Buch, Englisch, 139 Seiten, Format (B × H): 148 mm x 210 mm
From Raw Data to Validated Performance Indicators
Buch, Englisch, 139 Seiten, Format (B × H): 148 mm x 210 mm
Reihe: Dissertationen der Sportwissenschaft
ISBN: 978-3-658-53896-5
Verlag: Springer
This book presents a unified framework for automating performance analysis in soccer through machine learning. Rather than treating match segmentation, event detection, and performance indicator validation in isolation, it develops a sequential pipeline that transforms raw spatiotemporal tracking data into structured, analysis-ready representations. The work integrates three empirical studies with broader conceptual contributions—including a rigorous framework for performance indicator validity, a systematic clarification of Match Momentum distinct from psychological momentum, and a structured taxonomy of ML application domains in soccer. Embedded within a broader argument for democratizing soccer analytics, the book demonstrates how algorithmic automation can reduce reliance on costly manual annotation and proprietary data products, thereby lowering analytical barriers for resource-constrained clubs.
Zielgruppe
Research
Autoren/Hrsg.
Fachgebiete
Weitere Infos & Material
1. Introduction.- 2. Background.- 3. Application Fields of Machine Learning in Soccer.- 4. Predicting the In-Game Status in Soccer.- 5. Which Indicators Matter.- 6. How to Find Elementary Football Structures in Positional Data.- 7. Discussion.- 8. Conclusion.- 9. Outlook.




