No one gets lucky til luck comes along.
“It’s In The Way That You Use It”
Theme Song for The Color of Money (1986).
The greenish tint to the Olympic pool wasn’t the only thing fishy about the water in Rio last month: a “series of recent reports,” Patrick Redford of Deadspin reported recently, “assert that there was a current in the pool at the Rio Olympics’ Aquatic Stadium that might have skewed the results.” Or—to make the point clear in a way the pool wasn’t—the water in the pool flowed in such a way that it gave the advantage to swimmers starting in certain lanes: as Redford writes, “swimmers in lanes 5 through 8 had a marked advantage over racers in lanes 1 through 4.” According, however, to ESPN’s Michael Wilbon—a noted African-American sportswriter—such results shouldn’t be of concern to people of color: “Advanced analytics,” Wilbon wrote this past May, “and black folks hardly ever mix.” To Wilbon, the rise of statistical analysis poses a threat to African-Americans. But Wilbon is wrong: in reality, the “hidden current” in American life holding back both black Americans and all Americans is not analytics—it’s the suspicions of supposedly “progressive” people like Michael Wilbon.
The thesis of Wilbon’s piece, “Mission Impossible: African-Americans and Analytics”—published on ESPN’s race-themed website, The Undefeated—was that black people have some kind of allergy to statistical analysis: “in ‘BlackWorld,’” Wilbon solemnly intoned, “never is heard an advanced analytical word.” Whereas, in an earlier age, white people like Thomas Jefferson questioned black people’s literacy, nowadays, it seems, it’s ok to question their ability to understand mathematics—a “ridiculous” (according to The Guardian’s Dave Schilling, another black journalist) stereotype that Wilbon attempts to paint as, somehow, politically progressive: Wilbon, that is, excuses his absurd beliefs on the basis that analytics “seems to be a new safe haven for a new ‘Old Boy Network’ of Ivy Leaguers who can hire each other and justify passing on people not given to their analytic philosophies.” Yet, while Wilbon isn’t alone in his distrust of analytics, it’s actually just that “philosophy” that may hold the most promise for political progress—not only for African-Americans, but every American.
Wilbon’s argument, after all, depends on a common thesis heard in the classrooms of American humanities departments: when Wilbon says the “greater the dependence on the numbers, the more challenged people are to tell (or understand) the narrative without them,” he is echoing a common argument deployed every semester in university seminar rooms throughout the United States. Wilbon is, in other words, merely repeating the familiar contention, by now essentially an article of faith within the halls of the humanities, that without a framework—or (as it’s sometimes called), “paradigm”—raw statistics are meaningless: the doctrine sometimes known as “social constructionism.”
That argument is, as nearly everyone who has taken a class in the departments of the humanities in the past several generations knows, that “evidence” only points in a certain direction once certain baseline axioms are assumed. (An argument first put about, by the way, by the physician Galen in the second century AD.) As American literary critic Stanley Fish once rehearsed the argument in the pages of the New York Times, according to its terms investigators “do not survey the world in a manner free of assumptions about what it is like and then, from that (impossible) disinterested position, pick out the set of reasons that will be adequate to its description.” Instead, Fish went on, researchers “begin with the assumption (an act of faith) that the world is an object capable of being described … and they then develop procedures … that yield results, and they call those results reasons for concluding this or that.” According to both Wilbon and Fish, in other words, the answers people find depends not the structure of reality itself, but instead on the baseline assumptions the researcher begins with: what matters is not the raw numbers, but the contexts within which the numbers are interpreted.
What’s important, Wilbon is saying, is the “narrative,” not the numbers: “Imagine,” Wilbon says, “something as pedestrian as home runs and runs batted in adequately explaining [Babe] Ruth’s overall impact” on the sport of baseball. Wilbon’s point is that a knowledge of Ruth’s statistics won’t tell you about the hot dogs the great baseball player ate during games, or the famous “called shot” during the 1932 World Series—what he is arguing is that statistics only point toward reality: they aren’t reality itself. Numbers, by themselves, don’t say anything about reality; they are only a tool with which to access reality, and by no means the only tool available: in one of Wilbon’s examples Stef Curry, the great guard for the NBA’s Golden State Warriors, knew he shot better from the corners—an intuition that later statistical analysis bore out. Wilbon’s point is that both Curry’s intuition and statistical analysis told the same story, implying that there’s no fundamental reason to favor one road to truth over the other.
In a sense, to be sure, Wilbon is right: statistical analysis is merely a tool for getting at reality, not reality itself, and certainly other tools are available. Yet, it’s also true that, as statistician and science fiction author Michael F. Flynn has pointed out, astronomy—now accounted one of the “hardest” of physical sciences, because it deals with obviously real physical objects in space—was once not an observational science, but instead a mathematical one: in ancient times, Chinese astronomers were called “calendar-makers,” and a European astronomer was called a mathematicus. As Flynn says, “astronomy was not about making physical discoveries about physical bodies in the sky”—it was instead “a specialized branch of mathematics for making predictions about sky events.” Without telescopes, in other words, astronomers did not know what, exactly, say, the planet Mars was: all they could do was make predictions, based on mathematical analysis, about what part of the sky it might appear in next—predictions that, over the centuries, became perhaps-startlingly accurate. But as a proto-Wilbon might have said in (for instance) the year 1500, such astronomers had no more direct knowledge of what Mars is than a kindergartner has of the workings of the Federal Reserve.
