Bad light cricket rules: how umpires decide to stop play

Two umpires stand a few paces apart in the middle, one of them holding a small grey instrument up at chest height while the batters lean on their bats and a full house makes its feelings known. Nobody in the ground can see the number that comes back. The bad light cricket rules governing what happens in the next thirty seconds are among the least understood in the sport, largely because the decision looks like a verdict on whether it has got dark, and it is not that at all. It is a verdict on safety and good sense, taken by two officials whom the Laws require to reach it themselves.

The short answer, before any of the detail: play stops when the umpires together consider that conditions are either dangerous or unreasonable, those two words carry precise and separate meanings, the meter informs that judgement without settling it, and no captain or batter has a vote. Everything else that follows is machinery built around those four facts.

What the law calls dangerous, and what it calls unreasonable

Law 2.7 of the MCC code is the foundation, and the identical wording sits at clause 2.7 of the ICC playing conditions used for international cricket. It opens by handing the whole question to the officials: it is solely for the umpires together to decide whether ground, weather, light or exceptional circumstances make it dangerous or unreasonable for play to take place.

The next line does more work than any other sentence in the area. Conditions are not to be treated as dangerous or unreasonable merely because they are not ideal. That is the answer to almost every complaint made from a commentary box. A gloomy afternoon is not a legal ground for stopping. Nor is a batting side's discomfort, nor a fielding captain's preference, nor the fact that everybody would rather be indoors.

Two definitions then split the concept in half. Conditions are dangerous if there is actual and foreseeable risk to the safety of any player or umpire. Conditions are unreasonable if, although posing no risk to safety, it would not be sensible for play to proceed. Notice that these are alternatives rather than a sequence. A match can be stopped on either limb, and light tends to cross the second before it crosses the first.

The distinction has a practical shape on the field. A fast bowler operating in murk creates a risk to the batter's safety, which is the dangerous limb. A slow left-armer bowling to a batter who can see the ball perfectly well but where the fielders at slip and short leg are guessing creates something closer to the unreasonable limb: nobody is likely to be hurt, and the cricket has stopped being a genuine contest. Umpires weigh both, and the identity of the bowler in the attack is part of what they weigh.

That is also why a side can find the light suddenly acceptable the moment its captain throws the ball to a spinner. It is not a loophole. It is the safety half of the test easing while the sense half holds.

50%50%

• Dangerous: actual and foreseeable risk to a player or umpire50
• Unreasonable: no safety risk, but play would not be sensible50

The proportions are a reading aid, not a measurement. Both limbs stand on their own and either is sufficient to suspend play.

Show the numbers

The two limbs of the fitness-for-play test

Item	Value
Dangerous: actual and foreseeable risk to a player or umpire	50
Unreasonable: no safety risk, but play would not be sensible	50

The light meter measures the light, it does not make the decision

The instrument the umpires reach for is a hand-held light meter, supplied by the ICC for international matches and calibrated to a common standard so that a reading taken at one ground means the same as a reading taken at another. That standardisation is the point of central supply. Without it, the numbers would not be comparable between venues, and comparability is the entire function of the device.

The playing conditions are careful about its status. Umpires are entitled to use meter readings as a guideline for determining whether the light is fit for play, applied against the dangerous-or-unreasonable test rather than instead of it. The permitted uses are then listed in two lines: to determine whether there has been a deterioration or an improvement in the light at any stage, and to serve as benchmarks for the remainder of the match.

Read those two lines closely and the popular idea of the meter falls apart. There is no published figure below which play must cease. The regulation does not contain one, and no such figure is announced to broadcasters, because the meter is not a threshold instrument. It is a comparison instrument.

What the umpires actually get from it is a defence against their own eyes. Human vision adapts to falling light so gradually that two officials standing in the middle for ninety minutes are poor judges of how much has been lost. The meter does not adapt. It tells them whether the light in front of them now is better or worse than the light of twenty minutes ago, and that is a question they genuinely cannot answer unaided.

