Analytics Guides: Special Edition—Mark Cuban on the Mavericks’ Pioneering Use of Analytics
How analytics propelled the Mavericks and reshaped the league
This is the third volume of our Analytics Guides series, but this one’s a bit different—and special.
The Analytics Guides are a feature for paid digginbasketball subscribers where I dive into the history of NBA analytics, explore key data sources, and break down advanced concepts. The goal? To help you get the most out of our season coverage by making the stats and trends I’ll be sharing easier to grasp—and a lot more fun. Hopefully, you’ll also learn something new along the way.
Originally, this piece was intended to be a paid feature, but I’m making an exception this time. Why? Because I want more people to learn about this fascinating chapter in Mavericks history. And because Mark Cuban generously agreed to share some behind-the-scenes stories from that era with us.
I hope you’ll enjoy it, and if you do, consider supporting. These take a lot of time and effort to put together, and your support makes it all possible!
In Volume 1 of this series, we explored the early days of analytics, spanning from the 1950s to the 1990s, when Dean Smith pioneered possession-based analysis, laying the groundwork for modern basketball stats.
Volume 2 took us into the play-by-play era, which began in the late 1990s and opened up a whole new level of analysis. We introduced Dean Oliver’s Four Factors as the best way of systematically breaking down the game.
Today, we’ll focus on the next evolution: the 2000 to 2010 period, when the Mavericks—under their new, energetic, and unorthodox owner—emerged as the league’s analytics trailblazers, a journey that ultimately helped pave the way to their 2011 championship.
One other change from Volume 1 and 2: because of the special Q&A with Mark Cuban, we won’t be introducing a new key concept in this article (although WINVAL, one of the first advanced metrics, will be mentioned). Instead, we’ll dive deep into shooting—the most important of the Four Factors—in one of the next standalone Volumes.
From Underdogs to Innovators: Mavericks’ Decade of Data-Driven Change
As covered in Volume 2, the introduction of play-by-play data in the NBA during the 1996-97 season was a monumental shift. It meant teams could track not just the final box score, but every possession, substitution, and event in a game—providing a more detailed understanding of how basketball was played. This enabled analysis beyond traditional stats, allowing teams to study lineup combinations, player impact, and game flow—ushering in a new era of strategy and evaluation.
Seeing the potential, a new wave of analytics pioneers emerged, driven by the idea of using data to create an edge in basketball. Many were inspired by the growing field of sabermetrics in baseball, but the concept still drew plenty of eye rolls from traditional basketball minds.
That timing coincided with a turning point for the Dallas Mavericks franchise. In 2000, Mark Cuban, a young, tech-savvy entrepreneur, bought the team and was eager to use any innovative method that could give his new team an edge. That same year, Cuban ran into Wayne Winston, a decision sciences professor at Indiana University, at a Pacers-Mavericks game. Cuban, who knew Winston from attending his statistics classes at Indiana University, offhandedly posed a question that would shape the direction of the Mavericks franchise: “Do you have any way to make the Mavs better?”
Later on, Cuban brought Winston and his old MIT friend Jeff Sagarin on board to help transform the Mavericks using their analytical approach. Sagarin, a renowned sports statistician best known for his work in football and hockey rankings, had already spent years applying statistical models to various sports. Sagarin’s ratings became a staple in the USA Today sports section starting in 1985 and were used by the NCAA Tournament Selection Committee as far back as 1984.
Together, Winston and Sagarin began working on a new system to quantify player impact in the NBA. The duo developed WINVAL, one of the first iterations of adjusted plus/minus metrics that would later dominate the NBA analytics landscape. WINVAL and other iterations of adjusted plus-minus are derived using play-by-play data to keep track of all substitutions and possession-ending actions. The system measured a player’s overall influence by tracking how they impact their team’s scoring margin, effectively capturing the full impact of their presence on the court. With WINVAL in place, the Mavericks had a powerful tool that often contradicted conventional wisdom but provided insights that few other teams had access to. One standout feature was WINVAL’s “lineup calculator,” a software tool that measured the effectiveness of various player combinations on a game-to-game, weekly, and season-long basis. For example, during the 2003-04 playoffs, WINVAL showed that Shawn Bradley was significantly more effective against Portland than Sacramento. As a result, Bradley started all six first-round games against the Blazers but saw limited minutes in the following series versus the Kings.
Seeing the impact of this approach, Cuban didn’t stop there. In 2005, he came across another analytics enthusiast, Roland Beech, or rather his website. Beech, who had been involved with various Silicon Valley startups, had just launched 82games.com, a site that pioneered the use of plus/minus metrics and lineup analytics. Naturally, it quickly caught Cuban’s attention.
“Mark Cuban found it within two weeks,” Beech recalled in a 2016 interview. “He was one of the first hundred or so people to find the site.” Intrigued by Beech’s data-driven approach, Cuban brought him on board as a consultant during the 2005-06 season, and Beech quickly established himself as an important figure in the Mavericks’ decision making process.
