Artificial Intelligence in Sports

The film Moneyball, starring Brad Pitt, depicts one of the first success stories in applying data analytics to sports: in 2002, the Oakland Athletics baseball team, after years of poor results, qualified for the playoffs and achieved 20 consecutive wins, guiding their game strategy and signings based on mathematical analyses performed with data collected for that purpose.

The world of sports is closely linked to data. Traditionally, it has been one of the sectors where data has been most extensively used. Due to its everyday nature, we might not have noticed how abundant statistics are in sports reporting. For example, we know how many times a footballer has shot on goal in a match, but not only that, also how many times they have done so in the last ten years against the same opponent, how many fouls they have committed in the last five seasons, whether they have touched the ball more in the League than in the Champions League, and any other curiosities we might want to know.

Big Data and Artificial Intelligence in Sports

However, something has changed. The emergence and democratization of Big Data and Artificial Intelligence now allow for immense volumes of data to be collected, unmanageable without the aid of technology. And when we say unmanageable, we refer to hundreds of thousands of data points in a single match, for example, in team sports like football or basketball.

The automation of data collection, increased storage capacity, and the algorithms used for processing have made it possible to improve the information obtained, increase the speed and quantity of input acquisition, and enhance the precision of results. What does this mean? Minimizing the time dedicated to data collection and transformation, and gaining time for analysis and decision-making.

And where does this data come from? From cameras, sensors, lasers, wearables… which collect positions, speeds, movements, jumps, runs, touches, serves, acceleration, points won and lost, and everything else we can imagine. Once the data is collected, the magic of artificial intelligence and algorithms searches for patterns based on defined needs, transforming them into easily interpretable information to facilitate the analysis work of technical teams. Decisions are no longer based on subjective factors derived from human perception, but on objective results stemming from data science.

Benefits for Coaches and Athletes

To move from theory to practice, let’s look at two recent examples. Three-time world champion and Olympic badminton champion, Carolina Marín, has based her strategy, technique, and physical preparation on advanced analytics for 15 years, and given the results, the gamble, at a time when few spoke of the importance of data, proved successful. Sevilla Fútbol Club is another Spanish sports example linked to data science. It has a Research and Development department (with a budget of 500,000 euros in 2020), which analyzes the game and performance of its players, and evaluates new signings through statistical parameters.

As we can see, the goal is to gain efficiency and optimize available resources, thanks to profound, simple, intuitive, and neutral knowledge that helps coaches to:

• Improve athlete performance by detecting aspects of their work that need reinforcement.
• Prevent injuries through postural and overexertion analyses.
• Evaluate signings with objective assessments of athletes’ performance, their potential contribution to the team, their return on investment capacity, or cost-effectiveness.
• Analyze opponents to approach matches with in-depth knowledge of the adversary.
• Monitor matches in real-time to modify tactics or game strategies.
• Coordinate players with complementary characteristics to build balanced teams.

From the athletes’ perspective, it was recently reported that De Bruyne hired data analysts to assess his future at Manchester City and decide whether to renew his contract with the English team. He ultimately signed, so data science predicts titles for the team.

As we can see, data is now as important as training in sports. Professionals and matches are prepared based on information obtained through the use of Big Data and Artificial Intelligence in any discipline. And the results are highly profitable, as the data-sports relationship is becoming increasingly widespread. Sports analytics generated over 920 million euros in 2020, according to Grand View Research, a figure expected to multiply in the coming years.