The first installment of a planned football series argues finishing skill exists but is hard to isolate: goals above expected goals (xG) combine shooting, goalkeeping and luck, while xG baselines may reflect skilled scorers and seasonal results are highly noisy.
Finishing can mean the narrow act of striking the ball, or the broader ability to outperform xG through shot selection, decision-making and weak-foot use. A 2024 paper by Jesse Davis and Pieter Robberechts argued that xG training data can bake player skill into its baseline, distorting estimates.
For 80 non-penalty shots averaging 0.12 xG, expected goals total 9.6; the author says even an average finisher might score roughly 4 to 15. A 10% better conversion rate would mean about one extra goal, against roughly three goals of standard deviation. The writer plans to test a new model, while acknowledging it may be no more reliable than goals minus xG or conversion rate.
