🏴 AI improves corner kick tactics

Google DeepMind has collaborated with Liverpool FC to release the TacticAI model, which helps improve corner kick' tactics. When a corner kick is awarded, the game stops, and the team has the opportunity to plan their actions.

How does it work?

⚽ TacticAI was trained on data from 7176 corner kicks in English Premier League matches. The data includes various sources such as videos of players moving on the field, past game experiences, player profiles (height, weight), stadium information, and field dimensions. TacticAI successfully predicts corner kick play by applying a geometric deep learning approach.

🎱 📊 The model predicts which player will most likely receive the ball and whether a shot will be attempted. TacticsAI recommends position changes for players to increase or decrease goal-scoring chances. Experts preferred TacticAI's advice over existing tactics in 90% of cases.

How else can TacticAI be used?

In the future, TacticAI could be used to analyze throw-ins and free kicks. Researchers believe this method might be used in any sport where game stops allow teams to adjust tactics.

Impact of AI on sports

🏀 One of the most prominent examples is basketball. Big data research in the 2000s revealed that making a lot of three-pointers was the key to winning. If a team had players with a high shooting % from beyond the arc, scoring an extra point was worth the risk. Anyone who has watched the NBA over the last few years has noticed a revolution right in front of their eyes: individual abilities and breathtaking dunks were replaced by pragmatic passing and numerous three-pointers.

Sources: Google DeepMind, The Conversation, Contingencies, SmartData Collective

❤️ — Liverpool F.C.
🔥 — Arsenal F.C.
🏆 — Manchester City F.C.
😈 — Manchester United F.C.

#news