Winning the margins: who drives the Boston Celtics’ possession edge?


Joe Mazzulla often talks about his willingness to “dominate the margins and dominate the possessions.” But how?

The public often focuses on offensive rebounding when talking about the possessions battle. Yet, there are three other levers to this battle: defensive rebounding, forcing turnovers, and taking care of the ball. Because this possession battle seems to become more and more important, why not look at who’s the best at it? Who has the most influence on the possession battle on a basketball court?

This idea came after reading Owen Phillips’ excellent piece on the “Turnover Economy.” I wanted to dig deeper. What if we looked at this question from a more individual perspective—and included both turnovers and rebounding?

Using 2024–25 on/off data from Cleaning the Glass, I tried to estimate how players impact the number of possessions their teams gain or lose—both on offense and defense—when they’re on the court.

To do so, we need key metrics:

  • OFF POSS ADD (offensive possessions added) = (ORB% × 0.533) – TOV%

This tries to estimate how many extra possessions a player creates on offense, by grabbing offensive rebounds and limiting turnovers. The 0.533 factor comes from this season’s average miss rate—so it scales offensive rebounds to actual extra shot opportunities.

  • DEF POSS ADD (defensive possessions added) = Forced TOV% – (Opp ORB% × 0.533)

This estimates how many extra defensive possessions a player’s team gains thanks to their ability to force turnovers and secure the rebound after a missed shot.

  • TOT POSS ADD (total possessions added) = OFF POSS ADD + DEF POSS ADD

That’s the sum of both, giving a rough idea of how a player shifts the possession balance per 100 possessions played.

The reason we’re using on/off data is to avoid the bias of focusing solely on individual performance and rather look at the influence a player has on team performance. A great example of why individual rebounding metrics are biased is Andre Drummond. Despite being seen as one of the best rebounders of all time, his teams were sometimes better at rebounding when he was off the court.

The possession battle is a collective effort, and that’s why on/off statistics give us a better sense of individual influence. Yet note that on/off also has its own biases—such as who’s on the court with the player or who his backup is. No metric is perfect, but let’s dive into the numbers and try to understand who is the best Celtic at adding possessions on offense and defense.

A two way street

While some teams, like the Houston Rockets, are trying to win the possession battle from the jump till the end of the game, the Celtics seem to have another approach. Looking at the data, the Celtics appear split into two different teams.

In the bottom left, you can find all the starters. The players in that part of the chart are the ones less likely to have helped the team win the possession battle this season. On the other side of the line that cuts the chart in two, you can find all the players from the bench who played more than 300 minutes.

Watching the games, there was a clear feeling that with Luke Kornet and Payton Pritchard on the floor, the Celtics were crashing the offensive glass a lot more. But this two-way approach to the possession battle goes beyond that. On both offense and defense—through rebounding and turnovers—the Celtics were more intentional about winning the possessions battle with the bench unit than with the starting five. Why?

Winning the possession battle often comes from teams who lack offensive firepower or shooting, like the Houston Rockets. Therefore, it makes sense that the Celtics’ bench showed more willingness to fight and create more possessions to compensate for the more skilled players resting. But the offensive structure of the starting five also had an influence.

Indeed, the most influential players in the possession battle are often centers. Their size impacts rebounding, and the ground they can cover influences the ability to generate turnovers. Therefore, with Kristaps Porzingis spending most of his time on offense beyond the three-point line, it makes sense that the Celtics’ starting five is less likely to generate extra offensive possessions. However, what’s more interesting—and could have foreshadowed the Celtics’ struggles against the Knicks—is the starting five underperforming on defense in terms of generating extra possessions compared to the bench.

It’s no secret that the Celtics lost against the Knicks in part due to their inability to protect the glass and their struggles in the margins. Nonetheless, over the course of a full season, are the Celtics who help win the possession battle also the ones who help win games? Well—yes and no.

Is winning the possession enough?

For those who read a lot of my 10 takeaways and tactical deep dives, it won’t be breaking news that the Celtics were at their best with Kornet on the court. As we see on the chart below, the Celtics’ backup big man was on the floor when Joe Mazzulla’s team was winning the margins—and the scoring differential.

Also, some players were still highly positive despite not performing particularly well in the possession battle—such as Derrick White or Jayson Tatum. There were also players who helped the team dominate the possession battle but weren’t as impactful on net rating, like Pritchard or Neemias Queta. And then, there are three players who underperformed in both: Porzingis, Jrue Holiday, and Jaylen Brown. Over the course of the season, the Celtics were more likely to win the possession battle and have a better net rating when those three were on the bench.

Overall, this shows that the Celtics were able to play two styles of basketball: the five-out spacing with the starters, and the gritty possession battle with the bench. And when you mix it all, you end up with a great regular season. This also demonstrates that winning the possession battle is important—but not always enough to win. Also, note that a player’s impact on the possession battle depends heavily on context and role optimization. Today, Kornet is one of the best players at helping his team win that battle. Earlier in his career, in a different context with different assignments, he was one of the least impactful players on that end.

If you want to read more about the possession battle across the NBA, feel free to check out my recent Twitter thread on the topic:





Source link