In the NBA’s modern era of pace and space, small ball, and chucking away from three, it feels like there’s no more place for the lumbering 7-foot center who used to be the backbone of the league. But the burgeoning field of defensive analytics shows that this “dinosaur” might not be going extinct just yet. Ben speaks with Ivana Saric, data scientist for the Philadelphia 76ers, about how defensive analytics are changing pro basketball and the roles of the people who play it.
Research Updates From MIT SMR
Get weekly updates on how global companies are managing in a changing world.
Please enter a valid email address
Thank you for signing up
Ben Shields: Late in Game 4 of the second-round matchup between the Bucks and the Celtics this NBA postseason, Boston’s Jaylen Brown decided to attack the rim on a fast break. Though Boston had dropped the previous two games in the series, the Celtics had a surprising advantage in the points in the paint battle against Milwaukee’s top-rated defense, so this seemed like a sound strategy. But as Brown entered the key, he encountered the man that Milwaukee fans have lovingly dubbed Splash Mountain. Brook Lopez has earned this moniker thanks in main part to his emergence as a major shooting threat from downtown. But at 7 feet tall and 270 pounds, he certainly is a mountainous presence down low. Backpedaling toward the rim, Lopez got in the way of the drive and forced Brown to alter his shot.
Paul Michelman: The only problem, waiting behind Brown was none other than the “Greek Freak” Giannis Antetokounmpo. In his first season under head coach Mike Budenholzer, Giannis had unlocked his full defensive potential. With Lopez protecting the rim, Giannis could act like a 6-foot-11 Ed Reed, roaming the court in anticipation of a big play. Giannis swatted Brown’s shot attempt from behind, the ball careening off the backboard and igniting a fast break the other way, the Bucks eventually taking the game and the series. That play was a microcosm of the Bucks’ rise from the middle of the pack to the defensive elite. While Antetokounmpo gets the headlines for his highlight reel plays, it couldn’t have happened without Splash Mountain standing in the paint. I’m Paul Michelman.
Ben Shields: I’m Ben Shields, and this is Counterpoints, the sports analytics podcast from MIT Sloan Management Review. In this episode, we’re getting into our stance to look at defensive metrics and how big men like Lopez are still a valuable commodity in today’s NBA.
Paul Michelman: The big man is dead. Long live the big man. In the NBA’s modern era of pace and space, small ball, and chucking away from three, it feels like there’s no more place for what used to be the backbone of the league — the lumbering 7-footer. But in the burgeoning field of defensive analytics, what was thought to be a dinosaur might not be going extinct just yet.
Ben Shields: It’s easy to fall in love with the offensive revolution, but players like Rudy Gobert and Marc Gasol are helping their teams to success — not to mention playing their way into big contracts with their importance on the defensive end. And as teams ramp up their efforts to slow down high-scoring offenses, their value could still be rising.
Paul Michelman: In today’s interview, Ben speaks with Ivana Saric, data scientist for the Philadelphia 76ers, about how defensive analytics may be changing the way we look at the game.
Ben Shields: All right, it is a pleasure to welcome onto the show, Ivana Saric, who is a data scientist for the Philadelphia 76ers. Ivana, thanks so much for joining us.
Ivana Saric: Thank you so much for inviting me — it’s my honor to be here.
Ben Shields: Well, likewise. And we’re going to talk about this thesis of: Rim protectors still play a big role in the modern NBA. But before we do so — because part of our aim here with this podcast is educational — I wanted to know if you could give us some context on the rise of defensive analytics and specifically why defensive analytics have lagged behind offensive analytics to date in the NBA.
Ivana Saric: So for many years, we only had the box score data, and for offense that includes shots, rebounds, and assists. But for defense, it includes defensive rebound steals, which I [don’t think] is really a good indication of defensive performance. Well, you can argue that it’s more important to box out your opponent and keep them from getting offensive rebounds than which defender actually gets the rebound. So the defensive rebounds are not really that indicative. You can also argue, for the steals, that in some cases, players who get a lot of steals are also the ones who gamble a lot, and they get beat for it. And the last thing, for the block shots, you can say that players who get a lot of block shots are also the ones that are defending a lot of shots. So it doesn’t really tell you how good they are, even at blocking shots, but it tells you even less about their defensive performance before the shot happens. So until we had the player tracking data — which is from 2013, for the whole league — we couldn’t tell much about the defensive contributions of players, compared to offense, where we had a lot more metrics that are closer to telling players’ offensive efficiency.
