We Have A Trade
Earlier this year, on the first day of the Sloan Sports Analytics Conference, amateur and professional quants packed the Boston Convention and Exhibition Center’s largest room to watch five men talk for 75 minutes about advanced statistics in basketball. It was one of the weekend’s most anticipated panels. The audience consisted largely of people who might have killed someone—or at least modeled a cost-benefit analysis for killing someone—to parse through the same spreadsheets as some of the experts on stage. On one end was Mike Zarren, the Celtics’ assistant general manager, a Harvard Law School graduate who had clipped a shamrock pin to his lapel. In college, at the University of Chicago, Zarren had played quiz bowl, not basketball, and served as the team’s president the year it won the first triple crown of national championships in history. Twelve years later, he was sitting next to Mavericks owner Mark Cuban, who cut the imposing presence of a billionaire while wearing faded blue jeans and a ring-collared T-shirt that read “Talk Nerdy To Me.” These guys weren’t easily baffled by questions about NBA metrics. In fact, there weren’t and aren’t many people in the world with a better grasp on basketball numbers. But then someone in the audience stood up, took the microphone, and stumped just about everyone on stage.
“I was wondering,” he began, “if we could talk a little bit more about the analytical process during, like, draft trades, where you’re either comparing a more-known quantity, like a player in the league, with someone you’ve drafted or are potentially going to draft, or you’re talking about two players, either from Europe and one from the States—or, even if they’re from the States, where they’re from different conferences—and what analytics you have available to you for that kind of process?”
Whew! Almost immediately, the lightly bearded Zarren, who had been nodding at key words, removed his hand from his chin. “Make it tougher: include a future pick in there that you don’t even have any clue what it’s going to be valued at,” he said. He leaned back in his chair. “It’s just a really hard question to answer,” Zarren sighed.
Before long, on the other side of the dais, Kevin Pritchard chimed in. Pritchard was the only panelist who might identify himself as a basketball player. He helped Kansas to an NCAA tournament title in 1988, was drafted with the 34th pick in 1990, and labored through four seasons with five different NBA teams. But his jock bonafides mattered less here than his last job. After he latched on with R.C. Buford, the San Antonio general manager, Pritchard joined the Trail Blazers, first as their assistant general manager and then, from 2007 to 2010, as the general manager. Everyone in the audience knew of Pritchard’s resume. Some had probably memorized it. What they might not have realized was that Portland, under Pritchard, made more trades than any other team during the last five NBA drafts. “That, for me, is the hardest thing to do: trade a player that’s a known for an unknown, or vice-versa,” he said.
By that point in the weekend, I had met people with an alphabet soup of advanced degrees. I had glanced at poster displays with more Greek letters than English words, and I had scribbled down sentences that sounded so smart that I hoped I might one day understand what they meant. I had heard nonchalant observations that were more perceptive than anything I might ever say, all while waiting in line for the restroom. And yet Pritchard’s confession might have been the most curious revelation of all. Here was an NBA general manager, the executive who had been the most aggressive with his assets, basically admitting that draft-day transactions were still a blind spot for even the savviest front offices. He also said this in the blase tone of someone standing outside and talking about the weather.
Every basketball fan who knows LeBron James from Darko Milicic understands the stakes of the NBA draft. At its most basic level, the draft is the only opportunity for any team to improve solely by addition and for a small-market franchise to acquire the sort of superstar who only considers certain cities come free agency. But that’s just theory. The draft is a spectacle, really, because of all the arithmetic that might occur whenever David Stern moseys up to the lectern and says: “We have a trade.” A trade! Then comes a short silence. In this pause, for that split second, anyone can be traded anywhere. Not only is every team open for business on draft day, but business is also very, very good.
