Seth Partnow, author and former Director of Basketball Research for the Milwaukee Bucks, spoke with FanSided about his new book, The Midrange Theory.
Last month saw the release of Seth Partnow’s new book, The Midrange Theory. Throughout the book, Partnow, the former director of basketball research for the Milwaukee Bucks, explores the evolution of basketball over the last two decades and considers a number of questions about what it means to build a more successful and efficient basketball team.
It’s a consistently interesting book that readers new to the world of analytics will find accessible and those already well-versed in it will still find enlightening. I recently spoke with Partnow to further explore some of the questions discussed in his book and together we spoke about aesthetics, how to analyze decision-making, and what the NBA’s next trend could be.
FanSided: You begin your book by talking about what basketball analytics is not: what misconception about analytics would you most like to dispel?
Partnow: There’s a big and a small answer there. The small answer is that analytics means “shoot more 3s.” The big answer is that analytics is often seen as a reference material that’s apart from basketball rather than an approach to basketball that is based in data, but still is basketball. It’s not something different. That may sound like splitting hairs, but it’s less about the answers than it is the questions.
FanSided: What surprised you most during the course of your research?
Partnow: Since what’s ended up in the book is a career’s worth of research, you’re asking me for the most surprising thing I’ve ever found looking at basketball. One of those things that showed up in the book is, from a box score perspective, how similar Jimmy Butler and DeMar DeRozan have been over their careers. That was fortuitous for writing about the non-box score impacts that show up in the book.
Another big one, and this led to a lot of the discussions that led to the title chapter, is how constant the frequency of jump shooting in the NBA has been and how much the 3-point revolution is just the movement of certain kinds of jump shots to different areas of the court. That was not what I expected when I first started researching and you can see the notion that it’s a guy standing off the ball catching and shooting, that still hasn’t penetrated the broader discourse of the 3-pointer, shot selection, and the supposed death of the mid-range and so on.
FanSided: What bit of conventional NBA wisdom annoys you the most?
Partnow: I think the consistent one that annoys me the most — I don’t know if it’s conventional wisdom as much as a strategic thing — is “you don’t need a 3 here.” A concept that is constantly battled against by most people coming at sports from an analytic perspective is the bias towards losing slower. That’s probably the biggest one.
FanSided: Do you ever worry about a homogeneity arising throughout NBA thinking as analytics becomes more and more prominent?
Partnow: Is it something I worry about? Yes. Is it something that I think is happening now? No. I think a big part of doing this deeper analysis of the game is accurately describing what is happening. If we just look at “x number of shots were 3” and “every team is shooting 3,” therefore everyone is playing the same. If that’s all you look at, then superficially, yes. In the decade when the three came into prominence, there was a wide spread in team shot selection. Some of the early adopters, the Seven Seconds or Less Suns and the Moreyball Rockets, gained an advantage through shot selection and that gap has closed again, but I think that any deeper examination of how those shots occur illustrates the variety of styles.
But that doesn’t mean the worry about aesthetics is misplaced. That’s something that always has to be front and center, but I worry that by misdescribing what is happening you can prime people to see sameness when there isn’t any. The idea that two players who positionally play similarly and wind up with somewhat similar stat lines, like Nikola Jokic and Giannis Antetokounmpo, the notion that they have any real similarity in style is ludicrous. But if you reduce it to a certain level of abstractions — this guy dunks a little more, this guy passes a little more — that’s too reductive and it leads us to seeing sameness when there isn’t any.
FanSided: So even though the results may look the same in the box score, the methodologies, the ways the Jazz and the Warriors get their 3-pointers is very different.
Partnow: Yes. We have the tools to better describe what is happening and by better describing, I mean more akin to what we’re seeing. A lot of the stuff that comes out of tracking data and play charting are measures of style as much as achievement. How a team gets there, that’s the art of it. It’s not “oh they shot a 3 in this possession.” It’s the game that we’re experiencing. I think it’s the third chapter of the book where I start by describing one game in exhaustive detail. That’s the game we see and if you reduce it to the base level of stats, that 20 seconds of 10 players dancing, gets reduced to six words in a play-by-play. If it was just a ticker tape of those six-word descriptions, nobody would care so it’s the other thing that’s important. I think we have the tools to better describe that even if we do have to simplify it some because, again, as I talked about in that chapter you can’t describe an entire NBA season at that level of detail and expect anyone to take it all in.
FanSided: In the introduction, you talk about how imperative it is for the game to be aesthetically pleasing; do you think the league is in a good place in that regard and what should the league keep an eye on moving forward to make it better?
Partnow: Broadly speaking, I think the game is in a good place and that the rule changes from last year, reducing some of the more egregious foul-seeking behavior [have been good]. I think foul drawing is a skill, but there’s a line between clever foul drawing and, for a better term, bulls**t foul drawing and the league has done a much better job of recalibrating which of those is being rewarded with free throws this year.
