Remember Moneyball?
In 2002, Billy Beane’s Oakland Athletics team revolutionized baseball by minimizing subjective human input through the use of advanced statistics. This approach dumbed down the game into one that sacrifices traditional excitement – stolen bases, hit and runs, and pitchers going deep into games – for home runs and strikeouts in the name of “optimization.”
Since then, other American sports, most notably basketball and football, have undergone a similar transformation that drains character from the games. In basketball, physical play in the paint is essentially gone. Every team in the NBA now shoots more threes than they did in the entirety of the 2015-2016 Golden State Warriors season, who set the then record. In football, offensive and defensive coordinators spend more time staring at spreadsheets than the field. But is soccer, the beautiful game, immune to this homogenizing strategy that makes other sports feel duller?
Almost all professional soccer teams are now looking to xG (expected goals), xA (expected assists), and models that evaluate players on positions they take up throughout the game. More sophisticated teams have developed their own means of gathering statistics.
Liverpool, for one, hired theoretical physicist Ian Graham, who created a model that determines how much each player’s touch of the ball increases their team’s chance to score. This metric has been key for Liverpool – it inspired them to acquire Sadio Mané and Mohamed Salah, who are now widely regarded as some of the best players in the world.
Leicester City, a very small English team with similarly advanced analytics, identified a number of players before they’d proven themselves (eg, N’Golo Kanté, Riyad Mahrez, Ben Chilwell, and Harry Maguire), signed them, and then became the longest-shot Premier League champions in history.
However, teams struggle to incorporate soccer’s intangible factors into their models: tactics, the different functions of positions on different teams, opponent quality, leadership skills, team chemistry, and whether players will adapt to a new culture. While the leading three context factors may be solved by more advanced models, the others require humans to do more traditional scouting because every person is unique and builds relationships differently.
Take Paris Saint-Germain’s super team of Messi, Neymar, and Mbappé. Despite their talent, they couldn’t win the Champions League because the team just didn’t gel well. Messi didn’t enjoy Paris, Neymar played selfishly, and there was never strong chemistry. No model would’ve suggested that picking up the greatest player would hurt PSG, showing that statistics alone cannot predict success. Interestingly, some journalists predicted that the lack of pressing (putting pressure on defenders) between PSG’s stars and a culture shock for Messi could prove problematic.
While the success of teams like Liverpool and Leicester City shows that statistics can help teams identify the basic building blocks of success, analytics can’t define the relationships and flashes of genius that make soccer great. Soccer is a game that is too emotional and too fluid to be reduced to cold numbers. A back-heeled pass to create space or a run that draws a defender away from the ball might not be properly reflected in an algorithm, but can be decisive in the game. Elite managers understand this and have realized that genius is in defying the numbers, not relying on them.
Soccer is a game that excels in the unexplainable, and no computer program can ruin that. While fans of baseball barely recognize their game, basketball fans are getting bored of four quarters of a shooting contest, and football fans are mourning the loss of a “classic football game” full of tough hits and big runs, soccer fans are still enjoying the beautiful game.
Soccer is immune to advanced statistics because its soul lies in the immeasurable.
