Set piece coaching in football: the one rehearsed moment

The referee points at the corner flag and the match stops for forty seconds. Twenty-two players walk to positions somebody chose in advance. Two attackers form a line outside the six-yard box, then split. The taker places the ball inside the arc and waits, unhurried, because the Laws guarantee him nine and a bit metres of empty space in every direction.

That pause is the whole subject, and it is why set piece coaching in football has turned from a Thursday afternoon duty into a job with a name on the door. A corner is the only situation in the sport where a coach sets the starting position of every player on the pitch.

Everything else in a match is a system you can influence but not control. A restart is a system you can specify.

Set piece coaching in football exists because a dead ball is a different problem

Open play is continuous, and continuity is what makes it hard to coach precisely. A pattern rehearsed on Tuesday arrives on Saturday with the wrong spacing, an opponent standing somewhere unexpected, and a defender who has read it. Coaches deal with this by teaching principles rather than sequences: press when the touch is heavy, occupy the far side, stay compact. Principles survive contact with a moving game because they do not depend on the initial conditions being repeated.

A restart has no such problem. The ball is stationary at a known point. The clock, in every sense that matters to the players, has stopped. Both teams can see each other completely, and both can arrange themselves however they like. Nobody is out of breath in a way that changes the geometry. The initial conditions of a corner from the left in the eightieth minute are identical to the initial conditions of a corner from the left in the fourth.

That single property, repeatable initial conditions, is what allows a set piece to be rehearsed to completion rather than approximately. A team can practise the exact delivery, into the exact zone, against the exact structure the next opponent will use, at full speed, twenty times in a week. There is no equivalent for a counter-attack, and there never can be.

Three further things follow from the pause, and they are the reason a specialist can extract value that a generalist cannot.

The first is that the opponent's arrangement is knowable in advance. Defensive corner structures are stable across a season, because they are drilled and because changing them mid-campaign is expensive. A coach can therefore know, before kick-off, roughly where every one of the eleven defenders will stand, which zones will be covered, which attacker will be man-marked and by whom, and whether anyone will be posted outside the box. That level of foreknowledge does not exist anywhere else in the game.

The second is that the outcome space is finite. From a corner, the ball can be met at the near post, in the central zone, at the far post, or not met at all. It can be cleared to the edge, cleared long, or scrambled. There are perhaps a dozen meaningfully distinct things that can happen, which means they can be enumerated, assigned to players, and drilled individually. Open play cannot be enumerated.

The third is that the attacking side moves first. Once the defence has set up, the attackers choose the delivery, the runs and the timing, and the defence can only react. In a game where the initiative usually belongs to whoever has the ball moving at speed, this is a rare situation where the initiative belongs to whoever thought about it hardest during the week.

There is a cost, and it is the reason set piece work is an arms race rather than a solved problem. Because the situation repeats, the opponent also gets to study it. A routine used twice is a routine that has been filmed, catalogued and shown to eleven defenders. Specialist coaching is therefore not the accumulation of clever ideas but the management of a portfolio: several routines built off the same starting shape, so that the setup gives nothing away and the fourth corner of the match does something the first three did not.

• 9.15Metres opponents must retreat at a corner or a free kick
• 1Metres attackers must stand off a defensive wall of three or more
• 2Metres opponents must stand back at a throw-in
• 0Players who can be caught offside directly from a corner or a throw-in

These are the distances and permissions written into the Laws of the Game rather than any team's practice. They define the space a coach is working inside.

Man, zonal, and the hybrid nearly everyone actually uses

Defending a corner means solving one problem: an attacker arriving at speed into a space, against a defender who must get to the ball first or make the contact worthless. Every marking system is an answer to the question of what the defender should be attached to, the attacker or the space.

Man-marking attaches each defender to an opponent. Its virtues are obvious and real. Somebody is accountable for every dangerous attacker. Contact is made early, so the attacker cannot get a free run at the ball. If the delivery finds a specific head, there was a defender at that head.

Its weaknesses are equally structural. A man-marker cedes the initiative on movement, because the attacker chooses when and where to go and the defender can only follow. Following means turning, and turning means the defender ends up facing his own goal while the ball travels towards it. That is the worst posture in aerial defending: running backwards, watching a man rather than a ball, arriving at the contact point with no momentum. Man-marking is also the system that blocks and screens are designed to break, since separating one defender from one attacker for half a second is a much simpler engineering problem than disrupting a whole structure.

