Pass-Regime and Possession-Episode Modelling Data for CTMC/SAP Analysis of Football Passing Dynamics

This dataset contains derived pass-regime and possession-episode modelling data for continuous-time Markov chain (CTMC), stochastic approximation procedure (SAP), and Semi-Markov analysis of football passing dynamics.

The data were derived from An integrated dataset of spatiotemporal and event data in elite soccer by Bassek, Rein, Weber, and Memmert, originally published in Scientific Data and available via Figshare: https://doi.org/10.6084/m9.figshare.28196177. The associated data descriptor is available at https://doi.org/10.1038/s41597-025-04505-y.

Starting from the original synchronized German Bundesliga event and position data, pass-level source tables were extracted from the canonical event stream and aligned match metadata. Passes were classified into observable tactical regimes: circulation, progression, threat, shot/terminal, loss/turnover, and recovery/reset. These regime events were then transformed into possession-level sequences, within-possession transitions, between-possession links, fitted CTMC generators, SAP-refined generators, Semi-Markov companion models, and model comparison summaries.

This package includes the compact modelling inputs and outputs used in the repository, including source pass/event tables, derived regime and possession tables, fitted model JSON files, summary reports, and checksums. Full raw player/ball tracking tables are not included, because they are large upstream inputs and are not required to reproduce the modelling tables deposited here.