Football Stats Explained: Possession, PPDA, Progressive Passes and More

Football stats explained in one line: possession tells you who had the ball, expected goals (xG) tells you how good their chances were, PPDA tells you how hard a team pressed, and progressive passes and carries tell you who actually moved the ball towards goal. Each number answers a different question, and none of them tells the whole story on its own.
The trick is knowing which question each stat answers, and which it doesn’t. This guide walks through the numbers you’ll see on match graphics, in podcasts and on data sites, what they measure, how they’re calculated, and where they mislead people.
Quick answer: the key football stats
- Possession: each team’s share of the ball, usually calculated from passes rather than a stopwatch.
- Expected goals (xG): the probability that a shot becomes a goal, based on thousands of similar shots.
- PPDA: passes allowed per defensive action; the lower the number, the more aggressive the press.
- Progressive passes and carries: actions that move the ball significantly closer to the opponent’s goal.
- Expected assists (xA) and key passes: how much chance creation a player is responsible for.
- Field tilt: each side’s share of touches or passes in the final third, a better measure of territory than raw possession.
Why football stats matter (and why they’re tricky)
Football is a low-scoring game. A team can be much the better side for 90 minutes and lose 1-0 to a deflection. That’s what makes the sport so dramatic, and it’s also why goals alone are a noisy way to judge performance.
Modern football stats try to measure the process rather than just the outcome. If a team keeps creating good chances and keeps stopping the opposition from creating them, results usually follow over a season. Clubs use these numbers for recruitment and tactics, broadcasters use them to explain matches, and fans use them to win arguments.
The catch is that most stats come from event data: someone, or increasingly a computer vision system, logs every pass, shot, tackle and touch with a location. Different providers define things slightly differently, so the same match can produce slightly different numbers on different sites. That’s normal, not a sign that one of them is wrong.
Possession: the most quoted, most misunderstood stat
Possession is the stat everyone knows. It’s also the one most often used badly.
How possession is calculated
Most broadcasters and data providers calculate possession from passes. If Team A completes 600 passes and Team B completes 400, Team A’s possession is shown as 60%. Some systems instead measure time on the ball, which can produce a slightly different figure. Neither method counts the time the ball is out of play.
That means possession is really a measure of how much of the passing a team did. A side that plays lots of short sideways passes at the back will rack up possession quickly, even if very little of it happens near the opposition box.
What possession does and doesn’t tell you
High possession tells you a team was comfortable keeping the ball, and usually that the opponent chose, or was forced, to sit deeper. It doesn’t tell you whether that possession was dangerous.
The classic example is Spain against Russia at the 2018 World Cup. Spain had around three-quarters of the ball and made more than 1,000 passes, yet the match finished 1-1 and they went out on penalties. Russia were happy to defend deep and let Spain pass in front of them.
At the other end, Leicester City won the 2015-16 Premier League with some of the lowest possession figures in the division. They defended compactly, broke quickly through Jamie Vardy and Riyad Mahrez, and didn’t need the ball to hurt teams. You can see how they fit among the other champions on our Premier League winners list.
Possession and game state
Possession also depends heavily on the score. A team that goes 1-0 up early will often sit deeper and concede the ball, while the trailing team piles forward. So a 65% figure for the losing side can say more about the scoreline than about who played better. Analysts call this “game state”, and it affects almost every stat in this guide.
Expected goals (xG) explained
Expected goals is the most important advanced stat in football, and once you understand it, a lot of others make sense.
How xG works
Every shot is given a value between 0 and 1, representing how often a shot like that is scored. The model looks at thousands of historical shots with similar features: distance from goal, angle, whether it was a header or a kick, the type of assist, whether it came from a set piece or a fast break, and in better models where the defenders and goalkeeper were.
A tap-in from two yards might be worth 0.8 xG or more. A speculative shot from 30 yards might be worth 0.03. A penalty is usually valued at somewhere around 0.76 to 0.79, depending on the model. Add up all of a team’s shots and you get their xG for the match.
How to read an xG scoreline
If a match finishes 1-0 but the xG was 0.6 to 2.4, the losing side created far better chances and were unlucky, or ran into a goalkeeper in brilliant form. Over one game, that kind of gap happens all the time. Over 38 games, it tends to even out, which is why xG is a good predictor of future results.