Zonal marking attaches each defender to a piece of ground. The near post is covered, the six-yard line is covered, the central space in front of the goalkeeper is covered, and whoever comes into that space meets a defender who is already there. Zonal defenders face the ball throughout. They attack it from a standing start, jumping forwards into the flight rather than backwards away from it, which is worth a surprising amount of height. They are also, by construction, still standing in the most valuable areas of the pitch when the ball is half-cleared, which makes them better placed for what happens next.

Its weaknesses are also structural. A static defender meets a moving attacker, and momentum matters in the air. The height of the zonal line has to be right for the delivery, and getting it wrong by two metres means the ball drops behind everybody. Nobody is individually accountable, which is a coaching problem before it is a tactical one. And gaps between zones can be attacked deliberately by two attackers running at the same seam.

The hybrid is what almost every senior side actually plays, and it is not a compromise so much as a division of labour. A small zonal group holds the space that must not be conceded regardless of who runs into it, usually the near post and the line of the six-yard box. Individual markers are assigned to the two or three opponents who genuinely win headers. Somebody guards the short corner. Somebody stays high as a counter-attacking outlet. The system is chosen zone by zone rather than as a philosophy, and the fuller argument about how the two marking systems trade off against each other applies to open play as much as to restarts.

30%40%10%10%10%

• Zonal players holding the near post and the six-yard line3
• Man-markers on the opponent's genuine aerial threats4
• Screening the space at the top of the penalty area1
• Guarding against the short corner1
• Left high as the counter-attacking outlet1

A constructed hybrid arrangement with invented round numbers, not a measurement and not any particular team's setup. The point is that the allocation is a budget: every player given to the short corner or the counter is a player not defending the six-yard box.

Show the numbers

Worked example: how ten outfield defenders get allocated at a corner

Item	Value
Zonal players holding the near post and the six-yard line	3
Man-markers on the opponent's genuine aerial threats	4
Screening the space at the top of the penalty area	1
Guarding against the short corner	1
Left high as the counter-attacking outlet	1

That figure is the honest version of the argument. There are ten outfield defenders and rather more jobs than that, so every defensive corner setup is a set of decisions about what to leave uncovered. A team that posts two players outside the box against the counter is defending its penalty area with eight. A team that leaves nobody short concedes a free two against one thirty metres from its own goal. There is no arrangement that covers everything, which is exactly why the decision is worth a coach's attention rather than a shrug.

Zonal marking is blamed for goals it did not cause

This deserves its own section because the misunderstanding is so widespread that it distorts how clubs actually defend.

When a team marking zonally concedes, the still image is damning. An attacker rises unopposed, and nearby there are defenders standing in a line looking at the ball, apparently doing nothing about the man who has just scored. The picture appears to show negligence, and the conclusion writes itself: nobody was marking him.

When a man-marking team concedes, the same still image shows a defender beside the scorer, beaten. That reads as an individual error. The system is invisible in the frame, so the system escapes blame.

The two goals may have identical causes. In both, a defender misjudged a flight, or was screened, or jumped too early, or lost a duel to a better player. The difference is where the blame attaches, and it attaches by the composition of the photograph rather than by the mechanism of the goal.

This attribution asymmetry has consequences. A coach who switches to zonal marking and concedes twice in three weeks faces a specific kind of criticism that a coach conceding the same two goals from man-marking does not. The rational response, for anyone whose job depends on that criticism, is to defend in the way that produces excusable goals rather than the way that produces fewer. That is a genuine cost, and it is paid in points by clubs that never examine it.

The legitimate criticisms of zonal marking are narrower and more useful than the popular one:

• The height of the line is a real failure mode. Zonal players must be positioned for the delivery they will actually face, and an out-swinging corner drops further from goal than an in-swinging one. A structure set for the wrong delivery leaves the ball dropping behind the whole line, and this is a genuine, repeated, coachable error.
• Static defenders lose to moving attackers more often than the theory suggests. Facing the ball is worth a lot, but so is a running jump, and a zonal system that produces passive defenders standing and watching has thrown its advantage away.
• Communication has to be explicit. A zone is a responsibility, not a location, and defenders have to talk about who takes a ball dropping on a boundary. Silence in a zonal structure produces exactly the free header the critics describe.

