Friday, February 24, 2017

The Triple-Double Is Overrated. Kawhi Leonard Is Not.


Russell Westbrook of the Oklahoma City Thunder has received enormous acclaim for averaging a triple-double this season, a feat that has been accomplished only once in NBA history—by the all-time great Oscar Robertson 55 years ago. A New York Times reporter recently gushed that Westbrook’s triple-double run is  “a performance as aberrant as anything any of the primordial basketball legends (Wilt, Oscar, Bill Russell, Pistol Pete) ever did.”
 
But while Westbrook certainly has played great, let’s not get carried away with this specific achievement: in reality, the triple-double is overrated.

Nothing magical happens with the tenth point, rebound or assist. As the ninth is simply one more than the eighth, the tenth is just one more than the ninth. In that regard, Westbrook’s 31.1 points, 10.5 rebounds and 10.1 assists per game (as of the all-star break) barely differ from James Harden’s 29.2 points, 8.3 rebounds and 11.3 assists per game and are not much more than Lebron James’s 25.9 points, 7.7 rebounds and 8.8 assists per game. Indeed, James’s complete statistical line, which includes fewer missed shots and turnovers and a high proportion of assists on three-point shots, may be superior to Westbrook’s and Harden’s.

I would not rank Westbrook over Harden, James or others for reaching three unrelated numbers of no specific value; nor, in turn, should triple-double runner-up Harden rank over James and others for that reason; nor, should Westbrook or Harden necessarily rank above Kawhi Leonard, even though Leonard has never had a triple-double.

Why the fascination with the triple-double? Why the game-by-game Westbrook drumbeat?—did he or didn’t he?—and melodrama (“He passed Bird! Wilt is next!”….“primordial legends”)?

Tobias Moskowitz and L. Jon Wertheim (Scorecasting: The Hidden Influences Behind How Sports are Played and Games Are Won) explain the outsized allure of statistics like the triple-double: “we are slaves to round numbers.” They cite examples outside of sports—this is a widespread phenomenon—but their most pertinent discussion concerns baseball’s .300 batting average. A .300 average is essentially the same as .299, yet the authors found that “In the last quarter century, no player hitting .299 has ever drawn a base on balls in his final plate appearance of the season.” Not once! These hitters expanded their strike zone in order to keep alive the chance of hitting .300.

Furthermore, players hitting .300 on the season’s last day are much more likely to take the day off (to ensure their .300 average) than are players hitting .299. (Memorably, Ted Williams played on the final day of the 1941 season despite entering the game with a .400 batting average. He ended up at .406; his willingness to risk falling short went down in baseball lore as testimony to his greatness.) Not surprisingly, the same holds for batters who enter the final day of the season with 99 or 100 RBIs: the latter take the day off far more often than the former.

Most importantly for present purposes, players are paid more money after they hit .300 rather than .299, or drive in 100 runs rather than 99, or hit 30 home runs rather than 29. General managers over-and undervalue players who just reach or just miss round numbers.

The triple-double is not alone in failing to adequately rank players. All basketball statistics, from basic numbers (points, rebounds, and assists totals) to advanced analytics, have significant limitations.
           
Ideally, any statistical evaluation would eliminate all variables other than the performance of the athlete being measured. Unfortunately, basketball contains a myriad of important variables, including the considerable effects of the nine other players on the court, varied team paces and styles, different responsibilities assigned to players, and more. With that degree of player interaction and overall complexity, basketball simply does not lend itself well to numerical assessments of individual players. 
           
Context is key in evaluating players. For example, Kevin Durant’s exit led to the near-complete reliance of Oklahoma City on Westbrook—an NBA all-time high “usage percentage” —and to Westbrook’s increased scoring (also, turnovers and missed shots). In addition, Westbrook’s career high in rebounds owes in part to Durant and Serge Ibaka being replaced on the front line with poor rebounders; Westbrook filled a void. Westbrook deserves credit for everything he’s done this season—every last point, rebound, and assist—but he is not a better shooter, rebounder, or passer than last season. 
           
Similarly, although Durant is now shooting his highest percentage ever from the field, his shooting skills have not improved. Instead, he’s getting better shots from Golden State’s constant movement offense than he did with Oklahoma City’s isolation-emphasized style. On the other hand, sharing the offensive load with sharpshooters Steph Curry and Klay Thompson, he’s taking the fewest shots per game in his career and his scoring average is down.
           
