Statistics Glossary

A statistic normalised to a full match, so players with different minutes can be compared. A player with 5 goals in 900 minutes and one with 10 goals in 1,800 minutes both score 0.50 goals per 90.

Where a player ranks against peers, from 0 to 100. The 90th percentile means better than 90% of the comparison group. We compare players against the same position, in the same season — a centre-back's tackle numbers are never ranked against a striker's.

Players need at least 450 league minutes before percentiles are calculated. For metrics where less is better (for example dispossessed), the scale is inverted so a high percentile is always good.

The playing-time threshold — 450 minutes, roughly five full matches — below which a player is excluded from percentile rankings. Rate statistics from short cameos are noisy and would distort every leaderboard.

An algorithmic 0–10 performance score for each appearance, provided by our data supplier and averaged across the season. Brief substitute appearances with too few events to score are excluded from the average rather than counted as zero.

Matches in which the player was named in the starting eleven, taken directly from confirmed team sheets — not inferred from minutes played.

The probability that a shot results in a goal, given the quality of the chance. A penalty is worth about 0.79 xG; a speculative 30-yard strike might be worth 0.02. A player's season xG is the sum across all their shots — what an average finisher would have scored from the same chances.

Expected goals with penalty kicks removed. Penalties are high-value chances awarded rather than created, so npxG better reflects the chances a player earns through open play and set-piece movement.

Post-shot expected goals: the chance is re-evaluated after the ball leaves the boot, using where the shot was placed in the goalmouth. A shot arrowed into the top corner carries a much higher xGOT than its xG; a tame effort straight at the keeper carries less. Only shots on target have an xGOT value.

The average quality of chance a player shoots from. High values indicate disciplined shot selection — getting into good positions and declining low-value efforts.

Shot selection with penalties stripped out of both sides of the ratio — the cleanest read on the quality of positions a player works themselves into.

Finishing over- or under-performance. A positive value means the player scored more than an average finisher would from the same chances; a sustained positive gap across seasons is the signature of an elite finisher. Single-season values contain a lot of variance.

Shot placement quality. Because xGOT re-scores each shot on where it ends up in the frame, a positive difference means the player systematically improves their chances with placement — finding corners rather than hitting shots close to the goalkeeper.

What remains after accounting for both chance quality and placement — largely a measure of goalkeeper performance against the player's shots, plus deflections and fortune. Best read alongside xGOT − xG rather than alone.

Threat generated relative to involvement. Penalty-box poachers post high values from few touches; deep-lying creators post low values despite heavy involvement. Useful for separating player profiles, not for ranking outright quality.

All goal attempts. On target means the shot would have entered the goal without intervention — saved attempts count, blocked ones do not. Off target covers everything that misses the frame, including woodwork.

The share of a player's shots that test the goalkeeper.

The share of shots that become goals. Strongly influenced by shot selection — compare with xG per shot to tell clinical finishing apart from cherry-picked chances.

Failures to score from situations where a goal would reasonably be expected — one-on-ones, close-range finishes, open goals. High volumes can still indicate a player who gets into excellent positions; read alongside big chances received.

Shots that struck the post or crossbar without resulting in a goal.

Goals plus assists per 90 minutes — the bluntest but most direct summary of attacking output.

The share of attempted passes that reach a teammate, computed from season totals rather than averaged match percentages. Always read with volume and risk in mind: a centre-back recycling possession will out-score a No.10 attempting defence-splitting passes.

Passes that lead directly to a teammate's shot. An assist is a key pass whose shot was scored.

Key passes plus assists — every occasion the player's final ball produced a shot, regardless of whether it was converted.

Passes that put a teammate into a situation where they would reasonably be expected to score. A stricter, higher-value subset of chances created.

Passes threaded between defensive lines into a teammate's run behind the defence. We show attempted and completed volumes plus the completion rate — the highest-risk, highest-reward pass type in the game.

Passes of roughly 25 metres or more, attempted and completed. High accuracy on strong volume marks out players who can switch play and release runners — a distinct skill from short-range retention.

Balls delivered from wide areas into the penalty box, with the share that found a teammate. Cross accuracy is typically low league-wide (20–30%), so small differences are meaningful.

