +.051 -2.5 +.001
House is another player having the best season of his career after joining the Suns, and in this case Nash definitely deserves a good deal of the credit; House has been inefficient as he was through the first four years of his career when Nash has been on the bench.
Steven Hunter
0405 Rating TS% TORate FTA/FGA
-----------------------------------
with 110.7 .622 11.0 .709
w/o 96.2 .568 15.3 .777
-----------------------------------
Diff +14.5 +.054 -4.3 -.067
Hunter probably owes Nash some money from the not-inconsiderable contract he signed with Philadelphia last summer as a free agent. Hunter has always been a high-percentage shooter and shot 59.2% from the field even when Nash was on the bench last season. But the 63.2% he shot with Nash pushed his shooting percentage even further into the stratosphere. This year, Hunter is down all the way to 51.4% from the field and has languished on the bench lately with the 76ers, who tried to trade him to the Hornets before the deal fell through because of injury concerns.
Jim Jackson
0405 Rating TS% TORate FTA/FGA
-----------------------------------
with 118.7 .672 11.7 .068
w/o 81.7 .501 18.5 .093
-----------------------------------
Diff +36.9 +.171 -6.8 -.025
0506 Rating TS% TORate FTA/FGA
-----------------------------------
with 50.9 .291 12.7 .067
w/o 65.6 .424 22.6 .132
-----------------------------------
Diff -14.7 -.132 -9.9 -.066
Jackson shot the lights out alongside Nash last season in what looks suspiciously like his last NBA hurrah. He hasn't been able to make a shot in any situation this year, and sample size more than anything else probably accounts for the fact that he's done worse with Nash this year.
Casey Jacobsen
0405 Rating TS% TORate FTA/FGA
-----------------------------------
with 86.7 .497 12.8 .358
w/o 107.2 .614 12.8 .413
-----------------------------------
Diff -20.5 -.117 0.0 -.055
Jacobsen was the one Suns regular who did not see his Offensive Rating improve with Nash on the court last season, and I must say I'm genuinely mystified as to why. He wasn't just standing around on the perimeter, as his rate of free-throw attempts testifies, so I don't get it. Jacobsen had a genuinely bizarre statistical season in 2004-05 that continued in New Orleans; he managed to be a perimeter gunner who also got to the free-throw line as regularly as any guard in the league. He's now playing for Tau Ceramica in Spain alongside Peja Drobnjak, San Antonio draft pick Luis Scola, Toronto draft pick Roko Ukic and top prospect Tiago Splitter, amongst others.
Joe Johnson
0405 Rating TS% TORate FTA/FGA
-----------------------------------
with 109.1 .594 8.1 .143
w/o 84.8 .495 14.4 .169
-----------------------------------
Diff +24.4 +.098 -6.8 -.025
Johnson benefited from Nash's presence as much as anyone a season ago, going from 43.0% from the field and 30.5% from downtown to 46.1% and 47.8%, respectively. That's made it all the more impressive that Johnson is playing so well in Atlanta now that he's being forced to create almost all of his own offense. His 3-point percentage has come back down to pre-Nash levels, but his field-goal percentage has improved again.
James Jones
0506 Rating TS% TORate FTA/FGA
-----------------------------------
with 103.0 .539 4.5 .265
w/o 116.7 .616 5.4 .169
-----------------------------------
Diff -13.6 -.077 -0.9 +.096
Jones is another jump-shooting specialist who you'd figure to benefit from playing with Nash, but that hasn't been the case thus far.
Shawn Marion
0405 Rating TS% TORate FTA/FGA
-----------------------------------
with 105.0 .563 6.7 .198
w/o 94.9 .539 12.0 .261
-----------------------------------
Diff +10.1 +.024 -5.3 -.063
0506 Rating TS% TORate FTA/FGA
-----------------------------------
with 104.7 .564 7.1 .206
w/o 112.9 .609 7.3 .226
-----------------------------------
Diff -8.2 -.045 -0.2 -.020
Marion is one of the more interesting players to look at in that the "Nash Effect," so to speak, has reversed itself this year. What I suspect is that there might be a couple of opposing issues at play here. One is that (Marion), moreso than almost any swingman who scores as much as he does, is dependent on being set up; 77% of his shots this season have been assisted on. At the same time, Marion is much more effective when he uses his athleticism than spotting up on the perimeter, which he's often limited to alongside Nash, especially now that most of the Suns' pick-and-rolls with forwards are going through Diaw. It's also been theorized that we could be seeing a little "garbage-time" factor here; Nash doesn't play as many minutes in blowouts as Marion does to protect his back. This one deserves more observation.