Variables and context plague measurements of defense even more. Defense is always team-schemed and it’s nearly impossible to isolate a single player’s contribution (even though more cameras, shot charts, and other sophisticated strategies have been brought to bear on the attempt.)
           
Consider the publicly available overall defensive evaluations, which fail in two critical ways. First, they do not distinguish adequately among players. Second, they too often yield plainly ridiculous results.
           
For example, in Basketball-Reference’s defensive ratings (a complex analytics formula that yields the points allowed by a player per 100 possessions) the twelve rotation players on the Cleveland Cavaliers currently fall in the narrow range between 105 and 112. Measures of error are not provided but it may be that the formula finds no statistically significant difference between any of the players. (For comparison, Cavalier offensive ratings span from 97 to 127.) At best, any distinctions here are minimal.

Bizarrely, Kevin Love ranks as the Cavalier’s best defender. NBA cognoscenti (and anyone else who has ever watched a Cavs game) consider Love a poor defender.  His high defensive ranking indicts the methodology, as does Basketball-Reference’s assessment of Klay Thompson as the Warriors’ weakest defender. Worse yet, consider Harden, a notoriously poor defender; numerous internet videos document his self-acknowledged problem with concentration and his comic lapses into fugue-like states at the defensive end. Yet two seasons ago Basketball-Reference ranked him equal defensively to the hyper-energetic, focused, and more athletic Westbrook. Forensic pathologist Henry Lee quipped in the O.J. Simpson trial that if you find one cockroach in a bowl of spaghetti, you toss the entire plate; cockroaches abound in Basketball-Reference’s defensive ratings.
           
And since any composite statistical player evaluation must include defense as a significant component, none of them (i.e. WAR, VORP, and others) inspire any confidence in their value. Statistics alone, whether simple or complex, cannot adequately rank basketball players.
           
Most analysts still consider James the NBA’s best player, but his reign is nearing its end. At 32 years old, he now rests on some defensive possessions and will soon decline. Who is his heir apparent among perimeter players? With apologies to Durant and Curry, who have also hit their peaks and are now difficult to evaluate (playing together on a team that would win at least 60 games without either one), the major candidates for the NBA’s top dog over the next few years are Westbrook, Harden, and Leonard.
           
Westbrook and Harden both sport gaudier basic statistics than Leonard: more points, rebounds, and assists. But the inability of these statistics to establish Westbrook’s and Harden’s superiority becomes apparent when we imagine the three players switching teams. In the more egalitarian and far slower paced San Antonio system, Harden’s and Westbrook’s points, rebounds and assists would all decrease, with their points and rebounds nearing or reaching Leonard’s current numbers. Harden and Westbrook, as point guards, would still have more assists than Leonard, but also more missed shots and turnovers. Leonard shoots more efficiently than Harden and (especially) Westbrook. With these advantages cutting in both directions it would be hard to claim significant overall offensive superiority for any of the three over the others.
           
In contrast, we can make clear distinctions at the defensive end. To start, you don’t need anything more than your eyes to know that the reigning two-time Defensive Player of the Year Leonard dominates the blooper film star Harden, even if Harden has improved somewhat. Down goes Harden.
           
Furthermore, Westbrook is a solid defender but still falls short of Leonard, who is probably the NBA’s best perimeter defender since Scottie Pippen (or even longer). Westbrook cannot match Leonard’s defensive versatility, his ability to guard anyone from point guards to power forwards. Nor can he match Leonard’s extraordinary reach. Announcer and ex-coach Jeff Van Gundy appropriately calls Leonard “by far the best defender of the three.”
           
Thus, we can choose among these players simply by watching Leonard defend. That may seem odd in the modern analytic era, but his obvious defensive superiority trumps any fine distinction (statistical or otherwise) that can be made among the three at the offensive end.
           
Leonard is the least flashy of the superstars. He has no triple-doubles or other eye-catching achievements. There’s no statistical basis to affirm him as the NBA’s best player.  He won’t win MVP this year. Nevertheless, with his superb all-around play, I’d take Leonard over Westbrook and Harden.


Friday, February 10, 2017

Is Trump a Magical Realist?