Passes completed inside the attacking third of the pitch. The companion final-third pass rate — the share of a player's total passes played there — separates players who live in dangerous areas from those who circulate deep.

The percentage of a player's passes played backwards. Not inherently bad — recycling under pressure is a skill — but a useful texture stat for reading how progressive a passer's game is.

Every time the player contacts the ball. The base unit of involvement — most per-touch metrics divide by this to control for how much of the game runs through a player.

Attempts to carry the ball past an opponent, and the share that succeed. Elite volume dribblers usually sit between 45% and 60% — success rate must be read against attempt volume, since conservative dribblers protect their percentage.

In a small number of leagues the data provider under-reports failed attempts. Where our integrity checks flag this, dribble stats are withheld rather than shown inflated.

Times an opponent tackled the ball away while the player was in possession. Excludes miscontrols and failed passes — this is specifically losing a duel on the ball.

Every action that surrendered the ball to the opposition — failed passes, poor touches, losing duels, unsuccessful dribbles. The broadest measure of ball security; always higher for players asked to attempt risky actions.

Cheap losses of possession from a poor touch or being caught on the ball — a narrower, more culpable subset of possession lost.

Fouls won from opponents. Persistent foul-winners relieve pressure, earn set pieces in dangerous areas and draw cards — an under-rated component of ball-carrying value.

Challenges made to dispossess an opponent, whether or not the tackler's team kept the ball.

The share of a player's tackles that won the ball. Separates clean, timed tacklers from players who lunge into contact.

When a dribbler ran at this player, how often did the defender win? Combines tackles made with times dribbled past into a single duel-survival rate — one of the most honest one-on-one defending measures available.

Reading a pass and cutting it out before it reaches its target. Rewards anticipation and positioning rather than physical duelling.

Opposition shots blocked with the body before reaching the goal.

Kicking or heading the ball away from a dangerous area without securing possession. High volumes usually signal a team defending deep.

Winning back a loose ball — collecting possession where no duel was contested. A workrate and positioning stat that credits midfielders who consistently arrive first.

Total defensive workload in one number: tackles, interceptions, blocked shots and clearances combined, per 90.

Times an opponent carried the ball past the player. Judge against exposure: full-backs isolated against wingers face far more of these situations than central defenders.

All contested 50-50s for the ball — ground and aerial — with the share won. Computed from season totals, not averaged match percentages.

Contested headers, won and lost. Win rate matters most in the boxes, where a single aerial loss can concede a chance.

Duels contested on the floor — total duels minus aerials — with the share won. Isolates one-on-one strength for players whose game rarely leaves the ground.

Mistakes — a misplaced pass out from the back, a miscontrol, a slip — that directly handed the opposition a shot, or a goal. Rare events, but heavily scrutinised for good reason.

Tackles made as the final defender, with no cover behind. Low-frequency, high-value interventions.

Opposition attackers caught offside by the player's defensive line. A proxy for line discipline and coordinated stepping.

Tackles per 90, adjusted match-by-match for the team's share of possession. A holding midfielder at a dominant club is no longer buried below every defender from a low-possession side.

Interceptions per 90 on the possession-adjusted scale. Particularly revealing for defenders at top clubs, whose raw interception counts are suppressed by rarely being out of possession.

Blocked shots per 90, adjusted for how often the player's team is actually defending shots.

Clearances per 90 on the possession-adjusted scale, correcting for the deep-block inflation that dominates the raw leaderboard.

Ball recoveries per 90, adjusted for possession context.

Times dribbled past per 90, adjusted for how frequently the player's team faces attacking runs. The fairest beaten-in-the-duel comparison across teams of different styles.

The combined defensive workload metric — tackles, interceptions, blocks, clearances — on the possession-adjusted scale. Our headline number for comparing defensive activity across any two teams in any two systems.

The reverse correction, applied to on-ball volume. Touch and pass counts are deflated for players on ball-dominant teams and lifted for those starved of possession — revealing who would still be heavily involved on a team that saw less of the ball.

The share of the team's passes made by this player while on the pitch. Values around 15% or higher mark genuine hubs — typically deep playmakers and ball-playing centre-backs the entire build-up flows through.

