Football set piece routines: how each one is designed

A midfielder jogs to the corner arc, glances at the bench, and holds up two fingers. Four attackers who were spread across the eighteen-yard line converge into a tight line beside the near post. Nothing about that arrangement is comfortable or natural. It exists because somebody spent a Tuesday morning working out that this specific opponent's near-post defender steps out early, and that a line of four bodies makes stepping out impossible. Football set piece routines are engineered objects, and the interesting question is never whether they work but what each part of them is for.

Here is the anatomy in one paragraph, because everything below is an elaboration of it. A routine specifies four things: a target zone the ball is going to, a trigger that starts the movement, a mechanism for separating one attacker from one defender, and an arrangement for the ball that is not cleanly won. Delivery type, blocking, near or back post design, short options and throw-in patterns are all instruments serving those four decisions. Get the four right against the structure actually in front of you and the routine has done its job even if nobody scores, because a routine's purpose is to manufacture a free contact in a dangerous place, and the rest is finishing.

• 16.5Metres the penalty area extends from the goal line
• 5.5Metres the goal area extends from each goalpost
• 11Metres from the goal line to the penalty mark
• 9.15Metres opponents must retreat at a corner kick

Field dimensions and distances set by the Laws of the Game. A routine is a plan for moving bodies around inside these fixed shapes.

The four things a routine has to specify before anyone runs

Ask a coach to describe a routine and the description will not begin with the delivery. It begins with a place.

The target zone. Not a player, a patch of ground. Corners are aimed at areas roughly the size of a car: the space just in front of the near post, the central strip along the six-yard line, the back-post corner of the six-yard box, the penalty spot, or the top of the area for a deliberate pull-back. The reason it is a zone rather than a person is that people move and get blocked, whereas a zone is still there when the plan half fails. A taker aiming at a head has one chance of success; a taker aiming at a zone has as many chances as the number of players scheduled to arrive in it.

The trigger. Something has to start the movement at a moment both the runners and the taker agree on. It is usually the taker's raised arm, a shout of a number, or the first step of the run-up. Triggers matter more than they look, because a routine's whole value is timing: a block set half a second early is a defender who has walked round it, and a run started half a second late is a header taken standing still.

The separation mechanism. This is the engineering. Somewhere in the pattern, one attacker has to arrive at the target zone with nobody attached to him, and that separation is bought before the ball is struck, using bodies, angles and the fact that a marker cannot follow two people at once.

The residue plan. Where does the ball go if it is headed away, and who is standing there? A restart that ends the instant a defender wins the first contact is a restart designed by somebody who has watched highlights rather than matches.

Two constraints sit over all four. The first is legality, which is narrower than supporters assume in some places and much wider in others. The second is repeatability: any routine used twice is on video, and the entire discipline is a contest between preparation and counter-preparation.

The target zone is chosen first, and everything else is fitted to it

Zone selection is a reading exercise, and it is done in a video room days before the match.

Defensive corner structures are stable. A side that puts three players on the six-yard line and marks four opponents individually in September is very likely doing the same in February, because changing it requires retraining eleven people in a week when they also have a match to prepare for. That stability is what makes the whole enterprise possible. A coach can know, before kick-off, which patch of the penalty area is genuinely unattended.

The zones a defence cannot cover simultaneously are the ones worth aiming at, and there are usually two or three per opponent. A team with a deep zonal line concedes the near-post area in front of it. A team with a high zonal line concedes the space behind, at the back post. A team marking entirely man-for-man concedes any space it can be dragged out of, which is why it is beaten by movement rather than by geography. A team that leaves nobody outside the box concedes the short corner and the edge.

Once the zone is fixed, the delivery follows from it, the runs follow from the delivery, and the blocks follow from the runs. Coaches who design in the other order, choosing a clever pattern and then looking for somewhere to put it, produce routines that are impressive on a whiteboard and useless against a specific opponent. The valuation of any of this, which is a separate discipline with its own difficulties, is worked through in what a set piece expected goals model can and cannot see, and this piece deliberately leaves the numbers there.

