Quick, which five-team group was elite on defense last year, Group A or Group B?
|Team||Group A D-REB%||Group B D-REB%|
If you answered Group A, you are correct. These were the Defensive Rebounding rates of the top five defenses in the NBA as scored by Defensive Rating: the San Antonio Spurs, the Golden State Warriors, the Utah Jazz, the Atlanta Hawks, and the Miami Heat.
If you answered Group B, you are wrong. These D-REB% numbers, while similar to Group A, feature the five WORST defenses in 2016-17: the Minnesota Timberwolves, the Sacramento Kings, the Phoenix Suns, the Denver Nuggets, and the Los Angeles Lakers. I don’t blame anyone who answered Group B though. There are very few differences between each set of five numbers.
That leads me into this edition of Stat of the Week. A few days ago in the comments section, a discussion popped up about the Nuggets struggling on the defensive glass through the first two preseason games, namely Nikola Jokić. On one hand, the rebounding numbers were down for Jokić, and through watching game film, it’s easy to see where Jokić (and others) could improve upon while rebounding. On the other hand, the point was brought up that rebounding in general has very little correlation to winning and isn’t nearly as important as making shots or generating misses in the first place.
Both answers are correct. Jokić (and others) should strive to finish every play defensively and use their rebounding talents to the best of their ability. It is a competitive advantage, especially because Jokić isn’t going to generate a ton of misses through his quick feet and suffocating defensive nature. He in particular has to make the most of one of his best skills, even if the Nuggets as a team don’t necessarily need to grab every defensive rebound.
The above ten teams though? That is a really small sample size. What do the previous ten years look like? I averaged out the D-REB% for all 30 teams to get a better idea on correlation between the two statistics.
That’s a pretty strong correlation, especially for the second half of teams in defensive rating. The first half is more erratic, but truly the second best defensive team is generally weaker, and the third best defensive team is generally stronger. Those are the ten-year averages with the two largest residuals, likely meaning that there were some outlier years mixed in there. Overall though, it’s good to see strong correlation for one of the four factor statistics.
So, why did I include the first table? There was almost zero correlation for teams last season, and the correlation over the last five years has trended in a similar direction. Here is the same data, divided into five-year windows instead of ten-year windows.
First, 2007-08 to 2011-12 data:
Still a pretty strong correlation. This data is actually even more correlative than the 10-year data, which is really surprising.
Second, 2012-13 to 2016-17 data:
Yikes. That gets crazy quickly. The trend line follows the general flow of the data, but there’s no longer a trend there, as can be seen in the R-squared values for each time period.
|Time Period||R-Squared correlation|
|2008-12 (5 years)||0.8256|
|2013-17 (5 years)||0.3677|
|2008-17 (10 years)||0.8077|
For those that don’t know, the goal of R-Squared correlation is for the number to be as close to 1.0, or 100% correlation, as possible. The five-year data set from 2007-08 to 2011-12 is clearly lifting up the correlation for the ten-year data. The other five-year data set shows that the focus on defensive rebounding, while improving in volume, has meant less to a team’s overall defensive success than ever before. There is still correlation, but a 0.3677 R-Squared value is nothing to write home about.
So, what does this mean for the Nuggets? Well, they were the fifth best defensive rebounding team in 2016-17 but the second worst team using defensive rating. Because the team is playing a more aggressive defensive scheme, and because Kenneth Faried is likely to see less minutes, it wouldn’t surprise me if the Nuggets dropped in D-REB%, but experienced a boost in defensive rating. Paul Millsap isn’t an elite rebounder, but he’s not terrible either and certainly better than the departed Danilo Gallinari. While this team will still try to be a positive defensive rebounding team and not abandon the concept altogether, there’s reason to believe that they don’t need to be an elite team on the boards to be an elite (or at least improved) overall defensive team.
Let me know your thoughts on why you think the importance of defensive rebounding has lessened in recent years. I have various theories involving pace of play, increase in half court efficiency, and others that might explain it, but I will dig into the data another time.