Analysis
Progressive passes explained: what each provider counts
Three data providers count progressive passes three different ways. What each threshold means, what the filter throws away, and how the number gets inflated.
By CricketTaken EditorialPublished Analysis18 min read
A midfielder plays a ball forty yards up the touchline. On one analytics site it appears in his progressive passing column that evening. On another it does not appear at all. Neither site has made a mistake. They are running different rules over the same event, and the pass sits on opposite sides of two different lines.
That is the whole problem in one image, and it is why progressive passes explained properly has to begin with the arithmetic rather than the concept. The concept is uncontroversial: some passes move a team meaningfully towards the goal it is attacking, and counting them separates the players who do that from the players who keep the ball moving without advancing it. The trouble starts the moment somebody has to write down how much movement counts as meaningful, in which parts of the pitch, measured from where, and with what happening to the passes that miss.
The pass that counts on one feed and not on the other
Take an ordinary passage. A centre-back has the ball ten yards inside his own half. He plays it diagonally into a midfielder who has dropped short, twelve yards further forward and fifteen yards across. The ball has closed roughly twelve yards of the distance to the far goal.
Under a definition that asks for ten yards of forward gain, that is a progressive pass and it goes in the column. Under a definition that asks for thirty metres of gain while the ball starts and finishes in the passing team's own half, it is nothing at all. Under a definition that only looks at the attacking two-thirds of the pitch, it never entered the sample, because it began too deep to be assessed.
One action. Three verdicts. No disagreement about what physically happened.
This is not a flaw somebody forgot to fix. Each rule was written for a purpose, and each purpose implies a different threshold. The version with the low bar is trying to describe who moves the ball forward at all. The version with the high bar in a team's own half is trying to isolate the passes that actually break a defensive block rather than the ones that shuffle the ball between defenders standing in a line. The version restricted to the attacking two-thirds has decided that progression in front of your own box is not the thing worth measuring.
Once you know which question a column was built to answer, the number stops being confusing. Until you know, comparing two columns is comparing two questions.
Progressive passes explained: the three definition families in circulation
Strip away the branding and the published definitions fall into three shapes.
Fixed distance, measured against a moving reference. The definition circulated alongside the free public tables asks for a completed pass that takes the ball at least ten yards closer to the opponent's goal than its furthest forward point across the previous six passes, and it also counts any completed pass into the penalty area regardless of distance. Passes beginning in the defending forty per cent of the pitch are left out.
Zone-dependent distance. Wyscout's glossary defines a progressive pass as a forward pass attempting to advance the team significantly closer to goal, and then sets three separate bars: at least thirty metres of gain if the pass starts and finishes in the team's own half, at least fifteen metres if it crosses the halfway line, and at least ten metres if it starts and finishes in the opponent's half. A pass is successful where the next touch belongs to a teammate.
Proportional distance in a restricted zone. Opta's published wording is shorter and stricter: a completed pass in the attacking two-thirds that moves the ball at least twenty-five per cent closer to the goal.
Alongside these sits a fourth idea that is not a distance rule at all. StatsBomb's deep progressions count passes and carries that arrive in the opposition final third, which replaces the question "how far forward" with the question "which zone did it reach". It is a cleaner rule and it answers a narrower thing.
Why the bar has to move as the ball moves up the pitch
The zone-dependent rule looks fussy until you think about what a fixed rule does at each end of the pitch.
Deep in a team's own half there is a great deal of space and very little resistance. A goalkeeper rolling the ball to a centre-back who carries it forward and slides it to the other centre-back can accumulate forward distance without any opponent being inconvenienced. A single ten-yard bar applied there would reward the safest ball in football many times a match.
In the final third the geometry inverts. There is almost no room, the defensive lines are compressed into perhaps twenty-five yards of depth, and a pass that gains eight yards may have gone between two centre-backs and put a striker in on goal. A thirty-metre bar applied there would count almost nothing, because there is rarely thirty metres left to gain.