1. 1Watch the opponent's last six defensive cornersStructures change slowly, so six is usually enough to establish the shape: how many zonal players, where the line sits, who marks individually, who is posted outside the box, and where the goalkeeper starts.
2. 2Find the zone nobody ownsIdentify one or two patches the arrangement cannot cover at the same time. A deep zonal line leaves the near-post space; a high one leaves the back post; a fully man-marking side leaves whatever it can be dragged away from.
3. 3Choose the delivery that reaches that zoneFlight is dictated by the destination. An in-swinger for the near-post area and the goalkeeper's space, an out-swinger for the back-post contest, a flat driven ball for the six-yard line, a clipped ball to the edge for a shooting routine.
4. 4Assign the runner and a decoyThe runner attacks the zone. At least one other attacker runs somewhere else, hard, so that the defence has to resolve two problems and the marker on the real target has to make a choice.
5. 5Build the separationDecide how the runner gets free: a stack that splits, a static screen in the marker's path, a start from outside the penalty area, or simply a first step in the wrong direction. The half metre is bought here.
6. 6Set the triggerA raised arm, a called number, the first step of the run-up. Everyone in the pattern is timed off the same cue, because the block and the run have to happen in the right order.
7. 7Place the residueTwo players wide of centre on the edge of the box, one staying low near the penalty spot, one or two held back for the counter. Nobody is left unassigned.
8. 8Build two variants off the same shapeThe setup must look identical to a routine the opponent has already seen. The catalogue is the product, not the individual pattern, because the second time a shape appears it has already been analysed.

The order in which the decisions are actually made. Almost every step is settled before anyone stands on a training pitch.

Delivery types, and what each one asks of the attacker

There are four families of corner delivery in common use and they are not interchangeable. Each one shifts the burden of generating danger between the taker and the attacker.

The in-swinger. The ball curls towards goal. Its defining property is that it arrives already pointed at the target, so an attacker does not have to generate direction. A shoulder, a forehead, a glance that alters the line by a few degrees can all produce a goal. It attacks the space the goalkeeper occupies, which forces a decision: come and claim it, or stay and be beaten by a flick. A keeper who commits and misses has removed himself from his own goal. The cost is symmetrical, because a keeper who claims cleanly ends the attack entirely, which no other outcome does.

The out-swinger. The ball curls away from goal. The goalkeeper is unlikely to come for it, which removes him from the equation and turns the contest into an attacker against a defender in open space. The attacker must supply everything: the run, the leap, the direction and the power. Headers from out-swingers are struck rather than deflected, so this is the delivery for a side with genuine aerial ability and the timing to attack a ball moving away from them.

The flat driven ball. Hit hard and low into the near-post zone or along the six-yard line, this delivery removes time from everybody at once. Defenders cannot set, the goalkeeper cannot come, and a single touch changes the direction violently. It converts a jumping contest into a reaction contest, which suits a side whose attackers are quick rather than tall. It is also the highest-variance option, because a hard ball into a crowd frequently goes straight out for a goal kick.

The deliberate clip to the edge. Underhit on purpose to a player arriving at the top of the penalty area. This is a shooting routine rather than a heading routine, and it exists because defences pack the six-yard box and leave the space twenty metres out to whoever wants it. It also produces the cleanest second phase in football, because if the shot is blocked the ball is already in the area everyone else has vacated.

The choice is not a matter of style. It follows from the target zone, from the aerial ability available, and from the way the opposition's zonal line is set. A structure organised for an in-swinger and asked to deal with an out-swinger has its whole starting height wrong by two or three metres, and a ball dropping behind a defensive line is the most reliably conceded goal in the entire subject.

Why the taker's foot constrains the whole catalogue

This is the least glamorous constraint in set piece design and it shapes more routines than anything else.

An in-swinging corner from the right requires a left-footed striker of the ball. An out-swinger from the right requires a right-footer. The mirror applies on the left. A team therefore cannot simply decide which deliveries it wants; it can only choose from what its available takers can produce, and the number of players in a squad who can put a ball reliably into a zone the size of a car from forty metres is smaller than most supporters imagine.

The practical consequences run through team selection. A side with two right-footed takers is committed to in-swingers from the left and out-swingers from the right, whether or not the opponent's weakness happens to be on the correct side. A side that substitutes its only left-footed taker in the seventieth minute has quietly halved its corner catalogue for the closing twenty. And a manager who values a specific delivery highly enough will keep a player on the pitch for it, which is a real cost paid in a currency nobody records.