Quentin Richardson
0405 Rating TS% TORate FTA/FGA
-----------------------------------
with 100.2 .536 6.4 .162
w/o 87.0 .495 12.1 .208
-----------------------------------
Diff +13.3 +.041 -5.6 -.046
Pretty standard stuff here. Richardson lived on the perimeter last season and feasted on the open looks Nash created. Take Nash away from Richardson and have his back cause more problems and the result is a 31.9% 3-point percentage this season.
Amaré Stoudemire
0405 Rating TS% TORate FTA/FGA
-----------------------------------
with 115.3 .634 9.1 .603
w/o 100.8 .576 12.5 .577
0304 91.9 .536 14.4 .503
-----------------------------------
Diff +14.6 +.058 -3.4 +.026
Exhibit A in the case that Nash makes his teammates better. I've thrown in Stoudemire's 2003-04 numbers in these same categories to demonstrate that he improved independent of Nash, which we would expect from a 22-year-old preps product who made strides within the 2003-04 season. However, the magnitude of Stoudemire's improvement in two-point percentage (from 47.7% to 56.4%) was basically unprecedented. It might not even have been the area where Nash most helped Stoudemire; by putting him in positions where he was able to catch and shoot (dunk, basically), Nash slashed Stoudemire's turnover rate, which went from one of the worst in the league amongst big men to one of the best. The Suns' shooters deserve credit for this too, because opposing defenses weren't able to sag into the paint to help on Stoudemire for fear of giving up a 3.
An 82games study last year showed Stoudemire, along with Marion, was one of the NBA players most dependent on a single teammate (Nash, natch) for an assist. How much credit does Nash deserve for Stoudemire going from promising youngster to MVP candidate? A lot, I would say.
Kurt Thomas
0506 Rating TS% TORate FTA/FGA
-----------------------------------
with 97.3 .548 11.2 .239
w/o 59.2 .354 16.3 .243
-----------------------------------
Diff +38.1 +.194 -5.1 -.004
Thomas has only played 170 minutes with Nash on the bench, so this comparison might not meet the standards for statistical significance. Suffice it to say that Thomas has been pretty bad when he hasn't been paired with Nash, shooting 11-for-37 from the field (29.7%).
That wraps up the Suns regulars the last two seasons, but there's one other player I'd like to take at: Nowitzki. One of the counter-points frequently offered to the Nash makes his teammates better line of thinking is that it is disproved by Nowitzki's career season the year after Nash's departure. Well, take a look at the numbers from 2003-04:
Dirk Nowitzki
0304 Rating TS% TORate FTA/FGA
-----------------------------------
with 105.7 .571 7.5 .318
w/o 96.2 .535 10.2 .338
-----------------------------------
Diff +9.5 +.036 -2.6 -.020
Nowitzki had a relatively weak 2003-04 season, but that was despite Nash helping him out. Now what I might agree with is that, in a way, Nash's presence kept the Mavericks from fully taking advantage of Nowitzki's ability to create his own shot. In 2003-04, Nowitzki was assisted on 72% of his baskets; last year, that dropped all the way to 53%.
"His whole style of play is so much more aggressive," Mavericks lead assistant Del Harris told author David Friedman last spring, explaining Nowitzki's improvement. "Last year he depended on Steve for a lot of pick-and-rolls. This year he doesn’t get his points off of pick-and-roll that much. He gets them through various play actions, posting and driving the ball."