The proportional rule solves the same problem from the other side. Twenty-five per cent of the distance remaining is a big number when you are eighty yards out and a small one when you are twenty yards out, so the absolute gain required shrinks naturally as the ball advances. That is elegant, and it is also why a proportional definition and a fixed-yardage definition will disagree most violently in exactly the zone where the play matters most.
The practical consequence: a fixed-yardage column will be topped by deep-lying players, and a proportional column restricted to the attacking two-thirds will be topped by attacking midfielders and wide players. Neither list is wrong. They are lists of different jobs.
The reference point, and the trick hidden inside "furthest point in the last six passes"
The most-overlooked clause in any of these definitions is the one that decides where you measure from.
The obvious answer is the spot where the pass started. Most definitions do exactly that. One does not, and its reason is worth understanding, because it plugs a hole the obvious answer leaves open.
If gain is measured from the start of each pass, a team can manufacture progression by going backwards first. Play the ball twenty yards back to the goalkeeper, then twenty-five yards forward to the same player who passed it back. The second pass has gained twenty-five yards from its own starting point and qualifies, even though the ball has ended up five yards further forward than it was three seconds earlier. Repeat the loop and the column fills up.
Measuring instead against the furthest forward point the ball reached across the previous six passes closes that loop. The backward pass resets nothing. To score, the forward pass now has to beat the high-water mark the possession has already set, so the recycled pass is worth only its genuine net gain and usually falls short of the bar.
The clause has a cost. It punishes the legitimate reset. A side that works the ball to the right, finds the door shut, comes back through the goalkeeper and attacks down the left has done something intelligent and difficult, and the switch that unlocks the second attack may register as nothing because an earlier pass had already taken the ball further forward on the other side. Anti-gaming rules almost always cost you something real, and this is what this one costs.
What the filter throws away on purpose
Every progression count is a subtraction before it is an addition. These are the categories that are deliberately discarded, and each discard changes who tops the table.
Backward and square passes. They gain no ground, so they score zero by construction, whatever they set up two passes later.
Incomplete passes. Most definitions require a teammate to take the next touch. A defence-splitting ball that is read and intercepted counts as nothing, which means the metric quietly rewards caution. Two players with identical intent and different receivers will have different columns.
Passes from the deepest zone. The forty per cent exclusion in one definition and the attacking two-thirds restriction in another both exist to stop goalkeeper distribution and centre-back circulation dominating the sample. That also removes some genuinely excellent long passing from deep, which is a real category and one the metric simply refuses to see.
Everything about the receiver. No published definition asks whether the receiver was under pressure, facing forward, or able to do anything with the ball. Twelve yards into a midfielder's feet with an opponent tight behind him and twelve yards into the same midfielder in acres score identically.
The carry. A player who beats the same distance by running with the ball scores nothing in the passing column. That is correct, since it is a different action, and it is also why any judgement built on progressive passes alone underrates ball carriers badly.
- Forward but under the distance bar46%
- Qualifying and completed24%
- Qualifying but incomplete, so discarded12%
- Started in the excluded deep zone18%
An entirely invented set of one hundred forward passes for one invented player, constructed to show which categories a progression filter removes before it counts anything. No real player, match or season is described, and the proportions are chosen for legibility rather than measured.
Show the numbers
| Item | Value |
|---|---|
| Forward but under the distance bar | 46% |
| Qualifying and completed | 24% |
| Qualifying but incomplete, so discarded | 12% |
| Started in the excluded deep zone | 18% |
Where the credit lands, and why "passes received" exists
A pass has two ends, and the definitions hand the whole of the credit to one of them.
The passer gets the count. The receiver, who checked away from a marker, held his run onside, opened his body to receive on the half turn and made the pass playable in the first place, gets nothing in that column. Anyone who has watched a striker drop into the pocket to drag a centre-back with him knows how much of the progression that movement is responsible for.
