maskegger / quantifying-soccer-pressure

Project evaluating player skill under pressure using Statsbomb public event level data

Quantifying_Pressure

Project evaluating player skill under pressure using Statsbomb public event level data.

The main idea of this project is to use Expected Threat (xT), as introduced by Karun Singh here, to quantify how soccer players react to defensive pressure.

The GatheringEventDataFromStatsbomb.R script takes JSON match event files from the StatsBomb public repository and produces a csv with the events of interest (shots, passes, dribbles, turnovers). CreatingXThreat.R takes these events and calculates Expected Threat values for field locations and the xT-added for each event, found in xT_locations.csv and Statsbombevents_xTadded.csv, respectively. The Descriptive_Plots.R and PlayerEval.R scripts take the event and xT-added data and produce high level metrics and rankings for players.

See the QuantifyingPressure.pdf report for more detail.