There is a related instruction that catches out spectators watching a late-afternoon session. Umpires are told to disregard any shadow cast on the pitch by the stadium or by any permanent object on the ground. A pitch half in bright sun and half in deep stand shadow can look unplayable on television. In law, that stripe is not part of the assessment.

How one reading becomes the benchmark for the rest of the match

The benchmark mechanism is the part most worth understanding, because it explains behaviour that otherwise looks capricious.

Suppose the umpires decide, on the evening of the second day, that conditions have crossed the line. Before they take the players off, they take a reading. That figure now attaches to the match. It becomes the reference point for every subsequent light decision in the same game, at that ground, under that sky.

Two consequences follow, and they run in opposite directions.

The first is consistency, which is what the rule is for. Having gone off at a certain measured level on Friday evening, the umpires cannot permit play at a demonstrably worse level on Saturday evening simply because the game has become interesting or the crowd has become loud. The benchmark ties their hands in the direction of favouritism, which is exactly where officials most need their hands tied.

The second consequence is a ratchet, and it is the source of the strongest criticism of the system. A reading taken early in a match, in circumstances that may have included a genuinely awkward set of conditions, sets a standard for four more days. If that first stoppage was cautious, every later stoppage inherits the caution. The players cannot appeal it, the referee cannot revise it, and the match carries it to the end. Umpires know this, which is one reason the first suspension in a Test tends to be taken later than the spectators expect rather than earlier.

Once the players are off, monitoring becomes the umpires' own responsibility. They are required to inspect as often as is appropriate and to do so unaccompanied by any player or official, which removes the possibility of a captain walking out alongside them and making a case. The moment both umpires agree conditions are no longer dangerous or unreasonable, they must call the players back. There is no discretion to wait for a convenient moment.

1. 1The umpires begin comparing rather than lookingFalling light is invisible to adapted eyes, so the officials stop trusting their own impression and start taking readings to establish whether the light is genuinely worse than it was.
2. 2A judgement is made against a two-part testThe question is not how dark it is. It is whether there is foreseeable risk to somebody's safety, or whether play has stopped making sense even though nobody is likely to be hurt.
3. 3The ground authority is instructed to switch on the lightsWhere artificial lighting exists and the natural light is deteriorating towards an unfit level, the umpires are required to have it used so that play can carry on.
4. 4Play is suspended and a reading is recordedUnder the international playing conditions the two umpires must agree, having consulted the match referee. The recorded figure becomes the benchmark for the rest of the match.
5. 5The umpires inspect alone, at intervals they chooseNo player or team official accompanies them. The decision to resume is theirs and needs both of them, and it must be taken as soon as conditions allow rather than at a tidy moment.
6. 6The overs are reconciled against the clockTime lost is made up where the regulations allow it, the day's minimum over target is reduced where the loss is large enough, and anything still unbowled at the end of permitted overtime is written off.

The sequence set out in the ICC playing conditions for a Test match. Each step is a separate regulation rather than a matter of custom.

Why batters are no longer offered the light

For most of the game's history the umpires did not decide. They offered.

The old procedure was that when conditions deteriorated past a certain point, the officials would put the choice to the two batters at the wicket. If the pair accepted the offer, the players came off. If they declined, the match continued. Fielding sides had no equivalent right, which was the first oddity, and the batting side's answer was allowed to depend on the state of the game, which was the second and much larger one.

The consequence was entirely predictable. Light became a tactic. A pair sixty runs into a comfortable partnership would wave the offer away in conditions they would have accepted at once had two wickets just fallen. A side batting to save a match had every incentive to go off; a side batting to win one had every incentive to stay on; and the identical sky produced opposite answers depending on the scoreboard. Whatever else that was, it was not a safety regime.

The Laws were amended so that the offer disappeared and the responsibility sat entirely with the officials. The change took effect from October 2010, and it is the reason no modern batter is asked anything. The umpires confer, one of them signals, and the players walk.

