AI is behind Liverpool's top corner kick 'secret trick'

TacticAI, the result of a partnership between DeepMind and Liverpool, offers advantages beyond what existing data analysis methods currently offer in corner situations.

In the 2018-2019 Champions League semi-final second leg, Liverpool was 3-0 ahead in the first leg, but scored 3 goals to level the score.

In the 79th minute, a moment of genius appeared when Liverpool received a corner kick. Defender Trent Alexander-Arnold placed the ball, took 3 steps onto the field before running back and passing the ball to Divock Origi with a huge space between the 4 Barcelona players. The fourth goal finished off Barca and sent Liverpool to the final.

Corner kicks in soccer are a potential source of goals. However, setting an optimal habit when taking corner kicks depends on the combination of player intuition on the field and match design from the team's coaching staff.

However, Liverpool's analysis team turned what Trent Alexander-Arnold himself admitted was an instinctive goal into a product of calculation.

Reveal the secret

More than 5 years after Origi's goal, everything has been somewhat revealed. According to an article published in Nature magazine on March 19, DeepMind, a subsidiary of Google and one of the world's leading artificial intelligence laboratories, officially announced TacticAI.

According to DeepMind, this is an artificial intelligence (AI) system that can provide experts with tactical insights, especially about corner kicks , through prediction and synthesis capabilities. of computer.

How TacticAI synthesizes and suggests how to handle a certain corner. (Photo: DeepMind).

DeepMind , which famously brought its AlphaFold software to dominance in board games, says the patterns found on sports fields can also offer lessons on how to apply AI in other fields. others such as robotics and traffic coordination.

DeepMind's project is the product of 3 years of cooperation with Liverpool in implementing AI in penalty kick situations and predicting player movements.

Specifically, DeepMind's latest model uses geometric deep learning technology on a data set including 7,176 corners in the Premier League from 2020-2023.

According to analysis, corner kicks provide significant opportunities for attacking teams. Together with other set pieces, such as free kicks, they account for around 30% of all goals scored.

TacticAI analyzed the results from corners with various player arrangements, using criteria such as who would receive the ball and whether they could shoot.

The model then recommends placement improvements and evaluates their plausibility and usefulness in a case study with five experts in Liverpool, including three data scientists, one video analyst and an assistant coach.

The model is capable of creating many corner kick situations, helping coaches to check and make countless different options. (Photo: DeepMind).

According to DeepMind researchers, experts could not even distinguish AI-generated scenarios from real match situations, preferring TacticAI advice 90% more than strategies now available.

According to the article, this shows that TacticAI can 'easily provide useful, realistic and accurate recommendations'.

Outstanding potential

TacticAI's generative model also allows coaches to redesign corner kick tactics to optimize the probability of certain outcomes, such as reducing the probability of taking a shot to set up defense.

In addition, this tool also provides tactical suggestions to adjust the positions of all players in a team.

From these suggested adjustments, strategists can more quickly identify important patterns, as well as key factors that determine tactical success or failure.

Data analysis has influenced professional baseball and basketball in recent years. It has also begun to impact football.

In soccer, the chance to score a goal affects the match outcome much more than in other sports. The average number of goals in the Premier League is less than 3 goals per match as a specific example.

Liverpool was one of the first big clubs to introduce data analysis into coaching .

Liverpool is a club that benefits greatly from effective AI analysis. (Photo: DeepMind).

'Liverpool is the leader in the field of football data analysis ' , Ignacio Palacios-Huerta, Director of Athletic Club Bilbao and a professor at the London School of Economics, once replied to writer Simon Kupers. .

At Liverpool, their analytics team is headed by Ian Graham, who manages four members with doctorates ranging from astrophysics to physics and advanced mathematics. Most importantly, says Palacios-Huerta, 'they understand football'.

It is Graham's algorithmic insights, combined with football's data analysis, that has led to some of the shrewdest acquisitions in the Premier League.

According to Sudarshan Gopaladesikan, director of football analytics at top Serie A team Atalanta, projects like TacticAI are prime examples of how AI can develop in this sport.