The other issue that is forgotten when discussing Nowitzki's 2004-05 season is that virtually every high scorer in the league was better in 2004-05 than 2003-04 because of the NBA's new interpretation of rules prohibiting handchecking. I recently wrote about this for SI.com, and Nowitzki benefited almost as much as the speedy guards who are the stereotypical beneficiaries. Very few big men have the footspeed to contain Nowitzki on the perimeter, and he went from attempting 5.5 free throws per game in 2003-04 to 9.1 in 2004-05. I strongly suspect this would have happened even had Nash re-signed in Dallas.
If we look at all these different players, some very clear patterns emerge. For the most part, players tend to shoot better when Nash is on the court. The more important and more subtle effect, however, is that players slashed their turnover rate. (No regular improved their turnover rate alongside Nash the last two seasons; Jacobsen's was unchanged.) As I touched on in the discussion of Stoudemire, the reason for this is that Nash delivers the ball to players in position to shoot and without them having to do virtually any work for their shot besides getting open, as opposed to shooting off of a dribble or two.
If there is a common thread amongst Nash's recent teams, besides their ultra-efficient offenses, it is that they rarely if ever turn the ball over. The 2002-03 and 2003-04 Mavericks recorded the two lowest turnover rates since the league began tracking turnovers in 1973-74, and the 2001-02 incarnation ranks fourth. (The 2004-05 Suns are also in the top 25 of all time.)
What is interesting about this is that Nash is not particularly outstanding at avoiding turnovers. He averaged 3.3 turnovers per game last season to rank ninth in the NBA, and in Hollinger's Turnover Rate statistic (different from the one I calculated above because assists are included in the denominator, which helps Nash), he ranked a dismal 43rd amongst point guards. So Nash's teams are so good at avoiding turnovers not because he is good at avoiding turnovers, but because he keeps his teammates from committing them.
The other trend, one that was somewhat surprising to me, is that players tend to get to the free-throw line more often when Nash is on the bench. This makes sense when you think about it; all that creating offense that leads to turnovers also forces defenses to foul, and wing players tend to spend more time on the perimeter alongside Nash, spotting up for 3s.
As far as 2004-05 goes, the case for Nash making his teammates better is relatively unimpeachable. All four Suns starters and the team's top two reserves improved their offensive efficiency with Nash, four of them by at least 10 points per 100 possessions. This isn't, as it is for Bryant, a case of taking below-average teammates and improving them. Instead, Nash was doing this with an All-Star in Marion, a Rookie of the Year in Stoudemire, and promising young players in Johnson and Richardson. Making those guys better is pretty impressive.
How much of that is Barbosa? Conventional wisdom would say a lot. I don't think Barbosa is nearly as bad as he was made out to be last year, and nobody pointed out he actually played pretty well when Nash missed a second series of games last February, averaging 15.3 points as the Suns went 2-1. I think the issue was that the Suns forced him to try to be Nash when Nash was out of the game, and what backup guard in the league could do that?
Last year, the Suns' bench seemed to exist mostly to fill in for the starters; Hunter was Stoudemire-lite in the middle, while Jackson could spread the floor like Johnson and Richardson. Nobody could replace Nash and Marion. This year, Phoenix's second unit has its own identity. If anything, with the 6-8 Diaw often playing center, it's smaller and even more versatile than the Suns existing lineup. Having three players in the lineup with the ability to create for themselves and others (House, Barbosa and Diaw) has kept the open shots coming even when Nash is resting.
Additionally, defenses have put more pressure on Nash to score this season, and he struggled a bit with that during November and December, when he averaged 11.6 assists per 40 minutes -- still outstanding, but not quite his unreal 13.4 per 40 minutes from last season. Nash has struck a better balance over the last month and a half, and as his plus-minus has improved, so too, presumably, has the difference in his teammates' performance with him on the floor.
Nash is again having a great season, but I can't justify putting him in the MVP discussion. To me, the Suns' surprising success this year -- and I'll confess to thinking they were in for a drop-off even had Stoudemire remained healthy -- is much more a testament to the genius and creativity of D'Antoni and Colangelo in adding depth and improving the Suns' defense than it is to Nash's ability to make others better.
That said, given Nash's role in the fine seasons enjoyed in 2004-05 by Marion and particularly Stoudemire, it seems perfectly reasonable to me that Nash got the credit as the most valuable member of this terrific trio.
