How Effective Are NBA Teams At Finding Value in the Draft?

Ian LevyContributor IJune 25, 2013

The NBA Draft is one of the primary team-building tools available to a franchise. It provides the opportunity to snatch up cheap production and potential, while wiping away the unevenly distributed advantages of the free agent market—cap space, mythology and history, climate and regional income tax differences.

Unlike free agency, each team is in complete control of the choices they make and how they evaluate the available options. From the pool that is available they get to select the talent which appears to make the most sense for their team.

Talent is distributed extremely unevenly in the NBA. Just over 450 players entered an NBA game this season, but by nearly any statistical metric, the vast majority of production comes from just a handful of those players. There were 1,229 regular season wins this year. The statistical metric Win Shares (WS) attributes 76.5 of those wins, or 6.5 percent of the league-wide total, to just five players—LeBron James, Kevin Durant, Chris Paul, James Harden and Russell Westbrook.

268.8 wins, or 21.9 percent of the league-wide total, can be attributed to just 25 players. However, this talent chasm that exists in the league has been built up through multiple years of drafts and free-agent signings. The draft is a wellspring of hope because the talent disparity is usually much smaller and harder to actually gauge. If teams are prepared, thorough and lucky, they can find great value in unexpected places.

One way we can illustrate this is by looking at the average value which has been acquired at each draft position. To calculate that value I looked again at Win Shares, a metric which includes both offensive and defensive contributions. Because a player's rookie production doesn't always capture the true value of their abilities, I looked at the total WS they produced in their fourth season in the NBA. I looked at all drafts from 2000-2009 and calculated the average fourth-season WS numbers for players selected at each draft position. The results are shown in the graph below:

If teams did a perfect job of assessing talent, and selecting that talent relative to their draft position, we would see a smooth curve from the first pick in the draft down to the last, but the results are anything but. Luck and player development are both factors here, but some teams have been just plain better or worse than others at finding value in the draft. 

Knowing the average level of production for each draft slot actually gives us a way to measure how well teams have done at finding value from their picks. Looking at the same time span, from 2000-2009, I looked at all the draft selections made by each team and calculated the total fourth-year WS that we would expect to have been produced by those players. I then calculated how many WS those players actually produced in their fourth season. The difference, positive or negative, is called marginal value.

Marginal value is not a measure of overall production, it's a measure of production compared to what was expected. In evaluating draft success in this way we are putting teams who consistently draft at the top of the lottery like the Charlotte Bobcats on a level playing field with teams like the San Antonio Spurs, who always find themselves selecting at the end of each round.

A team with a positive marginal value drafted players who produced more WS than would have been expected for their draft position. A team with a negative marginal value found themselves in the opposite situation. We can also look at the marginal value provided by each player and each specific draft selection, something we'll dig into in the analysis below.

For the purposes of this investigation, players who were traded on draft night were counted as being drafted by the team they ended up with. The graph below shows the results:

The data here covers drafts from ten years, enough of a sample to minimize the effect of a few good or bad selections. The teams at both ends of the spectrum have shown some consistent patterns in finding or missing value. Here are a few of the team draft records which I found most interesting:

Oklahoma City Thunder

Expected Value: 51.2 WS | Actual Value: 74.0 WS | Marginal Value: +22.8 WS

Over the time span I looked at, the Thunder (Sonics) extracted more value from the NBA Draft than any other team. This is pretty incredible considering they have some serious lottery blunders on their record like Robert Swift (-1.9 WS marginal value) and Mouhamed Sene (-3.0 WS). But, for the most part, they've had sparkling success throughout the draft. They've scored at the top of the draft with players like Kevin Durant (+6.8 WS), James Harden (+6.6 WS) and Russell Westbrook (+1.8 WS). They've found gems later in the first round like Serge Ibaka (+7.5 WS), Desmond Mason (+2.4 WS) and Nick Collison (+1.8 WS). They've even unearthed some solid contributors in the second round like Bobby Simmons (+5.8 WS) and Earl Watson (+1.8 WS). This is the kind of draft success that fans and GMs both dream of—deep and sustained.

Orlando Magic

Expected Value: 34.0 WS | Actual Value: 54.9 WS | Marginal Value: +20.9 WS

It would be easy to look at the statistical drafting success the Magic have had, match that with the fact that they're back at the top of the lottery, and assume that Dwight Howard tells the whole story. But while Howard was an incredible value as a first pick, providing 12.8 WS in his fourth season against an average of just 6.1 for the top pick in the draft, he's not the only great pick the Magic made over this time span. Jameer Nelson (+3.9 WS), Zaza Pachulia (+4.1 WS), Marcin Gortat (+5.8 WS), J.J. Redick (+4.0 WS) and Brendan Haywood (+3.5 WS) were all huge values near the end of, or outside the lottery.

The problem is that by their fourth seasons, Nelson was the only player from that group still playing with the Magic. This represented a frustrating trend of trading value by the Magic. They made 18 draft picks over the time span of this study, 10 of which provided a positive marginal value. Of those 10, just four were still with the Magic by the end of their fourth season.

