Every player stat on the site, what it means, and how it's calculated. Hover or tap any metric label elsewhere on the site for the short version; this is the full reference. For the team-level models (power ratings, match odds, playoff odds), see About.
Almost every rate is per 90 minutes played, so a starter and a substitute compare on equal footing regardless of total minutes.
The colored pill is a rank against others at the same detailed position (a defensive mid vs other defensive mids), over a rolling three-season pool. Bright blue is elite, rose is below average.
Elite / Strong / Solid / Limited / Fringe grade overall value on an absolute scale vs positional peers (Elite ≈ a standard deviation above average). Early in a season a thin sample is regressed toward the player’s prior-season form (or a position baseline for players new to the league), fading as their minutes accumulate — so a proven player reads at their level in March, and a hot 3-game sample isn’t crowned “Elite.”
The familiar box-score numbers: goals, chances, and the expected-goals quality behind them. All per 90 minutes unless noted.
Total minutes played this season. More minutes mean a more reliable read on the rest.
League minutes played this season, summed across any in-season transfer. Nearly every other stat is a per-90 rate, so minutes is the sample size behind them — under about 450 minutes (~5 matches) a percentile is noisy and we flag the player as low-sample.
Goals scored, per 90 minutes.
Goals scored in league play (own goals excluded), divided by minutes and scaled to a per-90 rate so part-time and full-time players compare fairly.
Expected goals: the quality of the chances a player gets. A high xG means they're getting into good scoring spots.
Every shot is assigned a scoring probability from its location, type and context; summing those probabilities gives expected goals. Shown per 90, it measures the quality and volume of chances a player creates for themselves, independent of whether they finished.
Shots attempted, per 90 minutes.
Total shot attempts per 90 minutes.
Shots on target, per 90 minutes.
Shots that were on frame (forced a save or scored), per 90 minutes.
Passes that directly set up a shot, per 90 minutes. The playmaking number.
Passes that directly led to a teammate's shot, per 90. The raw playmaking-volume number, regardless of whether the shot was scored or assisted.
Assists (the final pass on a goal), per 90 minutes.
The final pass before a goal, per 90 minutes. Only the assisting pass counts, not earlier build-up.
Expected assists: how many assists the player's passes should produce based on the chances they create.
Each completed pass is credited with the xG of the shot it leads to (if any); summing gives expected assists, per 90. It rewards creating good chances for teammates even when they miss.
Finishing: goals minus expected goals. Positive means scoring more than an average finisher would from the same chances.
Goals minus expected goals, per 90. Positive means a player out-scored what an average finisher would manage from the same chances. Over small samples this is mostly luck and regresses toward zero, so read it as a long-run finishing signal, not a verdict on a handful of games.
All-in-one ratings built from goals-added. The headline g+ Total is what the quality tier (Elite / Strong / …) is cut from; the two-way split separates value on the ball from value winning it back.
Total goals added: every contribution (attack and defense) rolled into one value rating.
The sum of all six goals-added actions (shooting, passing, receiving, dribbling, interrupting, fouling) in one number — a player's total on-ball value in goals per 90. This is the headline rating the quality tier (Elite / Strong / …) is cut from, ranked against others at the same detailed position.
Goals added on the ball — dribbling, passing, receiving and shooting combined. How much a player adds going forward, ranked vs their position.
A grouping we compute: the four attacking goals-added actions (dribbling, passing, receiving, shooting) added together — a player's value with the ball, per 90, ranked against others at the same detailed position.
Goals added winning the ball back — tackles, interceptions, pressure and fouls combined. A holding midfielder's or defender's core job, ranked vs their position.
A grouping we compute: the two defensive goals-added actions (interrupting and fouling) added together — a player's value winning the ball back, per 90, ranked against the same detailed position. This is why a defensive midfielder is measured against other defensive mids rather than buried among attackers.
ASA's goals-added model breaks an outfield player's on-ball value into six action types. Each is the expected-goal value that action adds versus an average player at the same position, per 90.
Goals added from shooting: how much a player's shots help vs an average player.
From ASA's goals-added model: the expected-goal value a player's shooting adds versus an average player at their position, in goals per 90.
Goals added from passing: value created by moving the ball forward.
Goals-added from passing — how much a player's passes move the team toward (or away from) goal versus an average player at the position, per 90.
Goals added from receiving: value from getting open and taking the ball in dangerous spots.
Goals-added from getting open and receiving the ball in valuable areas, versus the position average, per 90.
Goals added from dribbling: value created carrying the ball past defenders.
Goals-added from carrying the ball past opponents, versus the position average, per 90.
Goals added from interrupting: defensive value from tackles, interceptions and pressure.
Goals-added on defense — tackles, interceptions and pressures that end opponent possessions — versus the position average, per 90. This is the main place a defender's or holding midfielder's value shows up in the model.
Goals added from fouling: net value of fouls won vs fouls committed.
Goals-added net of fouls: value gained from fouls drawn minus value conceded from fouls committed, per 90. Usually small in magnitude.
