Every report on this platform is full of short, technical-looking labels — xG, xT, PPDA, "progressive passes," "possession-adjusted." If you don't follow football data closely, they can look like noise. They're not — each one answers a specific, plain question about what a player or team actually did. This is that translation, written for someone who knows neither football tactics nor data science.

The "how good was that chance?" numbers

xG — Expected Goals. Not every shot is equally likely to go in. A tap-in from a yard out is far more likely to score than a strike from 30 metres against three defenders. xG rates every shot's quality on a scale from 0 to 1 — a chance rated 0.80 xG should be scored roughly 8 times out of 10 by an average finisher. Add up a player's xG over a season and you get a truer picture of the chances he's had than the raw shot count. If his actual goals are well above his xG, he's finishing better than expected; well below, and he's missing chances a typical player would score.

xA — Expected Assists. The same idea, for the pass before the shot. It measures how good the chance was that a player created for a teammate — regardless of whether the shot actually went in. A player can have a high xA even in a season with few actual assists, if his teammates keep missing the chances he creates.

xT — Expected Threat. The broadest of the three. Instead of just shots, xT scores every pass or carry by how much closer it moved the ball to a genuinely dangerous position — using a model of the pitch that knows some zones (like inside the box) are far more likely to lead to a goal than others (like your own half). A pass from the halfway line into the penalty box adds a lot of xT even if no shot follows immediately; a sideways pass in your own third adds almost none.

The "did he actually advance the game?" numbers

Progressive passes / carries. A pass or dribble that moves the ball meaningfully closer to the opponent's goal — not just any forward pass, but one that makes real territorial progress. This separates players who just circulate the ball safely from players who actually push their team up the pitch.

Line-breaking passes. A pass that goes past a line of opposition players — for example, splitting the midfield to find a striker, bypassing defenders who were positioned to stop exactly that. These are riskier and rarer than an ordinary pass, and far more valuable when they come off.

Box entries / final-third entries. Simply: how often does the ball (via this player) reach the penalty box, or the attacking third of the pitch? The closer to goal, the more dangerous — this counts how often a player is personally responsible for getting the ball into the danger zone.

Key passes. Any pass that leads directly to a shot by a teammate, whether or not that shot scored.

The "how does the team play?" numbers

PPDA — Passes Per Defensive Action. A measure of how hard a team presses. It counts how many passes the opponent is allowed to make, on average, before the team makes a tackle, interception, or challenge. A low PPDA (e.g. 7) means the team barely lets the opponent breathe before winning the ball back — an intense, high press. A high PPDA (e.g. 20+) means they sit off and let the opponent pass freely until deeper in their own half.

Field Tilt. What share of the play near either box happens in the opponent's third versus your own. A team with 65% field tilt spends most of the match camped in the opponent's half — it's a good single-number gauge of territorial control, separate from possession percentage (a team can have the ball a lot in its own half and still have low field tilt).

The "compared to what?" numbers

Percentile. Nearly every number in a report is shown alongside a percentile — e.g. "92nd percentile among centre-backs." This means the player ranks higher than 92% of other players in the same position and league on that specific metric. A raw number alone is meaningless without knowing what's normal for that role; the percentile supplies that context instantly.

Possession-adjusted (PAdj). A defender on a dominant, possession-heavy team simply gets fewer chances to tackle — the ball is at their opponent's feet less often. PAdj metrics correct for this by scaling a player's defensive numbers according to how much the opposition actually had the ball while he was on the pitch, so a defensive midfielder at a possession-heavy club isn't unfairly marked down next to one at a team that spends the whole match defending.

The platform's own layer

Style scores. Vision AI doesn't just show raw counts — it blends related metrics into named football qualities (Progression, Retention, Press Resistance, Creativity, Ball-Winning, and others), each scored 0–100 against players in the same position.

Archetypes. A short, scout-friendly label — "Ball-Playing Centre-Back," "Deep-Lying Playmaker," "Inside Forward" — assigned based on which combination of style scores a player's profile most resembles, with a confidence percentage attached.

The Playstyle Wheel. A team-level version of the same idea: fifteen metrics grouped into five identity domains — how much a team Possesses, Disrupts, Finishes, Presses, and Defends — each percentiled against the rest of the league, so a team's whole tactical identity is visible in one shape.

None of these numbers replace watching the game. They exist to answer one question honestly: given everything that happened on the pitch, what does the data actually show — and how confident should you be in it?