TMO rugby explained: how the video referee makes a call

The referee's arm is out, the try has not been awarded, and a stadium that spent two minutes screaming at a maul now watches a screen in silence. Somewhere in a truck or a booth, a colleague nobody in the ground can hear is scrubbing through four camera angles looking for a foot on a line. The players stand about. The clock does not.

That pause is the price of getting it right, and it is also the most complained-about feature of the modern game. Getting the television match official in rugby explained properly means putting the replay equipment to one side, because the interesting part is not the technology at all. It is the set of boundaries World Rugby has drawn around what the technology is permitted to do. The TMO is not a second referee with better pictures. The TMO is an official operating inside a narrow written mandate that says which offences may be reviewed, how far back in the match a review may reach, and how certain the evidence has to be before anything changes.

Almost every argument about a TMO decision is really an argument about one of those three boundaries.

What the television match official actually is

The TMO is a qualified match official, appointed to the game in the same way as the referee and the two assistant referees, who works from the broadcast feed rather than from the field. In elite matches the role is filled by a referee or former referee, not a technician, because the job is a refereeing job carried out with different information.

The TMO speaks to the referee on the same communications system the assistants use. Everything said on that channel is heard by the whole officiating team, and at most competitions it is broadcast to viewers as well, which is unusual among the major sports and which has done more than anything else to shape public understanding of the role.

What the TMO cannot do is award anything. No try is given by the booth, no penalty is awarded from a monitor, no card is issued remotely. The output of a review is information passed to the referee, sometimes with a recommendation attached, and the referee then makes and announces the decision. That arrangement is not a formality. It preserves a single point of accountability for the match, and it means the person who has to manage twenty-nine other players for the rest of the afternoon is the person who owns the call.

The other official who has appeared in the modern setup is the foul play review officer, who handles dangerous play away from the live feed. That role is separate from the TMO, it works on a different clock, and it exists for reasons set out further down.

• 2Phases the TMO may look back through for offside and obstruction
• 8Minutes a foul play review officer has to reach a decision
• 10Minutes a yellow card removes a player from the field
• 20Minutes a player is off before a replacement is allowed under the red card trial

Structural limits set out in World Rugby's protocol and competition guidelines, not measured averages. Numbers of this kind change whenever Council amends the protocol.

The clear and obvious test, and why everything hangs off it

The phrase that governs the entire protocol is "clear and obvious". An incident qualifies for intervention when it is of a kind that would not sensibly be refereed any other way once seen. It is a deliberately high bar, and it is doing far more work than most viewers realise.

Consider what the alternative would be. If the standard were simply "correct", the TMO would be obliged to keep looking until certainty arrived, and rugby is a sport where certainty is frequently unavailable. A grounding under three bodies, a marginal knock forward from a hand that also took contact, a foot that may or may not have brushed the touchline behind a blur of legs: these can be studied for four minutes and remain genuinely arguable. A protocol demanding resolution would produce long stoppages that ended in a coin-flip anyway.

The clear and obvious test converts that problem into a procedural rule. If the pictures do not settle the matter quickly, the pictures have not met the standard, and the on-field decision stands by default. This is why officials are trained to give an on-field call before going upstairs wherever they can. The on-field call is not a guess to be checked; it is the position the evidence has to overturn.

It also explains a pattern that irritates viewers who follow other sports. Rugby does not usually announce a review as inconclusive. It announces the original decision, because the original decision was never suspended in the first place. The review either displaces it or it does not.

What the TMO is allowed to look at

The scope of intervention is written down as a list, and knowing the list removes most of the confusion around what happens next during a stoppage.

For decisions about a score, the TMO may deal with grounding, with whether a player or the ball was in touch or touch-in-goal, and with the act of scoring itself. This is the oldest part of the mandate and the least controversial, because a try is the largest single event in a match and the cameras are usually pointed straight at it.

Beyond the act of scoring, the protocol distinguishes between two windows that reach different distances back into the play.

The wider window covers the whole final attacking passage before a score. Within it, the TMO may raise a knock-on, a throw forward or the ball having gone into touch. These are offences where the evidence is generally a single frame, the judgement is close to binary, and a miss produces a try that plainly should not have been awarded.

