From making Pokémon rankings to working for the most decorated franchise in the NBA

BarcelonaFrom creating rankings with Pokémon toys to working for the franchise with the most NBA titles, Adrià Arbués has realized his dream and has been working for the Boston Celtics as a Basketball Data Scientist for the past few weeks. "I consider myself very fortunate to have made it to the NBA, but doing so with the Celtics is special because of the franchise's history and what it represents. Boston has something that many other franchises lack: widespread continuity. This is quite unique in the world of sports, where there's often volatility in positions, personnel changes, and even changes in management. With this continuity, the tools we've created have undergone many iterations, making them incredibly valuable and tailored to the team's needs. Through time and countless questions and answers, we've created and established a data-driven culture."

Arbués' story is fascinating. "Since I was little, I've loved mathematics because it's an excellent tool for a wide variety of tasks: quantification, classification, prediction, optimization... As a child, I made my own classifications of the Pokémon toys I had, and later, when I discovered my mother's and father's notebooks—both of whom assigned numbers to things—I became a basketball fan, and I filled notebook after notebook with statistics of my favorite players.", explains.

The analytical side came later. "I suppose it was when I was finishing my degree. I already had a solid engineering background and enough experience as a basketball coach, and that's when it clicked. It also helped that, while reading research publications, I discovered the papers from Sports Analytics conferences, where they explained, for example, how going beyond the usual data and dedicating oneself professionally to data science and basketball wasn't a pipe dream," says Arbués.

In 2017, Arbués began a doctoral thesis while working as an assistant coach for Cornellà and, later, in Barça's youth academy. "Four years later, I decided to make the leap into the business world. The timings They played in my favor, and a group of highly reputable American academics had just launched a start-up "I joined Zelus Analytics, a company that aimed to outsource the data science processes for various sports teams as a consulting service. I joined their team as a data scientist, and then I was able to continue doing research with the benefit of having access to a wealth of data and with NBA teams as end clients," he recalls.

After four seasons doing consulting, Arbués felt that chapter was closed, as he was very much removed from the decision-making process with the people he wanted to work with. "The Celtics were my top choice." Sports analytics It's a very attractive field, but job opportunities are rather limited. That's why NBA teams operate more on "opportunistic signings" with people they know and who they know can be a good fit. I was able to meet with the people in charge of the Celtics during the summer, and luckily there was a complete understanding to finalize my signing a few months later," he reveals.

Billions of data points to analyze

Arbués can't go into specific details due to confidentiality issues, but his work is like that of any data scientist, but applied to sports. "In the NBA, there's an avalanche of data per game thanks to technologies that record the games and provide information about the players' positions and body parts 25 times per second. We're talking about more than 2 trillion observations per season, containing all the patterns of what's happening in the world of basketball, but hidden and with their secrets. These are called machine learning or deep learning—and now it seems everything is encompassed within artificial intelligence—that recognize these patterns and from which metrics can be extracted so that the people making decisions within the club have more tools to avoid mistakes, whether it's for making signings or determining how many minutes a player should play in context." The Catalan brings two worlds together. "Possibly, the most important part of my job is speaking both languages: the language of data science and the language of basketball." Being able to make people from different worlds understand each other is a very important task, which you can only accomplish if you have solid experience in both fields and understand the needs and language of each."

"The importance of data in basketball depends entirely on the decision-makers believing in its value and the people who create this data being able to communicate what they mean. Data communication is like a game of telephone where, in large organizations, many people are involved, and anyone who fails in this game causes the value to plummet. It's relevant because we're talking about a tool that can generate a clear competitive advantage at a reduced cost. The clearest example: identifying undervalued players who can contribute much more than their transfer fee allows you to reallocate resources to other areas," Arbués acknowledges.

According to Arbués, "offensive efficiency continues to skyrocket. Teams are increasingly creating better shots for their best shooters in specific situations, and that's thanks to the data. I'm not going to address the arguments that only three-pointers are shot now or that there's no defense in the NBA, because I think those debates are outdated. Today, games that focus on more than just dunks—and the NBA game has more dunks—are much more entertaining than many people realize. Data hasn't killed basketball, far from it; quite the opposite."

Catalonia boasts a handful of professionals working in the NBA. "I think there are three main reasons. The first is that it has a unique density of high-level basketball: there are competitive teams in practically every region. The second reason is that we have a really strong 'upper-middle class' of basketball players, much stronger than we realize. If you're 25 and want to be a coach, you'll sacrifice a whole weekend day to go to an arena and watch games. It's from these days, from getting together with other coaches and sharing perspectives, that you really learn. It gives everyone a very powerful network of contacts to help them forge their own path," he concludes.