The share of the team's shots taken by this player while on the pitch. Identifies focal-point attackers — and flags teams dangerously dependent on one shooter.

The share of the team's expected goals accounted for by this player's shots while on the pitch. A quality-weighted version of shot involvement: high xG share on modest shot share means the player takes the team's best chances.

Distinct from the team metric of the same name in section 11, which measures a team's share of all xG in its matches.

The share of the team's goals this player scored or assisted while on the pitch. The classic talisman measure — how much of the final product carries this player's fingerprint.

On-target shots stopped. Raw save volume mostly reflects how busy the goalkeeper is — a keeper behind a poor defence will always rank high. Rate and quality-adjusted measures below tell you more.

The share of on-target shots faced that were saved. League averages sit around 68–72%; sustained numbers above 75% are elite. Not adjusted for shot difficulty — see goals prevented for that.

Goals the keeper stopped relative to expectation, based on the post-shot expected goals (xGOT) of every on-target shot faced. Positive values mean the keeper conceded fewer goals than an average keeper would facing identical shots — the fairest single shot-stopping measure.

Shown only for matches where post-shot data is available; coverage varies by competition.

Saves made from shots inside the penalty area, and their share of all saves. Close-range shots are the hardest to stop, so a high inside-box share adds context to any save-percentage figure.

Crosses and high balls caught or claimed cleanly. Measures command of the area — the aerial dominance side of goalkeeping that shot-stopping numbers ignore.

Penalty kicks stopped. Around one in four penalties is missed or saved, so even small career tallies are meaningful.

Free kicks conceded. Some fouling is tactical and valuable; judge alongside the rate below.

How often defending tips over into fouling. Divides fouls committed by total defensive actions, separating clean high-volume defenders from players whose activity comes with a card risk attached.

Cautions and dismissals across the season. Suspension risk is a real selection cost for otherwise similar players.

Times caught offside. High counts usually accompany aggressive line-running forwards — a cost of the runs that also produce chances.

The share of their team's available minutes the player was on the pitch for across the season. The single best availability-and-trust measure: it combines selection, fitness and suspension into one number.

The share of the team's matches this player started, from confirmed team sheets. Read with minutes share to separate nailed-on starters from rotation options and super-subs.

The share of in-play time a team controls the ball. A style descriptor rather than a quality measure on its own — but the essential context for reading volume statistics, and the input that powers every possession-adjusted player metric in section 07.

The chance quality a team creates and concedes, summed per match and per season. The xG For minus xG Against balance predicts future results better than the actual scoreline does — it strips out finishing and goalkeeping variance on both ends.

A team's xG split by how the chance originated: from the run of play, or from corners and free kicks. The companion set-piece xG share — the percentage of total xG generated from set plays — flags teams whose threat is disproportionately dead-ball dependent, which behaves differently under fixture congestion and personnel changes.

Territorial dominance: the share of all dangerous attacks in a team's matches that belong to that team. Possession tells you who has the ball; field tilt tells you where the game is being played. A team can hold 55% possession yet lose the tilt battle if its possession is sterile and deep.

Game control measured by chance quality: the share of all xG in a team's matches that the team itself created. Above 50% means creating more than conceding; sustained values around 60% or higher are title-contender territory.

Distinct from the player metric of the same name in section 08, which measures one player's share of their own team's xG.

Directness: how many passes a team plays for each shot it takes. Low values mean fast, vertical attacking; high values mean patient build-up. Neither end of the scale is better — this is a style axis, useful for matchup analysis and squad-fit questions.

Total shots faced. The volume half of defensive quality — read with xGA per shot, which covers the quality half.

The average quality of chance a team concedes. A low value means opponents are forced into speculative, low-value shots; a high value means the defence leaks clean chances even if it limits volume. Together with shots against, this decomposes xGA into how often and how dangerously a team is opened up.

Saves as a share of on-target shots faced across the season — the goalkeeping line of the team's defensive ledger, aggregated over every keeper used.

The share of a team's shots taken inside the penalty area. A shot-diet quality signal: teams that work the ball into the box generate more xG per shot than teams that settle for range-finding efforts.

Fouls committed and yellow and red cards accumulated across the squad. At team level these read as much as a tactical signature — pressing intensity, tactical fouling — as a discipline problem.