The reform removed the tactical distortion and created a different complaint in its place, which is the one heard today: that a decision nobody in the middle wants can now be imposed on both sides at once, including on a batting pair who would happily continue and a fielding captain who would happily bowl. That is the trade the game accepted. A rule that cannot be gamed is a rule that cannot be waived either.

There is one more asymmetry inside the current wording that rewards attention. Under the MCC Laws, play must be suspended if either umpire considers conditions dangerous or unreasonable. Under the ICC playing conditions for internationals, the pair must agree together, and consult the match referee. Both codes then require both umpires to agree before play resumes. Stopping is easier than starting in every version of the rule, and deliberately so.

What the umpires are actually protecting against

The safety limb is not an abstraction, and understanding what it guards against explains why the officials are stubborn about it.

A batter facing genuine pace picks the ball up from the bowler's hand and then loses it for a fraction of its flight before reacquiring it off the pitch. That reacquisition depends on contrast between a dark red object and whatever sits behind it. As light falls, contrast falls faster than brightness does, and the reacquisition arrives later. A delay measured in hundredths of a second is enough to turn a stroke into a top edge and a bouncer into an impact.

The people at risk are not only the batters. Close catchers stand within a few metres of the bat with no time to react to a full-blooded shot, and their protection is seeing it early. The helmet and grille that a short-leg fielder wears assume he can see the ball at all. Cricket has spent decades improving the equipment that reduces the consequences of being struck, and the history behind that equipment is set out in the account of how cricket helmets developed and why. None of it substitutes for visibility.

That is why the sight of a spinner immediately changing the umpires' calculus is coherent rather than absurd. Remove the high-speed delivery and the primary safety mechanism in the argument is removed with it. What remains is the sense limb, and a match can carry on for some time on that alone. The rules on what a bouncer is and how many are permitted sit on the same underlying concern about a batter's ability to see and evade.

Floodlights: required, expected, and not always sufficient

Modern Test grounds have floodlights, and the playing conditions are not shy about them. If in the opinion of the umpires natural light is deteriorating to an unfit level, they are to authorise the ground authority to use the available artificial lighting so that the match can commence or continue in acceptable conditions. That is an instruction, not a permission, and it is why lights come on in Tests during afternoon sessions that look perfectly bright.

Three qualifications sit around it, and together they explain why the lights so often fail to save the day.

The first is availability. Home boards may, before a series begins, seek ICC approval to amend the clause so that artificial lights will not be used at specified venues. Not every Test ground has lighting to the standard required, and some that do have it under restrictions of their own. A venue can be operating, quite legitimately, without the option.

The second is failure. If the power goes or the lights malfunction, the provisions that govern delay and interruption for bad weather or bad light apply as if the lights had never existed. The regulation treats the loss of artificial light as identical in kind to the loss of the natural sort.

The third is the important one and has nothing to do with the lights themselves.

Why a red ball and artificial light do not combine well

Floodlights do not reproduce daylight. They deliver a different spectrum, from fixed positions, at an intensity that falls off across the playing area, and the object they are illuminating is a dark red ball that has been scuffed and dirtied for sixty overs.

Under a fading sky the red ball loses contrast against a dark background in a way that a white or pink ball does not. The lights raise the absolute amount of illumination without restoring the difference between the ball and everything behind it, and difference is what a batter's eye is actually tracking. This is precisely why day-night Test cricket uses a pink ball rather than simply putting the traditional one under lights, a design problem examined at length in the piece on why Tests under lights need a pink ball.

So the umpires can find themselves in an odd position that looks to a viewer like obstinacy. The lights are on, the ground is bright, the television picture is perfectly clear, and the officials still take the players off. Broadcast cameras have their own sensitivity and their own gain, and what a television screen shows is not evidence about what a batter can pick up at the crease. The umpires are judging contrast at ninety miles an hour, and it has gone even though the brightness has not.

The pink ball trial and where the ICC is heading

The governing body has stopped treating this as settled. At its board meeting in Ahmedabad on 1 June 2026 the ICC approved a trial that allows a Test match to switch from the red ball to a pink one when bad light threatens to halt play, so that the game can continue under floodlights. It is a trial rather than a rewritten law, and both participating teams have to agree to it in advance of the series.