There were 64 total trades in the last five NBA drafts, a period in which every team traded with another team, some with two other teams. These teams traded established NBA players, college players, and international players. They traded the No. 2 pick and the No. 60 pick. They traded for straight cash. They traded future picks in that year’s draft, future picks in the next year’s draft, and future picks in the drafts after the next year’s draft. Championship contenders traded for who they thought might be the player to bring them a title. Rebuilding organizations traded for what they thought might be the player (or future player, or something else entirely) to bring them a title in five years. In this time, Portland traded 16 times and Oklahoma City 12 times; Atlanta, Golden State, New Jersey, Sacramento, and Utah just once. Dallas made six trades and Miami eight. The league median was four.
But the remarkable thing about these transactions is that they happened at all. No general manager, whether he prefers Excel or swears by shorthand, denies that the draft reeks of risk on almost every level. Trades might be the riskiest part about it. They require front offices to forecast the potential of players who have never played in the NBA—in real time, no less—and then swap these prospects for assets that might be more certain, less certain, or much less certain. It’s no wonder that reliable metrics for the NBA draft, especially analytics that measure NBA draft trades, are thought to be so valuable even if no one knows what they are or how valuable they might be.
Or, as Steve Kerr, the Suns general manager of the Suns from 2007 to 2010, told me recently: “Every team in the league has some MIT whiz trying to make sense of all this. Frankly, I think it’s impossible. In the end, there’s no way you could ever quantify this stuff to a precise calculation.”
Maybe the best way to understand the analytics of NBA draft trades is to simplify. So let’s consider this year’s class of players as pieces of fruit. Start with Kyrie Irving and Kemba Walker. We might call them apples. (You don’t need Newton to see where we’re going here.) Irving and Walker are both scoring point guards. They both come from two of college basketball’s iconic programs. They both measure between 6-foot-1 and 6-foot-3 and weigh right around 190 pounds. They both will be two of the first three guards drafted this year.
The NBA organizations that rely on analytics—prominent among them are the Celtics, Mavericks, Rockets, Thunder, and Spurs—have their own internal models for forecasting professional potential. They protect these methods like modern state secrets, since there’s no incentive to share what could become competitive advantages. It’s only when general managers are out of jobs that they might speak openly, albeit in the rhetorical equivalent of a declassified memo, dark with redacted phrases. Even then, these executives know they’ll be back in a front office soon enough. What they’re willing to admit is that sophisticated teams prize different statistics—and that no system can eliminate risk altogether. “You have no idea how that college player’s going to contribute,” Memphis Grizzlies general manager Chris Wallace told me. “There’s no way to know if that’s going to translate to the NBA, or how quickly it would translate.”
So when a general manager confides that rebounding is something that does translate—an effective rebounder in college, for example, figures to be an effective rebounder in the NBA—he only does so because it’s common knowledge. In 2008, the ESPN columnist John Hollinger joined Basketball Prospectus’s Kevin Pelton in searching for college statistics that predict NBA success. In his annual report, which he calls the Draft Rater, Hollinger favors a regression analysis of 27 individual factors to project long-term potential. Pelton uses NCAA similarity scores, a system of 13 categories from height and weight to usage rate and per-minute winning percentage, to identify the NBA rookies with whom draft choices are statistically similar. But the rub is the same no matter the formula.
“Information is the key to making this thing work, and the more information we have, the better,” Hollinger wrote in his 2008 Draft Rater. He said this another way last year: “The Draft Rater has one other weak spot: It thrives on information and struggles when it lacks enough.”
Which brings us back to Irving and Walker, the two apples. Irving, the presumed No. 1 pick, played 11 games at Duke, leaving after a freshman season hampered by a toe injury. He also could be the star that Cleveland builds around for the next decade. Walker, whose stock soared this season at Connecticut, gave scouts 103 chances to see him in a three-year career that ended with 11 straight wins and a national championship. Irving’s statistical profile is like a high-school senior’s. The team that drafts Walker will know as much about him as maybe any other lottery choice. In other words, Irving is the better prospect but also the riskier one. He is a Red Delicious; Walker is a McIntosh. And if the 6-foot-9 sophomore Derrick Williams, who likes to shoot as much as he dunks, is a Granny Smith, then what does that make Kenneth Faried, the 6-foot-8 senior who has rebounded his way from Morehead State into the draft’s first round?