Players are judged the same way. A striker who scores 20 goals from 12 xG is either an exceptional finisher or riding a hot streak, and history suggests it’s more often the latter. A striker who scores 8 from 15 xG is getting into good positions and missing; that’s often a better sign for next season than it looks.
Non-penalty xG and xGOT
Because penalties are high-value and depend on who takes them, analysts often quote non-penalty xG (npxG) to compare players fairly. There’s also expected goals on target (xGOT), sometimes called post-shot xG, which is measured after the shot is struck and only for shots on target. It accounts for where the ball was placed in the goal. A shot curled into the top corner has a much higher xGOT than one hit straight at the keeper.
xGOT is the basis for most modern goalkeeping stats. If a keeper faces shots worth 10 xGOT and concedes 6, they’ve saved roughly four goals more than an average keeper would.
Expected assists, key passes and chance creation
Expected assists (xA) gives credit to the passer. In most models, the xA of a pass equals the xG of the shot it led to. So if a winger cuts the ball back for a striker to shoot from the penalty spot, and that shot is worth 0.35 xG, the winger gets 0.35 xA whether or not the striker scores.
This fixes a basic unfairness in assist counts. A playmaker who keeps setting up good chances for a wasteful forward will have few assists but plenty of xA.
Some providers use a slightly different “expected assisted goals” idea, and others measure the quality of the pass itself regardless of whether a shot followed. The principle is the same: reward the creator for the chance, not the finish.
A key pass is simply a pass that leads directly to a shot. Shot-creating actions, used by some data sites, go one step further and credit the last two attacking actions before a shot, which can include dribbles, fouls won and passes.
PPDA: how pressing is measured
PPDA stands for passes allowed per defensive action. It’s the standard stat for how intensely a team presses.
How PPDA is calculated
Take the number of passes the opposition makes in a defined zone, usually their own defensive 60% of the pitch. Divide it by the number of defensive actions your team makes in the same area: tackles, interceptions, challenges and fouls. The result is the number of passes you let the opponent make before you try to win the ball back.
A low PPDA, say around 8, means a team engages quickly and aggressively. A high PPDA, say 15 or more, means a team is happy to drop off and let the opposition pass in front of them.
What PPDA tells you
Jurgen Klopp’s Liverpool and Pep Guardiola’s Manchester City sides were known for very low PPDA figures because they hunted the ball high up the pitch. Teams that defend in a deep block, like many of Diego Simeone’s Atletico Madrid sides, tend to post higher numbers.
PPDA is useful, but it’s a style stat, not a quality stat. A high PPDA isn’t bad if your low block works. It also gets distorted by game state: a team protecting a lead will naturally press less.
Progressive passes, progressive carries and ball progression
Progression stats try to answer a simple question: who moves the ball towards goal?
Progressive passes
Definitions vary slightly by provider, but a progressive pass is generally a completed pass that moves the ball a meaningful distance closer to the opponent’s goal (often at least 10 yards compared with where it has been recently), or any completed pass into the penalty area. Passes from deep in a team’s own half are often excluded so that long goal kicks don’t inflate the numbers.
Centre-backs and deep midfielders who can break lines with passes, think of the likes of Trent Alexander-Arnold from full-back or Toni Kroos from midfield, rack up big progressive pass numbers.
Progressive carries
A progressive carry is the same idea when a player runs with the ball. Dribbling wingers and ball-carrying midfielders dominate this stat. A player can be excellent at progression without ever being involved in the final pass, which is exactly why these numbers became popular. They capture work that never shows up in goals or assists.
Passes into the final third and penalty area
Related stats count completed passes into the final third and into the box. They’re simpler and less dependent on definitions, which makes them handy for comparing players across leagues and data sources.
Field tilt, territory and other team stats
Field tilt measures each team’s share of passes or touches in the final third. If Team A makes 200 final-third passes and Team B makes 100, Team A’s field tilt is about 67%. It’s a much better measure of territorial dominance than possession, because it ignores harmless passing at the back.
Other team stats you’ll see often:
- Shots and shots on target: simple, but they don’t account for quality, which is why xG exists.
- Big chances: a provider’s judgement of situations where a player should reasonably be expected to score.
- Pass completion: easily inflated by safe passing; a centre-back on 95% isn’t necessarily better than a playmaker on 80%.
- Duels and aerials won: the share of physical contests a player or team wins.