The narrower window covers only the final two phases. Within it, the TMO may raise offside, obstruction at a maul, and whether a tackle had been completed. These are offences that require reading a passage of play rather than a frame, where reasonable officials differ, and where searching further back would produce an infringement in almost every attacking sequence in the sport.

Foul play sits outside both windows. Dangerous play can be brought to the referee whenever it is identified, whether or not a score followed, because player welfare is not a matter that expires after two rucks.

43%43%14%

• Reviewable across the whole final attacking passage3
• Reviewable only within the final two phases3
• Reviewable at any point in the match1

A count of the offence categories the protocol assigns to each window, not a measure of how often each is used. The categories themselves are what change when the protocol is amended.

Show the numbers

How many offence types sit inside each review window

Item	Value
Reviewable across the whole final attacking passage	3
Reviewable only within the final two phases	3
Reviewable at any point in the match	1

Where a phase begins and ends, and why the limit is two

"Two phases" sounds like a stopwatch instruction and is nothing of the sort. A phase, for these purposes, is bounded by a scrum, a lineout, a ruck or a maul. Play resets at each of those, and the count of phases is a count of those resets, not of passes, near-rucks, or seconds elapsed.

The final attacking passage is defined differently again. It runs from the last restart if possession has not changed hands, or from the last turnover if it has. A side that wins the ball back on its own line and moves ninety metres in one continuous sequence has produced a single attacking passage, however many phases it contains, and a knock-on at the start of that sequence remains inside the wider window.

Both definitions exist to answer the same question, which is the question every video review system in every sport eventually has to answer: how far back does the search go? Leave it open and you are asserting that a try scored after twelve phases can be cancelled by a marginal offside in the first, which is technically defensible and unwatchable in practice. Cut it too short and obvious errors survive into the scoreline.

Two phases is a compromise chosen because it is roughly the distance over which a defensive error still has a visible causal connection to the score. It is not a natural constant. It is a number that competitions and Council have moved before and can move again, and anybody arguing that a particular try should have been chalked off is usually arguing that the window is set too tight rather than that the protocol was misapplied. The mechanics of the resets that define a phase are covered in more detail in the guide to what actually happens at the breakdown and in the explanation of how a scrum is set and reset.

How a review starts, and who is allowed to start one

There are two routes into a review, and telling them apart is the fastest way to predict how long the stoppage will last.

The first is a referral. The referee is unsure, stops play, gives an on-field decision where possible, and asks a specific question. Referrals tend to be short, because the question has already been narrowed to one thing.

The second is an intervention. The TMO raises something the on-field team has not flagged. Interventions are restricted by design: the protocol expects the TMO to speak up unprompted only where serious foul play has been missed, or where a clear and obvious infringement has produced a try. That restriction was tightened deliberately after a period in which unprompted checking had become common enough to disrupt the rhythm of matches.

A third mechanism, added to reduce stoppages rather than to add them, lets the TMO monitor certain things without halting play at all. Where an incident can be assessed in the background and the answer is that play should continue, nothing happens and the crowd never learns that a check took place. Only when the answer changes something does the game stop.

1. 1Something is flaggedEither the referee asks a question, an assistant referee reports what they saw, or the TMO raises an incident falling inside the permitted grounds for speaking up unprompted. Anything outside those grounds is not raised at all.
2. 2The referee gives an on-field decisionWherever it is possible, the referee states what they believe happened before any replay is examined. This becomes the position that the footage must overturn, and it is the reason so many reviews end with the original call intact.
3. 3The question is narrowed to one thingThe referee specifies exactly what is being checked, such as whether the ball was grounded or whether a foot was in touch. Broad requests to have a general look are what produce four-minute stoppages, so the protocol pushes hard against them.
4. 4The footage is worked throughThe TMO selects angles and drives the replay, including any split-screen or zoom tools the competition provides. What the referee sees on the big screen comes from the officials' own station rather than from the broadcast director's cut.
5. 5The clear and obvious test is appliedIf the evidence settles the question at a glance, the review has met the standard. If officials are still weighing frames after several passes, it has not, and the on-field decision survives by default.
6. 6The referee announces the outcomeThe decision is stated to the players and, in most competitions, to the stadium. The TMO's part ends with the information; the announcement, the explanation and any card belong to the referee.
7. 7The match restarts under the appropriate lawA confirmed try goes to a conversion, a knock-on inside the passage goes to a scrum, an offside goes to a penalty, and a completed foul play sanction goes to whichever restart the offence dictates.