In the same fashion, Wilbon might point out about the swimming events in Rio, there is no direct evidence of a current in the Olympic pool: the researchers who assert that there was such a current base their arguments on statistical evidence of the races, not examination of the conditions of the pool. Yet the evidence for the existence of a current is pretty persuasive: as the Wall Street Journal reported, fifteen of the sixteen swimmers, both men and women, who swam in the 50-meter freestyle event finals—the one event most susceptible to the influence of a current, because swimmers only swim one length of the pool in a single direction—swam in lanes 4 through 8, and swimmers who swam in outside lanes in early heats and inside lanes in later heats actually got slower. (A phenomena virtually unheard of in top level events like the Olympics.) Barry Revzin, of the website Swim Swam, found that a given Olympic swimmer picked up “a 0.2 percent advantage for each lane … closer to [lane] 8,” Deadspin’s Redford reported, and while that could easily seem “inconsequentially small,” Redford remarked, “it’s worth pointing out that the winner in the women’s 50 meter freestyle only beat the sixth-place finisher by 0.12 seconds.” It’s a very small advantage, in other words, which is to say that it’s very difficult to detect—except by means of the very same statistical analysis distrusted by Wilbon. But although it is a seemingly-small advantage, it is enough to determine the winner of the gold medal. Wilbon in other words is quite right to say that statistical evidence is not a direct transcript of reality—he’s wrong, however, if he is arguing that statistical analysis ought to be ignored.
To be fair, Wilbon is not arguing exactly that: “an entire group of people,” he says, “can’t simply refuse to participate in something as important as this new phenomenon.” Yet Wilbon is worried about the growth of statistical analysis because he views it as a possible means for excluding black people. If, as Wilbon writes, it’s “the emotional appeal,” rather than the “intellect[ual]” appeal, that “resonates with black people”—a statement that, if it were written by a white journalist, would immediately cause a protest—then Wilbon worries that, in a sports future run “by white, analytics-driven executives,” black people will be even further on the outside looking in than they already are. (And that’s pretty far outside: as Wilbon notes, “Nate McMillan, an old-school, pre-analytics player/coach, who was handpicked by old-school, pre-analytics player/coach Larry Bird in Indiana, is the only black coach hired this offseason.”) Wilbon’s implied stance, in other words—implied because he nowhere explicitly says so—is that since statistical evidence cannot be taken at face value, but only through screens and filters that owe more to culture than to the nature of reality itself, therefore the promise (and premise) of statistical analysis could be seen as a kind of ruse designed to perpetuate white dominance at the highest levels of the sport.
Yet there are at least two objections to make about Wilbon’s argument: the first being the empirical observation that in U.S. Supreme Court cases like McCleskey v. Kemp for instance (in which the petitioner argued that, according to statistical analysis, murderers of white people in Georgia were far more likely to receive the death penalty than murderers of black people), or Teamsters v. United States, (in which—according to Encyclopedia.com—the Court ruled, on the basis of statistical evidence, that the Teamsters union had “engaged in a systemwide practice of minority discrimination”), statistical analysis has been advanced to demonstrate the reality of racial bias. (A demonstration against which, by the way, time and again conservatives have countered with arguments against the reality of statistical analysis that essentially mirror Wilbon’s.) To think then that statistical analysis could be inherently biased against black people, as Wilbon appears to imply, is empirically nonsense: it’s arguable, in fact, that statistical analysis of the sort pioneered by people like sociologist Gunnar Myrdal has done at least as much, if not more, as (say) classes on African-American literature to combat racial discrimination.
The more serious issue, however, is a logical objection: Wilbon’s two assertions are in conflict with each other. To reach his conclusions, Wilbon ignores (like others who make similar arguments) the implications of his own reasoning: statistics ought to be ignored, he says, because only “narrative” can grant meaning to otherwise meaningless numbers—but, if it is so that numbers themselves cannot “mean” without a framework to grant them meaning, then they cannot pose the threat that Wilbon says they might. In other words, if Wilbon is right that statistical analysis is biased against black people, then it means that numbers do have meaning in themselves, while conversely if numbers can only be interpreted within a framework, then they cannot be inherently biased against black people. By Wilbon’s own account, in other words, nothing about statistical analysis implies that such analysis can only be pursued by white people, nor could the numbers themselves demand only a single (oppressive) use—because if that were so, then numbers would be capable of providing their own interpretive framework. Wilbon cannot logically advance both propositions simultaneously.
That doesn’t mean, however, that Wilbon’s argument—the argument, it ought to be noted, of many who think of themselves as politically “progressive”—is not having an effect: it’s possible, I think, that the relative success of this argument is precisely what is causing Americans to ignore a “hidden current” in American life. That current is could be described by an “analytical” observation made by professors Sven Steinmo and Jon Watts some two decades ago: “No other democratic system in the world requires support of 60% of legislators to pass government policy”—an observation that, in turn, may be linked to the observable reality that, as political scientists Frances E. Lee and Bruce Oppenheimer have noted, “less populous states consistently receive more federal funding than states with more people.” Understanding the impact of these two observations, and their effects on each other would, I suspect, throw a great deal of light on the reality of American lives, white and black—yet it’s precisely the sort of reflection that the “social construction” dogma advanced by Wilbon and company appears specifically designed to avoid. While to many, even now, the arguments for “social construction” and such might appear utterly liberatory, it’s possible to tell a tale in which it is just such doctrines that are the tools of oppression today.
Such an account would be, however—I suppose Michael Wilbon or Stanley Fish might tell us—simply a story about the one that got away.