There are always tweaks that can be made. As far as what the NBA should be on the lookout for, one thing I worry about is, as the league has gone very pace and space, that you can get in a situation where there aren’t really in-game dynamics. By that I mean there’s a passage of play where the game is going slower and possessions take a little longer and maybe there’s a couple whistles. Then there are periods where teams are running up and down and shots are going up in six seconds every possession. I think the game is more enjoyable when there’s a flow between those types of periods in a game. Without those dynamics, it might be a little deadening to the viewer. The analogy that always struck me is one that Jimmy Page, the guitarist of Led Zeppelin, always used to talk about. He tried to balance light and shadow in his playing and I think that’s ephemeral and hard to define exactly but I think it paints a mental picture of what I’m getting at.
FanSided: You spend a good amount of time talking about why an individual’s defensive ability is so hard to evaluate: what would make it easier and is it possible?
Partnow: I think it’s possible for it to be better. As we develop better descriptions of style, and maybe move towards description and analysis of decision making, we can start to see the range of options and then the options not taken by an offensive player. The big challenge of measuring defense is how you count the things that didn’t happen. Well, if you start to know the possibilities of what could have happened in a given situation then you can maybe start to intuit something about how an individual defender has affected the decision-making of his opponents or the ability of his opponents to do certain things. And that starts to get us much closer to how a player, specifically a perimeter player, is impacting a game. It’s substantially easier to measure the individual impact of interior players. Defending shots at the rim is a demonstrable skill that is measurable. Even without tracking data, you can see the impacts of top rim protectors on opponents.
FanSided: You talk about analyzing decision-making — what does that look like?
Partnow: It’s complicated. With tracking data, you can start to identify possibilities of what could happen. You can see counterfactuals of other offensive actions and a lot of assumptions have to go into it but you can start to look at basic categories and obviously, it’s reductive to say in this spot he could [have done X]. When a player is running a pick-and-roll, he can shoot, he can pass to the roll man, he can pass to someone else. Those are the only three options. It’s a simplification, but we can see how often and in what circumstance a player makes each of those decisions. Obviously, it gets very complicated as you bring in the range of possible things that can happen on the floor, but that’s so much more than we could know about how a player played across an entire season than we could before.
FanSided: Yeah, because there are obviously those moments where you see someone make a bone-headed decision. It’s obvious they shouldn’t have done that but it’s more granular when you’re dealing with hypotheticals and asking what would have happened if he passed to the roll man instead of pulling up.
Partnow: Well, and the other part is he made a boneheaded decision that time. How often does he do that? How often does the league do that? The best players make dumb plays sometimes and the dumbest players sometimes make brilliant plays. The dumbest players at the NBA level, so among the 450 best in the world at doing this, he’s in the bottom quartile of that. So even then, obviously they’re going to do some things very well.
Frankly, I probably shouldn’t have even used a basketball analogy. The work I’ve seen done on it that’s more like in football, the quarterback has five receivers and he threw to that receiver at that time. What else could have happened on that play given where everyone was? That’s a pretty straightforward example and there’s all kinds of instances in basketball where you can approach that. I don’t think it’s ever going to be as clean as you could probably do with a football play because that’s the nature of the game and the difference between the two sports. But we’re not looking for perfect; we’re looking for better and I think we can get a better idea of who’s making good decisions, and once we have that, we can start to have an idea of which defenders are impacting those decisions.
FanSided: Were there any players your opinion changed on as you researched this book?
Partnow: I wrote the book pretty quickly, writing it in earnest took about six months and I tend to be a little sticky in my opinions on players so for most guys I’m not gonna change my opinion that much on them over the course of two-thirds of a season. So, in that sense, not really. In terms of players whose opinions have changed through research, Jrue Holiday is one of them. A guy who, early in his career, when he made an All-Star team with Philly, he was a guy who had the ball all the time and played huge minutes and took all the shots for a mediocre team and put up big stat totals without necessarily adding anything to his team. And then, over the bulk of his career, this is a guy who pretty massively impacts winning on both sides of the floor. It’s hard to say, did my opinion change? Did he get better? Probably some of both but that’s certainly one player that comes to mind where my opinion of them is very different from what it was earlier in their career.
FanSided: Now that practically every team in the NBA shoots a ton of 3s, what do you think the next strategic gambit to take the league by storm could potentially be?
Partnow: I don’t think there’s gonna be anything nearly as sweeping as 3-pointers. I know Daryl Morey has talked a lot about how he thinks offensive rebounding is a potential inefficiency. I agree with him, I think that’s gonna be a far lesser impact than going from few 3s to many 3s was. Part of that is that it is very situational. I am absolutely convinced that there are many situations in a game where NBA teams are leaving money on the floor by conceding rebounds to the defensive team. I think that’s very dependent on what the situation is when the shot goes up in terms of where the shot goes up from, whether the defense is in rotation or not, where the other offensive players are, all those other things. The tricky part is figuring out what some broad rules are for additional players to go to the glass basically. You can study it in great detail, but for it to be something that a coach could implement you’re gonna need some pretty broad rules that players can go to at game speed. And I think that it’s hard to come up with a 3-pointer that would be as damaging to a team as a hopeless offensive rebound chase can be.