The sequence set out in the protocol, described step by step. Individual competitions may add their own conventions on top, so the order below is the framework rather than a script.

The referee still owns the decision, and that matters more than it sounds

It is tempting to treat the division between advice and decision as a legal nicety. It is not, and two consequences follow from it directly.

The first is that a referee may decline to accept a recommendation. This happens rarely, and it happens most often where the referee holds information the camera does not: the position they were standing in, what a player said, what the pattern of the previous ten minutes has been. The protocol does not ask the referee to defer.

The second is that the standard of proof is applied by a person on the field with a match to manage, not by a person in a room with unlimited replays. Those two people would apply the same words differently if the words were the only constraint. The referee's ownership of the outcome is what keeps a review tethered to the game being played rather than to the footage of it.

This is also why the audio matters so much. Broadcasting the officiating channel was a choice that other sports have mostly avoided, and it has made rugby's review process unusually transparent and unusually exposed. Viewers hear the reasoning, which builds trust when the reasoning is clean and destroys it when officials talk themselves into a decision on air. The contrast with football, where the video assistant referee's deliberations stay largely private, is stark, and it accounts for much of why the two systems are received so differently by their own audiences.

Foul play, and the review that happens away from the pitch

Dangerous play created a problem the ordinary review process could not solve. Head contact cases in particular are decided by a chain of questions about contact point, degree of danger and mitigation, and working through that chain properly takes longer than a stadium will tolerate while twenty-nine players stand in the cold.

The answer was to move the decision off the live feed. Where the officiating team agrees an incident reaches at least the yellow card threshold but cannot resolve whether it goes further, the referee signals a review by crossing the arms and sends the player to the sin bin. The player is off for ten minutes regardless. A separate official, the foul play review officer, then examines the incident with the full range of angles and tools, in a room away from the match, with up to eight minutes to reach an answer.

Two outcomes are possible. The yellow card is confirmed, and the player returns after their ten minutes. Or it is upgraded, and the player takes no further part.

The design has a property worth appreciating. The sanction is not suspended pending review, so the offending side is a player short from the moment of the incident either way. That removes the incentive to argue, because nothing is gained in the short term from a favourable answer. It also means that if the review runs its full length, the decision arrives while the player is still in the bin rather than after they have returned.

Nothing in this route stops the referee acting alone. An act that is unambiguously a sending-off is still sent off on the field, immediately, without review. The off-field process exists for the genuinely finely balanced cases, and the concussion side of the same welfare problem runs on its own separate track through the head injury assessment protocol.

1. 1The incident happens and play is stoppedThe officiating team looks at the footage available on the field, typically no more than a couple of replays, and forms a view on whether the act is dangerous enough to warrant at least a yellow card.
2. 2The referee decides which route to takeAn obvious act of thuggery, of the kind listed under the foul play law, is dealt with there and then as a straight red. Anything the team judges to be at least yellow but arguably more is sent for review.
3. 3The arms are crossed and the player leavesThe crossed-arm signal is the formal marker that a review has begun. The player goes to the sin bin immediately and the ten-minute clock starts running whatever the review eventually concludes.
4. 4The game restarts and the review runs in parallelPlay resumes under the appropriate restart while the foul play review officer works. This is the point of the whole design, because the stadium is not made to wait for a decision that takes minutes rather than seconds.
5. 5The reviewer works through the head contact processThe relevant questions are asked in a fixed order: was there contact with the head, was there foul play, how dangerous was it, and does anything about the situation reduce the sanction.
6. 6An answer comes back inside eight minutesIf the yellow stands, the player rejoins at the end of their ten minutes. If it is upgraded, they are finished for the day, and whether the side may replace them depends on the card law that competition is operating under.
7. 7Any further punishment is handled after the matchA red card sends the case to a disciplinary panel, which sits separately and can impose a ban. The in-match sanction and the post-match sanction are decided by different bodies against different tests.

The foul play review sequence as published by World Rugby. The threshold for what a referee sends off directly rather than referring has been adjusted more than once and varies between competitions.

The twenty-minute red card and the line the referee keeps

Sitting on top of the review process is a change to what a red card does, and it has proved the most divisive item in rugby's rulebook in years.

Under the trial that elite competitions have been operating, a side that loses a player to a red card may bring on a replacement after twenty minutes rather than finishing with fourteen. The stated case for it is that punishment should fall on the individual rather than wrecking the contest for everybody who paid to watch it, with the real deterrent moved to the disciplinary hearing afterwards. The case against is straightforward: reduce the in-match cost of dangerous play and you reduce the incentive to avoid it. World Rugby moved the arrangement to a global trial in elite rugby, with a decision on whether it becomes permanent law resting with Council.

The important interaction for anybody watching a review is that the two systems have to be reconciled. If a red card no longer ends a side's afternoon, the question of which acts still deserve a permanent sending-off becomes live again, and the boundary has been redrawn more than once. The position competitions have settled towards is that deliberate acts of thuggery listed in the foul play law, striking, biting, stamping, contact with the eye area and the rest, stay with the referee as immediate and permanent dismissals, while dangerous but non-deliberate contact goes to the off-field reviewer to be graded.

Reasonable people disagree about where that line belongs, and the fact it has moved between seasons is not a scandal. It is what happens when a sport tries to price recklessness and deliberate violence differently using a single card. Expect the boundary to be adjusted again, and check the tournament regulations rather than assuming that what applied last season applies to the match in front of you. The Six Nations and the World Cup have both run under conditions that differed from one another in the recent past.

Why the protocol has been rewritten so many times

The TMO's mandate has not evolved gently. It has swung between wide and narrow, and each swing was a correction of the previous one.

The early expansion was driven by a simple observation: the cameras could see things the officials could not, and letting obvious errors stand looked increasingly indefensible on television. So the scope widened, the look-back distance grew, and reviews multiplied.

The correction came once the cost became visible. Matches were being interrupted repeatedly, tries were being cancelled by infringements the crowd had forgotten about by the time they were shown, and the officiating team's authority was diluted by the constant possibility of a second opinion. The amendments that followed pulled in two directions at once: they cut the circumstances in which the TMO could speak up unprompted, and they introduced the ability to check quietly in the background so that a check which changes nothing costs nothing.

The more recent movement has gone the other way again, but selectively. The wider window for knock-ons, forward passes and touch decisions reflects a judgement that these are cheap to check and expensive to miss, while the tighter two-phase window for offside and obstruction reflects the opposite. That is not indecision. It is a system being tuned offence by offence rather than by a single global setting, which is a more sophisticated approach than the one it replaced and a harder one to explain over a stadium public address.

Cameras, hubs and the standards arriving behind the scenes

A protocol is only as good as the pictures it runs on, and rugby has spent years discovering that the pictures are not consistent.

The problem is structural. The TMO watches a broadcast feed, and the broadcaster decides how many cameras to deploy, where to put them and which angles to cut to. A televised fixture in a major competition may have twenty-odd cameras; a smaller one may have a fraction of that. Officials have found themselves making identical decisions from wildly different evidence, and on occasion have been unable to see an angle the audience at home had already been shown.

World Rugby's response has been to write minimum standards rather than leave it to the production. Every angle available to the broadcast must be available to the TMO and to the foul play reviewer, the imagery the referee sees on the big screen must come from the officials' own station rather than from the director's cut, and the people operating the equipment must hold the relevant certification. Those requirements are being phased in for international-level competitions, with a compliance date in 2027, and club competitions are encouraged rather than compelled to adopt them.

Running alongside this is an experiment in centralisation. Instead of flying an official to every venue, competitions have trialled a remote hub where TMOs work from a single facility with a consistent bank of equipment. The arguments for it are cost and consistency; the argument against is the loss of a match official's physical presence at the ground. It is early, and the model has been tested at age-grade and second-tier events before anything larger.

What the TMO cannot fix

A review system is a correction mechanism, and correction mechanisms inherit the limits of the thing they correct.

The TMO cannot make a subjective law objective. Whether a tackler released, whether a player was supporting their own body weight, whether a maul was legally formed: these are judgement calls, and slowing them down does not convert them into facts. Watching a marginal breakdown decision fifteen times produces fifteen opinions rather than one answer.

The TMO cannot see through bodies. The commonest unresolvable review in the sport is a grounding under a pile of players, where no camera in the ground has a line of sight and the honest answer is that nobody knows. This is where the default to the on-field call does its most important work.

The TMO cannot recover time. Every review costs the match something, and the cost is not evenly distributed: a long stoppage before a scrum late in a close game changes the physical state of the contest in ways that a stoppage in the tenth minute does not.

And the TMO cannot restore consistency across competitions running different regulations. A supporter who watches domestic rugby, an international window and a knockout tournament in the same season is watching three slightly different rulebooks, which is the largest single source of the belief that officiating is arbitrary. The differences between the two rugby codes compound the problem further, as the comparison of union and league sets out.

How rugby's review compares with other sports

Placing the protocol next to its neighbours makes its choices look less arbitrary.

Football's video assistant referee operates on a much narrower list of reviewable events and a much higher bar for intervention, restricted to goals, penalties, direct red cards and mistaken identity. It looks back only to the start of the attacking phase, and its deliberations are conducted privately. The result is fewer interventions and considerably more public suspicion about how they were reached.

American football gives the decision to a coach's challenge and a strict on-field standard for overturning, with a booth taking over in the closing minutes and for scoring plays. The mechanics of challenges and booth reviews put a limited resource in a team's hands, which rugby has deliberately never done. Tennis went furthest in the other direction and removed judgement entirely for line calls, replacing officials with a ball-tracking system whose output is binary. Cricket built a hybrid, where a team spends reviews and an on-field decision survives marginal evidence through the umpire's call provision, which is the closest analogue in any sport to rugby's clear and obvious default.

Rugby's distinguishing features are the breadth of the offence list, the openness of the audio, and the absence of any team-controlled challenge. The first two make it feel more intrusive than football's system. The third makes it feel less fair to viewers who assume, wrongly, that a captain can ask for a look.

Why rugby has never handed captains a challenge

Every few seasons somebody proposes giving each side a limited number of referrals, on the model cricket and tennis use, and the proposal has never got far. The reasons are worth understanding because they say something about what rugby thinks its officials are for.

A challenge system transfers part of the officiating burden to the teams. That works where the reviewable events are discrete, infrequent and objectively resolvable: a ball in or out, a batter out or not out. Rugby's reviewable events are none of those things. They are continuous, they arrive in clusters, and a large share of them turn on judgement that a replay cannot settle. Handing a captain two challenges in a match where forty arguable decisions occur would not distribute correction fairly. It would reward whoever guessed best about when to spend.

There is also a tempo problem. A challenge is only useful if a team can stop play to use it, and rugby already struggles with the amount of time the ball spends out of play. Adding a team-controlled stoppage to a sport whose officials are simultaneously being told to review less would move in two directions at once.

The consequence is that rugby's captains have no formal mechanism at all. They may ask a referee a question, and a good referee will answer it, but the request has no standing in the protocol and generates no obligation. Whether the sport reviews an incident remains entirely a decision for the officiating team, which is exactly how World Rugby has chosen to keep it.

Reading a TMO review while it is happening

Six things tell you where a review is going before the referee announces it.

Listen for whether an on-field decision was given. If the referee has said try or no try before going upstairs, the footage now has to overturn a stated position, and the original call survives far more often than not. If no on-field decision was given, the review is genuinely open.

Listen to how the question is framed. A narrow question about one thing will be answered quickly. A request to look at the whole passage is a sign the officiating team is unsure what it is looking for, and those are the reviews that run long and end badly.

Count the phases yourself. If a possible offence happened three rucks before the score, it is outside the window for offside and obstruction, and no amount of replay will bring it back. Knowing this saves a great deal of shouting.

Watch for the crossed arms. That signal means the incident has left the live process entirely and is being graded elsewhere. The game will restart, the player is already gone for ten minutes, and the answer will arrive later.

Notice when nothing is announced. Silent background checks are part of the design. A stoppage that does not happen is the protocol working, not the protocol being skipped.

Check which regulations this competition is running. Card law and review scope differ between tournaments and change between seasons, and a decision that looks wrong is often a decision made correctly under a different rulebook to the one you were watching last month.

The protocol will be amended again. That is not a defect in it: a review system is a set of trade-offs between accuracy, time and authority, and the right balance moves as cameras improve and as the sport decides what it is willing to stop for. What stays constant is the underlying architecture, which is worth carrying into every match you watch. One person decides. The evidence has to be obvious. The search has a boundary. Everything else is detail.

More on the laws, the officiating and the arguments they generate is collected in the rugby archive, and the wider set of explainers across every sport sits in the blog index.