Posted by Neil Paine on March 5, 2009
Go ahead, admit it. The sight of Stephon Marbury in a #8 Celtics jersey getting picked by Will Bynum at midcourt and missing 3 shots against Detroit Sunday brought back some bittersweet memories of another Employee No. 8, right? No, no, I'm not talking about Scott Wedman; I'm referring to the unforgettable, roller-coaster ride that was the Antoine Walker era in Boston.
A polarizing figure for APBRmetricians everywhere during the early 2000s, Antoine could look like a versatile superstar one second, and a stubborn, selfish player with the worst shot selection in basketball history the next. Here's a great example from 2005 illustrating the difficulty we had in assessing Antoine's value to the Celtics (and, later, the Heat):
David Friedman: Based on past discussions here (actually, in APBR Analysis), Antoine Walker fares pretty poorly in most people's rating systems. Today there is an article about how much of a positive impact Walker has had on the Celtics since his return to Boston. I am interested if Walker's performance this year--or at least in this 6 game stretch--measures up differently than it did in previous seasons. Has Walker's game changed in some measurable way? Or does analysis of his game somehow miss his contributions to team chemistry? Or is six games too small of a sample size?
Justin Kubatko: I would choose (c). In those six games the Celtics have played three of the worst teams in the league: Utah, Charlotte, and Atlanta. Granted they also beat Phoenix in that stretch, but Steve Nash missed that game with a hamstring injury.
Mike Goodman: Sometimes I think I'm the only statistics person who doesn't hate Walker. He's a bit off his career rates this year, but not much. What people seem to hate most is his TS%, a career .479, which is right about where he is this year. [...] I'm not declaring Walker to be any kind of superplayer. His slippage in concentration might be seen in his FT%: from .74 just 3 years ago, to .55 this year. Almost nobody becomes a bad FT shooter at this stage in their careers.
Ben F.: I think what most people in the statistical community dislike about Walker is the abhorrent +/-. I know before leaving Atlanta he was worst in the league among players playing significant minutes with a -17.7 "Roland Rating". That's such a huge number it's hard to ignore. You'd think he was doing something to hurt the team. Then looking at the On/Off Court splits, the team played 5% worse eFG% defense with him on the court than off. [...] But he still confuses me. He's inefficient, turns the ball over, is an average rebounder for a PF, and is a horrible defender. Yet for some reason he seems to contribute to Boston.
Ben: I don't think he's that bad. He's got an above average PER this year too. I do think he wasn't the best fit with Dallas though. He's right, he's a "volume" player. He's better being a big fish in a small pond, he didn't seem to improve his efficiency when his role (usage) was scaled back (though not as much as it should have been).
Bob Chaikin: [...] let's just use some common sense to show that Antoine Walker is indeed one of the worst starting PFs in the league, and has been for quite some time. [...] is it any wonder that simulation shows the suns win 13 more games per average 82 game season with marion playing 40 min/g at PF than with walker? typically the best player at a specific position will win 15-16 more games per 82 games than the worst starting players in the league at that same position (duncan and kirilenko are even better than marion this season, by 2-3 more wins), which means walker is near the very bottom in terms of generating wins for his team as a starter at his PF position...
David Friedman: Bob, your analysis is thorough and logical--yet the Celtics have done well since acquiring Walker. [...]
Kevin Broom: I've long been befuddled by Walker's play. He sometimes make passes or plays that demonstrate to me that he sees the floor well, that he understands the game, and that he's a very smart player. Then two trips later he jacks up a 30-footer with a hand in his face and 22 seconds left on the shot clock. A few years ago, I honestly thought that he had the ability to be among the league leaders in assists and efficiency, if he'd just cut out a few of the silly shots and play with the smarts I thought he had. But that's not the way he's wanted to play. Puzzling.
...And so on and so forth. Get the idea? Walker had truly horrendous efficiency stats throughout his career, that much is undeniable. Here are his career ORtg/%Pos/DRtg numbers, normalized to this year's environment of 1.0855 pts/possession:
As you can see, not once was Walker at an average level of per-possession offensive efficiency. And some years he was well below the average mark of 108.6 -- including 2005 in Boston, when the excerpt above referred to his "positive impact on the Celtics". After arriving in the Hub, his efficiency level was more than 10 points/100 possessions worse than the NBA's average.
Of course, that's an overly simplistic interpretation of Walker's value. He was also above the NBA's average of 20% possession usage while on the floor in every season of his career, and sometimes he pushed his offensive burden into the lofty 27-30% range. How much of that was selfishness and how much was borne of necessity cannot be determined from the table above, but it's clear his extreme possession usage simultaneously drove down his own personal efficiency, and drove up that of his teammates. It's also clear that Walker was not as bad a defender as he was frequently made out to be throughout his career: adjusted to 2008-09, his lifetime defensive rating is 108.6 -- exactly the same as the league's average rating.