There is a related question about who takes a corner from a position that can also be crossed with the other foot. Some sides use two takers standing over the ball, one of each foot, precisely so that the delivery type is not announced by the walk to the arc. That looks like theatre and is not: it withholds information the defence needs in order to set its line height.

Near-post routines and the geometry of a flick-on

The near-post flick is the oldest designed routine in football and it survives because the geometry is genuinely favourable.

The mechanism is a redirection rather than a strike. An attacker attacks the ball at the near post and, instead of trying to head towards goal from an impossible angle, glances it on so that it continues across the six-yard box behind the defence. The value is that a single touch converts a cross into something no defender is facing: a ball travelling across the goalmouth, at head height, behind a line that has already committed forward, arriving at players running onto it rather than waiting for it.

Three things have to be true for it to work.

The delivery has to be in-swinging or driven, because a flick only redirects; it cannot reverse. An out-swinger flicked at the near post goes away from goal and out for a throw-in.

The flicker has to attack the ball in front of the defender rather than beside him. That is a timing problem and it is solved with a start from deeper than looks sensible, so the attacker meets the ball moving forward at pace while the defender is moving sideways.

And somebody has to be arriving at the back post on the far side of the goal, unmarked, with the whole width of the goal in front of him. The flick is worthless without the second runner, which is why near-post routines fail so visibly when the timing is off: the ball is glanced perfectly into a space nobody has reached.

The defensive counter is the reason zonal marking exists at all. A defender standing on the near post, facing the ball, jumping forward into the flight, beats a running flicker far more often than a man-marker chasing him does. That single duel is the origin of the whole marking argument, which is set out at length in how man marking and zonal marking trade off against each other.

Back-post routines, the overload and the late runner

Back-post design solves a different problem: how to get a free header at the far side of the six-yard box against a defence that knows it is the most dangerous place on the pitch.

The commonest solution is the overload. Attackers deliberately crowd one area, usually the centre or near post, so that defenders follow them there, and one player peels away into the space that has been emptied. Two attackers occupying three defenders in a small space is a profitable trade even though it appears to be the opposite, and the same logic that makes numerical concentration valuable in open play is explained in how overloads create advantages elsewhere on the pitch.

The second solution is the late runner starting from outside the penalty area. This is the hardest attacker to defend for a structural reason: a zonal system has nobody assigned to a player who is not in a zone, and a man-marking system has to decide whether to follow him out of the box and abandon the space he came from. Either answer is exploitable. The late runner also arrives at speed against defenders who are stationary, and momentum in the air is worth several inches of height.

The third is the double movement, where an attacker takes two hard steps towards the near post and then breaks back. Against a man-marker this manufactures separation on its own, because the defender's centre of gravity is committed. Against a zonal line it does nothing at all, which is a good illustration of why routines are chosen after the opposition's structure has been identified rather than before.

The delivery for all three is usually an out-swinger or a hanging ball with height, because the contact point is further from goal and the attacker needs time to arrive.

40%20%20%10%10%

• Runners attacking the three target zones4
• Screening or blocking roles2
• Waiting on the edge of the area for the second ball2
• Held back as counter-attack insurance1
• The taker1

A constructed illustration with invented round numbers, not a measurement and not any particular team's arrangement. The point is that every player given to the edge or to the counter is a player not attacking the six-yard box.

Show the numbers

Worked example: how ten attacking outfield players are assigned at a corner

Item	Value
Runners attacking the three target zones	4
Screening or blocking roles	2
Waiting on the edge of the area for the second ball	2
Held back as counter-attack insurance	1
The taker	1

Blocks and screens, and where Law 12 draws the line

Most of what a routine does is buy half a metre, and most of that half metre is bought with a body. The legal position is clearer than the arguing suggests.

Holding an opponent is a direct free kick offence, and referees are specifically directed to deal with holding at corners: warn players 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, which means moving into someone's path to obstruct them when you are not within playing distance of the ball, is an indirect free kick offence.

What is permitted is much larger. A player may occupy a piece of ground. He may stand still in a place that is exactly where a marker intends to run. He may move to a position and then stop. He may stand in front of the goalkeeper obstructing the view, provided he does not hold him or prevent him releasing the ball. Every one of those is legal, and all of them are coached, which is why penalty appeals at corners so rarely succeed and why supporters find the whole area maddening.

The reliable mechanisms therefore use geometry rather than hands.