What zonal marking is genuinely good at, and what gets almost no attention, is the phase after the first contact. Man-markers finish a corner scattered across the box in whatever positions their opponents dragged them into. Zonal defenders finish it in a shape, facing the right way, occupying the space where a half-clearance lands.

Blocking, screening and what referees actually allow

Attacking set piece design is largely the business of buying an attacker half a metre of separation. The tool for that is the block, and the interesting part is where the legal line sits.

The Laws are clear enough in principle. Holding an opponent is a direct free kick offence, and referees are explicitly directed to deal with holding at corners: warn before the ball is in play, caution the persistent offender, and penalise it if it continues. Pushing is an offence. Impeding an opponent with contact is a direct free kick; impeding without contact, simply standing in someone's path, is an indirect free kick offence only when the offender is not within playing distance of the ball.

The practical effect is a large permitted space. An attacker may occupy ground. He may stand still in a place that happens to be exactly where a man-marker wants to run. He may move to a position and stop, which is the entire mechanism of the block. He may position himself in front of the goalkeeper, obstructing the view, provided he does not hold him or make a challenge. None of that is an offence, and all of it is coached.

The illegal versions are the ones involving hands and force. Grabbing a shirt. Wrapping an arm. Shoving a marker as the ball is struck. Wrestling a man to the ground.

Referees, in practice, tolerate a considerable amount of mutual grappling in the penalty area and act on the obvious rather than the marginal. The usual sequence is a public warning to two players before the kick, then a penalty only for something flagrant. Video review changed this less than expected, because the threshold for intervention is a clear error rather than a technical offence, and most holding at corners is neither clearly one-sided nor clearly decisive.

Coaches design around that reality rather than against it. The reliable methods use bodies and geometry rather than hands:

The stack, or train. Attackers form a tight vertical line before the delivery, then break in different directions on a signal. Man-markers standing behind a line of four cannot get to their assignments without going round the line or through it, and the first two steps of the run are surrendered. No contact is required for it to work.

The static screen. One attacker positions himself where a marker must pass, and simply stands there. The marker collides with a stationary opponent. This is legal, it looks illegal, and it is the single most common source of penalty appeals at corners.

The goalkeeper screen. A player standing on the edge of the six-yard box in front of the keeper does not need to touch him. The keeper's decision to come or stay is made from a worse view and a fraction of a second later, and that fraction is often the entire margin.

The decoy run. The attacker who sprints hard to the near post and never receives the ball is doing two jobs: dragging his own marker out of the middle, and pulling a zonal defender forward so the space behind him opens for somebody else.

The late arrival. A player who starts outside the penalty area and enters at speed is the hardest to pick up, because a static structure has nobody assigned to him and a man-marking one has to make a decision about whether to follow him out.

In-swinging or out-swinging is a decision about who supplies the power

The delivery is not a matter of taste. In-swinging and out-swinging corners create different problems, demand different things from the attackers, and produce their second phases in different places on the pitch.

An in-swinging ball curls towards the goal. Its defining property is that it is already going in the right direction, so an attacker does not need to generate anything. A glance, a deflection off a shoulder, a touch that changes the line by fifteen degrees, all of these can score. The delivery does the work and the attacker only has to interfere with it.

That property produces the rest of the in-swinger's character. It attacks the goalkeeper's space, forcing him to decide whether to come, and a keeper who commits and misses has removed himself from the goal. It rewards near-post runs and flick-ons, because a flick on an in-swinger arrives in the back-post area still travelling goalwards. It punishes any defensive line that sets up too deep. And it carries a specific risk in the other direction: a keeper who does claim it cleanly has ended the attack completely, whereas nothing else on the pitch can do that.

An out-swinging ball curls away from the goal. The goalkeeper is unlikely to come for it, so the contest is between an attacker and a defender in open space, without a third party. The attacker must supply all the power and all the direction, which means he has to attack the ball with a running jump and meet it cleanly. Headers from out-swingers are struck rather than glanced.

That makes the out-swinger the delivery for a team with genuine aerial power and good timing, and the in-swinger the delivery for a team relying on chaos, deflection and volume. It also moves the second phase further from goal, because the contact point is deeper and clearances from it travel differently.