Keeper context numbers. Goals Saved (GSAA) is the headline because it adjusts for how hard the shots were; save %, goals allowed and shots faced give the raw picture.
Goals saved above expected: expected goals faced minus goals allowed, per 90. Positive means stopping more than an average keeper would from the same shots.
Goals Saved Above Average: the expected goals from the shots a keeper faced minus the goals they actually allowed, per 90. Positive means stopping more than an average keeper would from the same shots. The headline shot-stopping number because it adjusts for how hard the shots were.
Share of shots on target saved.
The share of on-target shots a keeper saved. Familiar and simple, but not adjusted for shot difficulty — use Goals Saved (GSAA) for that.
Goals allowed, per 90 minutes. Ranked so fewer goals allowed is a higher percentile.
Goals allowed per 90 minutes. Lower is better, so we invert the percentile — the stingiest keepers rank highest. Heavily influenced by the defense in front of the keeper, so treat it as context as much as keeper skill.
Shots on target faced, per 90 minutes — the keeper's workload (context, not skill).
On-target shots faced per 90 — the keeper's workload. Neither good nor bad on its own; it's context for the save numbers (a high GSAA on heavy volume is a busier, more tested keeper).
Keepers are scored on a separate goals-added model and ranked only against other keepers, across six goalkeeping actions plus the all-in-one total.
Goals added from shot-stopping: saves made vs what an average keeper would, given the chances faced. The core keeper skill.
From ASA's goalkeeper goals-added model: shots saved versus what an average keeper would save given the same shot quality, in goals per 90. Ranked against other keepers only.
Goals added from handling: ball-security on catches, parries and held shots vs spilling them.
Goalkeeper goals-added from secure handling — holding or parrying to safety versus spilling — per 90.
Goals added from claiming: coming off the line to win crosses and high balls.
Goalkeeper goals-added from coming off the line to claim crosses and high balls, per 90.
Goals added from sweeping: defending the space behind the back line as a sweeper-keeper.
Goalkeeper goals-added from defending the space behind the back line, per 90 — the sweeper-keeper skill.
Goals added from distribution: the value a keeper adds building out with their passing.
Goalkeeper goals-added from distribution — the value added building out with passes and goal kicks, per 90.
Goals added from fielding: clean handling of routine balls played back and into the keeper.
Goalkeeper goals-added from cleanly fielding routine balls played back to the keeper, per 90.
Total goalkeeper goals added: every keeper action (shot-stopping, handling, claiming, sweeping, distribution, fielding) in one value.
The sum of all six goalkeeper goals-added actions in one number — a keeper's total value in goals per 90, ranked against other keepers only. The headline rating a keeper's quality tier is cut from.
The columns on the standings and team pages: the table basics (points per game, goal difference), the expected-goals read (xGD, luck), and the PELE-lite team strength (attack, defense, overall rating, tilt).
Average league points per match (3 for a win, 1 for a draw) — compares clubs who've played a different number of games.
Total league points divided by games played. A fairer mid-season comparison than raw points when clubs are a game or two apart on the schedule.
Goals the team has scored this season.
Total goals scored in league play.
Goals the team has conceded this season. Lower is better.
Total goals conceded in league play.
Goals for minus goals against — the simplest one-number read on a season.
Goals scored minus goals conceded. A solid summary of dominance, though it can run ahead of (or behind) the underlying chance quality — compare it to xGD.
Chance quality created minus allowed. A truer read of how a team is playing than the scoreline.
Expected goals for minus expected goals against, built from the quality of chances each side created. Because it strips out hot/cold finishing and small samples, xGD predicts future results better than goal difference does.
Goal difference minus expected goal difference. Positive = scoring more than the chances warrant (likely to regress).
The gap between actual and expected goal difference. A big positive Δ means a team is outscoring the quality of its chances — often hot finishing or goalkeeping that tends to regress; a big negative Δ flags a team playing better than its results.
Goals a team is expected to score against an average opponent, in goal units.
The attacking half of the PELE-lite team rating: goals the team would be expected to score versus a league-average opponent on neutral ground, fit on a 70% xG / 30% goals blend with time decay.
Goals a team is expected to concede against an average opponent. Lower is better.
The defensive half of the PELE-lite team rating: goals the team would be expected to concede versus a league-average opponent. Lower is stingier.
Net team strength in goal units — attack minus defense vs an average team.
The team's overall PELE-lite strength: the attack rating minus the defense rating, in goals versus a league-average opponent. This is what the power ranking and the match predictor are built on.
Which way a team leans — positive is attack-tilted, negative is defense-tilted.
How a team's strength is distributed between attack and defense. A positive tilt means the team's edge is its attack; negative means it's built on defense. It says nothing about overall quality — only style.
Goals-added (g+) and the underlying data come from American Soccer Analysis. The two-way composites (possession value, ball-winning) and the position-relative tiers are computed by NWSL Labs from that data.