Football - Predict the Likelihood of a Goal Based on a Game Situation

Sports Operations Executive Summary
Ever wanted to be the coach of a football (soccer) team? Or to be a data scientist for a football team? Use AI/ML to predict what drives goal and build scenarios to get the optimized situations and be closer to scoring a goal
Build with Free Trial


Business Problem

Football is the most popular sport in the world by a vast margin. (maybe with an expectation in the US with football football) Played by a team of 11 players against another team of 11 players on a field, every team has one objective – to score as many goals and win the game. However, beyond players’ skills and teamwork, the situation for every event in the game, like the shot place, the body part, situation, location side, and more – can make or break scoring goals and win the game.

Games take place in different locations with different players combination. However, teams must have a game strategy and technique before they play — and that strategy must adapt to changing conditions throughout the game.

Intelligent Solution

Every fan wants their football team to win! AI and ML can help in the football field in predicting outcomes while transparently explaining how each variable contributes.

This raw dataset is from Kaggle – it has 10K past games (and 900k game events) to get insights on what drives goals and will try to predict the likelihood of a goal in every event situation. (The final dataset went through transformation and dataset joins, which can be found in the top right corner)

A total of five substitutions can be made in regular time (90 minutes plus added time). Still, each manager is only permitted three opportunities to make those changes, including halftime and hydration breaks.

After running through some basic questions and practices, the team can decide on the right strategy for the race:

  1. Which players are playing? What are their strength and weaknesses?
  2. Who is the opponent? How is their defense?
  3. Is it a home or away game?

After a quick consideration of these questions, the team must decide which players will play and plan scenarios for goal events.

ML-driven models can help the team develop optimized situations for scoring more goals by using the wealth of data available on historical games and their resulting outcomes. For example, event type: attempt, from the corner, using the left foot, has a high likelihood of a goal.

Bringing AI-driven intelligence to this event situation decision problem will give one side an advantage in this blink-of-an-eye conflict and potentially swing the balance of power in the contest.

banner purple waves bg

Experience the DataRobot AI Platform

Less Friction, More AI. Get Started Today With a Free 30-Day Trial.

Sign Up for Free
Explore More Sports Use Cases
Sports organizations use AI for player performance analysis, injury prediction, and fan engagement. AI tools also assist in game strategy formulation, ensuring competitive advantage and optimal outcomes.

Explore More Use Cases