aIvin adams wrote:i haven't ever spent much time looking at ESPN's Real Plus-Minus stat and i don't know anything about how it's calculated.
but by that metric, Kevin Love was the 6th best defensive PF in the league this year (third best starter). Amare Stoudemire was 9th. David Lee and Larry NAnce Jr were both better defensive PFs than Chris Bosh.
I dunno. Any stat will have its quirks that don't match the eye-test or don't square with common sense... but those are pretty strange.
i will acknowledge that Mirza might be ranked incorrectly. he should be last instead of second-to-last.
Real Plus-Minus is a pretty strong attempt to take plus-minus figures and dramatically reduce the "environmental static" that comes with raw plus-minus data. ESPN did not develop the formula, as the basic metric used to exist as Adjusted Plus-Minus going back several years.
http://www.82games.com/ilardi1.htmhttp://www.82games.com/ilardi2.htmAs I have noted before, it is not perfect—context still needs to be considered, and not all of the static can be eliminated. But in general, the results make sense. For instance, in those '07-'08 rankings, Steve Nash constituted the most valuable offensive player in the league, which makes a fair amount of sense, especially when considering the context. He was arguably the best shooter and passer in the NBA, and he played in a system that catered to and enhanced his strengths, with some outstanding finishers and shooters around him. Meanwhile, the system and roster magnified his value because his teammates were generally not adept at creating efficient offense on their own, and the team lacked a quality playmaking point guard behind him. For all those reasons, Nash was immensely valuable to the Suns' offense, perhaps more valuable than any other player in the league to his offense. (Of course, all offenses are not created equal, and if Nash had been playing in another offense with a different set of teammates, he might not have ranked number-one. After all, he left Dallas and the Mavericks did not miss him all that much).
However, Nash was a bad defender according to the metric, thus diminishing his overall value, although still leaving him as a top-ten overall player (yet not quite the top point guard, as he dropped behind Baron Davis, of all people, for while Nash trounced Davis offensively, Davis trounced Nash by an even greater margin defensively).
Again, all of these specific rankings can be argued with, but the general concepts make sense. There is some margin for error, as acknowledged in those tables, rather like in polling data. However, an overall top three of Kevin Garnett, LeBron James, and Tim Duncan made sense, as did the Celtics (who won a league-best 66 games and ultimately the championship) and Spurs (the defending champions, who reached the Western Conference Finals) both featuring two players in the overall top six. Kobe Bryant, the NBA's MVP recipient that year, made the overall top five. The top five offensive players were Nash, James, Chris Paul, Bryant, and Paul Pierce. The top ten defenders were Garnett, Kenyon Martin, Tyson Chandler, Yao Ming, Ron Artest, Tim Duncan, Ben Wallace, Shawn Marion, Marcus Camby, and Josh Smith, all of which made sense. The fact that the Suns' defense became much more of a sieve after trading Marion to Miami in the middle of that season hence made sense.
I always try to analyze matters visually and technically, and as I noted when I first cited the Real Plus-Minus figures in this thread, I wrote my initial assessment before looking up the data; the data just happened to completely confirm what I had written.
However, the concern with the "eye test" is that it is inherently subjective. Every "eye test" is potentially different, as no one person possesses a monopoly on truth. Many times, people are "watching" without "seeing," or they do not know or understand what they are viewing, with the end result being that they just regurgitate reputation and hype while passing it off as the "eye test." Even for the more scrupulous, knowledgeable viewer, completely shaking one's biases (not to mention prejudices) is very difficult. Take a baseball game, for instance. I often try to instantly evaluate the caliber of the contact made off the bat (visually and audibly) to see how well the hitter hit the ball, regardless of the actual result (which, in small samples, will often be influenced by randomness and luck). The better contact that he made (hard contact, line drives, potentially aiming up the middle, and so forth), the better of a job that he generally did and the better his long-term results will generally be. Inversely, a pitcher who induces weaker contact will, on average or over the long term, frequently fair better than one who yields higher-quality contact.
But eliminating one's biases is difficult. Often, one will misjudge the caliber of the contact made (forget results because of the influence of randomness) based on a host of biases: what that hitter did in his previous at-bat; what the previous hitter did; how the pitcher has been pitching; what kind of a pitch he made and what 'should' occur based on that kind of pitch or location; what one wants to happen (rooting interests); what one expects to happen; the type of swing taken by the hitter; something that the broadcaster has just said; and so forth. Any or all of these factors can cause one to misjudge the caliber of contact made—to instantly believe that something just occurred when something else actually just occurred. At least if one is scrupulous, one can recognize one's mistake almost as soon as it happens, but most people do not watch games that way. Reaching, and maintaining, a sense of blank objectivity is not impossible, but slipping from it is very easy.