The line that splits. Four attackers stand shoulder to shoulder in a vertical column before the delivery, with their markers behind or beside them, and break in different directions on the trigger. A defender standing behind a column of four cannot reach his man without going round it, and the first two steps of the run are conceded without anyone having touched anyone.

The static wall of one. A single attacker positions himself on the exact line a marker must travel and does not move. The marker runs into a stationary body. It is legal, it looks like a foul from every seat in the ground, and it is the most common source of the appeals that go nowhere.

Standing in front of the goalkeeper. No contact is needed. A keeper whose sight of the flight is broken for a fraction of a second decides late, and deciding late on a crossed ball is the whole difference between claiming it and watching it.

The runner nobody wants to follow. An attacker who sprints to the near post and never receives the ball is doing two useful jobs, pulling his own marker out of the middle and dragging a zonal defender forward so that the space behind him becomes available to somebody else.

The one genuinely restricted area is the defensive wall at a free kick. Where three or more defenders form a wall, attacking players must stay at least one metre away from it until the ball is in play, and an attacker who does not concedes an indirect free kick. That provision exists because attacking sides had become so effective at standing inside walls, blocking the goalkeeper's view and peeling away at the last instant that the Laws had to intervene, which is a rare case of a coached idea working well enough to be legislated against.

Short corners, and the arithmetic behind the booing

Short corners are jeered in most grounds and the reasoning behind them is better than the reaction.

Start with the defensive dilemma. A defending team has ten outfield players and must decide how many stand outside its own penalty area. If it puts everybody in the box, two attackers standing near the corner arc are entirely unopposed and can walk the ball into a crossing position at their leisure. If it sends one defender out, those two attackers have a two against one thirty metres from goal, and the defence has removed a body from the area where the danger actually is. There is no third option, and the attacking side gets something either way.

Then there is the delivery itself. A corner struck from the arc is a stationary ball from a fixed point at a known angle, and every defence in football has rehearsed against exactly that geometry. A cross played from eight metres infield and five metres up the touchline arrives from a different origin on a different flight, and a zonal line that has set its height for a ball from the corner has to reorganise while the ball is already in the air.

Then there is risk. A ball swung into a crowded box is a contested aerial duel whose loss produces a clearance and a counter-attack. A short corner keeps the ball with a team-mate in a controlled situation and lets the attacking side reset if nothing develops.

The costs are equally real. Two players standing short are two fewer attacking the delivery, so the eventual cross meets a thinner attack. The routine consumes time and lets a scrambling defence organise. And a short corner given away cheaply hands possession to a team already facing the right way, which is the worst possible outcome from an attacking restart.

The honest conclusion is that a short corner is a specific answer to a specific defensive arrangement. Against a side that posts nobody outside its box it is close to free value. Against a side that posts two it is a poor idea, and taking one anyway is a sign that the routine was chosen in the dressing room rather than read on the pitch.

Choosing a routine against the structure in front of you

Every defence answers one question: is a defender attached to an opponent or to a piece of ground? The answer determines which routines are worth running.

Against a man-marking structure, movement is the tool. Blocks work, because separating one defender from one attacker for half a second is a simple problem. Double movements work. Late runners from outside the area work, because someone has to make a decision about following them. Deliveries into congested areas work, because man-markers finish a corner facing their own goal, scattered, in whatever positions they were dragged into.

Against a zonal structure, geography is the tool. Blocks are close to useless, because there is nobody to block: the defenders are already standing where they intend to be. What works instead is attacking the seams between zones with two runners, and testing the height of the line by varying the delivery. A zonal line drilled for an in-swinger and served an out-swinger will let the ball drop behind it, and if that happens twice in a match it is a structural error rather than misfortune.

Against the hybrid that nearly every senior side actually uses, with a small zonal group holding the near post and the six-yard line and individual markers on the two or three genuine aerial threats, the target is the boundary between the two systems. That is where responsibility is ambiguous, where a ball dropping on a line produces hesitation, and where a runner can move from a marked situation into an unmarked one inside two steps.

There is a simpler heuristic that holds up well. Look at how many defenders are standing outside the penalty area and how many are facing the ball rather than an opponent. Those two counts tell you which family of routine the opposition is vulnerable to, and they are visible from the stand in the ten seconds before the kick is taken.