By Edna Aizenberg

The New York Times got it wrong. On the front page of its Sunday, January 29, 2017 edition, right under mega-headlines denouncing all manner of executive orders, fear-mongering, and unconstitutional edicts by our new president, it made a stab at categorizing the novice chief executive’s disconcerting behavior—Mr. Trump is a “magical realist” the Times declared, who makes “fantastic claims punctuated by his favorite verbal tic—‘Believe me.’”

In the article, written by reporter David Barstow, the editors of the venerable daily were laudably attempting—like so many other Americans—to decipher President Trump; or as we literary critics would put it in our lingo, they were trying to “represent” him, to find systems of meaning through which to understand and label the new resident of the White House.

The classifying phrase “Magical Realism” (MR) struck the Times as appropriate, since the term was taken to signify “fantastical claims,” “cascade of falsehoods,” and “alternate facts”—more simply and directly, lies. Mr. Trump was a magical realist because he was a skillful fibber, perjurer, fabricator, and so on and so on.

As a Latin Americanist specializing in Jorge Luis Borges, one of the initiators of the magical realism now identified with his continent’s literature, my attention was immediately piqued by the front-page combination of the Times, Trump and (un)Truth. Especially, I realized, because the distinguished publication had gotten it wrong. Magical realism is not lying, but the laying bare of lies; not fibbing, but the challenging of fibs; not a cascade of falsehoods, but a form that bears witness to the truth.

The only thing that the Times did get right in connecting Trump to magical realism was the political key, since one of magical realism’s central features is resistance to single-voiced political structures: it indicts, not supports political, cultural and social perversions.

Magical realism as an artistic concept was all the rage when I was a mere Latin American literature graduate student back in the antediluvian 1970’s. Everyone was suddenly gaga over an obscure novel coming out from, of all places, the scarcely-known Colombia, and entitled One Hundred Years of Solitude (Cien años de soledad, 1967) by one Gabriel García Márquez that broke with the realism of the traditional novel, and described the convoluted political and social history of South America, especially Colombia, through a blending of the magical and the mundane.

It was an artistic tour-de-force that earned its author the Nobel Prize for Literature in 1982. But in his Nobel lecture García Márquez warned against the misuse of magical realism, usually by Europeans and Americans, as precisely what it isn’t—an escape from reality and the justification for lying by governments and administrations.

Speaking of Latin America, García Márquez pointedly asked in his Nobel speech, “Why is the originality so readily granted us in literature so mistrustfully denied us in our difficult attempts at social change?” Notice the emphasis on social change. García Márquez, using MR, wants to retell a century of historical events in order to out the lies of official versions by displaying their outrageousness and absurdity.

The central occurrences in his book are the Thousand Days’ War (1899-1902) between Liberals and Conservatives in Colombia and the “banana massacre” of the workers striking against The United Fruit Company (1928). In opposition, García Márquez unearths the facts that are covered up—for example, his purposefully exaggerated portrayal of a heavy rain that falls on his town of Macondo relentlessly for five years to signal the destruction of the physical evidence of the brutal killing.

García Márquez was following in the footsteps of his precursor Borges, who in the 1940’s, while World War II was raging, attacked the genocidal totalitarians—Nazis, Communists, anti-Semites—who were questing global dominance through what he called the anti-human “rigor of chess-masters,” disintegrating our world by destroying peoples and nations as if they were no more than inert wooden pieces on a world-size chess-board, and by wiping out the past with the large scale provision of a fictitious past, outsizing the banana magnates of Macondo (he did this some years before Orwell in 1984.)

His masterful stories, “Tlön, Uqbar, Orbis Tertius” (1940), “El milagro secreto” (“The Secret Miracle,” 1943), and “Deutsches Requiem” (1946) propose topsy-turvy universes—imaginary planets in “Tlön,” stage plays where time moves backwards, and war crime trial proceedings constantly interpolated by the revelation of totalitarian mendacity.

Happily, we are not challenged on the magnitude that García Márquez and Borges were. But perhaps in the US the potential magical realists of our challenging moment are the masses gathered in Washington and elsewhere a week before the Times article for the Women’s March demanding gender equality in the face of Donald Trump’s glaring macho braggadocio.

Could the thousands of pink “pussyhats” they wore, adorned with cat ears, be not only a visual protest against the government, but the seed of a new fantastical strategy that will serve for future political purposes?

Edna Aizenberg is Professor Emeritus of Hispanic Studies, Marymount Manhattan College (New York), and the author, most recently, of On the Edge of the Holocaust: The Shoah in Latin American Literature and Culture