Now, the pears. These are the international players. It’s a good year for pears. Six foreign prospects could be first-round picks in the draft, including four in the top 10. Maybe the most intriguing import is Bismack Biyombo, a 6-foot-9 Congolese forward who picked up basketball four years ago, when he was 14, and moved to Spain to play professionally. But it wasn’t until a showcase in April that NBA scouts started salivating over him. In two days, this nobody from nowhere became a lock for the lottery. Enes Kanter is the lottery’s other mystery. Scouts know so little about the 6-foot-11 Turkish center that the discovery of grainy footage of Kanter dominating a high-school game in front of empty bleachers was hailed as a veritable breakthrough. The team that drafts Kanter will do so despite seeing him against legitimate competition on only a handful of occasions. But at least they’ll be drafting him to play right away. With the second-to-last pick of the 1999 draft, the Spurs took Manu Ginobili. He didn’t suit up for San Antonio until 2002. By that time, only one other player selected between the 48th pick and 58th pick remained in the NBA.
It’s challenging enough for scouts to assess overseas prospects when they’re matching up with guys who couldn’t even play for, say, Oak Hill Academy, a basketball powerhouse that doubles as an American high school. But that’s assuming they see them at all. In 2009, after Ricky Rubio thudded on the NBA scene at the Olympics, Kerr flew to Spain to watch the 17-year-old dynamo. The Suns were picking 14th. How tantalizing it must have been to envision Rubio as Steve Nash’s successor—except Kerr couldn’t watch Rubio when he went to Europe, because Rubio didn’t play that day. So, later that season, Kerr dispatched one of his trusted scouts on another trip to see Rubio. He didn’t play that time, either. The Suns couldn’t compare Rubio to the red apples or the pears, because they didn’t know if he was a red apple, a pear, a plum, or something else entirely.
Finally, the oranges: everything and everyone else. The oranges can be human beings (NBA veterans), imaginary human beings (future picks), or anything besides a human being (cash). Some teams with deep-pocketed owners, like Portland, often dole out money in exchange for second-round prospects. This type of trade serves dual purposes. It allows the Blazers to take a chance on a pick that probably won’t pan out at virtually no cost while also clearing salary-cap space for their partner in the trade. Other teams on the downward slope of success—Pritchard called it the “treadmill of mediocrity”—might want to stockpile picks for future drafts. It all depends on a team’s foundation, the owner’s aspirations, and a general manager’s tolerance for risk.
“You don’t come in with any preconceived philosophy,” Wallace said over the phone earlier this month. He had called me at first on a Thursday around 11:30 p.m., a few seconds after Dwyane Wade drilled a three-pointer and held his follow-through during Game 2 of the NBA Finals. We made plans to talk a few days later. “Some drafts,” he said then, “the opportunities are greater than others. Your team may be in a very different situation than the one two years from now. It’s determined by the state of a team. Not everyone is looking at their team with short- and long-term objectives in the same light. It’s a very difficult time of the year for us, in advance, just to make a blanket statement.”
Neither Wallace nor any other general manager denies that trading an NBA player for another NBA player is troublesome. So is sifting through draft choices every spring. But swapping a known entity for something or someone unknown—well, it’s like comparing apples and oranges. And pears. And also plums.
A few weeks ago, I asked Dean Oliver, the pioneer of basketball analytics, if he could recall any particularly memorable trades from recent drafts. “Certainly the Battier trade,” he said without having to recall for very long.
He was referring to the 2006 deal that sent Shane Battier from the Grizzles to the Rockets for the rights to Rudy Gay, whom Houston had selected with the No. 8 pick, and also Stromile Swift. No NBA player represents the budding industry of sports analytics better or more often than Battier. Michael Lewis, the author of Moneyball, profiled Battier in a 2009 New York Times Magazine cover story called “The No-Stats All-Star,” and Harvard Business Review detailed the trade in a case study with eight appendices. It’s been fetishized mostly because of what happened next: Battier became an unexpectedly crucial player for the mathematically minded Rockets, whose general manager, Daryl Morey, moonlights as the co-organizer for the MIT conference, where a photograph of Battier defending Kobe Bryant by covering his face with his hand served as the backdrop in the main auditorium this year.