Free kick routines from wide areas and from shooting range

Free kicks split into two entirely different design problems and the only thing they share is a stationary ball.

A wide free kick is a corner with better angles and a defence that cannot pack the six-yard box, because it has an offside line to hold. That single difference changes everything. Attackers can start from deep and time runs against a line rather than against a marker, so the routines look more like open-play crossing patterns: a run across the front to occupy the first defender, a hold at the back post, and a delivery into the space between the goalkeeper and the last line. The defending side's decision about how high to push its line is the whole contest, and the way an offside line is used and attacked is set out in the offside law and the two questions it asks.

A central free kick in shooting range is a wall problem. The wall is not there to block the goal. It is there to remove one half of it so the goalkeeper can commit to the other, which is why the player at the end of the wall nearest the near post must not move: the goalkeeper's whole position depends on the wall covering precisely what he was told it would cover.

Attacking routines against a wall come in four varieties. Over it, which requires the ball to rise and fall inside a window of a few centimetres and is the rarest skill in football. Around it, which the goalkeeper can usually see all the way. Under it, which produces more goals than its simplicity suggests, because a wall that jumps opens a gap at ground level for a fraction of a second, and which is why defending sides now lie a player on the turf behind the wall. And past it entirely, by rolling the ball square before the shot, which changes the angle by several degrees and forces the entire wall and the goalkeeper to reset while the ball is already moving.

The fifth option is not a shot at all. A runner sprinting past the end of the wall converts the free kick into a crossing situation with an offside line to hold, and defences that have organised entirely around blocking a shot are frequently very bad at that.

Throw-in routines, long and short

Throw-ins outnumber corners and free kicks combined, and for most of football's history nobody coached them. That has changed, and the two halves of the subject are worth separating.

The long throw into the penalty area is a delivery with two properties no other restart has. There is no offside offence directly from a throw-in, so an attacker may legally stand behind the last defender and receive it. And the ball is delivered with the hands, which for the right player means a flat, fast, accurate ball into the six-yard area that wind and spin do not affect.

Its reputation as a crude tactic gets the risk backwards. The thrower is standing still on the touchline holding the ball, so his side can be completely organised before the delivery, including the players held back against a counter-attack. A corner is struck by a player who must then sprint forty metres to recover; a long throw is not. It is a lower-risk delivery into a dangerous area, which is why it returned once somebody looked at it properly rather than at its associations.

The short throw matters more often and gets almost no attention. A throw-in taken in a team's own third, by a full-back with three opponents closing and a touchline behind him, is one of the most reliable ways to concede possession in a terrible position. Sides that coach it have rehearsed answers: a receiver coming short to set the ball back first time, a runner dragging a marker away to open the line, a designated long option, and an explicit rule about when to put the ball out of play rather than accept a fifty-fifty. None of it is exciting and it removes a category of goals conceded.

The one constraint to remember is that opponents must stand at least two metres from the point on the touchline where the throw is taken, which is a much smaller buffer than the 9.15 metres at every other restart. Space is tight at a throw-in by design, and routines have to be built for a crowded area rather than an open one.

1. 1The thrower preparesThe ball is dried, and the thrower takes a run-up along the touchline. Both feet must be on or behind the line and part of each foot must be on the ground, and the ball is delivered from behind and over the head.
2. 2The attackers take positions defenders cannot markThere is no offside offence directly from a throw-in, so an attacker may stand behind the entire defensive line. Defences have to mark man for man in an area where they would normally rely on the offside line.
3. 3A flick-on man is posted at the frontOne tall player attacks the ball at the front of the group. His job is not to score but to redirect, exactly as at a near-post corner, because the throw arrives flat and fast rather than dropping.
4. 4The runners arrive behind himTwo or three attackers time runs into the space behind the flick, where a defence facing the touchline is at its most disorganised. Nobody is jumping; they are running onto a knocked-down ball.
5. 5The defence clears under pressureA header away from a flat, fast ball is almost never controlled. It travels a short distance at an unpredictable angle, usually towards the edge of the area.
6. 6The second ball is contestedPlayers positioned outside the box attack the clearance. Because the defence is compressed against its own goal line, the space twenty metres out is empty and a first-time shot arrives through traffic.
7. 7The resetIf the ball is fully cleared, the thrower is still on the touchline in his own defensive shape rather than stranded in the opposition corner, which is the structural reason a long throw carries less counter-attacking risk than a corner.