There is a third family that sits outside the swing question entirely: the flat, driven delivery hit hard and low into the near-post zone. It removes time from everybody, defenders and goalkeeper included, and it converts the corner into a reaction contest rather than a jumping contest. It is also the delivery most likely to produce nothing at all, since a hard ball into a crowd goes out for a goal kick as often as it finds a head.

The choice interacts with the taker's foot in a way that constrains squads more than people notice. An in-swinger from the right corner requires a left-footed striker of the ball, and an out-swinger from the right requires a right-footed one. A team that wants in-swingers from both sides needs both feet available. A team with two right-footed specialists is choosing in-swingers on one side and out-swingers on the other whether it wants to or not, and any serious collection of set piece routines has to be built around who is actually available to take them.

The second phase is where a large share of the goals live

Ask a supporter to picture a set piece goal and they will picture a header from a cross. That is the smallest part of the story.

The first phase is the delivery and the first contact. The second phase is everything after: the header cleared only as far as the penalty spot, the ball bouncing loose off a defender's shin, the clearance dropping to a midfielder twenty metres out, the ball recycled to the wide player and crossed again. A large share of the goals that get filed as set piece goals actually happen in this phase, and it is the part that separates a coached attacking structure from a hopeful one.

The reason is arithmetic rather than mystery. A well-defended corner produces a clearance, and a clearance from a crowded six-yard box is not a controlled pass. It is a header, struck by a defender who is off balance and facing the wrong way, that travels somewhere between ten and thirty metres in an unpredictable direction. Whoever has thought about where that ball is going owns the next moment of the match, and that moment happens against a defence that is scattered, unset and still recovering from the first contest.

Coaching the second phase means answering four questions before the corner is taken:

• Who occupies the edge? Usually two players, positioned wide of centre rather than directly in front of goal, because clearances travel out at angles more often than straight. Their job is to win the ball or shoot immediately.
• Who wins the shin-height bounces? The scramble in the six-yard box has no shape and rewards whoever reacts first. Teams assign a player to stay low and alive in that space rather than jumping for everything.
• Where does the ball go if it is only half won? The recycled cross from the wide areas is a second delivery against a defence that has already jumped once, and jumping twice inside four seconds is harder than it looks.
• Who is not chasing it? The players held back for the counter do not join the second phase, and the discipline to stand still while a loose ball rolls past is the hardest instruction on the list.

Defensively, the second phase is where the marking systems separate most clearly. It is also the reason clubs that model restarts properly do not evaluate a corner by whether it produced a header. They evaluate the whole sequence, and the way set piece situations get valued shot by shot has to account for a second and sometimes a third attempt from the same restart.

1. 1The referee awards the cornerBoth sides get an unhurried setup. The attacking coach signals a routine from a small catalogue, usually with a hand or a shout of a number, and the taker walks to the arc. Nothing after this point is improvised.
2. 2The defence takes its shapeZonal players occupy the near post and the six-yard line. Man-markers find their assignments and get goal-side and touch-tight. One defender covers the short corner, one or two stay high as an outlet, and the goalkeeper sets his starting position for the expected swing.
3. 3The attackers set the decoy shapeThe visible arrangement is chosen to look like the last three corners. Attackers form a stack outside the six-yard box, one takes up a late position outside the area, and one occupies the space in front of the goalkeeper without touching him.
4. 4The signal and the splitOn the taker's cue the stack breaks. One runs to the near post, one blocks by standing still in a marker's path, one peels to the back post, and the late runner starts from outside the area. The separation is bought in these two steps, not in the air.
5. 5The deliveryIn-swinging into the near-post zone, out-swinging to the back post, or driven flat, chosen to match the runs and the defensive height. The taker is aiming at a zone, not at a player, because the player will be where the zone is.
6. 6First contactEither an attacker meets it, a defender clears it, or the goalkeeper claims it. Only the last of these ends the sequence. The other two open the second phase, and the attacking side already has players positioned for it.
7. 7The clearanceA defender's header travels ten to thirty metres from an off-balance jump. It is not a pass. Where it lands is close to random, which is why the attacking team covers angles rather than a spot.
8. 8Second phaseThe edge-of-the-box players attack the drop. A shot goes in first time, or the ball is worked wide for a second delivery against a defence that has already jumped. This is where a large share of the goals actually arrive.
9. 9The transitionIf the ball is genuinely cleared, the attacking side's held-back players become a defensive line and the taker sprints back. The opponent's outlet forward is already running. The corner has become a counter-attack, and the next ten seconds decide whether it was worth taking.