- High turnovers: possessions won close to the opponent’s goal, a good companion stat to PPDA.
Football stats explained: what each one measures
| Stat | What it measures | Good for | Watch out for |
|---|---|---|---|
| Possession | Share of passes or time on the ball | Understanding style and control | Says nothing about danger; heavily affected by score |
| xG | Quality of chances created | Judging performance beyond the scoreline | Single-game xG is noisy; models differ |
| xGOT | Quality of shots on target after they’re struck | Rating finishing and goalkeeping | Only covers shots on target |
| xA | Quality of chances a player creates | Rating creators fairly | Doesn’t credit the pass before the assist |
| PPDA | Pressing intensity | Identifying pressing teams | A style measure, not a quality measure |
| Progressive passes | Ball moved meaningfully towards goal | Spotting line-breaking passers | Definitions vary by provider |
| Progressive carries | Ball carried meaningfully towards goal | Spotting dribblers and runners | Doesn’t show whether the carry ended well |
| Field tilt | Share of final-third play | Measuring territory | Doesn’t account for chance quality |
Common mistakes when reading football stats
Treating possession as dominance
We’ve covered this, but it’s the most common error. Plenty of teams win comfortably with 35% of the ball. Ask what they did with it, and what the opponent did without it.
Reading too much into one game’s xG
One match is a tiny sample. xG is at its most useful over 10, 20 or 38 games. A team “winning the xG” by 0.3 in a single match tells you almost nothing.
Ignoring game state
A team that scores in the third minute and then defends for 87 will look worse on almost every attacking metric than it deserves. Always check when the goals went in.
Comparing raw totals instead of per-90 figures
A player with 2,500 minutes will have more of everything than one with 900. Analysts use “per 90 minutes” figures to compare fairly, though very small samples still need caution.
Forgetting context and role
A defensive midfielder in a dominant team will make fewer tackles than one in a struggling side, simply because the ball spends less time near him. Low tackle numbers can mean a player is excellent at positioning, not lazy.
Mixing data from different providers
Opta, StatsBomb and other providers use different models and definitions. Comparing one site’s xG with another site’s is like comparing two different stopwatches. Pick one source for any comparison.
How to use stats when watching a match
The best approach is to use stats as a check on what you’ve seen, not a replacement for it. If you thought your team was unlucky, the xG will tell you whether the chances backed that up. If a midfielder looked busy but ineffective, his progressive passes and passes into the final third will show whether he was actually moving the team forward.
It also helps to pair stats. Possession with field tilt tells you whether control was territorial. PPDA with high turnovers tells you whether pressing actually won the ball in useful areas. xG with xGOT tells you whether a team was creating chances and finishing them well.
And remember that some of football’s most famous nights defy the numbers. Manchester City’s title-winning comeback against QPR in 2012, which you can rewatch on our Aguero 93:20 highlights page, wasn’t a story about xG. It was about time running out and one shot going in. Stats explain patterns; they don’t take away the drama.
If you want official numbers for a particular league, the Premier League’s stats centre is a reliable starting point, and our own results pages give you the scorelines to put them against.
Frequently asked questions
What is a good xG in a football match?
It depends on the league, but most top-flight teams average somewhere around 1 to 2 xG per game. Creating more than 2 xG in a match usually means a team made plenty of good chances.
What does PPDA stand for?
Passes allowed per defensive action. It measures how many passes a team lets the opponent make in their own half before making a tackle, interception, challenge or foul. Lower numbers mean more intense pressing.
How is possession calculated in football?
Most providers calculate it from passes: each team’s completed passes as a share of the total. Some use time on the ball instead. Time when the ball is out of play isn’t counted.
What counts as a progressive pass?
Generally a completed pass that moves the ball a significant distance towards the opponent’s goal, or any completed pass into the penalty area. The exact distance threshold depends on the data provider.
Is xG accurate?
xG is reliable over a large number of shots and has been shown to predict future results better than goals alone. Over a single match it’s much less dependable, and different models can disagree.
What is the difference between xG and xGOT?
xG rates the chance before the shot is taken. xGOT rates the shot after it’s struck, based on where it’s heading in the goal, and only applies to shots on target. It’s mainly used to judge finishing and goalkeeping.
Ready to put the numbers to work? Head to our fan ratings to see how supporters rated the players, then check whether the stats agree.
Comments