Ben Shields: So that’s really interesting, because you have these counting stats like rebounds and steals and blocks that don’t really tell the whole story of a player or team’s defensive ability. Now, here comes player tracking data. It gives us a lot more information to better understand defense. Now, I was wondering if you could share with our listeners what some of the most important or relevant defensive metrics are today, especially in light of the player tracking data.
Ivana Saric: So on the team level, defensive rating has been widely used. But if you look at it on the player level, you would look at defensive rating of the player on the court. Of course, this metric is far from perfect. For example, if someone plays with the same teammates a lot, this rating will be affected by his teammates. Another sort of metric that’s been around for some time is a defensive shot chart. So it’s similar to the offensive or the regular shot charts that we’re used to, which include shots and efficiencies from different spots on the floor. Defensive shot charts contain your matchup shots, so you can see where the player’s matchups are taking shots from [and] how well they make them against the player. But again, we gain no information on the defense before the shot. But at least we can tell if the opponents are able to take good shots against this player or not.
Ben Shields: OK, so that’s helpful context, because I want to now dig into this thesis around rim protectors and the fact that they still play a big role in the modern NBA. And we do have some listeners out there that may not be as familiar with the ins and outs of basketball. So if you can — can you tell us what a rim protector is and why they’re important in the game?
Ivana Saric: So a rim protector is usually a big center who is, I guess, protecting the rim. It’s a player who tries to keep opponents from taking shots at the rim, which is a high-value shot, or from making those shots at a high percentage. So against a good rim protector, you would expect the opponents to make those shots at a lower rate. So the rim protectors are still important, because the shots at the rim are so valuable. Players will shoot them at 65% — so they’re even more valuable than 3-point shots, which players shoot at 37%, which is equivalent to 56% from a 2-point shot.
Ben Shields: So that’s interesting you mentioned the 3-point shot here. Because one of the reasons why I find this thesis so fascinating is because we are in this era of the 3-point revolution, where seemingly every NBA team is trying to upskill to shoot and make the three.
And yet, you’re making this point that bigs, and specifically rim protectors, still are important assets even in this era of small ball and 3-point shooting. So what does the data say as to why rim protectors are still so important in the game today?
Ivana Saric: So if you look at the shots at the rim, on average, players make them around 65%. So if a player takes 10 shots at the rim, an average player will score 13 points out of those 10 shots. So compare that to the 3-point shots which players shoot at around 37% on average. So if they take 10 threes, they will score 11 points. So you can see that the rim shots are still more valuable than the 3-point shots for an average NBA player.
Ben Shields: And in this era of trying to take the most optimal shots that will get you the highest value, I can see why the layups or shots close to the rim are still absolutely critical. And you’ve got to have a rim protector there to stop that from happening. Now let’s get into an example or two: What teams in today’s NBA are doing rim protection well and why?
Ivana Saric: So in order to be a good defensive team, you can do this in many different ways. The Milwaukee Bucks did it this year, and they were the top defensive team in the league. They were allowing 105 points per 100 possessions, and the league average is 109 points. One thing they did is they allowed the least shots at the rim. So out of all of the opponents’ shots, 31% of the shots were at the rim. And if you look at the second team in this category, it was the San Antonio Spurs, and they allowed 34%. So from Milwaukee to the second team there was a big difference. The opponents also made the rim shots at the lowest percentage — at only 54%. On top of that, Milwaukee was also the top defensive rebounding team this year. So if you [prevent] your opponents from getting offensive rebounds, you’re [preventing] more shots at the rim.
Ben Shields: How do teams defensively keep players out of the paint, away from taking those high-value shots?
Ivana Saric: So one way to keep your opponents away from the rim is to have a good rim protector, so that they would not attack once they see this person there. Another way is if your perimeter players don’t get beat a lot — either by being able to contain the drives or by creating turnovers and [in] that way eliminating shot attempts from another team.
Ben Shields: So I want to get into this question about — not to always be bringing it back to offense, because we do need to give defensive analytics its moments — but what are teams to do if they have bigs that protect the rim well but aren’t so effective on offense? What are the trade-offs involved in terms of having a great rim protector who may not be the best offensive player?
Ivana Saric: Well, when you’re on defense, you don’t really have a choice of what the opponent will try to do — if they’re going to shoot point threes, if they’re going to try to get to the rim. But on the other end, when you’re on offense, you can choose how you’re going to run your offense. You can choose to run it through different players. So if you have a big that is a really good rim protector but not as skilled on the offensive end, you will not run your offense through him, and you can use him as a roller or as an offensive rebounder.
