StarkPitch
Sign up

Concepts

8 min · 2026-08-18

Possession value without hype

Possession percentage says who had the ball. Possession value asks whether the ball was useful. Keep the question modest.

New here? This page uses terms like xG, PPDA, and progressive passes. The free Foundations lesson explains them in a few minutes. Continue on the Learn path →

Two different questions

A team can have the ball and create little. Field Tilt, progressive passes, and xG may all stay modest. That pattern is sterile control. Possession-value thinking stops “more of the ball” becoming “better attack.”

xT, VAEP, OBV, and related models try to value state changes. They are research tools with assumptions. They do not replace basic observation in open event data. State progression, shots, phases, and turnovers first.

How to practise without a vendor model

StarkPitch Match Analysis already gives useful proxies. The xG Timeline shows when shot value arrived. Half splits show whether the process changed. Progressive passes show whether the ball moved toward goal. The transition board shows whether recoveries were followed by advance.

That stack is enough for a calm possession-value paragraph. Example: “Spain moved the ball forward often, but the xG Timeline barely rose. Morocco’s recoveries sometimes progressed without a matching shot process.” You do not need a VAEP leaderboard for one knockout match.

Hype to refuse

Refuse rankings built from an unspecified “value” metric. Refuse using possession value to clean up a scoreline. “They lost, but value says they won” is too strong. Refuse tracking-level control claims from event data alone.

Allow “useful ball” as a teaching question. Allow disagreement between possession, progression, and xG. Say a model is approximate when you do not have its full specification.

Key points