The sequence a coached corner runs through, attacking and defending sides together. Every step here is an arrangement decided before the ball is struck.

The short corner is arithmetic, not cowardice

Short corners are booed in most grounds. The reasoning behind them is stronger than the reaction suggests.

Start from the defensive allocation. A defending team has ten outfield players and has to decide how many stand outside the penalty area. If it commits everybody to the box, two attackers standing near the corner arc are unopposed, and can walk the ball into a crossing position with no pressure at all. If it sends one defender out, those two attackers have a two against one, and the defender is now thirty metres from where he was needed.

Either way the attacking side has converted two players into a numerical advantage, and has removed at least one body from the penalty area. That is the first argument.

The second is about the delivery itself. A corner struck from the arc is a stationary ball from a fixed point at a known angle, which is the easiest thing in football to defend, because the defence has seen that exact geometry a thousand times. A cross played from ten metres infield and five metres up the touchline is a moving ball from a different angle with a different flight, and the defensive structure has to shuffle to meet it. Zonal lines in particular are drilled for a specific origin, and a delivery from somewhere else forces them to re-set while the ball is already travelling.

The third is risk. A corner swung into the box is a fifty-fifty in the most dangerous area of the pitch, and losing it produces a clearance that starts a counter-attack. A short corner keeps the ball in a controlled situation with a teammate, which is a safer state than a contested aerial ball, and lets the attacking side reset if nothing is on.

The costs are real too. Two attackers standing short are two fewer in the box, so the eventual cross meets a thinner attack. The routine takes time, which lets a stretched defence organise. And a short corner played badly hands the ball over in a position where the defending team is already facing the right way.

The reasonable conclusion is not that short corners are good or bad but that they are a specific tool for a specific defensive setup. Against a team that leaves nobody outside the box, the short corner is close to free value. Against a team that posts two defenders out there, it is a bad idea, and taking it anyway is the sign of a routine chosen before kick-off rather than read on the pitch. That is a coaching failure rather than a tactical one.

The throw-in is the restart everyone spent decades ignoring

There are more throw-ins in a match than corners and free kicks combined, and until fairly recently almost nobody coached them. That is the clearest evidence that attention to restarts was genuinely absent rather than merely informal.

Two properties of the throw-in are worth stating plainly, because both are exploitable. There is no offside offence directly from a throw-in, so an attacker can stand anywhere he likes, including behind the last defender, and receive the ball legally. The offside law and the situations it does not cover matter more here than at any other restart. And the ball is delivered with the hands, which for one player in a squad means a flat, fast, accurate delivery into the six-yard box that no wind or spin will affect.

The long throw returned for the reason most things return, which is that somebody measured it. A throw from the attacking third into the penalty area is a free delivery into the most dangerous area on the pitch, arriving at head height, with no possibility of an offside flag, against a defence that has to organise for it exactly as it would for a corner but usually has not rehearsed doing so. The only input required is one player with the shoulders and technique to produce it, which is a cheaper thing to acquire than almost any other source of goals.

It also has a defensive property that a corner lacks. The thrower is standing still on the touchline with the ball in his hands, so his team can be fully organised before the delivery, including the players held back against the counter. A long throw is therefore a lower-risk delivery than a corner, not a higher-risk one, which is the opposite of its reputation.

The other half of throw-in coaching gets less attention and is probably worth more points. A throw-in in a team's own third, taken under pressure by a full-back with three opponents closing, is one of the most reliable ways to lose the ball in a terrible area. Sides that coach it properly have rehearsed patterns for it: a receiver who comes short to set the ball back, a runner who drags a marker away, a designated long option, and a clear rule about when to simply put the ball out of play rather than accept a fifty-fifty. It is unglamorous and it removes a category of goals conceded.

The wall, the man on the ground, and the law that followed

A direct free kick within shooting range is the restart where the defending side does most of the arranging, and the wall is a structure with a specific job. It is not there to block the whole goal. It is there to remove one half of it, so the goalkeeper can commit to the other half instead of standing in the middle and covering neither.