The logic is direct. If the obstacle to playing under lights is the ball rather than the lighting, change the ball. The pink ball exists, it has been used in day-night Tests for years, and its visibility under artificial light is the reason it was developed.

The same board decision funded research into lighting technology for venues and match officials, carried out with the MCC, aimed at reducing the play lost to poor light. That is a tacit admission that the current arrangement is a compromise rather than a solution, and that both halves of the problem, the ball and the lights, are open to engineering.

Neither change alters the underlying law. Clause 2.7 and its two-limbed test survive intact. What the trial does is give the umpires a further option to try before they reach for the one that empties the ground.

• 90Minimum over target on a normal day
• 75Minimum overs on the last day, before the last hour
• 15Minimum over rate per hour
• 30Permitted overtime beyond the cessation time, in minutes

Structural figures taken from the ICC World Test Championship playing conditions. They describe the schedule, not the light.

What happens to the overs when the players go off

Losing light costs a Test more than atmosphere. It costs overs, and the regulations are specific about how many.

On days other than the last, play continues until the minimum target of ninety overs is completed or the scheduled cessation time arrives, whichever comes later, subject to a hard stop: play may not continue more than thirty minutes beyond the scheduled or rescheduled cessation time. That half hour is the permitted overtime. If the ninety overs have not been bowled when it expires, the over in progress is completed and the day ends. Those unbowled overs are gone. The regulation says plainly that they are not made up on any subsequent day.

Where an interruption is long, the day's target itself is reduced. If play is suspended for more than an hour in aggregate on a day, for weather, light or any other reason apart from the normal intervals, the minimum number of overs is cut by one over for each full four minutes of aggregate playing time lost. That four-minute figure is not arbitrary: it is the reciprocal of the fifteen-overs-an-hour rate the same regulations require, so the deduction is calculated at exactly the rate the players were expected to bowl at. The way that expected rate is policed in its own right is covered in the explanation of over-rate penalties and how they are applied.

20 minutes5overs

40 minutes10overs

60 minutes15overs

80 minutes20overs

100 minutes25overs

Arithmetic derived from the four-minute rate in the playing conditions, which applies once the aggregate loss on a day exceeds one hour. Illustrative of the rule, not a record of any match.

Show the numbers

Overs deducted from the day's minimum, by playing time lost

Item	Overs deducted
20 minutes	5overs
40 minutes	10overs
60 minutes	15overs
80 minutes	20overs
100 minutes	25overs

Extending the day, and pushing time into the next one

Against those deductions the regulations set two recovery mechanisms.

The first operates on the day itself. Where play is suspended for any reason other than the normal intervals, the playing time that day is extended by the amount of time lost, up to a maximum of one hour. That is the reason a Test that lost the first session can still finish close to a full quota, and the reason the players are sometimes still out there well after the advertised close.

The second operates across the match. If time is lost and cannot be recovered on the day, up to thirty minutes a day may be added to the scheduled hours on the following day, and on subsequent days as required, until as much of the loss as possible has been recovered. A match that loses two hours on the opening morning does not simply lose them; it claws them back in half-hour instalments through the rest of the week, which is why start times drift earlier as a rain-hit Test progresses.

There is also a specific provision for the ends of days. The umpires may decide to play thirty minutes of extra time, a minimum of eight overs, at the end of any day other than the last, if a captain asks for it and the officials believe it would produce a result. If they do not believe a result is achievable, no extra time is allowed. Once the decision is taken the whole period is played out, even if the possibility of finishing disappears halfway through it.

All of this recovery machinery depends on one thing that bad light removes. Time can only be added at the end of a day if there is usable light at the end of that day. Rain that clears at four o'clock can be recovered. Light that fails at four o'clock cannot, because the extension runs into a sky that is getting darker rather than clearer. That asymmetry is the reason bad light irritates captains more than rain does, and it also complicates the timing of a declaration and how a captain judges it on an overcast final afternoon.

