Now is a very good time to be a seamhead. Ever since the publication of Moneyball, even the most casual fan has known that there's truth in them there advanced baseball statistics. Billy Beane's teams, and then Theo Epstein's Red Sox, backed up theory with results. By this point, the movement has transformed itself from an esoteric experiment to a haven for clear, sane thinking about baseball. Going after sacred cows, debunking myths, calling bullshit on the folkways of the game with the truth on their side, sabermetricians have earned the right to drop all pretenses and laugh at the losers. Witness the blog Fire Joe Morgan, where the message of baseball's new, rational path is delivered via jabs and jeers at the ignorant media. You don't need a weatherman to know which way the wind blows, but knowing does grant you a lifetime right to sarcasm.
"The stats vs. scouts battle in basketball circles is a circus of the absurd."
But the related battle in basketball circles is a circus of the absurd, a never-ending hostage crisis where small-school economist Dave Berri fights it out with international detective Chad Ford. Berri, whose laughable The Wages of Wins posits an ideal style of play in which teams never shoot but always rebound, is a joke to most. Berri gleefully decrees that Dennis Rodman had a more productive season than Michael Jordan in 1995-96, and that Allen Iverson was no better than the league's 91st best player the year Philly went to the Finals. Meanwhile, Ford spread the gospel of shadowy Euros like Darko Milicic, Nikoloz Tskitishvili, and Pavel Podkolzin on the strength of some very patchy workouts and observations. He owes the people of Detroit money. He also stuck a guy (John Riek) in his preliminary Top 10 for the 2008 draft based solely on Riek's 7'2" frame. The average fan can readily mock either camp, and no one for a second believes thatas is the case with baseballthere's any sort of authority hanging in the balance. The war goes on, but history stands still.
It's no mystery why quantitative basketball analysis has lagged. Baseball is at its core a series of one-on-one battles with a defined set of possible results. The action in basketball is a series of complex interactions, where interdependency rules, and nothing happens in a vacuum. Intangibles are more than just an old scout's tale; at the same time, the number of quantifiable events that aren't officially tallied is staggering, including potential assists on missed shots, or shots defended, or charges taken. Baseball sabermetrics caught on because of baseball's almost perfect data to test theories and develop models; basketball fans are confronted daily by the insufficiency of the box score. Just as baseball lends itself naturally to radio broadcast, and basketball is failed utterly by anything but television, stats in basketballwhere the "how" is as important as the "what"are understood as painting an incomplete picture.
Thus, when it comes to the powers of insight and prediction in basketball, no one's ever going to be able to tell the whole story. For while that sport's relationship with data is tricky, sometimes arbitrary, there's nothing more notoriously fallible than the human gut. Numbers may not lurk in the very soul of basketball as they do baseball, but wherever there's production, accumulation, and outcomes, you can't not give quantitative analysis a try. The problem, though, is that in basketball stats can often present just as much "noise" as the loud-mouthed old codgers who repeat, ad infinitum, that "you build around size" and "slow the game down to protect your lead." A scorer who tallies 20 points per gameone of the sport's yardsticks for individual excellencecould easily be the result of a gunner going for self on a fast, bad squad. The same can be said of a blocked shot or a steal, which can represent nothing more than one's single-minded devotion to blocking shots or stealing the ball at the expense of all other defensive matters. Basic stats in baseball can be imprecise, but rarely, if ever, are they reliably misleading and illusory.
" 'Good work' doesn't sell to Joe ESPN.com Reader. You need to be bold and say something to get paid."To advance the state of this game, the stathead and the scoutman should be friends. Watching games matters. Better numbers matter. It's just not practical to sever the game from, well, the game. At the same time, close observation remains vulnerable to all sorts of human prejudice. What we need, then, is both: Stats guys who try to account for the game's eternal fluxits slippery, systemic fluidityand who accept the futility of absolute power. Conversely, scouts need to accept that just because traditional stats can be laughable, there's no reason why scouts can't translate their observations on the ground into some kind of helpful metrics.
There's plenty of good work out of the last half-decade, especially on matters of efficiency, pace, and the vagaries of the rebound. But "good work" doesn't sell to Joe ESPN.com Reader. You need to be bold and say something to get paid. ESPN's John Hollinger has some great theories and analysis, but he gets put into dangerous situations. For instance, the draft: Hollinger's old Pro Basketball Prospectus books barely dealt with college translations, because it's extremely difficult to do given spotty college (never mind high-school) statistics and wildly divergent roles and opponents. How do you manipulate a spreadsheet to figure how LeBron's 30/10/5 at St. Vincent matches with Carmelo's 22/10 in the Big East? But at ESPN, Hollinger can't just skip draft season. He has to apply his assigned angle. In the end we get questionable analysisHollinger's system seems devoid of backing theory, and it must be reinvented (or "tweaked," as he says) every year because of the number of laughable mistakes it producesthat hurts the reputations of both Hollinger and the greater metric set. Dean Oliver's seminal Basketball on Paper is an underground classic, but probably no more than four NBA executives have read it. 82games.com's advancements are four years old, and its proprietors have turned their focus to private research for teams.
The answer of the future might lie with the likes of the Rockets' Daryl Morey, the Blazers' Kevin Pritchard, and the Thunder's Sam Prestiall young, open-minded general managers who keep an eye on the stats revolution, while being the shrewdest executives in the room. They don't just invest their faith in stats or scouting; they see both as complementary approaches. They've got no ideological bone to pick other than winning. When they do that, this new merger of stat research and keen-eyed observation will truly become the only true path to take.
Of course, Morey's now banking his all on a traditional superstar triad of T-Mac/Yao/Artest, and Presti spent his first season just trying to make the Sonics as paltry as possible. But in Portland, where Pritchard has assembled a Murderer's Row of young talent that presents any number of plausible combinations, we've got the chance to see a team guided by this dual philosophy. And if the Blazers can follow through on their potential and go deep into the playoffs, they'll have set a new standard for running a basketball team in the 21st century. Then we can call this hybrid approach simply "smart."
Related in Gelf
•A call for a stats-scouts truce in baseball.
•Aaron Schatz is at the vanguard of the stats-based revolution in football.
•Why NFL coaches aren't numbers-savvy.