That division of labour explains everything about how walls are built. The number of players is set by the distance and the angle: more from central positions, fewer from wide ones where the shooting angle already does the work. The end of the wall nearest the near post is anchored by a player who must not move, because the keeper's position depends on the wall covering exactly what he was told it would cover. The jumpers are placed to deny the ball over the top. The whole thing must stand at least nine and a bit metres away, a distance the referee now marks so that the argument cannot be had.

Attacking sides have three ways past it. Over, which requires the shot to rise and fall inside a very small window and is the most difficult skill in the game. Around, which means bending the ball outside the end of the wall and back, and which the goalkeeper is positioned to see all the way. Under, which is the one that produces the most goals relative to how hard it looks, because a wall that jumps leaves a gap at ground level for a fraction of a second and a hard low shot arrives during that fraction.

The answer to under was a player lying on the ground behind the wall. It is inelegant and it works: the goalkeeper can only cover one half, the wall covers the top, and a body across the turf covers the bottom of the space the jump opens.

The attacking response was to interfere with the wall itself. Attacking players began standing in it, or at the end of it, to block the goalkeeper's sight of the ball, to disrupt the jump, and to peel away at the last instant so that the wall broke as the ball was struck. It was effective, it produced a lot of pushing and shoving in the seconds before a kick, and it created an obvious refereeing problem.

The IFAB dealt with it directly. Where three or more defenders form a wall, all attacking players must now stay at least a metre away from it until the ball is in play, and an attacker who does not is penalised with an indirect free kick to the defending team. That is a rare and instructive case of a coached set piece idea being effective enough to change the Laws of the game rather than merely the way it is played.

Two smaller pieces of free kick craft are worth knowing because they are visible from the stand. The ball rolled square before the shot changes the angle by a few degrees and forces the entire wall and the goalkeeper to re-set while the ball is already moving. And the runner who sprints past the end of the wall converts the kick into a different problem entirely, since the defenders now have an offside line to hold as well as a shot to block.

Defending the counter from your own corner

The most expensive thing a set piece can do is concede one, and this is the part of restart coaching that supporters see least and coaches worry about most.

The situation is close to the worst defensive scenario in football. Six or seven attacking players are inside or around the opposition penalty area. The goalkeeper is at the far end of the pitch. The ball is about to be cleared by a defender who is aiming for distance rather than accuracy, which means it will travel a long way at pace, often to a forward the opposition deliberately left upfield for exactly this reason. Nobody on the attacking team is facing their own goal.

The counter-measures are all decided before the corner is taken.

The held-back players. Two, sometimes three, positioned wide of the centre circle rather than in it, because a clearance that goes central can be attacked from either side but a clearance into a channel needs somebody already in that channel. They do not join the second phase, whatever happens.

The first stopper. One player designated to close the receiver of any clearance immediately, before the ball is controlled and turned. This is the same logic that makes attacking a disorganised opponent immediately after a turnover so valuable, running in the other direction. Whoever is nearest goes, without waiting to see whether the situation looks dangerous, because by the time it looks dangerous it is.

The taker's recovery. The player who swung the corner is, at the instant it is cleared, the furthest attacking player from his own goal and the least involved in the box. His sprint back is the difference between three against two and three against three, and it starts before the outcome of the corner is known.

The goalkeeper's height. Standing on the edge of his own area rather than his line converts a long clearance into a ball he can come and claim rather than a foot race his defenders will lose.

There is a symmetrical decision on the other side, and it is genuinely a trade. A defending team that leaves two forwards upfield has a real counter-attacking threat, and forces the attacking team to hold back three players, which takes three bodies out of the penalty area. A team that brings everybody back defends the corner with eleven and has no outlet, so every clearance comes straight back. Neither is obviously right, and which one a coach picks says a lot about how the rest of the game is going.

What set piece coaching in football actually buys a club

The case for a specialist is an economic argument dressed as a tactical one, and it survives scrutiny better than most things clubs spend money on.

Begin with the shape of the problem. A club has a fixed budget, and almost every route to scoring more goals is expensive: better forwards cost transfer fees and wages, a better build-up structure takes years and a squad rebuild, and both are contested by every other club in the division. Set piece improvement is one of the few sources of goals that is bought with attention rather than money. The players are already there. The training time required is small. The competition for the resource is weak, because the resource is somebody thinking carefully about a situation everyone else treats as an interlude.

