NBA Stats

Jun 13 2013 4:59PM

A New Way To Look At Defense And Its Impact On The Finals

In Game 1 of the NBA Finals, Tony Parker sealed the victory for the Spurs by making a twisting, leaning, bank shot after having slipped and lost his balance. The shot was incredible, especially because the defense was just about perfect. The Heat had forced the Spurs into a broken possession that ended in Parker's wild shot. But because it went in, we forget that the Heat's defense deserves some credit for its effort. If the Heat were to play defense like this the entire series, Parker and the other Spurs would be forced to take difficult shots, most of which they would miss. Instead, the Heat have allowed the Spurs to take too many relatively easy shots, like wide open threes.

Scenarios like these have encouraged the NBA to work on a new statistic called the Effective Defensive Rate (eDEF%) to better measure defensive contribution. At its core, the statistic represents the percentage of defensive possessions that end in good outcomes for the defense when a given player is on the floor. Good outcomes for the defense are things like missed shots and turnovers, while the bad outcomes are made shots and free throws. If a player has an eDEF% above 50%, then he forces more good outcomes (for his team) than bad. The higher the eDEF%, the more good defensive possessions the player produces and the stronger he is as a defensive player.

But the statistic is more than a simple ratio; it weighs defensive outcomes based on historical implications. That is, eDEF% realizes that LeBron James, Chris Bosh, and the other Heat players on the floor did a good job of forcing Parker to take a tough shot. Had Parker's shot missed, the Heat players would have been given more credit than if they had allowed Parker to snake his way into the paint but missed a layup, or allowed him to kick the ball to Danny Green for an open corner three, that he missed (don't worry, we'll get to this). Since Parker's shot went in, we still penalize the Heat, but more often than not, that shot would miss, and we would give the Heat praise for good defense. Essentially, this statistic determines which defensive possessions lead to stops more often, and weighs those possessions accordingly.

The creators of this metric at the NBA explain that the weights for the respective possessions were calculated with a regression model, utilizing eight years of data going back to the 2005-06 season, to determine how valuable the different defensive outcomes are to overall team performance. Some positive defensive outcomes are listed here, in order from most valuable to least:

  • Steals
  • Blocks (assuming they're rebounded)
  • Missed mid-range attempts
  • Missed attempts in the paint (outside the restricted area)
  • Defensive rebounds
  • Missed restricted area attempts
  • Missed three-point attempts above the break
  • Other turnovers (traveling, bad passes out of bounds, etc.)
  • Missed three-point attempts from the corners
  • Missed free throw attempts
  • Many conclusions can be taken from these rankings -- like the fact that steals are the most valuable outcome for the defense because they prevent the opponent from even attempting a shot, and often lead to points in transition.

    Revisiting the example above, the Heat forced Parker to take a mid-range shot, so the only thing they could have done better would have been to not let him shoot at all, by blocking his shot or stealing the ball. As mentioned above, it would have been worse for Miami had they allowed Parker to get to the rim for a layup, since forcing restricted area attempts is less valuable to the defense than forcing mid-range shots. The worst situation for the Heat would have been for Parker to find an open shooter in the corner, since forcing corner three-point attempts is among the least valuable of defensive possessions.

    While the model is still being tested, I asked the NBA to provide some preliminary results to see if there were any surprises. All the talk after Game 3 has been about the Spurs' three-point shooting, which deserves praise. However, a big problem for Miami was their inability to gain any sort of offensive flow. Which San Antonio players deserve the most credit for this? A good guess would be Kawhi Leonard, since he was tasked with guarding LeBron James, the world's best player, and has miraculously held James to under 20 points in all three games. But here's a list of the eDEF% for all the Spurs so far in the Finals. As a reference, the league average is just under 50%.

