xG over time

By match orders each club’s covered fixtures chronologically across the selected season range. The x-axis is the club’s match number, not the scheduled matchday. Per match shows that fixture’s xG and xGA; cumulative total adds all matches to that point; cumulative average divides the running total by matches played to that point. Clubs without coverage have gaps excluded, and shorter series stop at their last covered fixture. The 5-match rolling average uses the current fixture and up to four preceding covered fixtures within the selection, including across season boundaries. Early points average only the matches available so far. Minute mode pools the selected seasons by covered matches.

Managerial hire markers are optional in By match. They show the first covered league fixture on or after a recorded appointment, with the name and appointment date. Summer hires belong to the upcoming campaign; earlier appointments and spells without a covered fixture are omitted. Caretaker appointments are included. Dates come from the historical manager register, with current appointments checked against the Premier League manager list (5 November 2026 snapshot).

Each shot contributes its xG to the attacking team and the same value as xGA to its opponent. Penalties are included. Cumulative average sums the recorded minute bins and divides by covered matches. Cumulative total sums those bins across all covered matches without division. Per-minute mode shows each bin separately, divided by covered matches. Averages include matches where a team recorded no shot at that minute.

Shots are ordered by half: 1–45, 45+1 onward, 46–90, then 90+1 onward. Added-time minutes retain their own bins; the half-time break is omitted. Own goals have no xG and are excluded. Timing corrections move the original xG into its recorded minute; they do not recalculate chance quality. The audit records the original clock, corrected clock, player, xG, evidence and matching method. Where a missing event falls between two independently assigned shots in the same half, its period follows that bracket; these cases are marked as inferred in the audit. Manual report checks establish the half; where a report does not specify the exact minute, the original Understat elapsed clock supplies it. Selected seasons combine both sources into one series per team. Understat and FotMob use different xG models; values are pooled as supplied, without recalibration.

Historical first-half added time

Shots and original xG allocated to individual 45+ minutes. Season totals are preserved.