ABSTRACT
The use of more localised spatiotemporal data is commonplace in sports, particularly in sports performance analysis. The distinction between spatial and temporal aspects is made, although the temporal aspects often involve changes in spatial data over time. Vast volumes of GPS data and other player- and ball-tracking data can be generated quickly and used to analyse tactical movement as well as work rate within sports. This chapter discusses a range of approaches to analysing such data from algorithmic methods to neural networks for identifying tactical patterns. Such algorithms and neural networks have limitations and identify many false positive events as well as fail to identify actual events within player and object movement data. Therefore, there is a clear role for human verification during data analysis as well as decision making.
