Interview: Dean Oliver on His New Book ... and What's Next

The NBA analytics godfather recently returned to ESPN to develop basketball metrics

Dean Oliver’s eventful 2024 has included a reunion with the ESPN Analytics team he helped establish more than a decade ago — and the publication of his latest book, Basketball Beyond Paper: Insights into the Game’s Analytics Revolution.

---

Here at 🏀5x5, Dean previewed the new book last last year and followed that up with a series of NBA pieces written from the perspective of basketball analytics … and beyond.

Today Dean is back for a Q & A about the book and what he’s working on now at ESPN:

Royce: This book is less about the math of basketball and more about how to understand basketball analysis and blend it with the personal, emotional side of the game. What’s an example of that blend?

Dean: I wrote Basketball on Paper 20-plus years ago to illustrate how to better measure the performance of teams and players. I wrote Basketball Beyond Paper to illustrate how to change the performance of teams and players.

The distinction is important. I had a belief more than 30 years ago that players were pretty constant except for how they aged, getting better early on and getting worse as their career came to an end. What I experienced working for NBA teams was that this belief was often challenged, not necessarily by my boss or by coaches, but just by watching closely and listening to what was happening.

For example, I was in Denver when Carmelo Anthony wanted to be traded away. I saw how that affected all his teammates as they heard their names constantly in trade rumors. I saw how much better they got when the trade finally happened.

Royce: What else is different about your perspective since Basketball on Paper?

Dean: In recent years I worked on the Wizards coaching staff and not only saw but felt the need to improve our players. I got a lot of credit for helping Kristaps Porziņģis restore his game in Washington. My role was only part of it, but I got to use data to look at the things that were making him worse, think about what was possible to change, then communicating with him.

That kind of thinking was only remotely in my head when I wrote the first book.

To do all of that, there was a lot of math involved and there was a lot of complex data — the new tracking data that was introduced a decade after my first book. I explain some of how I thought about that data in the new book, but the math of how to use it — this wasn’t the book for that. I wanted this book to tell stories of how to blend the analytics with the personal relationships.

---

---

Royce: What insights have you developed about the role of relationships in NBA analytics?

Dean: Mark Warkentien was the general manager in Denver and he gave me progressively more responsibility as time passed, but almost no one keeps a job long in the NBA, so when his contract was allowed to expire, he and I both had to move on. Math could do nothing about that. Better relationships potentially could have. One could be “on the side of truth” when it comes to analytics, but that doesn’t mean that the new boss is going to let you practice that, not without getting to know each other, getting to trust each other.

Relationships are critical not only in getting analytics implemented, but in coaching. Coaches spend a lot of time thinking about the psychology of the game. I asked a good number of coaches how much they thought about psychology and the smallest number was 40%. Managing the ego of the best players and the anxiety of the role players takes effort.

No one told me that when I started this work, but I was around it constantly on the Wizards coaching staff. Analytical tools exist to help with that — emphasizing what stars do poorly and role players do well, for example — but they’re just one part of the tool box.

Royce: Starting in 2004, you worked with the Seattle SuperSonics, followed by stints with the Nuggets, Kings and Wizards. If I ask you what statistical secret sauce you’ve discovered over the past couple of decades, what comes to mind?

Dean: Analytically, one of the most useful tools for me has been the player metric called net points.

If a team lost a game by seven points, net points allocates that seven to different players and can explain why (to varying degrees, depending on the quality of the data). Being able to say that DeMarcus Cousins was bad in a game because he committed three offensive fouls, that provided some way to effect change in how he played. It could illustrate that, without just one of those offensive fouls per game, we would have won an extra three games in a season — that drove home a point.

---

---

Royce: You recently rejoined the ESPN Analytics team to focus on NBA metrics. What are you working on?

Dean: We’ll be putting out net points at ESPN in January. It won’t be the fully-detailed version because there is work to bring in all the data that the NBA provides, so it will get better with time, eventually accounting for play types, matchups, help defense, and more. The January version will have estimates for those things and it will be good, just not where it will end up.

That will say who has played the best, not necessarily who will. It can do it at a game level or at a season- level or at the level of clutch time or when Steph Curry is in the game or in transition. That ability to break down net points to different situations can lead to better predictions, for sure.

Predictions are good not only for a fan’s gambling portfolio, but also for making the right prescription to improve a player, which gets us back to what the book is about — the ability to change performance.

Royce: You played the game at the college level, and you’ve played a lot of pickup basketball as well — you and I have played ball together. How has your playing experience intersected with your work?

Dean: If I knew 40 years ago what I know now, my own basketball game would have been better. I played through my sophomore year in college and never had a jump shot, but there is no doubt that analytics would have made me a better player. Just measuring performance better has helped make a lot of current players better.

It’s not going to heal the knee and ankle problems I’ve had, but maybe the next book is on the analytics of injuries.