Home/Soccer Analytics/Pressure Event Data Structure

Pressure Event Data Structure

Event Data ModelsLevel 1 — Novice

Unlocks

10-Second Ball Path After PressureBall Receipt Under Pressure & Intent TrackingEvent Type Pressure Rates (~120/game benchmark)Pass Completion Under Pressure ParadoxPressure Duration & Direction VectoringTwo Types of Pressure (Angle-Block vs. Physical)

What It Is

A pressure event is a discrete record capturing when a defending player actively closes down, tracks, or blocks the passing angle of a ball-carrier. Each event contains the pressing player's identity, a timestamp, pitch position, and a duration field marking how long the presser is engaged. Every on-ball action (pass, shot, dribble, ball receipt, carry) that occurs within that duration window is tagged "under pressure" in the event log.

Correct Execution

When working with pressure-tagged event data correctly: every on-ball action in the dataset has a boolean under_pressure field; pressure records are linked to the presser (not the pressured player); duration is measured in seconds from engagement start to disengagement; approximately 120 events per match will carry the under pressure tag in a typical top-division game.

Progression Levels

Diagnostic Tree

Coaching Cues

  • "~120 under-pressure actions per match is your sanity-check benchmark." — Thom Lawrence, 2018
  • "Duration is the window; every action inside that window gets the tag." — Thom Lawrence, 2018

Common Errors

  1. Using presser-side records to analyze the pressured player: Pressure event rows describe the presser. To analyze the pressured player, use the under_pressure flag on their action rows → join back to pressure records for intensity context.
  2. Assuming one pressure event = one pressured action: Duration windows can span multiple actions; one presser can generate multiple tagged actions.
  3. Ignoring multi-presser scenarios: When 2+ defenders press simultaneously, the pressured player has multiple overlapping pressure records — aggregate before modeling.

Sources

  • Thom Lawrence, StatsBomb Data Launch presentation, YouTube, 2018-05-23 — introduced the pressure event schema, duration field, multi-presser design, and 120/match benchmark
  • Ted Knutson, Barcelona Coach Analytics Summit, YouTube, 2018-11-18 — introduced the term "AUPs" (Actions Under Pressure) as the collective label for pressured on-ball events; described using AUPs for in-game tactical analysis and halftime adjustments, not just player profiling
  • Ted Knutson & Siqur Arshad, WFS 2019 StatsBomb presentation, YouTube, 2019-10-02 — identified pressure data as a critical data quality differentiator between providers; without pressure, defensive activity visualizations show fundamentally misleading pictures (e.g., Real Madrid's defensive shape changes dramatically when pressure events are added vs. removed)