A few weeks back, I wrote extensively about the idea that everyone in the NBA is playing offense the same way, and how — this is a technical term — it’s a bunch of bullshit. I’ve been thinking about a way to explain and visualize the same concept for NBA defenses, and to do so, I’m turning the clock back a few years.
One of my favorite defenses of all time was that of the early Nick Nurse-era Toronto Raptors. They were shapeshifters, dramatically altering their tactics on any given night to fit the opponent. To explain what they were doing and what made it so unusual, Krishna Narsu and I created two new metrics and wrote about them for FiveThirtyEight.
Using Second Spectrum’s tracking data to identify which coverages teams play against five offensive actions — pick and rolls, isolations, post-ups, off-ball screens and dribble handoffs — we created Aggression+: an index statistic that shows how aggressive a team’s defense is relative to the average team in the league that season, both against every individual action and overall.
The league average is set at 100, and each point above or below indicates that team is 1 percentage point more or less aggressive than the average team in the league. Teams that blitz pick and rolls and dribble handoffs, double-team isolations and post-ups, and jam or switch off-ball screens are considered more aggressive, while those that drop their big man back into the paint on screens and rarely send help at one-on-one plays are considered less so.
…
By calculating the variance in each team’s Aggression+ for each play-type and overall in every game they played, we created Variance+: an index statistic that shows how often a team changes up its coverages on a night-to-night basis, both against every individual action and overall, relative to the average team in the league that season. Again, the league average is set at 100, and each point above or below indicates a team is 1 percentage point more or less willing to vary defensive strategies than the average team.
To recap:
Aggression+ measures how much more or less often your defense sends extra defenders at an offensive action, compared with the league average team
Variance+ measures how much more or less often your defense changes its night-to-night tactics, again compared with the league average team
If you check out the front page, you’ll see a new tab for Aggression, with each team’s Aggression+ and Variance+ plotted against each other. This shows you which teams are most likely to send extra defenders at specific opponent actions (the Nets are by far the leaders in Aggression+) and which teams are most likely to change tactics on a night-to-night basis (the Heat are the runaway leaders in Variance+).
Brooklyn’s defense has been so aggressive this season that is actually checks in seventh among player-tracking-era teams in Aggression+, while Miami has been so likely to vary its strategy that its Variance+ ranks fourth out of the 360 team-seasons included in the database.
The Jordi Fernandez Nets fit the mold of several different teams near the top of the Aggression+ charts in that they are undermanned on a talent level and throwing bodies at the problem. That’s as good a way as any to (try to) generate stops when you don’t have the personnel to do it on the basis of scheme and execution alone.
The team at the top of those rankings, though, is the 2013-14 Heat. That was the year after their first title, when they just absolutely ran roughshod over the league and smothered opposing offenses with their blitz-and-switch-happy system anchored by LeBron James, Chris Bosh, Dwyane Wade, and even guys like Shane Battier.
Erik Spoelstra is also probably the coach who is most willing to change defensive tactics of any that has been in the league over the last decade-plus. We’ve seen this with all different kinds of rosters. When you look at the list of teams with the highest Variance+ over the years, you can find five different Heat teams in the top 30: this year's squad, the 2014-15 team (aka the final Heatles team), the 2018-19 Heat team that spent the entire season just searching for answers, the 2013-14 team that romped its way to the title, and the 2020-21 team the year after their run to the Finals in the bubble.
Spo is just like that.
Of course, aggression and variance aren’t the only ways to exemplify the ways in which defenses are different. In the same way we did with offenses, we can take a look at which things different teams prioritize when it comes to getting stops.
Aggression+: Explained above
Variance+: Explained above
Turnovers: Opponent turnover rate
Defensive Rebounding: Defensive rebound rate
Shot Prevention: Opponent Moreyball Rate
Foul Avoidance: Free throws per field goal attempt
Just as was the case in the post about offense, I converted each of those measurements into percentiles so they could be scaled together on radar charts, mostly because I like the way those charts look. And in those, too, you can see that there are so many teams playing basketball so differently from each other.
Consider the league’s top three defenses, for example. You can see in the chart above that both Oklahoma City and Houston are on the high end of the Aggression+ and Variance+ scales, while Orlando is at the exact opposite end of the spectrum. But the three teams also each go about achieving their results differently.
The Thunder prioritize forcing turnovers to the near exclusion of almost all else. The Magic also generate a ton of turnovers, but also make a point of hitting the defensive glass and do a better job of preventing their opponents from getting the best shots. And the Rockets are a bit more balanced than both of them when it comes to their priorities.
We can do the same thing with other teams. Look at the top of the East: The Knicks are quite static and passive, prioritizing defensive rebounding and limiting fouls. (Duh, they are coached by Tom Thibodeau.) The Celtics, meanwhile, are more aggressive and look to prevent opponents from getting the best shots. (Obviously; they’re coached by Joe Mazzulla.) And the Cavs fall somewhere in the middle.
Update: On a suggestion from Kristian Palotie, I added data for ball pressure to the defensive style charts. His theory was that it would better capture the Grizzlies’ aggressiveness, and he was extremely correct.
All of this unsurprisingly carries over to the ways that defenses approach trying to stop specific opponents. Because why would you try to defend the Bucks and the Grizzlies the same way, when they have totally different personnel and even more divergent offensive philosophies?
Again, we can use Aggression+ and Variance+ to visualize these differences.
Opposing teams have been more aggressive this season against Minnesota and Phoenix than against any other teams, while playing more passively against Washington and Memphis than anyone else.
Intuitively, this makes sense. With Mike Conley struggling and questionable decision-makers elsewhere on the floor, playing aggressively against the Wolves can lead to more turnovers and forced shots. With the Suns dealing with injuries and chemistry issues and having a caretaker point guard in Tyus Jones alongside three perimeter-oriented star contributors, ramping up the aggression can be helpful.
On the opposite end of the spectrum, the Wizards simply don’t have anybody that merits increased aggressiveness, while the Grizzlies play such a wildly different style than everyone else that teams revert to a more conservative shell against them because that might be the easiest way to deal with it.
Meanwhile, nobody sees more variance in strategy against them than the Bucks. Again, this makes sense. Everybody is going to have different ideas of how to deal with Giannis Antetokounmpo and Damian Lillard. And nobody sees less varied looks than the Lakers, because how else would you deal with defending a LeBron James-and-Anthony Davis team that surrounds those guys with not all that much in the way of shooting or playmaking?
We can also look at this another way: How often do teams deviate from their own defensive strategies against specific opponents? Basically, which teams force their opponents into using bespoke game plans?
This is something I looked at as it related to the Thunder last season, when Mark Daigneault talked about the way teams were throwing the kitchen sink at his team’s offense in order to generate stops.
There’s a really interesting list here. The teams that force defenses outside of their comfort zone most often are Atlanta, Minnesota, New York, Phoenix, and San Antonio. That’s a list of teams that includes Trae Young; Anthony Edwards; Jalen Brunson and Karl-Anthony Towns; Devin Booker, Kevin Durant, and Bradley Beal; and Victor Wembanyama. It’s no wonder that defenses venture outside of the norm against those squads. You don’t want to give any of them anything familiar, because they’ll probably beat it with relative ease.
The opposite end of the spectrum makes me feel similarly. The teams that see the least divergence from opposing defenses’ usual strategies are the Grizzlies, Wizards, Raptors, Jazz, and Kings. The Raptors (especially this season, with all of their various injuries) and Jazz fall into the same bucket as the Wizards in that they don’t really have players who force opponents to adjust in a major way; while the Kings are somewhat of an unusual offense due to their pace 0and high volume of hand-offs and off-ball screens, which makes them somewhat similar to the Grizz. Again, this all makes a lot of intuitive sense.
So, don’t let anybody tell you that all NBA defenses are the same these days, either. Everyone is trying different things, and prioritizing different aspects of defense in order to achieve their results.