Predicting Team Performance with Spatial Temporal Graph Convolutional Networks
This addresses sports analytics challenges like coaching and opponent modeling, but it is incremental as it applies a hybrid method to a specific domain.
The paper tackles predicting team performance from agent behavioral traces using Spatial Temporal Graph Convolutional Networks (ST-GCN), achieving superior performance over other classification techniques in predicting game scores from short segments of player movement and game features.
This paper presents a new approach for predicting team performance from the behavioral traces of a set of agents. This spatiotemporal forecasting problem is very relevant to sports analytics challenges such as coaching and opponent modeling. We demonstrate that our proposed model, Spatial Temporal Graph Convolutional Networks (ST-GCN), outperforms other classification techniques at predicting game score from a short segment of player movement and game features. Our proposed architecture uses a graph convolutional network to capture the spatial relationships between team members and Gated Recurrent Units to analyze dynamic motion information. An ablative evaluation was performed to demonstrate the contributions of different aspects of our architecture.