AI Sports (It’s in the Game)

Technology has always played a role in the evolution of sports — think of the first radio broadcasts of baseball games in the 1920s or the introduction of the modern replay review into the NFL in 1999. In many ways, AI represents the next step in this evolution. We now live in a world where AI seems to be making its way into everything, and sports are no exception. This is an
important moment to ponder the questions these technological innovations pose about the future of the sporting world. Here are three spaces we’re seeing AI integration already.

Scouting and Performance Metrics

One of the most prominent ways AI is being used in sports is in the analysis of player performance, what we might call “Moneyball with AI.” In the same way that Oakland A’s General Manager Billy Beane, as told in the book-turned-movie Moneyball, revolutionized baseball scouting by introducing data and statistics in new ways, the scouting world is now embracing opportunities to leverage all of that data using AI. Uplift is a new tool for scouting and performance optimization, and it is available for baseball as well as other sports. As explained in a recent Wall Street Journal profile of the company, Uplift takes video footage of athletic prospects and translates it into what its creators say are predictions of a player’s future performance and risk of injury.

New technologies like this provide lots of insights that weren’t possible before, and they provide opportunities to help players as well as scouts. However, more complex metrics also have their limitations. One common issue with AI in any context is its lack of interpretability. AI technologies tend to involve complex calculations that are beyond human comprehension. Because of this, if scouts rely more on AI models, in which computers take large amounts of data and spit out numbers without clear explanations of where those numbers are coming from, they will be giving up some of their ability to intuitively understand the factors that predict future success.

Scheduling

A less visible yet still important use of AI in sports is in the scheduling of professional leagues. Imagine the enormous logistical hurdles these leagues face in creating their schedules each year — they have to consider travel distances, the availability of stadiums, the balance of home and
away games for each team, and many other factors. One company using AI to provide solutions to these challenges is Fastbreak, who recently signed a deal with the NBA and WNBA. Their tool is also used by the NHL and the SEC for scheduling.

Aside from alleviating the logistical challenges of scheduling, using AI in this way also has the potential to make positive changes in professional sports in several ways. For example, travel schedules could be designed to minimize a league’s carbon footprint, and the timing of games could be set based on the ideal amount of rest players need to minimize injury. But there are many different factors professional sports leagues could optimize for, so prioritizing environmental sustainability and player well-being will have to be an intentional choice.

Equipment Development

AI has also made its way into the research and development process of equipment manufacturing for various sports. This is probably most notable in golf, where Callaway, a massive golf club manufacturer that makes the clubs used by many professionals, first used AI to help design their drivers released in 2019. Their most recent technological innovation, what they call the Ai Smart Face, feeds a massive amount of real-world swing data into a computer algorithm, which then finds the optimal face shape for the golf club based on that data. Other golf manufacturers have used similar technologies in their equipment design as well.

The baseball manufacturer Rawlings has also begun using a similar technology for its metal bats. According to the company’s explanation of the technology, they used “thousands of different computer simulations before arriving at a completely optimized barrel design.” While this technology has not yet made it into the design of wooden baseball bats used by professionals, Mohammad Mazloomi and Philip Evans, in a 2021 research paper lay out the possibilities of using AI (Parametric Modeling and Genetic Algorithms, to be specific) to design wooden bats. Many wooden bats used by big-league players are handmade to the players’ preferences, so it’s an open question whether hitters at the top of the game would be willing to let a computer algorithm take over a design process that has always been just as much an art as a science.

Too early to tell

There are many ways A.I. is already being used in sports, and the data-rich nature of modern athletics makes it an area full of opportunity for future development in this regard.

However, we are in the early stages of implementing many of these new tools into the sporting world. While it is still human athletes that play the games and humans that will be deciding how the technology is used and what goals it is used to achieve, the capabilities of A.I. increase by the day.