The sequence a rehearsed long throw runs through. The offside exemption in step two is what makes the whole pattern possible.

Designing for the second phase and the exit

Two arrangements are made before the ball is struck and neither of them concerns the first contact at all.

The second phase is everything after the initial header or clearance, and a substantial share of the goals credited to set pieces arrive there rather than from the delivery. The reason is mechanical. A well-defended corner produces a clearance, and a clearance from a crowded six-yard box is not a pass. It is a header from an off-balance defender facing the wrong way, and it travels some distance in a direction close to random. Whoever has planned for where it lands owns the next few seconds of the match, and those seconds happen against a defence that is scattered and has already jumped once.

Planning for it means answering three questions in advance. Who stands on the edge, and where: two players positioned wide of centre rather than directly in front of goal, because clearances leave the box at angles more often than straight. Who stays low near the penalty spot, because the scramble at shin height has no shape and rewards whoever reacts rather than whoever leaps. And what happens if the ball is only half won and ends up in a wide area, since a second cross against a defence that has just jumped is a better proposition than the first one was.

The exit is the other half, and it is the part that keeps coaches awake. At the moment a corner is cleared, six or seven attacking players are inside or around the opposition penalty area, the goalkeeper is a hundred metres away, and the ball is travelling at pace towards a forward the opposition deliberately left upfield. That is close to the worst defensive situation in football, and it is entirely foreseeable, which means it can be arranged for: players held back wide of the centre circle rather than in it, one player designated to close the receiver immediately without waiting to see whether it looks dangerous, the taker sprinting back before the outcome of his own delivery is known, and the goalkeeper starting on the edge of his area rather than his line so that a long clearance becomes a ball he can claim. The reason all of this is worth the trouble is set out in why a disorganised opponent is the most valuable thing in football.

Why a routine has a shelf life

The uncomfortable truth about this whole discipline is that its products expire.

A routine used once is an idea. A routine used twice is a clip in somebody's video folder, catalogued, timestamped and shown to eleven defenders on a Friday. Everything that makes restarts coachable, the repeatable initial conditions, the stable structures, the finite outcome space, works equally well for the side preparing to defend them.

That is why serious set piece work is portfolio management rather than invention. The observable requirement is that the setup gives nothing away: several different routines have to start from the same visible arrangement, so that a defence recognising the shape has learned nothing about what will follow. The fourth corner of a match should do something the first three did not, and the first three exist partly to make the fourth work.

It also explains a pattern that looks like incompetence and is not. A side runs an obviously telegraphed near-post routine that achieves nothing, twice, and then in the eightieth minute the same shape produces a completely different delivery to the back post against a defence that has stopped watching that side of the box. The two failures were the price of the third attempt.

The wider argument for staffing this properly, and for measuring it by process rather than by goals, is made in what a dedicated set piece coach actually buys a club.

What to watch once the corner flag comes into shot

The next corner you see, ignore the ball for ten seconds and watch the other twenty-one players, because everything worth knowing happens before anyone kicks anything.

Count the bodies outside the penalty area. Both sides. The attacking team's held-back players tell you how much they fear the counter. The defending team's forwards left upfield tell you whether they intend to break or simply survive. That count is the clearest single statement of how each side is approaching the game.

Look for defenders facing the corner flag rather than an opponent. Those are the zonal players. How many there are, and how far off the goal line they stand, tells you which delivery they are expecting, and whether the delivery that arrives is the one they prepared for.

Watch for a column of attackers that breaks apart. That is a routine running. Which direction the first player goes, and whether anybody simply stops moving while a marker runs into him, is the separation mechanism, and it is far more visible live than it ever is on a replay that follows the flight of the ball.

Find the player who does not jump. There should be one lurking near the penalty spot and one or two waiting at the top of the area. They are the second phase, and whether they exist is the single best indicator of whether a team has been coached or is simply hoping.

Follow the taker after the delivery. If he is running back towards halfway before the first contact has been made, his side has an exit plan. If he is watching, it does not.

Ten corners read that way will tell you more about how two teams have prepared than ninety minutes of watching the ball. More tactical writing, on restarts and on everything around them, sits in our football section, and the full run of explainers across every sport we cover is in the blog.