This more favorable assessment of Walker's contribution is corroborated by his statistical +/- record:
A career SPM of 0.09 (0.48 if we throw out his last two awful seasons in Miami and Minnesota) speaks fairly well of the guy, considering his rather brutal offensive efficiency marks. But then there's his raw on-/off-court data at 82games (which goes back to 2003):
It's also worth noting that in the '06 playoffs, when Miami won it all, Walker was at -3.7, basically the same level of impact he had during the regular season. At the same time, Hall of Fame coach Pat Riley played 'Toine 37.5 MPG during those playoffs, including the 2nd-most minutes (36.5 MPG) of any Heat player during the Finals, knowing that at least some portion of his legacy rode on the outcome of that series, since many felt he'd unfairly forced Stan Van Gundy into a "resignation" earlier in the season.
How can we reconcile the varying perceptions of Walker? Let's break down what the stats say about each facet of his game, and perhaps come to a clearer understanding of Walker's true ability level.
|Category||Walker||Lg Avg||F Avg|
Walker never really lacked for physical tools -- he had the prototype size & strength, plus better quickness than you'd expect from a 4. The "energy index" (invented by John Hollinger) is simply the cube root of (OR/40)*(BK/40)*(ST/40), and serves as a rough proxy for "athleticism" when compared to other players of the same size/position. Walker was known for being overweight & out of shape late in his career, but he posted good energy indices (1.58, 1.62, & 1.48) in his 1st 3 NBA seasons; it was only in 2000 that you started to see a drop-off in his effort, athleticism, and conditioning, culminating with what we saw in Miami and Minnesota.
(Note: "NBA" is the entire league avg., "Fwd" is just the avg. for forwards)
This is where the classic Walker M.O. starts to appear. Simply put, Antoine was not good at shooting the rock -- sure, he had range, but his jumper was just flat-out inaccurate, as evidenced by his horrendous percentages. You name the category (2-pointers, 3-pointers, free throws, etc.), and Walker was below par; he once claimed he shot so many threes because "they don't have fours," but I'm guessing he wouldn't break average on those, either.
He had no problem creating shots, though, often taking around 30% of his teams' FGAs when in the game, and that alone allowed him to score at an above-average rate in all but two of his NBA campaigns. But even if you had never seen Antoine play, you could look at these numbers and envision the (de-?)evolution of his career... As the years piled on, his FTr, which was never even average to begin with, dropped precipitously and his 3-point tendency rose proportionately. This tells you all you need to know: Antoine started settling for more and more jump shots, which of course is why his percentages were so low -- he was taking the majority of his attempts from areas of the floor where he simply wasn't very good at making shots. Frankly, if he had worked harder on a post-up game instead of forcing so many ill-advised jumpers, he could have been a much more valuable player over the course of his career.
One of Antoine Walker's strengths as a player was his ability to handle the rock and see the floor; the Celtics even used him sometimes as a "point forward" to distribute the ball to teammates. He was also good at getting out in transition and finishing or making passes in the open court, because of his nice combination of size and dribbling skill. The downside to putting the ball in Walker's hands that much was that with too much freedom, 'Toine had a tendency to make bad turnovers and call his own number too often. In other words, Walker's greatest strength (his floor game) could lead to his greatest weakness (shooting/shot selection) if he wasn't kept on a tight leash.
Antoine was frequently called out for a lack of intensity on the glass, but as you can see, he wasn't a particularly bad rebounder, especially early in his career. As the years went on, he stopped working hard on the offensive boards, but his defensive rebound rate was consistently above average until his final 3 seasons with Miami and Minnesota.
Defense was arguably Walker's biggest flaw as a player: he'd fall asleep on the weak side all the time, give up too many easy buckets simply for lack of hustle. He was also undersized in pure post-up situations against true 4s, and didn't give you the shot-blocking you'd want out of a frontcourt defender. But at the same time he was surprisingly decent at stealing the ball early in his career, and his prominent playing time on some tough defensive teams under Jim O'Brien in Boston suggests that he wasn't as horrendous a defender as the popular perception held. I'm not saying anyone would confuse him with Tim Duncan down low, but he was probably no worse than an average defensive player during his prime.
|Nice passing ability & court vision||Laid a lot of bricks from the outside|
|Good ballhandling skills||Struggled at the free throw line|
|Unique, versatile skill set||Amazingly poor shot selection|
|Could create his own shot||Bad habit of drifting to the perimeter|
|Possessed 3-point range||Unable to block many shots|
|Clearly a talented basketball player||Poor conditioning habits|
|Underrated rebounder & defender||His focus would come and go|
|The "shimmy"||The "shimmy"|