The last hour, when the overs stop following the clock

The final hour of a Test is governed separately, and it is the passage in which bad light does the most damage to a contest.

Once one hour of scheduled playing time remains, the umpire at the bowler's end signals the start of a minimum of fifteen overs, and that period is called the last hour whatever its actual duration turns out to be. Uninterrupted, play continues until the end of the over in progress when sixty minutes have been played, or until fifteen overs have been bowled, whichever comes later. The overs, not the clock, become the binding constraint.

Interruptions during the last hour reduce that fifteen, at the same rate of one over for every complete four minutes lost. There is a detail here that differs from the rest of the day: where there is more than one interruption, the minutes are not aggregated, and each interruption is calculated separately. The rounding therefore works in the batting side's favour when the breaks come in several short bursts rather than one long one, since each is truncated to complete four-minute blocks on its own.

None of this can be recovered afterwards, because no time is made up in respect of any interruption that begins after the last hour has been signalled. A side needing thirty runs with an hour left and a side needing thirty runs with fifteen overs left are in different positions, and a light stoppage at that moment can move a match from the first situation into a much narrower version of the second.

Why the argument about lost playing time keeps recurring

The complaint is a standing feature of Test cricket in northern latitudes, and it comes in several forms that deserve separating.

The first is that the standard has become too cautious. Batters and captains regularly say they were happy to continue, and the ratchet built into the benchmark rule means one careful decision propagates through a match. The counter is straightforward. A safety judgement that can be relaxed at the request of the people at risk is not much of a safety judgement, and the players asking to stay on are asking on behalf of a batting pair whose interests may not match the fielding side's.

The second is that the sport is unusual in stopping at all. Football continues under lights; so does baseball. The reply is that both are played with a larger, brighter object that no one is required to face at ninety miles an hour from twenty-two yards without an appeal to a governing authority. The comparison flatters cricket's obstinacy and understates the specific problem.

The third complaint is the strongest, and it is about scheduling rather than safety. A Test played at the end of a northern summer, in a five-day window, in a country where the light fails early, will lose overs. The playing conditions can redistribute that loss but they cannot manufacture daylight. Every mechanism described above is a way of moving lost time around inside a fixed envelope. Widening the envelope is a fixture question rather than an umpiring one, and the way those windows are chosen is set out in the piece on how the international cricket calendar is put together.

The fourth is the one the ICC has now acted on: that the game had accepted a technological constraint without testing it. The pink ball trial and the lighting research are both responses to the argument that the light itself has not been the limiting factor for some years, and the ball has.

What to watch for when the covers are off but the players are not

A few practical markers will let you predict the next twenty minutes better than the commentary usually does.

Watch which bowler has the ball when the umpires start conferring. A change to spin at the start of an over, in poor light, is a fielding captain making an argument in the only language the officials have to listen to. It often works, and it is not sharp practice.

Watch for the meter coming out rather than for the umpires talking. Conversations in the middle happen constantly. A reading is a formal act, and where it is taken as the players leave, it fixes the standard for the rest of the game.

Watch the scoreboard's over count against the clock during the final session. If the day's minimum has been reduced, the players will leave earlier than the advertised time, and the reduction will have been calculated in four-minute blocks against the aggregate time already lost. Following that count is much easier if the abbreviations are familiar, and the guide to reading a cricket scorecard sets out what each column is doing.

Watch for both umpires walking out together after a stoppage, unaccompanied. That inspection is the mechanism by which the game restarts, and the requirement that they go alone is what stops it becoming a negotiation.

And treat the television picture as the least reliable evidence in the ground. A camera with its gain wound up will show a bright, sharp image of an evening in which a batter genuinely cannot pick up the seam. What the umpires are protecting is not the view from the stand or the broadcast feed; it is the half-second in which a person standing twenty-two yards away has to see, judge and move.

For more on how the officials are appointed, assessed and supported, see the pieces gathered under cricket, including the profile of the ICC elite panel of umpires, and the wider archive of explainers covering the rest of the sport's rules.