Update to its Sportian Performance product adds new layers of analysis and engagement
Sportian, the sport division of Globant, has updated its Performance platform with a number of AI agents for football analysis.
The tools provide real time analysis into players’ attacking potential, overall performance, and the goal probability of certain situations. Another analyses betting market data to detect possible match fixing. The data can be sent to commentators for use in broadcast, and from August, the company will be able to send video clips related to the action for use in broadcast.
AI Media News is the newly-launched publication from Broadcast, Screen Daily, and Global Data covering everything AI in the media industry. You can follow its Linkedin account here, and X here, and sign up to its weekly newsletter here.
For example, if the AI agent saw that a team had gone behind, it could send videos of goals scored when behind in a game to the production team. These will be available to the match director almost immediately as they become relevant.
Sportian AI agents
● Attacking Potential: A metric that offers a new way to measure attacking players’ impact beyond goals and assists. By analyzing movement trajectories, positioning tendencies, and passing dynamics using deep learning models trained on extensive match footage, this tool offers a more holistic view of a player’s offensive contributions, thereby helping to quantify their impact on the game.
● Real-Time Performance Charts: This agent processes live match data to deliver new content visualizing shifting game states, detecting a change in team formations, passing patterns, or player workload for example. The agent will share this information with coaches to help them respond tactically, and can also share approved content with fans via social networks.
● Betting Controls: Using deep learning, this agent continuously analyzes betting market data alongside match events to detect potential anomalies indicative of match-fixing. Through anomaly detection techniques and historical data comparisons, the system provides match organizers with early-warning indicators of suspicious activity.
● Goal Probability: This evolution of the xGoal metric assesses historic and game state variables—including shot angle, player scoring records, defensive pressure, and goalkeeper positioning—to calculate a more accurate probability of a goal being scored. Trained on a deep repository of historical match data, this predictive model dynamically adjusts based on hundreds of in-game conditions, generating a % metric within match broadcasts. Already implemented in LaLiga, this metric offers a new level of real-time game analysis.
The tech is available to all football competitions, with minimal training time needed to adapt. LaLiga is currently working with Sportian, which used to be known as LaLiga Tech before Globant bought a controlling stake, and its president, Javier Tebas recently previewed the future capabilities the league will be taking advantage of next season at the ISDE Sports Convention 2025.
He said: “We’re working on using AI to improve our broadcasts. If there is a goal in a game, AI will automatically detect a similar, or almost identical, goal in the LaLiga archive and we will insert it.”
No comments yet