Towards Automated GIS-based Analysis of Scoring Attempt Patterns in Association Football

This article is about GIS-based football game analysis in general and focusses on an approach towards an automated spatiotemporal analysis of scoring attempt patterns based on football-specific geo data, which were kindly provided by Prozone Sports. The main objective of this paper is to present our developed Analyse Scoring Attempts tool, which corresponds to a Python script but is designated for the execution as a custom tool for ArcGIS for Desktop 10.x. In this context we attempt to portray the functionality of that tool in as much detail as possible. We also demonstrate that by the application of that tool it is possible to gain the following information automatically: First, the number of players involved in a scoring attempt sequence. Second, the length of that sequence. Furthermore, the duration between the capture of the ball and the final shot. Moreover, both the initial as well as the finishing position of each sequence are intersected with special pitch zones. In addition to that the developed Spatial Scoring Probability Index (SSPI) is determined for the shooting position. This pure theoretical index indicates the probability of scoring a goal and is composed of weighted values for distance and angle of the shooting position in relation to the goal. Besides, the tool’s textual and graphical outcome are presented. Finally, this paper is concluded with a brief discussion about upcoming challenges.