Visual Analytics for Soccer Analysis
Dienstag, 25. April 2017, 16:15 Uhr bis 17:30 Uhr
Soccer is one the most popular sports today and research-wise very challenging. For development and alignment of tactics and strategies, professional soccer analysts spend up to three working days manually analyzing and annotating professional soccer matches. Our aim is to improve soccer analysis workflows by providing visual-interactive and data analysis support. We discuss a system for analyzing high-frequency position-based soccer data at various levels of detail, allowing to interactively explore and analyze for movement features and game events. Our Visual Analytics method covers single-player, multi-player, and event-based analytical views combined with annotations of important types of soccer match elements, such as player interaction spaces, free spaces, and pass options. We present a Visual Analytics approach enabling the analyst to find interesting events based on classification. The feedback loop in the Visual Analytics pipeline helps to further improve the classification results. We evaluate our approach by analyzing real-world soccer matches and several expert studies. Quantitative measures show that our annotation methods for pass options and free spaces perform significantly better than na├»ve solutions. As a result, our research lays the foundation for innumerable further and deeper analysis tasks.
Raum: Y25 H92