Photo: AP Images
As far as our two eyes can see, Carmelo Anthony is an excellent basketball player. The 6-foot-9 forward instantly stands out for his ability to put the ball in the basket.”LeBron is a better all-around player,” NBA trainer and scout David Thorpe recently said. “But Anthony’s the better bucket getter.”
Yet by most advanced statistical measures Anthony is little better than average. Casual fans might gawk at his gaudy scoring totals, but efficiency enthusiasts are unimpressed by his 25 points because they come on too many shots. Advanced statistical measures from Wins Produced to PER show Anthony to be an inefficient scorer and therefore an overrated commodity. He is a superstar name, but statistical evidence is growing that Anthony lacks a superstar game.
This weekend a contrarian analysis that favours ‘Melo arose when Nate Silver found that despite Anthony’s own inefficiency, his teammates actually become more efficient when they’re on the court with him. Those numbers validated what logic implies: defenses are so aware of Anthony, that all five opposing players fail to focus on their teammates.
It’s a sound analysis, but one that has its fair share of flaws and spurned another debate over Anthony’s value. Over at TrueHoop, Henry Abbott detailed all of the ways – both on the court, and off of it – by which Anthony’s value need be measured. It’s a layered debate.
But if there’s one thing that’s become perfectly clear, it’s that statistics have a long way to go in basketball.
What makes advanced analysis so much more widely accepted in baseball, is that each game can be deconstructed into finite, independent units – innings, at bats, and at the most basic level, pitches – whose outcome is decided almost entirely by the pitcher throwing the pitch and the hitter to whom the pitch is thrown. Each pitch contains two essential outcomes, either an out, or not an out.
Basketball, on the other hand, is far more fluid. Each possession breaks down into infinite permutations, and each shot isn’t a two-pronged, shot–no-shot decision, but rather a decision not to pass to any one of four teammates, not to hold the ball in hopes of a better passing opportunity presenting itself, and not to put the ball on the floor to try to get a better shot attempt. We can’t chart each of those actions or in-actions, as we can each pitch in a major league baseball game. With so many lurking variables it’s vastly more difficult to correlate individual performance with winning, and therefore impossible to put any weight into over-arching statistics, as many baseball fans do with WAR (wins above replacement-level player).
David Berri, the statistician who invented the Wins Produced statistic that attempts to quantify how many wins each player is responsible for, highly overrates rebounders. That’s why David Lee (17.4 wins produced) was nearly twice as good as Kobe Bryant (9.9 wins produced) in 2010 by Berri’s calculations. Yes, Berri correctly notes that advanced statistics work to debunk conventional wisdom – so just because results are unexpected, doesn’t mean they’re inherently wrong – but there’s little reason to abandon conventional wisdom when the alternative suggests a team of Ben Wallaces would rank among the best of all time.
The more accepted advanced statistic, John Hollinger’s PER, fails by overvaluing scoring. A player’s PER increases as he takes more shots, provided he converts them above historically-bad rates.
The flaw in all of these statistics is that none can quantify roles. They assume all basketball players are on the court trying to achieve the same goal. But that’s not the case. Only one player at a time is trying to score, while only one or two is trying to set a pick, grab a rebound, or stretch a defence by setting up for a jump shot. Anthony is relied upon to bail out his team on offence, so he shouldn’t be penalised for shooting too much, just as rebounding isn’t Kobe’s major responsibility so he shouldn’t be penalised for a deficiency in that area.
And at the end of the day, the rarest of those specialised skills is the ability to score at will. Many players can knock down an open shot, set a strong screen, learn proper off-the-ball rotations in a coach’s offence, and box out an opponent, but few can take the ball against an opposing team’s best defender and put it in the basket. Anthony does this better than anyone else.
So while the New York-area media tries to determine whether it’s worth trading for ‘Melo, they’re too busy looking at field goal percentages and efficiency ratings, but not spending enough time considering how he would actually fit on the Nets and Knicks rosters. Do they have the players that thrive without the ball, to warrant acquiring a high-usage talent? Do they have players that can defend where ‘Melo can’t? Will they make the most of Anthony’s rebounding prowess?
That’s not to say statistics can’t provide an accurate picture of a player’s performance or help us determine a players’ strengths and weaknesses. But unlike baseball, calculating a conclusive statistic is futile. A players’ decisions on the court are a function of his teammates, and there’s no way to control for a players’ performance against his teammates’ strengths and weaknesses.
Melo’s scoring wouldn’t play well on a team like the Mavericks, who already possess a go-to scorer, multiple second options, and are strong on the backboards. On the other hand, he might be the missing piece for the Sixers, who need a go-to option in crunch time. His basic statistics, efficiency ratings, and wins produced might not change on either team, but his marginal utility certainly would.
If we were going purely by advanced statistics, this discussion would be moot. The Nets and Knicks would be vying to mortgage their future for Nene (21.24 PER) rather than ‘Melo (19.96 PER).
Business Insider Emails & Alerts
Site highlights each day to your inbox.