Skip to content

Repository files navigation

xG Model

Repository Structure

xG_Project/
├── README.md
├── requirements.txt
├── LICENSE
├── .gitignore
├── 1_importing_data.ipynb
├── 2_splitting_data.ipynb
├── 3_preprocessing.ipynb
├── 4_exploratory_data_analysis.ipynb
├── 5_model_training.ipynb
├── selected_features.md
│
├── data/
│   ├── preprocessed/
│   ├── raw/
│   └── split_data/
│
├── src/
│   ├── __init__.py
│   ├── constants.py
│   ├── get_data.py
│   ├── xG_constants.py
│   ├── xG_evaluation.py
│   ├── xG_models.py
│   ├── xG_preprocessing.py
│   └── xG_visualization.py
│
└── docs/
    ├── Open Data 360 Frames v1.0.0.pdf
    ├── Open Data Competitions v2.0.0.pdf
    ├── Open Data Events v4.0.0.pdf
    ├── Open Data Lineups v2.0.0
    ├── Open Data Matches v3.0.0.pdf
    └── StatsBomb Open Data Specification v1.1.pdf

Description

Feature Selection

Dropped Tables

Tables Used

Events

Frames

The frames table contains the positions of each player during each event for competitions covered by the StatsBomb 360 data. The competitions included are:

  • 1. Bundesliga 2023/2024 - Bayer Leverkusen
  • UEFA Euro 2020
  • FIFA World Cup 2022
  • UEFA Euro 2024

References

About

An Expected Goals (xG) model built with StatsBomb open data. The project covers data scraping, cleaning, feature engineering, model training, and evaluation to estimate goal probabilities from shot events.

Topics

Resources

Stars

2 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages