How VAR works in football, and what it cannot review

The ball is in the net, the away end is halfway down the stairs, and then everything stops. Not because of the assistant referee's flag, which never moved, but because the referee has a finger pressed to his ear and is standing very still on the halfway line. In a room some distance away, a colleague of his is watching the same goal from eight angles and deciding whether to say anything at all.

That pause is the part everyone hates, and it is the only visible part of the system. Ask how does VAR work in football and most answers describe the pause. The pause is the exception. It runs continuously whether anybody notices or not, and nearly everything the video assistant referee does happens without a stoppage, without a signal and without a single person in the ground being aware that it happened.

What follows is the mechanism underneath: what may be reviewed and what may not, what the intervention threshold really is, who sits in the video operations room and what each of them does, how an offside decision was assembled before it was automated and what automation actually changed, and which of the standard complaints survive contact with the protocol.

What VAR can review in football, and why the list is short

There are four categories. That is the whole remit.

Goals, and offences in the build-up to them. Not just whether the ball crossed the line, but whether anything in the immediate passage of play that produced the goal was itself an offence.

Penalty decisions, and offences in the build-up to them. Both directions: a penalty given that should not have been, and a penalty missed that should have been. The same build-up principle applies.

Direct red card incidents. Serious foul play, violent conduct, denial of an obvious goalscoring opportunity, spitting, and the rest of the sending-off offences. Explicitly not a second caution. If a referee shows a player a second yellow, the video assistant referee has no standing to intervene, however plainly wrong the first or the second may have been.

Mistaken identity. The referee cautions or sends off a player, and it was the wrong one. This is the least glamorous category and the least argued over, because it is purely a matter of who was where.

Everything else sits outside. A corner given the wrong way, a throw-in, a foul in midfield, a free kick on the edge of the box that leads to nothing, a handball that does not produce a goal or a penalty: none of it is reviewable, no matter how clearly the cameras show the error. The VAR can see a hundred mistakes in a match and is permitted to speak about roughly none of them.

The shortness of the list is deliberate, and it is the single most misunderstood design decision in the whole system. The protocol was built on a principle its authors state plainly: minimum interference, maximum benefit. Every extra category added to the list would create a new class of stoppage, a new thing for players to appeal, and a new reason for the game to stand still. A referee who can be corrected on throw-ins is a referee nobody obeys on throw-ins.

The four categories were chosen because they share one property. Each of them either scores a goal, creates the chance to score one, or removes a player for the rest of the match. They are the decisions from which there is no recovery. A wrong throw-in near the halfway line can be absorbed by the next ninety seconds of football. A wrongly awarded penalty cannot be absorbed by anything.

• 4Categories of decision VAR may review
• 3Routes from a check to a final decision
• 0Second yellow cards inside VAR's remit
• 1Officials who take the final decision

Every one of these is set by the protocol itself rather than by any competition's practice.

The check nobody sees, and the review everybody does

The two words that would end most arguments about this system are check and review, and they are not the same thing.

A check is automatic. Every goal, every penalty appeal the referee has ruled on either way, every incident that might be a straight red, every case of possible mistaken identity is checked, in real time, as a matter of routine. Nobody has to ask for it. The video assistant referee is watching the broadcast feed with the incident already isolated, comparing what he can see with the decision the referee made on the pitch.

The overwhelming majority of these checks end in silence. The VAR sees nothing that meets the threshold, says "check complete" or its equivalent into the headset, and the game carries on. The referee may not even acknowledge it. In many cases the check is finished before the ball is next out of play, and the only trace it left is a line in a log.

This is why the system appears to do almost nothing and then suddenly does something enormous. It is working constantly. You only find out about it when it fails to stay quiet.

A review is what happens when a check does not end in silence. The VAR tells the referee that he believes the decision was a clear and obvious error, or that a serious incident was missed entirely, and describes what he has seen. He does not announce a verdict. He describes evidence. From there the referee has two ways to reach a final decision.

For a factual matter, the referee may accept the information without leaving the pitch. Whether a player was in an offside position, whether the ball had gone out, whether the contact happened inside the penalty area or a foot outside it: these have answers that do not depend on interpretation, and the referee does not need to see a screen to accept one. This is usually called a VAR-only review, and it is why a great many overturns happen with no trip to the monitor and no theatre.

For anything involving judgement, the referee goes and looks. He signals a rectangle with his hands, jogs to the referee review area at the side of the pitch, and watches the incident himself.

1. 1IncidentPlay continues. The video assistant referee is already on it, because every goal, penalty decision, straight red card and possible mistaken identity is checked automatically.
2. 2Silent checkThe VAR watches the broadcast angles as they arrive and compares them with the decision given on the pitch. Most checks end here, often before the ball is next out of play, and nobody in the ground knows one took place.
3. 3RecommendationIf the VAR believes the decision was a clear and obvious error or that a serious incident was missed, he says so and describes what he has seen. He offers evidence rather than a verdict.
4. 4The referee chooses how to lookFor a factual matter he may accept the information where he stands. For anything requiring judgement he signals the television rectangle and walks to the monitor.
5. 5On-field reviewAt the pitchside screen the replay operator plays the angles the referee asks for, at the speeds he asks for. The referee is being shown the incident, not a conclusion.
6. 6Final decisionThe referee signals the rectangle again and gives the decision, which is his and is recorded as his. The time lost is added on at the end of the half.

Most incidents stop at step two and the game never notices.

What "clear and obvious error" actually means in practice

This phrase carries the entire weight of the system, and it is routinely read as though it meant "wrong". It does not.

The VAR is not asked whether he would have made the same decision. He is asked whether the decision the referee made falls outside the range of decisions a competent referee could reasonably have reached, looking at that incident, in that instant, from where he was standing. If it sits inside that range, the original decision stands. Even if the VAR disagrees. Even if a majority of qualified referees would disagree.

That is a threshold, not a standard of correctness, and the distinction produces most of the frustration around the system.

Consider two penalty appeals in the same match, near enough identical in every respect that matters. In the first, the referee points to the spot. In the second, he waves it away. Both are checked. Both survive, because in each case the referee's reading of the contact was defensible. The result is an afternoon in which the same action was and was not a penalty depending on which end it happened at, and every fan in the ground concludes the technology has failed.

The technology did exactly what it was built to do. The inconsistency was on the pitch, and the protocol was designed to protect it.

Whether that design is right is a real argument and a good one. The case for it is that the alternative is a system in which the on-field referee is a formality, every judgement call is re-made in a room by somebody who was not there, and the game's decisions migrate to a screen. The case against is that fans do not care much about referees' authority and would happily trade all of it for consistency, and that a threshold nobody can define in words will inevitably be applied differently by different people on different weekends.

There is a further wrinkle that gets almost no attention. The threshold applies to the intervention, but the referee at the monitor is not bound by it. Once he is standing at that screen looking at slow-motion replays, he is making a fresh decision with better evidence than he had the first time. In practice this means that whether the incident gets to the screen at all matters far more than what happens once it does. On-field reviews that end with the original decision unchanged are the rare ones, which tells you where the real decision was made. It was made when a VAR in a room decided this one crossed the line.

How a VAR review works in football when the referee goes to the monitor

The referee review area is a monitor on a stand at the side of the pitch, positioned so that the referee's back is not to the crowd and the cameras can see his face. Everything about its placement is a communication decision.

He could be told the answer through the earpiece. It would be quicker. The reason he is not is that the protocol insists the final decision belongs to the referee, and a decision announced from a room by somebody the crowd cannot see is not obviously the referee's. So he walks over, he watches, and he signals. The walk is not a formality wasting your time. It is the mechanism by which the authority stays in the middle of the pitch.

What he actually sees at the screen is chosen for him. The replay operator queues the angles, and the referee asks for what he wants: this camera, that camera, from the touchline, from behind the goal, at full speed, then frame by frame. Refereeing practice has settled on a rough rule for how to watch. Full speed for anything about force, intensity or recklessness, because slow motion makes every collision look brutal and every arm look deliberate. Slow motion or a freeze for anything about position and point of contact, because that is what the human eye genuinely cannot resolve in real time.

That distinction matters more than the number of angles. Play a fair shoulder-to-shoulder challenge at a quarter speed and it acquires a menace it never had. Play a stamp at full speed from a poor angle and it disappears. The referee's job at the monitor is partly to resist what the replay is doing to him.

Two things he is not doing. He is not being shown a preferred angle by somebody with a view on the outcome, at least not by design; the replay operator's job is to supply what is asked for. And he is not consulting the VAR about the answer while he watches. The VAR's contribution ended when the recommendation was made.

Then he signals the rectangle again, gives the decision, and the game restarts in the manner the new decision requires. If a goal is disallowed for an offence in the build-up, the restart is the free kick for that offence, not a drop ball or a kick-off. Getting the restart right is a surprisingly common source of error in itself, and one of the few genuine embarrassments the system produces that has nothing to do with cameras.

Who is in the video operations room, and what they each do

The room is not one person watching a television.

The video assistant referee is a referee. In most senior competitions that means a current or recent referee from the same pool as the person in the middle, which is a structural fact worth holding on to when you get to the criticism section. His job is to run the check, decide whether the threshold has been met, and speak to the referee. He watches the broadcast feed and the isolated camera angles as they are cued.

The assistant VAR watches the live game. This sounds trivial and is not. The moment the VAR starts studying a replay, he stops seeing the match, and the match does not stop for him. If something happens in the seconds after an incident while the VAR is buried in a replay, somebody has to have seen it. In competitions running more than one assistant, a second AVAR may take the offside work specifically, so that the offside process runs in parallel rather than in a queue.

The replay operator is not a match official at all. This person knows the camera plan, knows which of the cameras in the ground was pointing at the far post, and can find and cue an angle in seconds rather than minutes. The competence of the replay operator determines how long you stand in the cold more reliably than anything the referees do. When a review feels interminable, the delay is usually being spent hunting for a usable angle that may not exist.

They are all on one communication channel with the referee and the assistants, and every word is recorded. In some competitions the room is at the stadium; in others it is a single centralised hub handling several matches at once, on the reasoning that a consistent room with a consistent set of operators produces more consistent decisions than a different portakabin every week.

The failure modes of a room like this are the ordinary failure modes of any team working at speed. The wrong incident gets checked because two things happened in the same passage. The angle that would have settled it is the one camera that was blocked. Somebody says "the number nine" when they mean the other number nine. None of these are technology failures, and all of them are the reason the recorded audio matters so much more than the graphics.

Semi-automated offside: what the tracking actually measures

Offside is judged at the moment the ball is played by a team-mate. Every word of that sentence is load-bearing, and the argument about the lines was never really about the lines.

Under the old system, a human being did four things in sequence. They found the frame in which the pass was played. They chose the point on the attacker's body that was furthest forward and legally counted. They chose the equivalent point on the second-last defender. Then software drew two lines, calibrated to the plane of the pitch, and told you which was in front.

Three of those four steps were human judgement dressed as measurement, and the first was the worst of them.

Broadcast cameras deliver discrete frames. The pass happens at some instant that is almost never the exact instant a frame was captured. So the operator picks the closest frame, which is either just before the kick or just after it. And in the gap between two frames, everybody on the pitch keeps moving.

• Sprinting at 8 m/s
• Jogging at 4 m/s

One frame out16cm8cm

Two frames out32cm16cm

Three frames out48cm24cm

Constructed arithmetic with invented round numbers, not a measurement from any match. Centimetres travelled if the chosen frame sits one, two or three frames away from the true moment of the kick, at fifty frames per second.

Show the numbers

Worked example: how far a forward moves while the kick frame is uncertain

Item	Sprinting at 8 m/s	Jogging at 4 m/s
One frame out	16cm	8cm
Two frames out	32cm	16cm
Three frames out	48cm	24cm

Take a forward running at eight metres a second. That is a made-up figure, chosen because it is round and unremarkable for a striker chasing a through ball. At fifty frames a second, one frame lasts twenty milliseconds, and in twenty milliseconds that forward has travelled sixteen centimetres. Two frames of uncertainty is thirty-two centimetres. Add the defender moving the other way and the gap between two candidate frames can comfortably exceed the margin the whole decision turns on.

Which means the famous complaint about the lines had the wrong target. The lines were drawn accurately on the frame they were drawn on. The question was whether that was the right frame, and nobody watching at home could tell, because the graphic never showed you the alternative.

Semi-automated offside technology attacks precisely that problem. Dedicated cameras fixed around the stadium track many points on every player's body, many times a second, and turn them into three-dimensional positions in the real space of the pitch rather than pixels on a flat broadcast image. Separately, the moment the ball is played is established from the ball itself or from the tracking data, at a sampling rate far finer than the gap between broadcast frames.

What the system actually measures is worth being precise about, because the marketing tends to blur it. It does not measure offside. It produces two things: the position of every tracked body point on every player at every sampled instant, and the instant at which the ball was played. The offside judgement is then a geometric comparison, made by software, on inputs that a human still has to accept.

1. 1Track the bodiesDedicated cameras fixed around the stadium follow many points on each player and convert them into three-dimensional positions in the real space of the pitch.
2. 2Find the kickThe instant the ball is played is taken from the ball or from the tracking data rather than from an operator scrubbing through broadcast frames. This is the step that used to be guesswork.
3. 3Compare the geometrySoftware takes the relevant limb positions at that instant and compares the attacker with the second-last defender. The comparison is arithmetic rather than a line drawn by hand.
4. 4Alert the video roomThe system flags a possible offside to the video officials. It does not send anything to the referee and it does not decide anything on its own.
5. 5Validate and informA human confirms the system has the right kick, the right attacker and the right defender. Only then is the referee told, and only then is the three-dimensional animation generated for broadcast.

The last step is the one people forget. The animation is produced after the decision, not in order to make it.

Two things follow from that sequence, and both cut against the way the technology is usually described.

The first is that the human has not been removed. The human has been moved. Where an operator used to construct the measurement, an operator now validates one, which is a genuinely easier task and a far more auditable one, but it is not the same as automation and nobody should pretend otherwise.

The second is that the animation everyone sees is a rendering of a decision that has already been made. It is a communication tool, not evidence. This is a real improvement on the old two-line graphic, which looked like evidence and was not, but it is worth knowing which way round the causation runs.

What automation changed, and what it merely moved

The best thing semi-automated tracking did was change the subject.

Under the old system, every marginal offside produced an argument about the measurement. Was that the right frame. Is that line straight. Why is the line on his armpit. Are those two lines even parallel. The argument was unresolvable because the evidence needed to resolve it was never shown, and the people arguing had no way to inspect the inputs.

Now the argument is almost entirely about the law. Should a toe count. Should an armpit count. Should there be a tolerance band, so that anything within a few centimetres is given to the attacker. Is a system that can resolve a decision to a couple of centimetres actually measuring something the law was ever written to care about.

Those are much better arguments, because they are arguments people can actually have. They are about what the game wants rather than about whether a graphic was drawn honestly, and they can be settled by lawmakers rather than by shouting. The offside law was written when the naked eye was the only instrument available, and it contains no tolerance for the simple reason that nobody imagined an instrument precise enough to need one. That question, and the surprisingly long history of the law being rewritten whenever it stopped producing the football people wanted, is worth taking on its own terms in the full explanation of the offside law.

What automation did not change is the part of offside that involves judgement. Being in an offside position is not an offence. Becoming involved in active play is, and involvement includes interfering with an opponent, which the law describes in qualitative terms: challenging for the ball, clearly obstructing an opponent's line of vision, making an obvious action that clearly affects an opponent's ability to play the ball. No camera resolves any of that. A tracking system can tell you a striker's shoulder was eleven centimetres beyond the last defender. It cannot tell you whether the goalkeeper could see through him.

Factual and subjective, and why one of them was never going to work

Here is the distinction that explains almost every VAR argument you have ever had.

The question	What kind of question	What the video adds
Did the ball cross the line	Factual	A definitive answer
Was the attacker in an offside position	Factual	A definitive answer, given the right kick moment
Was the contact inside or outside the area	Factual	A definitive answer
Which player committed the offence	Factual	A definitive answer
Did contact occur at all	Nearly factual	A much better view, and occasionally still doubt
Was that contact a foul	Subjective	Better evidence for a judgement that stays a judgement
Was the arm in an unnatural position	Subjective	Better evidence, and a new argument about slow motion
Was the challenge reckless or serious foul play	Subjective	Better evidence, and a strong pull towards over-punishment
Did the offside player interfere with an opponent	Subjective	Almost nothing

Read down the two halves of that table and the public argument about VAR resolves itself. On the factual questions the system is close to unimpeachable, and it has genuinely removed a category of injustice that used to decide seasons. On the subjective questions it cannot do what people want it to do, because what people want is for disagreement to end, and disagreement about whether a shove was a foul is not a technical problem.

The trap is that the four reviewable categories mix the two freely. An offside decision starts factual and can end subjective. A penalty decision is usually entirely subjective except for the location of the contact, which is factual. A red card is a judgement about force and intent, informed by footage that systematically distorts both.

Handball is the extreme case, and the reason it generates more heat than anything else in the sport. The law's own terms are qualitative: whether the arm has made the body unnaturally bigger, whether the position is a justifiable consequence of the player's own movement, whether the ball came off the player's body first. Those are not questions a camera answers. They are questions a person answers, and a camera can only supply a better picture for people to disagree over. When a handball decision infuriates you, the video assistant referee is usually the last link in a chain whose weak point is the wording of the law.

Football is not the only sport that ran into this wall, and it is instructive to see where the others chose to put their boundary. Rugby union built its television match official into the referee's spoken decision-making process, broadcast to the crowd as it happens, which is a completely different answer to the transparency problem and comes with its own costs, as the breakdown of how the TMO operates sets out. Cricket rationed the thing entirely, giving each side a fixed number of player-initiated reviews and building an explicit tolerance into the ball-tracking so that marginal calls are handed back to the on-field umpire, which is the closest any sport has come to writing "we accept some error" into its own rules; the explanation of how DRS works shows why that choice was made and what it costs. Tennis went furthest and removed the human from line calling altogether, which was only possible because the question is purely factual and nobody has to judge intent. The mechanics of automated line calling are worth reading precisely because they show what a sport can do when its decisions have no subjective component at all.

Football has the hardest version of the problem. Its most consequential decisions are also its most subjective ones.

The attacking possession phase, and why a goal can be pulled back

A goal goes in. Ninety seconds earlier, at the other end, a defender was fouled and the referee played on. Can the goal be disallowed?

Not for that. The check is bounded, and the boundary is the attacking possession phase.

The idea is simple enough. If the attacking team should never have had the ball in the first place, or committed an offence while it had it, then what it did with the ball is tainted. So the check runs backwards from the goal to the start of the phase in which the attacking team gained possession, and stops there.

The phase ends, going backwards, when possession changed hands or when play was stopped and restarted. A goal scored directly from a corner has a very short checkable phase: essentially the corner and what followed it. A goal scored at the end of two minutes of patient build-up carries a long one, and every touch inside it is in the window.

That asymmetry is not widely understood and it produces an unfairness of a subtle kind. Two teams score identical goals. One came from a turnover and a three-pass break. The other came at the end of a long spell of possession. The second goal has spent far more time exposed to a check, simply because there is more of it to examine. Nobody designed that. It falls out of the rule.

The phase boundary also explains the disallowed goals that feel most arbitrary. A foul six passes earlier, in the same phase, that nobody appealed for and the referee did not see, can remove a goal that has already been celebrated. The offence was real. The evidence is clear. It simply arrives in a form the game has no emotional vocabulary for, because for a century and a half an unpunished foul in the build-up was gone the moment play moved on.

There is a further consequence that changed how the game is actually played. Because an offside can be checked after the ball is in the net, assistant referees in senior competitions are now instructed to keep the flag down when a goalscoring opportunity is developing and the offside is not obvious. The reasoning is sound: raising the flag kills a legitimate goal that cannot be given back, while delaying it costs a few seconds if the player really was offside. The effect on the pitch is that defenders stop, attackers keep running, and everybody plays several seconds of football whose status nobody knows. Late tackles have been made on players who were already offside. Goalkeepers have been hurt saving shots that should never have been taken. It is a real cost, it was accepted deliberately, and it barely features in the arguments about delay.

What VAR does to the ninety minutes, and to the ground

The Laws require the referee to add on the time lost to reviews, and referees do, which is why added time has grown and why the complaint that VAR steals football is close to backwards. Measured in ball-in-play time, a reviewed match is not obviously shorter than an unreviewed one.

The cost is not measured in minutes. It is measured in what happens to the shape of the experience.

A goal used to be an event with no gap between the thing and the feeling. That is the whole architecture of the sport: a low-scoring game in which the release is enormous precisely because it is rare and instantaneous. A review inserts a delay into the one moment the game exists for. Plenty of supporters have learned to hold back, to celebrate at half volume and wait for confirmation, which is a small and permanent loss that no amount of accuracy compensates for. A goal disallowed three minutes after it was scored hurts far more than a goal never given, because the joy was real and had to be taken back.

Inside the ground the problem is worse than on the sofa, and this is the most under-discussed part of the whole business. The viewer at home gets the replays, the graphics, a commentator explaining what is being checked and, in some competitions, the referee's own audio. The person in the stadium, who paid considerably more for the privilege, historically got a man standing still with his hand on his ear and no information whatsoever. Announcements over the public address, replays on the big screen and referees explaining the outcome to the crowd have all been trialled or adopted in various competitions, and every one of them exists to close a gap that should never have been allowed to open. A system that makes the paying customer the least informed person in the transaction has a design fault, however accurate it is.

There is one small and genuine gain in the other direction. Surrounding the referee to demand a decision achieves less than it used to, because the referee is no longer the only person looking. The compensating behaviour, players making the rectangle gesture to demand a review, is itself a cautionable offence, which is at least a rule the game can enforce. The stakes attached to all of this run a long way beyond a single afternoon, since one overturned penalty in May can settle who stays up, which is the thin end of a very large financial wedge covered in the piece on how promotion and relegation works.

Where the criticism is fair, and where it is aimed at the wrong thing

Some of the standard complaints are correct.

Transparency was handled badly and the damage is lasting. For years the conversation that produced a decision was recorded and not released, or released selectively, long after anybody still cared. Rugby made the referee's reasoning audible to everybody in the ground in real time and has a fraction of the suspicion to show for it. A system that asks for trust while withholding the material that would justify it should expect exactly the reception it got.

The threshold is applied inconsistently, and it cannot easily be otherwise. "Clear and obvious" is a judgement about whether somebody else's judgement was defensible. That is a judgement squared, made by a small pool of people who work together every week, about the competence of a colleague. There is a structural conflict sitting in that arrangement which no amount of training removes.

The delayed flag has a physical cost. Players are running into contact in passages of play that will be erased. That is a safety trade the game made quietly and has never really justified in public.

Reviews are too slow when the evidence is poor. And the evidence is often poor, because the cameras were installed to make television, not to adjudicate. A stadium's camera plan is designed around narrative, not around covering every square metre from two angles.

Some of the complaints are aimed at the wrong target.

"VAR ruined the offside law." The offside law is what it is. Precise measurement did not change the rule; it revealed that the rule, applied exactly, produces outcomes the game finds distasteful. If a toe should not count, change the law. That is a lawmaking question, and the video officials have no vote in it.

"VAR should not be deciding subjective calls." It does not decide them. It refers them to the referee, who decides them at a monitor. What people usually mean by this complaint is that they disagreed with the referee, which was also possible before.

"VAR takes forever." The judgement almost never takes long. What takes long is locating a usable angle, establishing where the attacking possession phase began, and fixing the exact moment the ball was played. Those are evidence problems, and the answer to them is better cameras and better tracking, which is precisely the direction the sport has gone.

"There were fewer controversies before." There were fewer visible controversies before. A wrong decision used to be weather: bad luck, unknowable, forgotten by Tuesday. Now every error is attributable, documented, shown from four angles, and arrives with the knowledge that a person had the evidence in front of them and got it wrong anyway. The volume of injustice went down and the volume of complaint went up, and both are consequences of the same change.

That is the deepest thing to understand about the whole system. It did not introduce error into football. It made error legible. A sport that spent a century agreeing not to look too closely at its own decisions now looks very closely indeed, and discovers, every weekend, that the underlying judgements were always this contestable. The cameras did not create the disagreement. They stopped it being possible to pretend the disagreement was not there. The rest of the machinery those decisions sit inside, tactical and financial, runs through the wider football coverage.

What to actually watch for during a review

Next time the referee stands still with his hand on his ear, four questions will tell you more than any amount of studio argument.

Which of the four categories are we in? If the answer is none of them, nothing is going to happen, whatever the replay shows. A great deal of televised outrage is spent on incidents that were never reviewable in the first place.

Is the question factual or subjective? A factual question has an answer and will produce a quick, undramatic outcome. A subjective one is going to end with somebody unhappy whatever is decided, and the review will not change that.

Did the referee go to the monitor? This is the single most informative signal available to you. A referee who stays on the pitch has accepted a fact. A referee walking to the screen has been told his judgement may have fallen outside the acceptable range, and the original decision is far more likely to change than to survive.

Where did the attacking possession phase start? For any disallowed goal, this is the whole game. Find the moment the attacking team won the ball or the play restarted, and everything from there to the goal is inside the window. Everything before it is not, however egregious.

Get those four right and you will be ahead of most of the people shouting, and occasionally ahead of the commentary. You will also find that the decisions genuinely worth the anger are rarer than they feel, and that the ones that deserve it most are usually about a law somebody wrote, not a camera somebody installed.