Problem
Research questions and friction points this paper is trying to address.
Spatio-temporal tracking data
Match phase classification
Football tactical analysis
Multi-player interaction modeling
Automated Detection of Match Phases in Football from Spatio-Temporal Tracking Data Using Graph Neural Networks
🤖 AI Summary
This study addresses the challenges of modeling complex multi-player interactions and automatically classifying tactical phases in soccer matches. To this end, it proposes a spatiotemporal modeling framework that integrates Graph Neural Networks (GNNs) with Long Short-Term Memory (LSTM) networks. Specifically, the method innovatively employs Delaunay triangulation to construct dynamic graph structures and introduces a novel SEAConv layer that injects edge attributes into node representations, enabling second-by-second tactical recognition from spatiotemporal data. Experimental results demonstrate that the proposed approach outperforms baseline models by 4.6% in macro F1-score on a seven-class tactical phase classification task. Furthermore, incorporating Integrated Gradients attribution analysis enhances model interpretability. Overall, this work provides an effective solution for fine-grained tactical analysis and automated annotation in sports analytics.
📝 Abstract
Spatio-temporal tracking data has opened new possibilities for detecting complex tactical patterns in football, yet modeling the interactive movements of multiple players remains challenging. This paper proposes a framework combining graph neural networks (GNNs) with a sequential model to classify match phases on a second-by-second basis across a seven-class taxonomy. Match phase classification is tactically meaningful, and the availability of rule-based labels across 203 matches makes it a suitable testbed for a systematic comparison of adjacency constructions and message-passing layers, a question that has received limited attention in existing research. Our selected GNN-LSTM model outperforms all aggregated-feature baselines, including XGBoost and a Long Short-Term Memory (LSTM) network, as the strongest baseline scores 4.6% lower in macro F1. Graph representations using a domain-informed Delaunay triangulation that approximates passing lanes, paired with a custom Spatial Edge-Augmented Convolution (SEAConv) layer that injects edge attributes directly into messages, achieve the best performance by capturing spatial dependencies while limiting uninformative messages from redundant edges. Integrated Gradients attributions indicate the model uses spatial player configurations, particularly horizontal positioning, in combination with possession and ball-status indicators. This work offers an automated solution for fine-grained tactical analysis, reducing the need for manual tagging and providing deeper insight into dynamic team behavior.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Graph Neural Networks
Spatio-Temporal Tracking Data
Match Phase Classification
Delaunay Triangulation
Spatial Edge-Augmented Convolution
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
No related jobs found.
