How Liverpool are using artificial intelligence to improve their corner kicks
As Arne Slot settles into his role as Liverpool’s head coach, a host of exciting new ideas are emerging from his new coaching staff.
The new training methods of assistant coaches, Sipke Hulshoff and John Heitinga. The lively voice of the head coach physical performance, Ruben Peeters. The creative coaching exercises of head goalkeeping coach, Fabian Otte.
Off the pitch, Liverpool have developed an assistant coach who could help them tackle setbacks next season… and it’s not even a human being.
Liverpool’s analytics department is known for its groundbreaking work, led by research director William Spearman, who took over from Ian Graham in 2023. Their most recent venture has seen them enter into a partnership with Google DeepMind, using artificial intelligence (AI) to determine corner kicking strategies.
The collaboration resulted in an article published in Nature Communications — TacticAI: An AI assistant for football tacticsThe project, led by researchers Zhe Wang and Petar Velickovic, used data from 9,693 corner kicks collected from the 2020-21, 2021-22 and 2022-23 Premier League seasons (up to January 2023), providing information on each player’s height, weight, starting location and movement during the corner kick routine.
The information from each player allowed the researchers to predict the outcomes that were most likely to occur within a given corner formation. For example, which player is likely to receive the ball? Will the order result in a shot attempt?
Once the sequence has been played, the analysis can build a picture to determine whether similar routines have been successful in the past. Crucially, TacticAI can use this analysis to make suggestions that improve the outcome of the corner. For example, by changing the position or body position of players to reduce the chance of a shot being given away from a specific corner.
The image below shows four suggestions from TacticAI to adjust player positions when defending a corner kick.
As evidence of the model’s capability, experts in Liverpool say — including assistant coaches, video analysts and data scientists — were unable to distinguish between the output of TacticAI’s suggestions and real corner routines, with the proposed routines from the original corner series being preferred in 90 percent of the cases, highlighting how effective the model can be in making improvements within the tactical structure.
As you can see in the chart above, the suggested adjustments may seem subtle, but joint lead researcher Velickovic emphasizes that these suggestions are appropriate for modern football. Small changes in distances, timing and decision-making can make the difference between winning and losing.
“The idea is not to tell you that a player has to move two or three meters to the left or to the right,” Velickovic said The Athletics“You make small adjustments to a player’s location, orientation or speed, all of which are relatively small, and that was intentional.”
One of the main aims of Liverpool’s project with Google DeepMind is to provide coaches and analysts with a tool to help them with their workflow. Opposition analysts often watch hundreds of videos in the build-up to a match, which is labour-intensive. TacticAI’s ability to comb through similar opponent routines and curate defensive strategies is powerful.
It is more efficient to draw conclusions in a fraction of the time and this work offers an objective tactical approach without bias, supported by thousands of examples.
“This tool is designed to accelerate a coach’s ability to recognize patterns,” Velickovic said. “Coaches are looking at complex situations with 22 players and they have to figure out what are the key components that made or broke a particular strategy — and which players were responsible.”
“With a system like this, where adjustments are made instantly to all defensive players, you can focus your attention and immediately see if, for example, a specific defender is doing something wrong.”
“If this happens to your defender in multiple situations, you can try to resolve it in your coaching. If it’s an opponent, you can develop strategies to exploit that weakness.”
Trent Alexander-Arnold takes a corner against Atalanta last season (Jonathan Moscrop/Getty Images)
It’s worth noting that this research assesses the potential use of such an approach and has not yet been rolled out to Liverpool’s matchday analysis. Nevertheless, the rise of dedicated set-piece coaches shows how many clubs are waking up to this key part of the game – with 28 per cent of all goals scored from dead-ball situations in the Premier League last season.
GO DEEPER
Permanent trainers, analysts… and chefs: your club staff member’s secret weapon
For Velickovic, the marginal impact this work can have could prove crucial.
“We can never predict exactly what happens in a corner,” Velickovic said. “But if you increase your chances of scoring or reduce the chances of conceding by even one percent, it can make a huge difference over the course of a season.”
Corners have been responsible for some of the most important moments in Liverpool’s recent history. Jurgen Klopp’s last trophy as manager came courtesy of Virgil van Dijk’s header from a Kostas Tsimikas corner to secure the Carabao Cup against Chelsea in February.
The unlikely Alisson Becker scored an iconic last-minute header to convert a Trent Alexander-Arnold corner against West Bromwich Albion in 2021 to keep Liverpool’s hopes of a top-four finish alive.
And Liverpool fans won’t need a second invitation to relive Alexander-Arnold’s swiftly taken corner for Divock Origi in the 2019 Champions League semi-final against Barcelona – one of the most memorable goals in the club’s history.
“Quickly taken corner… ORIGIIIII!!!!” 😱😍
Does Divock, who made an incredible comeback against Barcelona, get your vote for our greatest goal ever?
— Liverpool FC (@LFC) June 22, 2022
Ultimately, the delivery that Alexander-Arnold provides from set pieces is the crucial part of ensuring that a choreographed routine is executed. As set-piece specialist Gianni Vio previously explained The Athletics“The one who takes the ball is the most important player in set pieces.”
The statistical models can be instructive for coaches, but without good presentation the series collapses.
GO DEEPER
Gianni Vio on the art of set situations: ‘They are a game within the game’
Looking at last season, the figures suggest that Liverpool could improve their corner conversion compared to the rest of the Premier League. Their 4.2 goals per 100 corners was their worst rate since 2018-19, with their total number of goals scored and conceded per 100 corners almost perfectly in line with the league average.
This project was the latest milestone in Liverpool’s multi-year journey with Google DeepMind, but the collaboration demonstrates a continued desire to push boundaries and find an edge in the use of analytics in football. It’s a partnership that dates back to 2021, where their work saw them publish two further papers relating to AI in football tactics and its use in penalty analysis.
The constant evolution of football tactics means that corner analysis is a dynamic process that needs to be constantly updated. An optimal strategy in 2015 might look different today. The best for the German Bundesliga might be different than the best for the Premier League.
The potential for this AI model to grow over multiple seasons and larger competitions is where analysts can add an extra dimension to their tactical approach and strategic decision-making.
At its core, AI is the ability of a computer to perform tasks that we typically associate with humans. The growth of AI — specifically Generative AI via platforms like ChatGPT, Claude, and Google Gemini — is visible across multiple industries worldwide.
GO DEEPER
‘Will Harry Kane be a good signing for Bayern?’: The rise of generative AI in football scouting
While these platforms are fantastic tools for coaches and analysts to use as tools, they should never replace human expertise. Data’s place in football is secure, but the combination of objective and subjective analysis is where staff can optimise their decision-making during the game. We are not entering a world where football is played by machines.
At least, for now.
(Top photo: Playmaker/MB Media/Getty Images)
