In a world full of data, Decision Support Systems (DSS) based on ML models have significantly emerged. A paradigmatic case is the use of DSS in sports organisations, where a lot of decisions are based on intuition. If the DSS is not well designed, feelings of unusefulness or untrustworthiness can arise from the human decision-makers towards the DSS. We propose a design framework for DSS based on three components (ML model, explainability and interactivity) that overcomes these problems. To validate it, we also present the preliminary results for a DSS for rival team scouting in basketball. The model reaches state of the art performance in game outcome prediction. Explainability and interactivity of our solution also got excellent results in our survey. Finally, we propose some lines of research for DSS design using our framework and for team scouting in basketball.
© Copyright 2024 Machine Learning and Data Mining for Sports Analytics. MLSA 2023. Communications in Computer and Information Science. Veröffentlicht von Springer. Alle Rechte vorbehalten.