Answers are grounded in this paper and cite their sources.
This study examined how shooting profiles varied jointly by match status and match period in Europe’s five major domestic leagues. The sample included 17,965 matches and 404,386 regular-time non-penalty shots from 2015/16 to 2024/25, of which 384,822 had linked Understat expected goals (xG). Pre-shot match status was classified as drawing, leading, or trailing, with regular time divided into six 15-min periods. Poisson regression estimated exposure-adjusted shot rate; shot-level models estimated mean xG, the probability of xG ≥0.10, and shot composition. Point estimates were largest in the final 15 min; cross-period tests supported this except for trailing-team opportunity quality. Relative to drawing, leading was associated with fewer shots (−2.70 per 90 min) but higher mean shot xG (+3.97 per 100 shots) and probability of xG ≥0.10 (+13.94 percentage points). Trailing was associated with more shots (+2.60) but lower mean shot xG (−0.99) and probability of xG ≥0.10 (−4.13). In that period, leading shots were more often inside the penalty area, big chances, or fast breaks, whereas trailing shots were more often set pieces or headers. These directions persisted across sensitivity analyses. Shooting profiles should therefore be interpreted by separating exposure-adjusted shot rate from per-shot opportunity quality and composition.
No takes yet. Share an insight, caveat, or question.
Peng et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: