DEV Community

Tyson Cung
Tyson Cung

Posted on

How AI is Revolutionising Football — From Scouting to Tactics

A video went viral recently about AI in football, and I went down the rabbit hole. Turns out, the beautiful game is being quietly transformed by machine learning — and not just at the top clubs.

Here's what I found.


The EPL is Leading the Charge

Liverpool × DeepMind

Liverpool partnered with Google DeepMind to build TacticAI — an AI system that analyses corner kicks and set pieces. It predicts which players to target, where to position, and which defensive setups are most vulnerable.

In blind tests, football experts preferred TacticAI's tactical suggestions 90% of the time over existing approaches.

This isn't replacing the coach. It's giving them a tactical edge that's invisible to the opponent.

Manchester City — The Data Machine

City's analytics department reportedly tracks 10 million data points per match. Their AI systems analyse:

  • Player positioning and movement patterns
  • Pass completion probability in real-time
  • Fatigue indicators from GPS data
  • Opponent weakness detection across multiple matches

Pep Guardiola has spoken publicly about using data to inform squad rotation — a key reason City can compete on 4 fronts every season without burning out players.

Arsenal — Injury Prediction

Arsenal invested heavily in injury prediction models after years of losing key players to long-term injuries. Their system analyses:

  • Training load vs. match intensity
  • Historical injury patterns per player
  • Biomechanical data from wearables
  • Sleep and recovery metrics

The goal: flag a player as "high risk" 48 hours before an injury actually happens. Pull them from training early, adjust the session, prevent the 6-week absence.

Brighton — Punching Above Their Weight

Brighton's recruitment model is arguably the most impressive AI success story in football. They use machine learning to:

  • Scout players from lower leagues and undervalued markets
  • Predict which players will improve (not just current ability)
  • Find statistical profiles that match their playing style

Result: they bought Moisés Caicedo for £4.5M and sold him to Chelsea for £115M. Marc Cucurella, Alexis Mac Allister, Leandro Trossard — all identified by data-driven scouting.


Beyond England

Juventus (Serie A)

Juventus partnered with tech companies to build AI-powered match analysis. Their system breaks down every match into thousands of micro-events and identifies patterns that human analysts miss.

AC Milan

Milan uses AI for player wellness monitoring — tracking nutrition, sleep, psychological state, and physical metrics to optimise performance windows.

Barcelona (La Liga)

Barça's famous La Masia academy now uses AI to track youth development. The system analyses positional play patterns and compares young players' development trajectories against historical data from graduates who made it to the first team.

Real Madrid

Madrid invested in AI-powered stadium operations (the new Bernabéu) and uses predictive analytics for transfer market valuations.


The Five Key AI Applications in Football

Application What It Does Who's Leading
Tactical Analysis Set piece optimisation, formation weaknesses Liverpool (DeepMind)
Injury Prediction Flag high-risk players before injury occurs Arsenal, AC Milan
Scouting & Recruitment Find undervalued players with ML models Brighton, Brentford
Performance Tracking Real-time GPS, biomechanical, fatigue data Man City, Bayern
Youth Development Track development trajectories over years Barcelona (La Masia)

The Numbers

  • 10M+ data points tracked per Premier League match
  • 90% preference rate for TacticAI vs. human-only tactics (DeepMind study)
  • £115M — Brighton's return on a £4.5M AI-scouted signing
  • 48 hours — how far ahead injury prediction models can flag risk
  • $4B+ — estimated global sports analytics market by 2027

What's Coming Next

The next frontier is real-time tactical AI — feeding live match data into models that suggest substitutions, formation changes, or pressing triggers during the game.

Imagine a manager getting a notification at half-time: "Switch to 3-5-2, move X to left wing. Their right-back has covered 2km more than average and his defensive response time has dropped 0.3s."

That's not science fiction. Teams are building this now.


The Takeaway

AI isn't replacing managers or scouts — it's making the good ones better and exposing the ones who refuse to adapt.

The clubs that invest in data infrastructure today will dominate the next decade. The ones that don't will keep overpaying for players who don't fit their system.

Football has always been about margins. AI just made those margins measurable.


I made a full video diving deeper into each club's AI setup — watch it above! ⬆️

Follow me for more on AI, tech, and the occasional sports rabbit hole. I post daily.

Top comments (0)