But the trade was fascinating long before Battier moved to Houston. In fact, the prior uncertainty may have been the most fascinating thing about it.
In the four drafts since, only once more have two teams basically straight-up traded an NBA veteran for a lottery pick, the ultimate example of apple for orange. It occurred again the next year, when Golden State traded Jason Richardson for the eighth pick, Brandan Wright. (That deal also involved the No. 36 pick, Jermareo Davidson. He and Swift, the other player in the Battier deal, have since combined for 30 percent of Battier’s scoring output—and Battier is known for not scoring.)
In the 2006 offseason, Houston was taking aim at the NBA title with Tracy McGrady, Yao Ming, and, of course, Battier. Over the next three years, they averaged 53 wins, but ill-timed injuries prevented them from advancing past the Western Conference semifinals. The Grizzlies, meanwhile, were falling fast. They won 22, 22, and 24 games the next three years, and their mediocrity was rewarded with the No. 2 pick in the 2009 draft, which they wasted on Hasheem Thabeet. But Memphis also used the drafts to accumulate valuable young players like Gay, Mike Conley and O.J. Mayo, the last in a draft-day blockbuster. At this season’s trade deadline, before Memphis stunned San Antonio in the playoffs and finished one win from the Western Conference finals, the Grizzlies prepared for their playoff push with one last move. They sent their most recent draft mistake, Thabeet, and a future first-round pick to the Rockets. In return, they acquired none other than Shane Battier.
It wouldn’t have been unreasonable for two equally astute analysts to land on either side of the original Battier trade. In the end, even though neither team has won an NBA championship and Battier ended up right back where he was, the Rockets improved right away, and the Grizzlies laid the foundation to improve over time. And neither team has shied away from draft deals since then. In the next four drafts, Houston and Memphis combined for nine more trades—and one with each other.
Last month, on an off day during the NBA’s conference finals, the Warriors announced that they had hired a new member for their executive board. It was Jerry West, the shadow of the NBA’s logo and, more recently and more in the flesh, the general manager who won seven championships with the Lakers and led the Grizzlies to their first playoff berth. It didn’t take long for West to start consulting on basketball operations for Golden State. Just a few days after he was hired, he was already scouting draft prospects. West is fond of telling his proteges that they should pride themselves in being right about 51 percent of their front-office decisions. That success rate might seem pessimistic or optimistic, but either way, it’s indicative of the overwhelming uncertainty in NBA front offices. This is why even wildly revered general managers are so loathe to take on all that inertia. If the risk is too risky in the first place, it won’t matter if it’s worth the reward.
In March, near the end of the panel in Boston, Zarren offered an estimate that wasn’t much different from West’s. “The worst thing is that if you’re taking something with a 60 percent chance of succeeding, that means you have a 40 percent chance of it not succeeding. You have a 16 percent chance of it not succeeding twice—which isn’t ridiculously small,” Zarren said. “But if you screw up two drafts in a row, you’re not making a third draft.” Cuban made it known that he agreed, and that it was time for another question, by staying silent.
The session ended not long afterward, and the crowd spilled out into the convention center’s hallways, which were lined with stale breakfast pastries, leftover boxed lunches and coffee pots that were constantly being refilled. In this chaos, I tried to find Pritchard, who stands 6-foot-3 and seems even taller. I still couldn’t track him down. I assumed there was a long line to talk to him, anyway.
I soon found out that milling around, even purposefully, didn’t guarantee a conversation. The next day, I spotted Zarren in the hallway between sessions. He was standing by himself—this was like seeing a rock star without a crew of roadies—so I walked over to introduce myself. I was planning to ask him more about his draft-day philosophy, as it were, but his phone began to ring just when I reached him. It was Danny Ainge, the Celtics’ president of basketball operations. Zarren excused himself very politely and took the call.
A version of this story was originally published in 2011 on Grantland.