New Orleans Pelicans

Expected Value: 34.8 WS | Actual Value: 48.2 WS | Marginal Value: +13.4 WS

Like every team on the list, the Pelicans have had their share of first-round misses. Do the names Cedric Simmons (-2.2 WS), Julian Wright (-1.0 WS), Kirk Haston (-1.7 WS) and Hilton Armstrong (-2.7 WS) ring a bell? Probably not, which should give you an idea of the kind of production they provided from the middle of the first round. However, the Pelicans also pulled J.R. Smith (+0.7 WS), Marcus Thornton (+1.7 WS), Darren Collison (+1.4 WS), David West (+2.2 WS), Brandon Bass (+2.7 WS) and Jamaal Magloire (+4.7 WS) out of the latter stages of the draft. Combined those players provided a marginal value of +13.4 WS in their fourth seasons in the NBA. But the real bright spot on the Pelicans draft record is Chris Paul. With 18.3 WS in his fourth season, Paul provided a marginal value of +12.2 WS, the highest of any drafted player from this era. 

Cleveland Cavaliers

Expected Value: 37.2 WS | Actual Value: 40.3 WS | Marginal Value: +3.1 WS

The Cavaliers have a modest positive marginal value of the time span of this stretch, but as with everything else he does, LeBron James has messed up the curve. James produced 13.7 WS in his fourth season, a marginal value of +7.6 WS for a top pick in the draft. If you remove him from the equation, the Cavaliers total marginal value falls to -4.5.

Positives on the Cavaliers' draft record are Daniel Gibson (+1.5 WS), Shannon Brown (+1.9 WS), Carlos Boozer (+2.3 WS) and Danny Green (+5.3 WS). Unfortunately, the last three were no longer with the team in their fourth seasons. On the negative side, they have Desagana Diop (-2.4 WS), Luke Jackson (-3.0 WS), Chris Mihm (-1.0 WS), DaJuan Wagner (-3.5 WS) and Christian Eyenga (-2.9 WS) on their record. At least things are looking up for them with their more recent selections, as Kyrie Irving, Tristan Thompson and Dion Waiters all look like positive values for their draft position.

San Antonio Spurs

Expected Value: 23.8 WS | Actual Value: 25.6 WS | Marginal Value: +1.8 WS

The San Antonio Spurs have a reputation for finding gems at the end of the draft, but their overall record is not quite as impressive as their reputation may indicate. Manu Ginobili was drafted in 1999, and so wasn't included in these numbers, but the 8.8 WS he put up in his fourth season gave the Spurs a marginal value of 8.2 WS, which would have been the second-highest marginal value of any player in my sample, behind Chris Paul. Tony Parker was a gem as well, providing a marginal value of +5.5 WS from the 28th pick in the draft. However, after that the record gets a little bit shakier.

DeJuan Blair (-0.2 WS) is usually discussed as a 2nd round steal, but he actually provided a negative marginal value in his fourth season compared to the average 37th pick. We also find a ton of misses, especially on those European players they liked to stash overseas. Tiago Splitter and Luis Scola (+4.8 WS) turned out to be NBA rotation players, but not so much for Robertas Javtokas (-0.7 WS), Sergei Karaulov (-0.6 WS) and Viktor Sanikidze (-0.7 WS). Altogether just seven of the 18 Spurs' draft picks from this time span were playing in the NBA by their fourth season, and just three with the Spurs. They swing for the fences in San Antonio and the home runs they've hit make it easy to forget about all the strikeouts.

Charlotte Bobcats

Expected Value: 29.1 WS | Actual Value: 21.0 WS | Marginal Value: -8.1 WS

The Charlotte Bobcats hold the 4th pick in this year's draft and an absolutely atrocious draft record. They've had moderate success with players like Gerald Henderson (+0.7 WS), Emeka Okafor (+0.6 WS) and Jared Dudley (+4.3 WS). But things didn't work out as well with D.J. Augustin (-1.8 WS), Bernard Robinson (-1.6 WS), Alexis Ajinca (-1.7 WS), Kyle Weaver (-0.7 WS) and Sean May (-1.3 WS). Then, of course, there is the whiff of whiffs, Adam Morrison, who was out of the league by his fourth season. A 3rd pick in the draft like Morrison produces, on average, 6.1 WS in their fourth season in the league. But of course, the Bobcats saw none of that. 

Chicago Bulls

Expected Value: 75.9 WS | Actual Value: 72.4 WS | Marginal Value: -3.5 WS

Given the Bulls' recent draft success with players like Taj Gibson (+0.3 WS), Joakim Noah (+4.4 WS), Luol Deng (+0.5 WS) and Derrick Rose, it's easy to forget the years of missteps. Go back a little further and you'll find other solid picks like Tyson Chandler (+3.1 WS), Kirk Hinrich (+6.2 WS) and Thabo Sefolosha (+2.4 WS), but also some monumental miscalculations like Marcus Fizer (-5.6 WS), Tyrus Thomas (-3.5 WS), Eddy Curry (-1.8 WS), James Johnson (-2.2 WS) and the unfortunate, injury-shortened career of Jay Williams (-5.2 WS). But the Bulls are as good a reminder as any that when it comes to measuring draft success, it's all about what you've done lately.