Over the last years, more and more sport related data are being collected, stored, and analyzed to give valuable insights. Football is no exception to this trend. An important way of identifying a team`s "style" of play is through analyzing passing sequences. However, passing sequences either concentrate on the specific players involved or the structure of passes and ignore where these sequences took place. In this paper, we focus on identifying frequent passing zone subsequences that lead to created or conceded goal scoring opportunities. We partition the pitch into a set of disjoint zones and apply sequential pattern mining. Our experimental study on the 2020/21 Danish Superliga season shows that our method is able to predict goal scoring opportunities better than random subsequences that occurred, in median, 99.5% of the cases.
© Copyright 2022 Machine Learning and Data Mining for Sports Analytics. MLSA 2022. Communications in Computer and Information Science. Published by Springer. All rights reserved.