Ben Shields: How do teams protect the rim when they don’t necessarily have a 7-footer, if they don’t have someone that fits that prototype? Are there ways for teams to get around maybe some physical limitations of their personnel in order to protect the rim from a team strategy standpoint?
Ivana Saric: If the teams don’t have a big, you would assume that players are more mobile, so they’ll be able to play more aggressive defense with a lot of rotations and try to create turnovers to get advantage in a different way.
Ben Shields: What would you say is probably the most relevant or important metric for measuring effective rim protection? Is it the metric that you shared earlier around the field goal attempts around the rim, or is it something else? Is there [a] most important metric that we should keep in mind in evaluating effective rim protection for teams?
Ivana Saric: Oh, well it’s really a combination of how many shots teams take at the rim and how well they make them. So you can really translate that in terms of points at the rim per 100 possessions. So it includes how many shots they’ll take at the rim and how well they’re going to make them, in one metric.
Ben Shields: Well that’s great. That’s very helpful. And especially for our listeners to evaluate their own team’s rim protection, you’ve given us a nice metric to take a look at. All right, because defensive analytics are increasingly having their moment in the league, I wanted to talk to you about some broader trends. It does feel like we’re scratching the surface in our understanding of effective defenses. So especially from a fan standpoint, Ivana, how do you think this line of analytical work will evolve going forward?
Ivana Saric: Well, the metrics we talked about so far are really still only measuring how well teams are making shots against this defense. Also, in terms of defensive shot charts, we can see where the shots come from and what [are] their efficiencies. But [with] both of these metrics, like defensive rating, it’s still difficult to interpret which aspects of defense are responsible for it. Is the team or the player or the lineup either good or bad because of the shot defense? Or because they force turnovers? Or because they keep their opponents off the glass? So we worked a lot as of late in terms of understanding what actions create offensive advantages and how to better stop them. But there’s still a lot of room to grow.
Ben Shields: Yeah. And that makes it very exciting. And you know, you mentioned, of course, the player tracking data, which has been instrumental in the evolution of defensive analytics. What data do you not have today that you wish you had?
Ivana Saric: The player tracking data really just gives us player locations in terms of X-Y coordinates, but we know nothing about their body positioning — for example, which way they are facing, where are their hands, how are their feet positioned. So there’s still a lot more data that we could get that will help us evaluate the defense.
Ben Shields: Well, and you’re at the forefront of that, Ivana. And it’s going to be very exciting to see how this field evolves going forward. Before we let you go, I did want to ask you a little bit about your background and skill set, because what you do is fascinating. And I’m curious, how did you get your job? And any advice for people trying to break into sports analytics roles?
Ivana Saric: I played basketball since I was 7 years old, and I played at NJIT during my college years. And after college, I decided to continue my education, and I got a PhD degree in computational fluid dynamics. And then in the middle of my graduate school, I realized the NBA teams are hiring data scientists to look at this player tracking data. So I started learning data science in my free time, and then I just applied for the job when it was open. But I would say everybody has a different path to how they get positions like this. But there’s definitely a minimum [amount of] technical skills that is required. And I would say a passion for basketball is very important, and also being able to translate mathematical terms to basketball — and the other way around; to translate basketball into terms that we can form mathematically.
Ben Shields: Yeah, that point you make about translation sounds particularly important. What do you focus on in your own communication skills to ensure that the analysis that you’re doing gets through to all the stakeholders that you’re working with?
Ivana Saric: Yeah, I would say communication is just as important as technical skills, if not more important. You can do the best analysis, you can have the perfect model, but if you can’t communicate it to people for making decisions, it really makes no difference. So communication is really, really important. And being able to translate technical terms into normal human language is really important.
Ben Shields: Well, and on that note, we appreciate you coming onto our show and translating the rise of defensive analytics and its importance to the game, and specifically the role of rim protectors in the modern NBA. You made it clear and interesting and compelling, and we appreciate your time.
Ivana Saric: And thank you so much for inviting me.
Paul Michelman: This has been Counterpoints, the sports analytics podcast from MIT Sloan Management Review.
Ben Shields: You can find us on Apple Podcasts, Google Play, Stitcher, Spotify, and wherever fine podcasts are streamed. And if you have an idea for a topic we should cover or a guest we should invite, please drop us a line at firstname.lastname@example.org.
Paul Michelman: Counterpoints is produced by Mary Dooe. Our theme music was composed by Matt Reed. Our coordinating producer is Mackenzie Wise. Our crack researcher is Jake Manashi, and our maven of marketing is Desiree Barry.