More and more football clubs are relying on artificial intelligence when searching for new players. This can create results for each player individually – with sometimes surprising results. How good is the technology?
It's every club's wish: to finally sign a new star player not only costs little, but also represents a direct boost. The highest possible return with the lowest possible risk. Football is becoming more and more like mathematics. Scouting is crucial to minimizing risk when signing new players. The better a player is analysed, the better he can be evaluated.
This is where artificial intelligence comes into play. Quite simply, the AI evaluates the data and then evaluates the players. This is exactly the mission of scouts. A human scout sets certain criteria before monitoring a player. The AI was also taught certain rules at first. For example, it is better for a player to win a duel than to lose it.
Data about each player
Since professional players have been constantly monitored and evaluated during training and matches for years, AI already has a lot of data. On this basis, the tool can then make an assessment of the players. Software company SCOUTASTIC has been working with Bundesliga clubs for some time.
SCOUTASTIC's Christian Rumke explains that the AI system not only uses pure player data, but can also evaluate texts: “A lot of smart scouts walk around and write reports every weekend,” says Rumke. It is difficult to follow up and know what was written in the report three years later.
Better than human scouts?
Plaier has also developed artificial intelligence aimed at helping football clubs find the right player. Jan Wendt, co-founder of Plaier, even claims that AI is correct in player evaluations more often than humans: “If you pick ten players with us, we will be correct 8.5 times, which is much better, as if you were doing it by mankind.” Wendt means that the AI will be right at least eight times if you choose ten players.
Many scouts will surely like such a stake. According to Fendt, it depends on the individual situation, some clubs scout well, others less so. He stresses that it is not a matter of classifying scouts into right and wrong. The tool is a support for scouts. “But if you constantly make sure that players have to go through our filter and only include players who we see as suitable and who will make the club better, the scouting rate will improve,” says Wendt.
AI score for each player
The Plaier tool assigns a so-called AI score to each player. This is calculated through data collected by artificial intelligence about the player in question. “A striker is of course evaluated using different criteria than a central defender,” explains Wendt. In general, player performance is measured by more than 200 parameters and the player's influence on the team's performance will also be included in the evaluation, explains the Plaier president. The score can then be used as a comparison parameter to compare players around the world.
In an ideal scouting scenario, Wendt's AI shows that Stuttgart striker Deniz Undav is rated better than Erling Haaland. Wendt explains that according to the AI, Undav is a slightly more complete striker, achieving better values, especially when working against the ball.
Limits of artificial intelligence
Even if the artificial intelligence comes to the conclusion that Undav is better than Haaland, this of course cannot be fully proven. They both play in different clubs and in different leagues. For example, a coach or player's personal environment can also affect performance.
This means that if Undav moves to another club and another coach, he will not be able to perform well. “The AI is simply reaching its limits. And then we're really saying that the player fits perfectly into this system, with this coach, in this environment that he feels comfortable in, there are a lot of factors that AI can certainly really support.” But for now, It doesn’t bring everything together,” explains SCOUTASTIC’s Rumke.
Human scout support
Despite everything, it can be said that artificial intelligence is changing the approach to football scouting. Having a decision aid that is entirely data-driven and does not include personal preferences is definitely an advantage. It also relieves clubs and scouts of a lot of work and, above all, can help discover players in the minor leagues abroad who may not be on the club's roster.
Both Plaier and SCOUTASTIC stress that AI does not replace any scout in the world. “We see ourselves as more supportive and helpful in making informed decisions,” Rumke said. “We do not have the right to make transfers of players.” In the end, the decision still has to be made by humans, and it's not just about data and sports links. Rather, it is based on a human feeling and a personal assessment of whether the player fits into the team or not.
“Total coffee aficionado. Travel buff. Music ninja. Bacon nerd. Beeraholic.”
More Stories
Most meteors appear on these nights
Almost a million prize money – Nobel Prize for research in the field of global prosperity
AI often knows the right answer, even if the answer is incorrect