Thread by @FlorisgoesF

Tactical behavior in #Football has a spatial and a temporal component, and results from interaction with the opponent. It’s key to account for all these aspects in data-driven tactical analysis, as well as to respect the complexity of the temporal and spatial dimensions 🧵

Two years ago I published a systematic review in @EurJSportSci on using big data in #soccer for tactical performance analysis that illustrates the associated challenges and provides a data-driven scientific framework. #DataScience tinyurl.com/mrxky6ca

The most common analysis issue is the fact that spatial and/or temporal complexity is not respected. For example by aggregating data over multiple minutes, or constructing spatial features aggregating 11 player positions into a single variable.

Another issue is the focus on point-in-time KPIs in outcome assessments. A game of #football is not a serie of static snapshots but rather a time-series of collective and individual behavior and should ideally be treated as such.

I proposed a more fine-grained approach. Constructing spatial features over small subsets of players, and aggregating data over small tactical game phases, like a single attack. Check for example my publication in @JSportsSci tinyurl.com/2s3a9ysf #DataScience #Analytics

Finally. Tactical behavior in #Football emerges from interaction between players / teams. Years of research have shown two teams in a game are strongly coupled on all dimensions, and behave as complex dynamical systems.

This means teams typically move in “in-phase” or “anti-phase” synchrony, but rarely asynchronous. In other words, every action sparks an opposing reaction. If you space the field, the opponent will follow to some extent, and the other way around.

This coupling can be assessed through relative phase analysis, which is the dominant framework in sports #science literature on tactical behavior in #Soccer. See for example @benedictlow review. pubmed.ncbi.nlm.nih.gov/31571155/

Curious how to conduct relative phase analysis and how to do statistical analysis on these time-series? Later this week I’ll share threads on relative phase analysis in #Soccer as well as circular statistics. #DataScientist #PhD