    PlayerMineDEF%
    Splitter,Tiago69.557.25%
    Leonard,Kawhi99.056.39%
    McGrady,Tracy15.055.05%
    Green,Danny94.051.85%
    Parker,Tony100.051.12%
    Duncan,Tim97.550.35%
    Mills,Patty14.049.82%
    Blair,DeJuan14.049.71%
    Neal,Gary67.549.22%
    Joseph,Cory33.048.93%
    Ginobili,Manu72.044.95%
    Bonner,Matt28.537.61%
    Diaw,Boris20.534.14%

    So according to this metric, Tiago Splitter and Kawhi Leonard have been extremely important on defense for the Spurs. It's been clear that the Heat have struggled to get to the rim during the Finals -- only 28% of their shots have come from the restricted area in the Finals compared to 33% during the regular season -- and Splitter is a big reason why, even more so than Tim Duncan. Not only are they getting to the rim less often, the Heat are also shooting a slightly worse from the restricted area in the Finals (65.7%) than they did during the regular season (67.1%).

    Turning to the Spurs' three-point shooting, the Spurs made an NBA Finals record 16 threes in Game 3 and are shooting a scorching 44% from three in this series. Who's to blame for all the open threes? First, let's look at how Miami's players have defended overall using eDEF%:

    PlayerMineDEF%
    Chalmers,Mario84.055.67%
    Haslem,Udonis41.554.18%
    Bosh,Chris101.052.28%
    Wade,Dwyane101.549.80%
    James,LeBron122.047.85%
    Miller,Mike58.546.69%
    Allen,Ray70.545.63%
    Anthony,Joel14.544.00%
    Andersen,Chris38.041.71%
    Jones,James11.040.38%
    Lewis,Rashard11.040.36%
    Battier,Shane20.540.05%
    Cole,Norris58.037.46%

    Of the players getting big minutes (more than 10 per game), Norris Cole, Ray Allen, and Chris Andersen are at the bottom of the Heat's eDEF% list, while Mario Chalmers, Udonis Haslem, and Chris Bosh are at the top. We should keep in mind that Allen's role is more important on offense, so the Heat are more accepting of his poor defense than that of Cole or Andersen.

    Although we know how each player has performed overall, it would be even better to know which areas of the floor each player excels in, or struggles at defending. Another interesting part of this analysis is (because we've broken down every defensive possession for a given player) we can determine which areas of the court the team defends well while the individual player is on the floor, and which areas the team defends poorly.

    Although we don't know which specific players are defending a particular area of the floor on each possession, we do know how well the Spurs as a team defend that area when that player is on the floor. For example, let's take a look at how the individual Spurs players have guarded the restricted area:

    

    Based on our analysis, Leonard and Splitter have been the ones protecting the basket for the Spurs, while Diaw and Bonner have not (although like Ray Allen, Diaw and Bonner are not really in the game because of their defense). We expected to see Splitter and Duncan at the top of this list, but Leonard has actually been the most influential defender of the basket area for Spurs in this series, while Duncan has only been slightly above average. Note that this is a measure of Leonard's effectiveness while he's on floor, so his influence is most likely coming from his ability to prevent Miami's drivers from getting into the lane comfortably. Another interesting note is that none of the players getting big minutes for the Spurs have been poor defenders in the paint.

    We can break down Miami's defense of the three point line in the same manner:

    We mentioned earlier that Cole, Allen and Andersen are among the lowest ranked Heat players in overall eDEF%. When we look at specific areas of the floor, we notice that the Heat are allowing more corner threes with Cole on the floor and the Spurs are hitting more threes from above the break when Andersen and Allen are on the floor.

    They've also struggled defending threes above the break with Mike Miller on the floor, although he's made up for it with his own excellent three-point shooting (9-10 3FG). It will be interesting to see if Haslem gets more minutes, since Miami has been defending the three-point line at its best when he is on the floor.

    The model is still being tested, but it has a lot of promise. And we'll see if Miami makes adjustments to guard the three-point line better in Game 4. Also, it's time to recognize Tiago Splitter and Kawhi Leonard for their fantastic work on the defensive end of the floor. Both of these players rated very highly throughout the season based on eDEF%, and those results are being magnified in the Finals.