For similar reasons, eyewitness accounts in criminal cases are notoriously—notoriously—unreliable. For similar reasons, racial profiling is very difficult to entirely eradicate, especially without proper professional training. And in basketball, too, the brain's biases can deceive or influence the eyes.
Now, the more that one visually describes and technically breaks down what is occurring, the better an "eye test" might be. From what I have read, many of the people on this board are probably pretty good at it, but we likely do not represent the average fan or talk show host who claims that he "watches the games" yet has little idea of what he is talking about and actually just regurgitates media soundbites, presumptive reputations, and promotional hype without even realizing what he is doing.
At the end of the day, though, the careful use of sound empirical measures should complement and confirm, rather than contradict, knowledgeable visual assessments. Again, the problem is that most "visual assessments" from fans and media members are pretty superficial and reflect notoriety more than technical knowledge and wisdom.
But again, Real Plus-Minus is not perfect. The specific rankings that you cite may represent "statistical error." On the other hand, there may be other possibilities at play. Kevin Love's defense may be underrated, and it may have improved in Cleveland given that he is playing with LeBron James (someone who sets a certain tone) and for a championship. Additionally, playing in Cleveland has reduced Love's offensive burden and surely granted him greater energy for the defensive end of the court. Before Charles Barkley arrived in Phoenix, Kevin Johnson constituted a good defender—sometimes very good and capable of great defensive plays—but Barkley's presence reduced Johnson's offensive load. The result was that K.J. redirected his energy and transformed himself into arguably the best defensive point guard in the NBA in '92-'93, playing the position defensively about as well as anyone has ever played it. His defense proved so great that year that the Suns eventually ended up sticking K.J. on Michael Jordan for much of the NBA Finals, even though K.J. gave up about four inches and Jordan seemingly enjoyed a walking mismatch. See the 1:09-5:59 section of this video:
[youtube]https://www.youtube.com/watch?v=udxsmKTaDUQ[/youtube]
Or perhaps Love really fits the Cavaliers' defensive scheme. Again, Real Plus-Minus attempts to isolate the value of individual players and remove the contingencies of other players, but it cannot completely liquidate context.
Larry Nance Jr. may well have been a better defensive power forward than Chris Bosh this past season. Like his father and namesake, Nance is a rangy leaper with wingspan and instincts for the ball.
As for David Lee and Amar'e Stoudemire, a few points need to be considered. First, Lee and Stoudemire played many, and perhaps the majority, of their minutes this season matched up against centers, yet ESPN is ranking them with other power forwards. As "centers," Lee and Stoudemire offer mobility, and since few centers (or players generally) post up any longer in the NBA, their potential deficits in that regard are minimized.
Additionally, Lee and Stoudemire did not play major minutes, and ESPN did not sort its rankings based on minutes thresholds. Guys who play fewer minutes should perhaps be grouped together, because they are of less significance, and playing with high energy defensively is much easier in short bursts.
More significantly, since Lee and Stoudemire are aging players whose offensive talent has diminished and who now need to fight for every minute of playing time, they quite possibly have improved their defensive commitment. In his last full season as Phoenix's head coach, '91-'92, Cotton Fitzsimmons saw the offensive slippage and decline in speed of a thirty-two-year old Tom Chambers, and he implored Chambers to compensate by improving his attention to defense. Chambers did not take Fitzsimmons' criticism well (years later, he admitted that he should have absorbed what Fitzsimmons was telling him), and late in the regular season, the coach moved Chambers to the bench in favor of a twenty-two-year old, second-year Cedric Ceballos. Then, in Game Two of the 1992 Western Conference First Round versus San Antonio, Fitzsimmons benched Chambers entirely.
http://www.basketball-reference.com/boxscores/199204260PHO.htmlBut while Chambers did not respond effectively on the defensive end, perhaps Lee and Stoudemire—similar players to Chambers in many ways—did heed the message, especially since their playing time had already plummeted prior to this season.
Or, alternatively, we may be looking at statistical error. There are different potential explanations for those rankings, but they at least encourage us to not rely on assumptions, reputation, and precedent, for something may well have changed or some old beliefs may have been mistaken. We must therefore question, explore, and watch more closely.
Anyway, here is an explanation of Real Plus-Minus from 2014:
http://espn.go.com/nba/story/_/id/10740818/introducing-real-plus-minusThe metric may have undergone some refinements since the earlier Adjusted Plus-Minus, but the basic concepts and methodology seem to be the same.
