An Introduction of Explainable Artificial Intelligence to Winter Sports Performance Analysis

In recent years, the utilization of machine learning techniques has revolutionized performance analysis in the field of sports. However, a crucial drawback of these algorithms is their inherent black box nature, which hinders stakeholders' ability to comprehend the underlying rationale behind their decisions. This lack of transparency presents challenges for coaches, trainers, and athletes who heavily rely on analysis outcomes to shape their strategies and training approaches. To address this limitation, this paper introduces the concept of Explainable Artificial Intelligence (XAI) to sports performance analysis. We employ SHAP (SHapley Additive exPlanations), a model-agnostic explainable method, to provide a clear and interpretable representation of the contribution of different features towards the final model prediction. Ski jumping is chosen as the initial test case to illustrate the effectiveness of this approach.