This paper summarizes the current results of a 3-year project that applies data science methods in the context of a traditional team ball game. Based on a simple model of the game it will be described how data was collected and analyzed, including a description of the concepts that have been used. An important aspect of the paper is the new perspective of looking at team-handball data. We use sport-specific events like goals and passes to analyze the stream of indicators rather than abstract concepts like the time or just the simple sequence of events. Thus, the discovered patterns become actionable for coaches.
© Copyright 2020 Veröffentlicht von Duale Hochschule Baden-Württemberg Stuttgart. Alle Rechte vorbehalten.