By 2010, Beech had become the first "stats coach" in the NBA, sitting behind the bench during games and providing instant analytical feedback to head coach Rick Carlisle. His impact culminated in the Mavericks’ 2011 title run. It wasn’t just about using data to identify the best lineups or rotations—Beech’s analysis extended to play calls, defensive adjustments, and mid-game strategy tweaks that helped the Mavericks optimize their approach against each opponent. As Cuban noted after the championship:
“Roland was a key part to all of this. I give a lot of credit to Coach Carlisle for putting Roland on the bench and interfacing with him, and making sure we understood exactly what was going on. Knowing what lineups work, what the issues were in terms of play calls and training.”
The Mavericks’ embrace of analytics set a new standard in the NBA, proving that data could be more than just a tool for the front office—it could be integrated into every facet of coaching and player evaluation. As early pioneers and adopters of these methods in the early 2000s, they paved the way for other teams and analysts to build on and expand these concepts. Figures like Dan Rosenbaum (during the research for this article, I found out he shared a hotel room in May 2004 at the Chicago pre-draft camp with Dean Oliver and Roland Beech) later expanded on the work of Sagarin and Beech with his Box Score Plus-Minus model while working for the Cleveland Cavaliers.
John Hollinger, who developed the Player Efficiency Rating (PER), Justin Kubatko, the creator of Win Shares, Kevin Pelton, who became known for his advanced statistical projections, and Daryl Morey, who launched the MIT Sloan Sports Analytics Conference and became the NBA’s most famous analytics-driven General Manager, were among those who took these ideas and continued pushing the boundaries of what analytics could achieve in the NBA.
Now, onto the promised part—the best way to truly understand how the Mavericks became analytics pioneers is to hear directly from the man who made it all happen.
Here’s my Q&A exchange over email with Mark Cuban, where he shares his thoughts and stories from that era in Mavericks history.
Mark Cuban Reflects: How Analytics Shaped the Mavericks’ Success
When you bought the Mavericks, how much did you know about analytics in sports and the NBA?
A little. But only what I had learned at Indiana.
Your professor at Indiana University, Wayne Winston, was one of your first analytics hires. Can you explain how you came up with the idea of bringing him on board?
When I was a freshman at Indiana University, I snuck into an MBA class for stats that was taught by Wayne. He used sports for all the examples, which made it easy to learn. So, that was my introduction to analytics and Wayne. Fast forward to right after I bought the Mavs, I was at our Pacers game when I heard someone shouting my name. I looked up and there was Wayne.
The crazy thing is that the only reason I recognized him 20 years after I took his class was that I had randomly seen him on Jeopardy! He killed it. So I kept watching. Which led to me recognizing him and that lead to a conversation. And that led to me hiring him.
Winston and his colleague Jeff Sagarin developed a system called WINVAL, which was the precursor to adjusted plus/minus and other advanced stats. Were the Mavs the first to use WINVAL? How did it impact decision-making?
Yes. It didn’t help the first year or two for player selection, but after a couple years the data became more valid. I could see players that had consistently good ratings for impact, plus/minus and more. That led to us using it as an input for player selection.
Where it really had an impact was with best lineups. We got to see what combos worked together and which didn’t. It was a big help when we got to the playoffs. That gave us enough data vs our opponents to have an expectation that it would be accurate.
You’ve always engaged with smart analytics people. Roland Beech was another example. He claimed to be the first “stats coach” to sit behind an NBA bench. How did that come about?
Rick Carlisle wanted as much information as he could get. He was great at synthesizing data from any number of sources. So he wanted Roland behind the bench.
I read that Roland provided lineup advice to Coach Rick Carlisle during the 2011 title run. Did the decision to start JJ Barea come from that? Can you share how that process worked?
It’s hard to say what Rick used to make the final decision. But that was definitely the input.
Which other teams did you see as the most forward-thinking in analytics from 2000 to 2010?
There weren’t a lot. Miami was good. Then Houston got in the mix. Back then analytics were a huge competitive advantage.
Aside from player and lineup evaluation using variations of adjusted plus/minus, what were some other areas where analytics made the biggest impact for the Mavs?
We tried to take data inputs wherever we could create them. One of my favorites was on game shooting arc. The other was tracking shooting percentages in practice and shootarounds. One of the things we did was hire former college players as interns to always be available to help players recreate in game speed and situations. I think that also helped a lot.
How did tracking data change the analytics landscape? I understand the Mavs were one of the first teams to adopt it.
We are always looking for an edge. I think it allowed us to take deeper in game data and look for patterns or situations to give us an edge.
At what point did you feel like the league caught up to the Mavs, and analytics became less of a competitive advantage? Or do you think it still is?
I think analytics have become a very efficient marketplace. We saw the big change probably around the time the last TV deal came into place. That allowed more teams to invest in it.
What do you see as the next big developments in NBA analytics? What’s the next milestone?
Applying AI to game footage and letting it tell us what we are missing.
These analytics articles are great reads! thanks for putting these together
Great piece had no idea you and mark were buds 😅