Drafting by the Numbers

[MavKikiNYC]

Keeping Score
It's Not Where N.B.A. Teams Draft, but Whom They Draft

By DAVID LEONHARDT
Published: June 26, 2005

The 29 teams trying to catch up to the Spurs may do well to keep this in mind during Tuesday's draft: San Antonio has not drafted a single high school player or college freshman in the past decade.

They have chosen college veterans, including Tim Duncan, and international players like Tony Parker. But every time they have had a chance to fantasize about the sweet potential of a high school player, to see flaws as nothing more than trifles waiting to be fixed by an N.B.A. coach, the Spurs have said no thanks.

On Tuesday, N.B.A. executives will get one last chance to indulge their worst draft-day instincts. The new collective bargaining agreement, which takes effect next season, bars any player who is less than a year out of high school from being drafted. The change will help level the playing field between the league's savviest drafters and its incurable optimists. In the last seven drafts, teams have taken 26 high school players and 21 college freshmen, and most of them have been chosen far too early.

This pattern is the clearest lesson to emerge from a new analysis of the draft, done by three members of basketball's small crew of "Moneyball"-like analysts. Wayne Winston and Jeff Sagarin, statisticians who work as consultants to the Dallas Mavericks, and Nate Medland, a student of Winston's at Indiana University, looked at every draft since 1998, comparing the spot where a player was taken with his N.B.A. performance.

The average high school draft choice was picked seven spots above what his play ended up warranting. College freshmen were almost as overvalued. Players with two years of college experience, like international players, were drafted roughly where they should have been, on average. The typical college junior or senior was undervalued by two spots.

So there is an elegant relationship between big-time experience and performance. Many of the teams that have done the best drafting relative to their position, like the SuperSonics, the Spurs, the Heat, the Nets and the Grizzlies, have avoided the question marks that hang on teenagers (though the Knicks have taken mostly upperclassmen and have still been the league's worst drafters since 1998). "It is very difficult with a high school kid or a kid with one year of college," Ed Stefanski, the Nets' general manager, said last week. "The body of work that has been done is not large."

High-profile exceptions - Kevin Garnett, LeBron James, Jermaine O'Neal - are easy to name. And their success is one reason Martell Webster and Gerald Green are projected to be chosen in the first half of the first round Tuesday. But it wasn't long ago that executives spoke with similar excitement about Leon Smith, DeSagana Diop, Dajuan Wagner, Kwame Brown, DeShawn Stevenson and Jonathan Bender.

The uncertainty attached to young players is one reason for their relative lack of success. But the scant playing time that most teenagers get in the N.B.A., at an age when their games should be improving rapidly, also matters. "If you're going to tell me guys are going to get better sitting on the bench, I don't believe it," said Tony Barone Sr., the player personnel director for the Grizzlies, who ended up with Pau Gasol of Spain in the 2001 draft rather than one of several tempting high schoolers. "There's nothing like game competition."

The study by Winston, Sagarin and Medland will not be the last word on drafting. They looked at two windows to measure performance, a player's first three years and his first five. With more data, somebody else may discover that college freshmen are good draft-day deals as long as a team holds on to them for close to a decade.

"A four-year span," Randy Pfund, the general manager of the Heat, said, "is not enough time for the Jermaine O'Neals of the world to get better."

If the N.B.A. ever develops better defensive statistics, a future analysis could also examine whether certain players - those from the Big Ten, for instance - tend to have specific skills. For their study, Winston, Sagarin and Medland relied on a plus-minus statistic that gives players credit for being on the court when their team does well. By that measure, Duncan was this season's best player.

But the analysis does establish a framework for thinking about the draft, especially because other researchers have recently studied drafts in other sports. In the N.B.A., the 16th pick is typically worth about half as much as the first pick, and a pick early in the second round - in the low 30's - is worth about half as much as the 16th pick. The quality of N.B.A. draft choices seems to drop off at about the same rate as that of baseball picks but much faster than N.F.L. picks.

Still, the decline is gentle enough that trading down can make a lot of sense. Four years ago, the Nets traded Eddie Griffin, the seventh pick, to Houston for the 13th, 18th and 23rd picks. Winston said the Nets got 40 percent more draft-pick value than they gave up, just based on the pick numbers, not the players.

But the Nets did something else smart. In Griffin, they traded a forward who has turned into an N.B.A. bench player. They then used two of the three lower picks on Jason Collins and Richard Jefferson, who became core parts of their Eastern Conference championship teams.

Collins played four years at Stanford. Jefferson played three at Arizona. Griffin, a Seton Hall freshman when he was drafted, was short on experience, but he sure did have a lot of potential.