The naive assumption is that being pressed makes passing harder, so pass completion rates should drop under pressure. In practice, the difference is less than 1%. The reason: pressure changes what players choose to attempt, not just how well they execute. When pressed, players self-select shorter, safer, lower-difficulty passes — so the average completion rate stays stable because the denominator shifts to easier attempts. This is one of the most important lessons in soccer analytics: selection effects can mask true difficulty.
Correct interpretation: a player's raw pass completion rate under vs. not under pressure is an unreliable difficulty signal. Players who look equally good under pressure may be taking completely different types of attempts — one plays a simple sideways ball, the other plays a 30-yard diagonal under a press. The right question is not "did they complete more passes?" but "what did they choose to attempt, and does that choice reveal strength or conservatism?"
under_pressure as a feature to pass completion models and expecting a big coefficient: It won't be — you need attempt-type controls first.Raw pass completion drops less than 1% under pressure because players self-select shorter, safer passes. Pressure doesn't reduce execution quality — it changes attempt type. The real signal is the shift in what players attempt, not whether they complete it.