Flexible marked spatio-temporal point processes with applications to event sequences from association football

We develop a new family of marked point processes by focusing the characteristic properties of marked Hawkes processes exclusively on the space of marks, allowing a separate model specification for the occurrence times. We develop a Bayesian framework for their inference and prediction that can naturally accommodate covariate information to drive cross-excitations, offering broad flexibility for real-world applications. The framework is applied to in-game event sequences from association football, resulting in inferences about previously unquantified characteristics of game dynamics, extraction of event-specific team abilities and predictions for event occurrences, such as goals or fouls in a specified interval of time.