"Humans tend to filter the world to confirm what we already believe." —Jonah Lehrer.
That quote is from Jonah Lehrer's appearance on the Colbert Report in May 2009 when he was promoting How We Decide, a well-regarded book on human decision-making as informed by his background in neuroscience. It's an important point and a driving force behind the statistical analysis movement in sports, which latter's goal is to eliminate subjective biases, confirmation and otherwise. More recently, Lehrer has penned a sloppy defense of intangibles in sports at Grantland, an article to which I won't provide a link because I'm quite sure its primary intent is to piss off stat nerds, who will happily direct their ire — and, more importantly, their traffic — to Bill Simmons's sports-meets-pop-culture offshoot of ESPN.com. If you insist on reading it, you can Google "jonah lehrer being an asshole" (no quotes) and it'll be the first search result. You can also read the cogent responses by our own Matthew Callan earlier today, Bill Petti at Beyond The Boxscore, and Colin Wyers at Baseball Prospectus; you'll find a link to the original Lehrer piece contained therein.
I could spend hours fisking Lehrer's article from top-to-bottom, but I have neither the time nor the inclination to do so. Instead, I'll just list a bunch of things that irked me about the column.
The car metaphor. Lehrer spends the first three paragraphs trying to force sabermetrics into a really bad car metaphor. Something about how car-buyers focus on things they can quantify like horsepower and gas mileage while ignoring things like comfort and structural integrity (?) that actually make a difference to the overall enjoyment of car ownership. This is supposed to be an analogy to the way statheads (I'm going to be using "statheads" to mean "sabermetricians" or "people interested in statistical analysis" or "mom's-basement-dwellers" or whatever you want to call them/us) fixate on "statistics" while ignoring "the inherent mystery of athletic talent," which latter is just a fancy way of saying "grission".
Straw men abound. Lehrer suggests that statheads are "pretending that the numbers explain everything." The problem is that nobody really thinks that. Not Billy Beane, not Sandy Alderson, not Theo Epstein, not Paul DePodesta, not Tom Tango, Rob Neyer, Bill James, nor any other credible stathead. To whatever extent the effects of intangible qualities on on-field performance can be measured — and therefore become tangible — statheads do their best to measure them. Things which are as-yet-unmeasurable aren't simply swept under the rug. It's frustrating that we can't reasonably measure things like leadership, chemistry, streakiness, and grission, but nobody is pretending they don't exist.
He also asserts that "[t]he goal of these new equations is to parse the complexity of people playing together, finding ways to measure quarterbacks while disregarding the quality of their offensive line, or assessing a point guard while discounting the poor shooting of his teammates" (emphasis mine). Lehrer has this backwards, actually. I'm not really up on advanced football statistics, but the problems with traditional football stats like quarterback wins, touchdowns, rushing yards, and so forth is that they have always disregarded the quality (and qualities) of the team — and other circumstances outside the player's control — while highlighting the individual accomplishments of the player. The same is true of traditional baseball statistics like pitcher wins and runs batted in. The point of sabermetrics (or whatever you'd call it in football) is not to discount outside influences, but rather to normalize or control for them. In other words, how would Player X perform given an average team and average circumstances.
Making shit up. Lehrer says, "Because it translates sports into a list of statistics, the tool can also lead coaches and executives to neglect those variables that can't be quantified." No coach or executive does this. I don't even know of a journalist or blogger who does this, so to presume that someone in charge of decision-making on a professional sports team does this is either silly, intellectually dishonest, or both.
Reductio ad absurdum. Lehrer relays an analogy made by author Philip Roth in a May 2000 issue of The New Yorker in which Roth, responding to misinterpretations of his 1997 novel American Pastoral, posited a young boy at a ballgame whose father insisted that he (the boy) watch the scoreboard instead of the field. Asked afterward how the game was, the boy replied, "It was great! The scoreboard changed thirty-two times and Daddy said last game it changed only fourteen times and the home team last time changed more times than the other team. It was really great! We had hot dogs and we stood up at one point to stretch and we went home."
Lehrer caustically suggests that "[i]f that little kid were around today, he'd be obsessed with sabermetrics." Statheads like stats, therefore they prefer the objective side of the game to the subjective side, therefore they'd rather monitor the scoreboard than take in the glorious action on the field. This characterization of statheads as fans of numbers first and the game second (if at all) is wrong-headed, idiotic, and insulting.
J.J. Barea. I don't watch basketball and I have no idea who J.J. Barea, but based on Lehrer's description and lionization I can only assume he's the NBA equivalent of David Eckstein. Someone please confirm.
More straw men. Lehrer continues: "[C]oaches and fans use the numbers as an excuse to ignore everything else, which is why our obsession with sabermetrics can lead to such shortsighted personnel decisions." Coaches don't do this. Executives definitely don't do this. Some fans might do this, but no credible, peer-reviewed stathead does this. Sorry, it just doesn't happen. Do we have difficulty quantifying intangibles? Absolutely. Do we think they probably aren't as important as traditional sportswriters, baseball insiders, and sports-radio callers make them out to be? Yes, probably. Do we think they have no value and pretend the don't exist? No chance.
Statheads aren't robots. "For reasons that remain mysterious, some teammates make each other much better and some backup point guards really piss off Ron Artest." This is really just a placeholder for the idea that, no matter how much we know about how and why things happen in baseball (and other sports), sometimes odds are defied, the best team doesn't win, and nobodies (read "Scott Brosius") become World Series MVPs. I love that about sports. We all love that about sports. How dare you tell me that I don't appreciate these things?
Putting a bow on it. "This is the paradox of sports statistics: What the math ends up teaching us that is that [sic] sports are not a math problem." When you begin with a false premise — that anyone truly believes that sports are about statistics and that statistics tell us all we need to know about sports — what you get is this sort of flaccid, misinformed piffle masquerading as pseudo-intellectual journalism.
Jonah Lehrer seems like a pretty bright guy, and his writing on how the human brain processes decisions sounds like the sort of thing I'd enjoy reading. In other words, I think he's probably better than this Grantland article would lead you to believe. Whether he truly believes the puzzling assertions he makes in his column or he's just trolling statheads for pageviews is anybody's guess, though I'm not sure one is preferable to the other.