In the top sports many times the development of events is completely unpredictable – this is one of the reasons that create such strong emotions in us. Coaches make decisions based on instinct and influence games in a pivotal way. But now artificial intelligence (AI) may change this condition. And many are wondering: Is sport under threat? Maybe with TN he will become more predictable and therefore more unpredictable?

For years, the data collected at sporting events has played a very important role in the analysis of matches – data that has become more and more accurate and extensive over the years. Coaches often fail to take full advantage of all the data. And this is where TN comes in.

“Artificial intelligence is a technology applied by computers to perform tasks that, while requiring human intelligence, are usually too complex for humans,” says Daniel Memmert of the Sports University of Cologne. “Artificial intelligence systems can use data and algorithms to objectively recognize and evaluate game patterns.” This objective assessment is also their difference from the subjective assessment of a coach, as the sports scientist says.

Will people be replaced?

Artificial intelligence has revolutionized data analysis and is increasingly influencing tactics, injury prevention or training management. Thus, the staffing of sports clubs also changes.

People with as much experience as possible in each sport are no longer hired, but more programmers and IT scientists. “In 2017 Barcelona hired 15 data and video analysts,” says Memmert, adding that “the use of artificial intelligence systems will not reduce the number of people involved in professional sports, but increase them.” Because experts who understand the sport in depth are still needed to know how to manage the information offered by artificial intelligence, explains Memmert.

TN in Schalke’s scouting department

Artificial intelligence can be used in many different fields. At Schalke 04, artificial intelligence is used in the scouting of new players. “We have about 40 evaluation categories in the scouting department and in three of them we use the help of artificial intelligence,” such as the accuracy of the players’ passes, says Schalke sporting director Peter Knebel on Schalke’s second channel German television station ZDF.

This way it is easy to see if this footballer is of the same quality as a player who has left the team. But again, the final decision still rests with the scouter and the coaches. Artificial intelligence is just an additional way to evaluate players more effectively.

“Artificial intelligence is like a small child”

Such technologies are also used in handball, where players wear so-called “wearables”, which are like sports vests with sensors, through which all kinds of data are recorded during matches, such as the distances the athlete runs or the speed of.

The system is fed more and more data and, as Flensburg-Hadewitt’s Mikael Döring explains, “it’s like a small child in the beginning, who starts to understand the game better through the data it receives.”

Calculation up to and including goal probabilities

Dering and his colleagues can even use TN to determine which player and in which position can create the greatest chances for goals. “Where should the player stand? How should we start the attack? Such questions suddenly become more important,” the sports scientist tells DW.

Dering, who has been a handball coach, is the one who passes the data on to the team coach. After all, a selection must always be made of those information that can actually be useful in training or the team’s tactical approach.

The more complex, the more efficient the AI

However, the use of AI-based data is still at an early stage because the system’s computational capabilities are limited to the data provided by humans.

But it is certain that such technologies will continue to develop at a rapid pace, as well as “digitize” and change sports even more. “The more players there are on a field, the more complex the data generated in each match. Therefore, the more people are active in a sport, the greater the amount of data and the more efficient it is to process this data through intelligent algorithms,” concludes Memmert.