Corner Threat Index (CTI)

A Graph Neural Network-based system for evaluating corner kick quality in soccer using tracking and event data.

---

🎓 Course Context

This project was developed as the final project for the Deep Learning and AI in Sport course by Twelve Football.

⚠️ Data Availability & Privacy Notice

The tracking and event data used in this project was provided by SkillCorner exclusively for educational purposes as part of the Twelve Deep Learning course. It is not publicly available and cannot be shared. Access to this data was granted solely for the duration and scope of this course.

---

📊 Project Outputs

Team Performance Summary

Offense vs Counter-Attack Risk

Tactical Patterns (NMF Features)

---

📚 Documentation

The documentation has been organized into the following guides:

• 01_Technical_Guide.md: Complete technical framework, architecture, and implementation details.
• 02_User_Guide.md: Step-by-step guide for installation, running the pipeline, and troubleshooting.
• 03_Methodology_and_Insights.md: Detailed explanation of the CTI formula, models, and coaching applications.
• 04_Results_and_Reports.md: Summary of results, reliability/calibration reports, and validation metrics.

---

🚀 Quick Overview

The Corner Threat Index (CTI) quantifies the offensive value of corner kicks by combining five key components:

$$CTI = y_1 \cdot y_2 - \lambda \cdot y_3 \cdot y_4 + \gamma \cdot y_5$$

• y₁: Probability of a shot within 10s.
• y₂: Expected Goals (xG) of that shot.
• y₃: Probability of a counter-attack.
• y₄: Risk (xG against) of the counter-attack.
• y₅: Territory gained (ΔxT).

For detailed instructions on how to run the code, please refer to the User Guide.

---

🙏 Acknowledgments

A special thank you to the course instructors and mentors for their guidance and support throughout this project:

• David Sumpter: For his incredible contributions to the football analytics community and for leading this course with such insight.
• Pegah: For her excellent administration and smooth organization of the course.
• Dimitris: For being a fantastic mentor, providing invaluable advice, technical considerations, and support during the development of this project.

---

Author: Tiago Date: 2025-12-06