This is why progressive passes received became a separate column, and it is a genuinely useful one. It measures how often a player is on the end of the ball that broke a line, which is close to a measurement of positioning and movement in the spaces that matter. Wide forwards and second strikers tend to lead it, and a high figure with a low progressive passing figure describes a specific kind of player very precisely.
The two columns are not symmetrical, though, and it matters. Progressive passes made are constrained by how often you have the ball in a position to play forward. Progressive passes received are constrained by whether anybody in your team can find you. A forward in a side with no line-breaking midfielder will show a low received figure that says almost nothing about the forward.
Read together, they describe a chain. Read separately, each one flatters or damns a player for something the other end of the pass was doing.
Worked example: one invented move, counted three ways
The cleanest way to see the definitions diverge is to run a single passage through all of them and watch the totals separate.
- Goalkeeper to left centre-back, 8 to 14Six units of gain, deep in the team's own half. Fixed-yardage rule: below the ten-yard bar and inside the excluded deep zone anyway, so nothing. Zone-based rule: far below the thirty-metre bar for an own-half pass, so nothing. Attacking two-thirds rule: out of sample. Running totals 0, 0, 0.
- Left centre-back to right centre-back, 14 to 15A square ball across the back line. It gains one unit. Every rule scores it zero, which is the one thing all three agree on.
- Right centre-back to holding midfielder, 15 to 34Nineteen units of forward gain, both ends inside the passing team's own half. Fixed-yardage rule: comfortably over its bar, but the origin sits in the excluded deep zone, so still nothing. Zone-based rule: nineteen is under the thirty-metre own-half bar, so nothing. Attacking two-thirds rule: still out of sample. Totals 0, 0, 0.
- Holding midfielder back to right centre-back, 34 to 16Backwards. Zero everywhere. Note what it has done to the reference point: the possession has now reached 34, and the fixed-yardage rule will measure the next forward pass against that mark rather than against wherever the ball happens to be.
- Right centre-back to right winger, 16 to 47Thirty-one units of gain from the ball's current position, and it crosses the halfway line. Zone-based rule: over the fifteen-metre cross-half bar, so it scores. Fixed-yardage rule: measured against the high-water mark of 34, the net gain is thirteen, which still clears ten, so it scores too. Attacking two-thirds rule: the pass begins at 16, outside the zone, so nothing. Totals 1, 1, 0.
- Right winger to attacking midfielder, 47 to 61Fourteen units, and the ball is now inside the attacking two-thirds. Fixed-yardage rule: fourteen past the mark of 47, so it scores. Zone-based rule: the pass crosses no halfway line and does not sit wholly in the opponent half, and fourteen falls short of what a cross-half pass needs, so on a strict reading it does not. Proportional rule: fifty-three units of distance remained, fourteen closes about twenty-six per cent of it, so it scores. Totals 2, 1, 1.
- Attacking midfielder to striker inside the box, 61 to 88Twenty-seven units into the penalty area. Fixed-yardage rule: scores twice over, on distance and on the completed-pass-into-the-box clause. Zone-based rule: both ends in the opponent half and well over ten metres, so it scores. Proportional rule: thirty-nine units remained and twenty-seven closes about sixty-nine per cent, so it scores. Totals 3, 2, 2.
- The same move, three answersThree progressive passes, two, or two, depending on whose column you are reading. The move did not change. Every discrepancy comes from a threshold, a zone restriction or a reference point, and every one of them was a deliberate choice by whoever wrote the rule.
A wholly invented seven-pass move, used to show which rule catches which action. No real match, team or player is described. The pitch runs from 0 at the passing team's own goal line to 100 at the goal it is attacking, and the distances are chosen to sit near the thresholds on purpose.
The four ways a progressive passing number gets inflated
None of what follows requires anybody to cheat. Systems produce the numbers they are asked for, and players are coached into shapes that happen to produce them.
Volume through territory. A side that spends the match camped in the opposition half has more chances to play a qualifying forward pass than a side pinned in its own third. Its deep midfielder can accumulate a large count while doing something a good under-eighteen player could do, because the passing lane he uses is unguarded.
Recycling under a start-point rule. Where gain is measured from the start of each pass, the back-pass-then-forward-pass loop described earlier is free progression. This is precisely the reason the high-water-mark clause was written, and it is why a count from a provider without such a clause needs reading more sceptically.
The safe forward pass into space. A ball played into the gap in front of a deep-lying opponent is often uncontested. It gains ground, it completes at a very high rate, and it beats nobody. A player told to play more of them will move up the table without improving.
Being in a bad team. This one is the least intuitive and it catches people constantly. A defender in a side that concedes territory takes possession deep and often has an enormous amount of pitch in front of him. Long clearances that a teammate happens to win, and hopeful diagonals into a wide area, can clear a distance threshold routinely. High progressive passing figures from a struggling defence are frequently a symptom of that defence being under siege.
The common thread is that a distance count cannot see difficulty. Beating an opponent and finding empty grass look identical to a rule written in yards.
Volume, accuracy and share: three different columns and three different meanings
A raw count answers "how much", which is a role question. Two derived figures answer better questions.
Completion rate on progressive attempts separates the player whose forward passes land from the player whose forward passes are a lottery. Read alone it is misleading in the opposite direction, since the safest progressive passer in the league will lead it. Read next to volume it starts to describe a style: high volume with high completion is a player finding available passes, high volume with lower completion is a player attempting harder ones.
Progressive passes as a share of all passes normalises for how much a player touches the ball. This is the figure that most changes rankings when you apply it, because it removes the advantage of the deep midfielder who has ninety touches to the winger's forty. A player who plays forward on one pass in five is telling you something about intent that a per-ninety count cannot.
Progressive distance, where a provider publishes it, sums the yards gained rather than counting the passes. It rewards a single sixty-yard switch far more heavily than three qualifying ten-yard passes, which is right for some purposes and badly wrong for others. Long diagonals into a full-back inflate it easily.
Any of these is more informative than the raw count, and none replaces watching where the passes actually went. If you want the tactical picture underneath the columns, the mechanics of playing out from the goalkeeper explain why the deepest progressive passes look the way they do.
Progressive passes and progressive carries are not substitutes
Two players move the ball the same distance up the same channel. One passes it, one runs with it. The first appears in the progression column. The second appears in a different column, and if you are only reading the first, he appears to have done nothing.
The distinction is not academic, because the two actions have opposite risk profiles and opposite tactical effects. A pass covers ground instantly and leaves the passer where he was. A carry covers ground slowly, drags at least one opponent out of position, and leaves the carrier in an advanced position with the ball at his feet. A carry that draws a defender out is often the thing that makes the next pass a line-breaker.
Any honest ball progression assessment adds them, and the sensible providers publish both. Where only passing figures are available, the players most badly underrated are the inverted full-backs and carrying centre-backs whose whole job is to advance the ball on the dribble until somebody commits to them. The role is set out in the piece on the inverted full-back and what the position is for.
The mirror-image error is also common. Progressive carry counts flatter players in transition-heavy teams, where the pitch is open and running forward is easy, and undervalue those in sides that face a compact block with nowhere to run.
- 9Player A, progressive passes per 90
- 8Player A, share of his passes that qualify
- 9Player B, progressive passes per 90
- 21Player B, share of his passes that qualify
Two entirely invented players, constructed so that an identical headline count sits on top of two completely different profiles. Nothing here is measured, and no real player is described.
What the metric cannot see at all
Some things are not measured badly. They are outside the definition entirely.
The pass that was available and not played. No count of what happened describes what was declined. A midfielder who ignores the forward option four times a half is invisible to a column that only sums the passes he did make.
The pass that created the pass. A switch of play gains no ground and scores zero, then the winger receives with a full-back isolated and the next ball goes into the box. Credit for the sequence lands entirely on the last action. The way teams build a numerical advantage on one side to attack the other is covered in the piece on creating and using overloads.
Pressure on the ball. Whether the passer had four seconds or half a second changes the difficulty of the same pass by an order of magnitude, and no distance rule contains a term for it.
What the pass was worth. Ten yards gained down a blind alley and ten yards gained into the space behind a full-back have the same value in the column and nothing like the same value in the match. This is the gap that possession-value and expected-threat models exist to fill, and it is the same conceptual move that produced the shot-quality model behind expected goals: stop counting events, start valuing states.
That last point is the honest ceiling on the whole family of counting metrics. They are a description of activity. Anything that claims to describe contribution has to attach a value to each action, and the moment you do that you are no longer counting, you are modelling.
How progressive passing is actually used inside clubs
Recruitment departments rarely rank players on a progression count, for the reasons above. They use it as a filter and a descriptor.
As a filter, it narrows a list. If a club needs a deep midfielder who can break a first line, a minimum threshold on progressive passes per possession removes several thousand players in one line of code, and the remaining shortlist is watched. Nobody signs anybody on the strength of the filter.
As a descriptor, it separates similar-looking players. Two number sixes with matching pass completion and matching minutes can have entirely different progression profiles, and the difference tells a coach which one fits a system that wants the ball moved forward early. That is a real use and a modest one.
The third use is diagnostic and applies to teams rather than players. A side that has plenty of possession and very little progression is circulating without penetrating, which usually means the receiving movement in front of the ball is not happening. That pattern shows up in the numbers before it shows up in the results, and it is a different problem from the one a side has when it progresses well and creates nothing.
Analysts also read progression against pressure. A team that progresses fine against a mid-block and stalls against a high one has a specific weakness, and the shape of the opposition press is described in the piece on the PPDA pressing metric and what it counts.
Set pieces, throw-ins and the passes the definitions leave ambiguous
The edge cases are where two feeds most often part company, and the published wording usually does not say which way they fall.
A corner is a completed pass into the penalty area more or less by definition. Under a rule that automatically qualifies any completed pass into the box, an inswinging corner headed clear by a defender but touched first by an attacker could satisfy every stated condition. Providers handle this with implementation choices that never make it into the one-line glossary entry, typically by excluding dead-ball deliveries from the progression sample altogether. If they did not, the taker of every corner would sit near the top of the table and the column would be measuring set-piece duty.
Free kicks are messier still, because they come in two kinds. A short free kick played out of defence is functionally an ordinary pass and most rules treat it as one. A ball hoisted into the box from forty yards is a delivery, not a progression, and yet it can satisfy a distance threshold comfortably. Where the line is drawn is an implementation detail that changes defenders' totals noticeably, since defenders take a lot of long free kicks.
Throw-ins sit in a third category. They are not passes in the legal sense and several event feeds record them under a separate event type, which means they are often excluded automatically without anybody deciding to exclude them. A long throw into the box therefore scores nothing under most rules, while a five-yard pass into the same area might score, which is arbitrary but consistent.
Goalkeeper distribution is the last of them. A drop kick that clears fifty yards obviously advances the ball. Some definitions cut it out through a zone restriction, some through an explicit goalkeeper exclusion, and some do not cut it out at all, which is how a keeper occasionally appears in a progression leaderboard and everybody assumes the data is broken. The data is not broken. The rule simply did not exclude him.
What happens to your figures when a provider rewrites its own rule
Definitions are not fixed points, and this catches people who built a view on last season's numbers.
Providers revise their glossaries. A threshold gets moved because analysts complained that the old one caught too much noise. A zone restriction gets added because the previous sample was dominated by defenders. A completion requirement gets relaxed because a customer wanted attempts as well as successes. Every one of those revisions is defensible and every one of them silently changes historical rankings.
