Revolutionizing Football: How AI is Transforming Transfers, Training, and Scouting at Top Clubs Like Spurs and Liverpool

As the World Cup final hangs in the balance, a crucial piece of tactical advice comes from England’s most trusted assistant coach, steering the team toward victory.

Observing the match, the assistant notices France’s left back showing signs of fatigue and suggests exploiting this with long diagonal passes. This insight is promptly communicated to the players, and the strategy proves effective. The decisive goal for England is crafted from the right wing, leading the team to a long-awaited triumph and reclaiming the prestigious trophy.

This tactical maneuver is not merely a stroke of luck; it stems from advanced technology. A sophisticated device tracks the game in real time, utilizing artificial intelligence to identify exploitable patterns. Recently, Daily Mail Sport highlighted Arsenal’s use of machine learning to enhance their match preparations.

The ambition now is to analyze games as they unfold, offering an in-game advantage. While it might not be widespread at the next World Cup, this cutting-edge approach is likely to become standard practice by the 2030 tournament.

The next target is to deconstruct games as they are happening and gain an edge that way. It may not be the norm at the next World Cup, but there is every chance it will be before the 2030 tournament.

At the Hudl Performance Insights event at Craven Cottage last week, experts from across the industry gathered at Fulham’s ground to discuss the future of the game. Prominent on the agenda, inevitably, was the influence of AI on football.

Many clubs use AI models routinely to analyse physical data and produce lists of transfer targets from specific criteria.

Now things are about to move a step further, as Ed Sulley explains. Having spent 19 years with Bolton Wanderers and Manchester City, Sulley is now director of customer solutions at global sports technology company Hudl.

‘There are now AI models being built which, within a matter of hours, would be capable of analysing more matches than have ever been played in the history of football,’ he tells Daily Mail Sport.

‘So the golden ticket from a coaching perspective would be live intelligence coming from AI. It’s live tracking of all the data but also understanding the game you’re trying to play and the tactics you’re trying to deploy.

‘If we notice the opposition have set up in a certain way to try to block our aims, we want to be able to spot that quickly but also have several suggestions about what to do next. It’s the next frontier from a technology perspective.

‘At Hudl we’re invested in the idea of a connected stadium – cameras, tracking technology, wearable technology. We already have the technology that can spot trends like players staying wide or others who can deliver line-breaking passes.

‘Next it will be about the technology giving suggestions – what if we do this instead? Or, the technology might figure out that certain players have switched off from the previous four corners. Then the coaches can get that message to their players.’

Increasingly, AI helps to recruit these players, too. Brentford owner Matthew Benham and his counterpart at Brighton, Tony Bloom, developed data models to sign players who have helped turn them into established Premier League clubs. With the progress of AI systems, such machines are becoming more and more sophisticated.

Arsenal were ahead of the game when they bought analytics company StatDNA in 2014. Other clubs to follow that path include Norwich, who are in talks to buy src ftbl (pronounced ‘source football’), and Birmingham, who bought Real Analytics earlier this year. Premier League clubs, including Chelsea, are trying to perfect their own systems.

Clubs are not the only ones to travel this route, with top agencies creating their own data and analytics departments, informed by AI. Raiola Global Management, the supporting commercial agency to Team Raiola established by the late super agent Mino, have built a model that uses multiple data points to determine which clubs would suit the players on their books. 

In this way, Tottenham was identified as an ideal home for Micky van de Ven, and Liverpool for Ryan Gravenberch.

‘It took almost three years to develop a model like that,’ explains Mark Nervegna, managing director at Raiola. ‘Models like this will affect transfers a lot and also help to determine the actual worth of players. 

‘We have built club rating models, league rating models, the tactical and technical characteristics of certain teams and certain players. Then it’s about analysing the coach and the club overall.’

This is where budget planning will be key. Just as Brighton and Brentford have managed to stay ahead of the game with their smart use of data, clubs of a similar level may choose to invest in cutting-edge AI innovations rather than, say, a new centre forward.

Lincoln City have one of the lowest budgets in League One but are second in the table thanks to clever work off the pitch, including the use of AI. ‘For the last two seasons, we have been working on a specific type of long throw,’ their manager Michael Skubala says. 

‘My coaching team and sports scientists have been involved in the process and we’ve used AI to inform our decisions too. Now suddenly, it’s coming back at the highest level – but we brought in back two seasons ago. There is a strong rationale behind where and why we do it.’

The role of scouts is certain to change. Over time, the machines will become so efficient at sifting information and watching matches, that the ‘eye in the stand’ may become obsolete.

Yet AI cannot do everything. It still cannot assess a player’s family situation, their childhood or how they might react to certain situations. The models are unable to discover whether a player likes a night out, how careful they are about nutrition or if they are careless with money.

Asked whether scouts may become more like private investigators, Sulley partly agrees. ‘The role may evolve from watching players in action, into checking details that cannot be measured by data,’ he says. 

‘Scouts have contacts throughout the game who enable them to make the background checks that are vital whenever clubs are researching a transfer.

‘You can’t know through data whether a player is likely to fall out with others or what their family is like. The data could look great but it’s no use if you end up with a character who could destroy your environment overnight, and then it is very difficult to sell that player. 

‘But the people making these checks will still need to understand football. They can study behaviour around a training ground or understand how to react around a disappointed player.’

With technology advancing at bewildering speed, those who do not board the train now risk being left behind permanently. While the excitement is understandable, there are those who sound an important note of caution.

During four years as game insights lead at the FA, Chris Markham helped overhaul England’s approach to penalty shootouts, and was thanked by Gareth Southgate after the Three Lions’ memorable win over Colombia in the 2018 World Cup. Markham’s spell at the FA was sandwiched by stints at Huddersfield and Bolton.

‘Hopefully, the speed with which AI models process information will allow us to have better conversations and ultimately take better decisions,’ Markham argues. ‘But the idea AI is without risk is a fallacy. The models are only ever as good as the information that is fed into them.

‘About 20 years ago, data started to have a huge influence and you had to get on board with it. The same is true of AI now. 

‘It is still about the players on the pitch and the coaches in the dugouts, though, who are the key performers and sometimes, they are the most difficult people to convince when it comes to modern technology. That is why it is important for us all to be as well-informed as possible.’