A few posts back, I introduced the concept of defensive usage and attempted to quantify it. That iteration of the metric was rough and did not include a number of events on defense that I’ve now added to the statistic. This version now incorporates the following:
- Defensive errors
- Forced turnovers
Which means that the more subtle elements of team defense are still being overlooked. “Deterrence” is probably the most difficult factor to qualify: when a good rotation deters a pass, when a good rotation deters dribble penetration and nothing much comes of it or when good on-ball defense deters a shot. Then there are the degrees of basic defensive principles: how well a player closes out, how quickly a player makes a correct rotation, how good on-ball defenders are at preventing penetration and the ability to hedge screens.
So, this is not a perfect measurement of defensive contribution. But it is a fairly relevant one and a good starting place considering the general paucity of defensive information in basketball analysis.
Using this updated methodology, we now see the following positional breakdown by DUsg:
A few comments:
- The sample is over 2943 possessions involving 233 players on 22 teams
- Defensive Usage still does not sum to 100% due to reasons previously explained
- There is essentially no difference in turnovers forced between positions
This gives us a fairly clear picture that interior players, on average, shoulder a larger defensive role than perimeter players. It is important to note that this is not a true mirror statistic to offensive usage because they are measuring different things. Neither is comprehensive, and in this case neither is measuring an equal portion of offensive and defensive events.
Some people might argue that what this methodology is missing overlooks wing players more than bigs. After all, with today’s NBA rules perimeter players move freely and dominate on offense. The ability to deter movement and stop penetration is incredibly important. Then again, all of the subtle positioning and help action by bigs in the paint likely offsets this to a large degree.
Finally, these are averages – outliers exist. It is possible for certain players to buck the trend. In the future I will post individual leaders using this methodology. As of writing this, the highest DUsg from a perimeter defender is 19.3% (Luol Deng, Chicago, in 359 possessions).