Project 07 · Football analytics

Football Probabilities · Expected Threat (xT)

✅ publisheddata: 2022

Probabilistic Expected Threat model in football using public StatsBomb data. It estimates the probability of a goal in the next 5 actions.

Python
Pandas
Football
Data science
Data analysis
Data storytelling

CONTEXT AND OBJECTIVE

Measure the threat generated by a football action by estimating how goal probability changes within a short possession window.

DATA AND DIAGNOSIS

  • Public StatsBomb event data is used as the source.
  • Spatial and contextual variables are built to represent each action.
  • A future goal probability is estimated and the incremental contribution of actions is read.

DELIVERY AND LEARNING

The delivery translates sports events into a probabilistic reading, connecting data preparation, modeling and visual explanation of generated threat.