Then measure it properly, which means not measuring goals. Goals from restarts are rare enough that a season's worth is mostly noise, and a coach evaluated on them would be judged almost at random. Clubs that take this seriously measure the process instead:

• The share of deliveries that land in the intended zone.
• The share of first contacts won, attacking and defending.
• Whether the intended attacker was actually free at the moment of the delivery, regardless of what happened next.
• The share of second balls recovered.
• Shot volume and shot quality from restarts, compared against what a neutral delivery into that area would be worth, which is the same logic that underpins what an expected goals model is really counting.
• Goals conceded on the counter from the team's own attacking restarts.

Those numbers accumulate fast enough to be meaningful within a handful of matches, and they are the ones a coach can actually change.

Corners taken across a season200

Deliveries meeting an attacker cleanly60

Attempts on goal, first and second phase30

Goals6

Constructed arithmetic with invented round numbers on stated assumptions. Not a measurement from any competition. A hypothetical side takes 200 corners in a season; three in ten produce a clean attacking contact, half of those produce an attempt on goal counting first and second phase, and one in five of those attempts is scored. The point is the shape of the funnel, not the values.

Show the numbers

Worked example: the funnel from corners taken to goals scored

Item	Value
Corners taken across a season	200
Deliveries meeting an attacker cleanly	60
Attempts on goal, first and second phase	30
Goals	6

Run that constructed funnel twice. If a specialist improves the delivery accuracy and the movement enough to raise clean contacts from sixty to seventy-five, and the rest of the ratios hold, the same two hundred corners produce a little over seven goals instead of six. Do the same on the defensive side and the swing doubles. Against the cost of a transfer fee that would produce a comparable improvement in goal difference, a coach's salary is not a close call, and this arithmetic is why the role spread across football so quickly once one club made it visible.

The job itself is mostly not what people imagine. It is a video job. Opposition defensive structures are catalogued in advance, and because they change slowly, the catalogue is reliable. Routines are designed against a specific opponent's specific weakness: a zonal line that sets up too deep, a marker who consistently loses his man on a back-post run, a team that leaves nobody at the near post against out-swingers. The pitch work is short and frequent, because a routine needs repetition rather than duration, and because eleven players standing about while a coach explains a diagram is the least efficient use of a training pitch ever invented.

There is a limit worth stating. A set piece programme is a marginal-gains operation, and marginal gains are only decisive when the rest of the team is close to competitive. A side being outplayed for ninety minutes every week does not fix that with corners. What restart work reliably does is convert a team that draws matches into one that wins them, which is a narrower claim than the enthusiasts make and a much more useful one.

Watch the setup, not the ball

The next time a corner is awarded, ignore the taker completely for ten seconds and look at the other twenty-one players instead. Everything worth knowing happens before the ball moves, and none of it is in the camera shot that follows the flight.

Count who is outside the box. Both teams. The attacking side's held-back players tell you how worried they are about the counter. The defending side's outlet forwards tell you whether they intend to break or simply survive. That count is a coaching decision, and it is the clearest single signal of how each side is approaching the game.

Find the zonal players. Look for defenders standing on the six-yard line facing the corner flag rather than watching an opponent. How many there are, and how deep the line is, tells you what delivery they expect. If the line is deep and the corner comes in out-swinging, the ball will drop behind them, and if that happens twice, it is a structural error rather than bad luck.

Watch the stack break. Attackers who form a line and then split are running a routine. Which way the first one goes, and whether anyone simply stands still while a marker runs into him, is the whole mechanism of the block, and it is far more visible than replays make it look.

Note who does not jump. Somewhere near the penalty spot there should be an attacker staying low, and somewhere near the top of the area there should be one or two waiting rather than joining. Those players are the second phase, and their positioning is a better guide to whether a team has been coached than anything that happens in the air.

Follow the taker after the delivery. If he sprints back towards halfway before the header has been won, his team has a rehearsed answer to the counter-attack. If he stands admiring the outcome, it does not.

Ten corners watched that way will tell you more about how two teams are prepared than a full match of watching the ball. The whole point of a restart is that the preparation is legible: it is written on the pitch, in the arrangement of twenty-two people, in the two or three seconds before anybody kicks anything.