Soccerment: The Intelligence Layer of Football
Aldo Comi, co-founder and CEO of Soccerment, explains how football has evolved with the greater use of data and analytics and how it is using AI to change the way clubs scout and analyse matches.
The role of data in football has evolved hugely in recent years and industry is now at a point by which it’s turning that information into genuine football intelligence.
“For twenty years the industry has been describing football to computers one event at a time, a pass here, a shot there, then adding tracking data, then adding more of everything,” explains Aldo Comi.
“Each wave increased what a club owned and none of them changed how its Monday worked, because every tool stopped at the same place: a dashboard.”
Comi suggests that the collection and storage of that data meant there was a great depth of resource but the time and cost to query and make use of it had still not been worked out.
“Somebody still had to know where the data lived, how to query it and how to turn the answer into something a coach would read,” he continued. “The cost of asking a question is the reason most of the data a club pays for has never been looked at.
“The industry is now at the moment where that cost collapses. A football department is intelligent when its questions are cheap, and the measure I care about is the distance between a question and a decision. Today that is a morning. It is about to be a minute.”
Soccerment has recently launched SICS Atlas, with AI built into the platform from the outset. The platform brings video, event, tracking and physical data from more than 310 competitions and over 90,000 players into one system, with a natural-language interface for building shortlists, clip playlists and reports.
“We did not want to build another chatbot that answers questions from an event table, because football has enough of those already,” says Comi who suggests most programmes use AI as a language model on top of an existing database.
By building the intelligence in from the start the data layer is designed around the structure of the game, sequences, space and phases rather than isolated events, with AI agents sitting on top of that whose job is to do the work rather than talk about it.
SICS Atlas is built with the ambition to set a new benchmark for scouting and match analysis connecting video, events, tracking data, tactics and trends into one unified intelligence platform for football, designed from day one as a connected system.
“For the person using it every day, the difference is that they stop operating a platform and start delegating to it,” he stresses.
Due to AI’s ability to reduce the time analysts and scouts spend gathering and processing information it will change their day-to-day roles. But is there a risk of clubs using it to reduce headcount?
“I expect the number of people working in football clubs to go up, not down,” he says. “What changes is what they do with their week. Today, a serious share of an analyst’s time goes on tasks that are automatable by definition: tagging footage, cutting clips on request, rebuilding the same opponent report every Tuesday.
“Machines will take all of that, and nobody who does it will miss it. What grows is the part where somebody exercises judgement, and that work is worth hiring for. This is not a story about a department getting smaller. It is a story about the same department producing something a Chief Executive would pay more for, and the clubs that understand that first will hire ahead of the others.”
Key to the platform is that the data complements rather than replaces the human qualities that have always shaped football – experience, instinct, talent and judgement. Comi believes that judgement in football is not threatened by data, rather it’s threatened by what it is currently spent on such as retrieval, formatting and waiting for the report.
“Free a scout or a coach from that and you get more judgement, not less,” he says.
“Where data earns its place is in correcting the eye, which is never neutral. Take academies. Every one of us, watching a fourteen-year-old, mistakes early physical maturity for talent, which is why academies systematically keep early developers and release late ones.
“Data adjusted for biological maturity does not replace the scout. It removes a bias from the scout, and the judgement that follows is better for it. No model has ever signed a player, and none should. The right division of labour is simple: machines do the counting, people do the deciding and because the counting is now free, people get to decide far more often.”
Empowering more people freeing up time for better decision-making shouldn’t be reserved for football’s well off. Comi says the next generation of football intelligence should be accessible across the wider pyramid; from academies, women’s football and clubs with smaller analysis departments.
For Comi, SICS Atlas is key to this transition but how does he see the next three to five years developing?
“A club with fifteen analysts already has the manpower to absorb the cost of asking a question,” he says. “A club with one analyst, or none, does not, and for them an intelligence layer is not a productivity tool, it is the analysis department.
“Women’s football is the clearest case. It has been underserved on data for years and coverage is now improving, and the analysis departments being built there today should not have to repeat twenty years of dashboards to make use of it. They can start from the question.
“Academies are where time matters most, because the two things that compound in this business, structured data history and staff habit, cannot be bought later. A small club that starts building its academy data now has something in three years that no budget buys retrospectively. That is worth more to the small club than to the large one.”
Scale and accessibility across all levels of the game will ensure that clubs can continue to progress and develop not just the game but players with genuine value for the club.
Building a genuinely intelligent football department which has access to innovative technologies and working practices will become part of everyday football.
“First,” he says, “analytics, video and wearables will converge into a single relationship per club by the end of the decade: the largest suppliers have spent the last two years buying each other, and subscale point solutions do not survive the next funding cycle.
“Second, models trained on the game itself will replace models bolted onto data tables, and the word “dashboard” starts to sound the way “fax” sounds now.
“Third, maturity-adjusted measurement will become standard in academies, and we will look back on ranking fourteen-yearolds by age group the way we look back on smoking in dressing rooms. Fourth, more people will work in football clubs, not fewer.
“What becomes ordinary is the thing that seems remarkable today: asking a question about your own team in plain language and getting an answer you can use before the coffee is cold. If in 2030 the typical top-division club still runs three separate suppliers and its analysts still build the same report by hand each week, I was wrong, and you are welcome to print this paragraph again.”
To find out more about Soccerment and SICS Atlas, visit: www.soccerment.com/products/sics-plus/atlas
Images: Soccerment
