In an extremely close vote, Kawhi Leonard of the Spurs was selected as Defensive Player of the Year over the Warriors’ Draymond Green and the Clippers’ DeAndre Jordan. The voting was quite close among those three players, with Green actually receiving more first-place votes than Leonard, but the reigning NBA Finals MVP won the award by 16 voting points.
The selection of Leonard isn’t shocking because he was undeserving; it’s more surprising because of how unlikely it was that he would win based on past voting trends and the perception of his candidacy. According to the ESPN Forecast panel, which has been pretty accurate with award predictions the past couple of years, Leonard was given only a 1 percent chance of winning the award.
As others have noted, the evaluation of defense has improved considerably in recent years. Instead of focusing almost exclusively on traditional defensive stats like blocks and rebounds (which would heavily favor Jordan this year), more in-depth measures are being considered.
• How specific opponents perform vs. the player in question, using manual video tracking like Synergy or automated player tracking like SportVU.
• How a team’s defense performs with a player on vs. off the court, using play-by-play data.
• More sophisticated, statistically based measures like defensive real plus-minus (RPM), which looks at on-off court stats and adjusts for teammates and opposition, among other things.
In the case of this year’s race, Synergy notes that Leonard allowed the fewest points per play when defending pick-and-roll ball handlers among the 71 players who faced at least 200 such plays this season. That speaks to how well he performs in his defensive role, going beyond the basics (like Leonard leading the league with 2.3 steals per game).
On-off data also supports Leonard’s candidacy, as the Spurs' defense was about five points per 100 possessions better with him on the court than off -- and that difference took them from about average without him on the floor to elite when he was playing. The difference was similar for the Warriors' defense with and without Green, but the Clippers’ defense was actually slightly better with Jordan off the court.
Specific to the case of Jordan, back in March Tom Haberstroh did a great job of explaining why the Clippers center shouldn’t have been a DPOY candidate. That article got a lot of attention, in large part because Doc Rivers blasted it -- while admitting he hadn’t read it. Perhaps that influenced voters to look beyond the basics that would traditionally favor the big man’s candidacy over the others.
Finally, the defensive RPM numbers are very high on Leonard, saying that his defensive impact is worth nearly five points per 100 possessions to the Spurs, controlling for who he plays with and against, among other things. Green’s impact is of a similar magnitude, while Jordan is much further back.
Evaluating defense in basketball is quite difficult for many reasons, but vast improvements have been made to NBA defensive metrics over the last few years. Leonard winning the Defensive Player of the Year award may be kind of shocking in terms of historical precedent, but it shows that a sizable portion of the media is beginning to pay attention to those better methods of evaluation and making smarter choices as a result.