The practical damage takes two forms. The first is comparison across seasons: a player whose count rises by a third between one campaign and the next may have improved, or the bar may have dropped. Unless the provider republishes the whole archive under the new rule, the two seasons are not the same measurement wearing the same label.
The second is the borrowed number. A figure quoted in an article, screenshotted into a thread and then reused two years later carries no version stamp. It looks like a fact. It is the output of a specific rule at a specific moment, and the rule may no longer exist in that form.
The defence is simple and slightly tedious. Take your numbers from one provider for any comparison that matters, check the glossary entry rather than trusting your memory of it, and be suspicious of any cross-season claim that rests on a metric with a threshold in it. Structural facts about football, such as how many players are on the pitch or how wide the goal is, do not move. Analytical definitions move constantly, and they move without an announcement anybody reads.
The same warning applies to every derived measure built on top of a progression count. If the underlying filter changes, everything downstream of it changes too, including the composite scores that some sites present as a single rating out of a hundred.
Reading a progressive passing figure without being had
Five questions, in order, before you let a number tell you anything.
Whose definition is this? If the page does not say, treat the figure as unusable for comparison against any other page. This is not pedantry: the same player under two rules can differ by a factor that would change your opinion of him.
Per what? Per ninety minutes, per hundred passes, or per possession are three different rankings. Per ninety rewards players who touch the ball a lot. Per hundred passes rewards intent.
From where? A count with no zone breakdown hides the difference between a defender clipping balls into the channel and a midfielder threading them between lines. If the provider publishes progression by starting third, use it.
Completed only? If so, remember that the column is quietly punishing ambition, and look at attempt volume alongside it.
What happened next? Progression is a means. A side or a player who progresses often and produces nothing is doing the easy half of the job. Set the progression figure against shot volume and chance quality before you decide it means anything.
The metric is worth having. It picks out a real skill, it is cheap to compute, and it is far more informative than pass completion, which mostly measures how far backwards a player is willing to pass. It simply cannot survive being quoted without its definition attached, and most of the arguments it causes online are two people reading two different rules and assuming the other one is lying.
More on how the numbers behind the modern game are put together sits across our football coverage, and the rest of the explainer library is on the blog.
Common questions
What is a progressive pass in football?
It is a pass that carries the ball a defined distance closer to the opponent's goal, where the distance required is set by the data provider rather than by the laws of the game. Every provider agrees on the idea and disagrees on the number. That is why the same match can produce three different progressive pass counts for the same player.
How many yards does a pass need to travel to be progressive?
There is no single answer. One public definition uses ten yards of gain measured against the ball's furthest forward point in the previous six passes, plus any completed pass into the box. Another uses thirty metres inside a team's own half, fifteen across the halfway line and ten in the opposition half. A third asks for a twenty-five per cent reduction in the remaining distance to goal, counted only in the attacking two-thirds.
Do backward passes ever count as progressive?
No definition in common use rewards a pass that moves the ball away from the goal being attacked. Sideways passes are excluded on the same logic, since they close none of the remaining distance. This is a deliberate choice and it is also the reason the metric undervalues the switch of play that creates the space for the next pass.
Are incomplete passes counted?
Usually not. Most published definitions require the ball to reach a teammate, which means an ambitious pass that is cut out scores nothing while a short safe one that lands scores fully. Ball progression models that use expected values rather than counts handle this better, because they can credit the attempt and charge for the risk separately.
Why do two websites show different progressive pass totals for the same player?
Because they are running different filters over different event feeds. The distance threshold, the zone restrictions, the reference point used to measure gain, the treatment of the penalty area and the handling of set pieces all vary. Compare a player only against others processed under the same definition.
Is a high progressive pass count a sign of a good midfielder?
It is mostly a sign of a role and a team. Deep-lying players in sides that dominate the ball accumulate the most, because they touch the ball often in the zone where forward passes are available and low risk. Read the number as a description of what a player is asked to do, then look at where the passes went and what happened next.
Filed under Football·football · analytics · passing · data · scouting · midfielders