Top-class sport stands for emotions and unpredictable events. Coaches make decisions and influence games based on their gut feeling. It has always been like this. But something could change: Is the use of “artificial intelligence” (AI) threatening to make sport more predictable and emotionless?
Although data has been used to analyze games for years, it is precisely when it comes to evaluation that coaches are increasingly reaching their capacity limits. The reason: The collected game facts are becoming more and more precise and extensive. This is where “AI” will help in the future.
“Artificial intelligence is a technology that is carried out by computers and machines to complete tasks that normally require human intelligence but are usually too complex for humans,” says Daniel Memmert from the German Sport University in Cologne. “These AI systems use data and algorithms to autonomously recognize and objectively evaluate game patterns.” And it is precisely this objective evaluation that makes the difference to the often subjective assessments of a trainer, according to the sports scientist.
Does AI replace humans?
AI has revolutionized data-based analysis and is increasingly influencing decisions about tactics, injury prevention or training management. “Sports clubs have brutally expanded their use of AI,” explains Memmert. “The clubs have realized how much they can do with the data collected.”
Sports clubs are also expanding their competence teams. People who have gained as much experience as possible in a sport are no longer hired, but rather programmers and computer scientists. “In 2017, FC Barcelona hired 15 new data and video analysts and data scientists,” says Memmert, adding: “The use of AI systems will not lead to fewer people being used in competitive sports, but more.” Because there still have to be experts who understand so much about the respective sport that they know how to deal with the information and metrics created by AI, says the sports scientist.
Schalke scouts with the help of AI
Artificial intelligence can now be found in many different areas. At FC Schalke 04, the scouts are supported by an AI in finding new players. “We have defined around 40 values in our scouting and we let artificial intelligence run along with three of them and also decide whether these are values that make the difference for us,” Schalke sports director Peter Knäbel told ZDF. “For example, you define a player’s free-running behavior, i.e. his passing accuracy. And you put artificial intelligence on top of it.”
In this way it is easy to see whether this footballer has a similar value to a player who has moved away. Nevertheless, the final decision about a commitment still lies with the scout and the coaches. In this case, the AI is just another way to better assess potential Schalke players in advance.
Michael Döring: “AI is like a little child”
Intelligent systems are also used in handball. For several years now, players in the handball Bundesliga have been wearing so-called “wearables”, i.e. vests or sports bras into which a sensor can be inserted. During a game, several million pieces of data are recorded: mileage, speed, biometric data or, for example, player position data.
“We then feed the computers with this information so that they learn how a handball game works,” says athletic trainer Michael Döring. “In the beginning, the system is like a small child and only learns to understand the game better with the data it receives.”
The probability of a goal is increased
In the past two years, the AI has been fed around 600 handball games. The system used by SG Flensburg-Handewitt has now accumulated a lot of knowledge about the sport of handball. When analyzing data, AI can now predict goal probabilities.
Döring and his colleagues can see which player in which position has the greatest chance of scoring a goal. “Where should the circular runner be? How should we start the second wave? These are suddenly things where we have added value,” the sports scientist tells DW. “It’s like an additional guide. And we can use this data-based approach to make our system more successful.”
Döring, who used to be a handball coach himself, acts as an intermediary between the data and the coach in the Flensburg team of experts. Because not all of the information that is determined with the help of AI is helpful for training or tactics in a game. “We take the questions from practice into theory. And we have to package the knowledge that we then get through the AI-based analysis in such a way that we can do something with it in practice.”
In the future, handball players hope that the use of artificial intelligence will provide support with tactical questions, but also with putting together a team that has the highest possible probability of success.
The more complex, the more sense the AI makes
Nevertheless, the use of AI-based data is still in its early stages. Currently, the system is only as intelligent as the information supplied by people. However, it will continue to develop rapidly and continue to digitize and change the sport. “The more players there are on a playing field, the more complex the patterns that occur. So the more people are active in a sport, the larger the amount of data emerges and the more it makes sense to have this data processed by clever algorithms,” says Memmert .