Athletics false start rules, and the 0.100 second line

An athletics false start is not only what it sounds like. Leaving the blocks before the gun is a false start, and everybody agrees it should be. But the athletics false start rules also disqualify a sprinter who left after the gun, provided they left too soon after it, and that second category is where every argument in the sport lives.

The threshold is 0.100 seconds. React inside a tenth of a second of the report and the rule treats you as having guessed rather than heard, and the disqualification follows without any human being deciding you did anything wrong.

That number is not a measurement. It is a hypothesis about human physiology that was written into a rulebook and then given the force of an automatic sanction, and once you see it that way the rest of the sport's difficulties with the start make sense.

• 0.1Reaction time threshold, in seconds
• 0False starts allowed in an individual race
• 1False starts allowed per athlete in combined events
• 1Maximum permitted timing delay, in milliseconds

The rule is a presumption, not an observation

Nothing in the starting blocks can tell whether a sprinter heard the gun. There is no sensor for intent, and there never will be.

So the rule does what law does when it cannot see inside somebody's head: it picks an observable proxy and attaches an irrebuttable presumption to it. The proxy is the interval between the gun signal and the moment force rises in the blocks. The presumption is that any interval under a tenth of a second cannot have been a response, and must therefore have been a movement already underway when the gun happened to fire.

That presumption is doing enormous work. A sprinter with a recorded 0.099 has not been found to have moved early. They have been found to have moved at a speed the rulebook says is impossible to achieve honestly, which is a completely different finding, and one no official in the stadium has to endorse or explain.

The chain being modelled runs: sound reaches the ear, the auditory system registers it, the signal travels to the motor cortex, a command travels back down to the legs, muscle fibres begin to contract, and force appears against the footplate. Each link takes time. Add them up and you get a floor below which a genuine response cannot fall.

The trouble is that the floor is a distribution rather than a wall, that different laboratories put it in different places, and that the equipment measuring the last link in the chain does not measure it the same way everywhere.

The blocks measure force, and that is not the same as measuring movement

A start information system does not watch the athlete. It measures what the athlete is doing to the block.

Each footplate carries a sensor, usually a force transducer or an accelerometer mounted behind the plate, sampling continuously. The system is triggered by the starter's signal and the rules require that trigger to be effectively instantaneous, with the delay between the report and the start of timing held to a thousandth of a second or less. From that instant the system watches the force trace, and reaction time is recorded as the interval until the force crosses a threshold above the athlete's baseline.

Three consequences follow, and none of them is obvious from the outside.

A sprinter can move without being detected. Lifting the head, rocking the shoulders, shifting weight forward: if it does not change the force on the footplates past the threshold, the system records nothing. Those movements are caught by the starter's eyes or not at all.

A sprinter can be detected without moving. Pressure builds against the plates before the body travels anywhere. The force trace rises during the isometric phase, well before there is anything to see. This is precisely why the electronic system exists, and it is also why a disqualification can look, from the stand, like nothing happened.

The threshold is a setting. Where you place the force level that counts as the start of the push determines the reaction time you record. Set it low and you catch the response earlier and report a faster time. Set it high and you report a slower one. Different approved systems make that choice differently, and the number that appears beside a sprinter's name is a product of the equipment as much as of the athlete.

That last point is the quiet scandal of the whole business. The rule is enforced to the millisecond against a measurement whose precision is real and whose accuracy depends on a calibration decision nobody in the stadium can see.

1. 1The starter firesAn electronic pistol sends a signal simultaneously to the timing system, the loudspeakers and the start information system. The delay before timing begins is limited by rule to one millisecond.
2. 2Sound reaches the athleteSpeakers behind each set of blocks deliver the report to every lane at the same instant, which is why the gun is no longer fired from beside lane one alone.
3. 3The auditory system registers itThe signal is transduced in the ear and travels to the brain. This link is fast and reasonably consistent between people.
4. 4A motor command is issued and travels to the legsNerve conduction down the spinal cord and out to the muscles of the hip and thigh. This is the largest and most variable part of the interval.
5. 5Muscle fibres begin to develop tensionForce starts to rise against the footplate before any part of the body has visibly travelled anywhere.
6. 6Force crosses the system's thresholdThis is the only event actually measured. The interval from the gun to this crossing is the reaction time that appears on the screen.
7. 7The reading is compared with 0.100 secondsBelow the threshold, the system flags a false start and the athlete is disqualified. Above it, the same push is a legal start and a very good one.

Only the final step is measured. Everything above it is inferred from the timing of that measurement.

Everybody has to hear the gun at the same moment, and once they did not

For most of the sport's history the starter stood beside the track and fired a pistol. Sound travels at roughly 343 metres per second in ordinary conditions, which is fast enough to seem instantaneous and slow enough to matter at the scale the rule now operates on.

Here is the arithmetic, worked from the rule-defined lane width of 1.22 metres and the speed of sound. It is a calculation, not a measurement of any real race.

A sprinter in lane eight is seven lane widths from a sprinter in lane one, which is 8.54 metres. At 343 metres per second the report takes about 0.025 seconds to cover that distance. Twenty-five milliseconds is a quarter of the entire margin the rule allows between a legal reaction and a disqualification.

Lane 23.6ms

Lane 410.7ms

Lane 617.8ms

Lane 824.9ms

A worked calculation, not a measurement. Lane width is 1.22m by rule and the speed of sound is taken as 343 metres per second in still air at about twenty degrees. The figures show why loudspeakers are now placed behind every set of blocks.

Show the numbers

Extra time for the gun's report to reach each lane, if fired beside lane one

Item	Value
Lane 2	3.6ms
Lane 4	10.7ms
Lane 6	17.8ms
Lane 8	24.9ms

The fix is a speaker behind each set of blocks, fed from the same electronic signal that starts the clock, so the report arrives everywhere at once. It is a good rule and an unglamorous one, and it exists because the sport tightened the tolerance on the start until the physics of sound in air became a competitive disadvantage for the outside lanes.

Worth holding on to as you read the rest: the sport was prepared to redesign its equipment to remove a twenty-five millisecond inequity, and it has not moved the hundred-millisecond line that the same equipment is used to enforce.

Where the published measurements actually disagree

The threshold is usually defended as settled science. It is not.

The literature splits, and it splits in both directions, which is why the argument has never been resolved by simply reading more papers.

On one side, Lipps, Galecki and Ashton-Miller analysed reaction times recorded at the Beijing Olympics and argued that a legal response faster than about 120 milliseconds is very unlikely to be detected by approved systems. Their conclusion is that the threshold is in the wrong place because it is too generous, and that the sport is permitting some genuine anticipation to pass as a legal start. The same paper reported that the fixed limit did not sit equivalently for men and women, with their analysis suggesting a materially lower limit, on the order of a fifth lower, would be needed to treat female sprinters the way the current one treats male sprinters.

On the other side, Pain and Hibbs used instrumented blocks with their own detection algorithm and reported athletes producing valid responses faster than 100 milliseconds. If that is right, the current rule is disqualifying people for being quick.

Both cannot be correct as stated, and the reason they can both be published is the measurement problem above. Each study defines the onset of the push differently. Change the definition of onset and you change every number in the dataset without anybody's leg moving any faster.

Later work has pushed the same point harder, reporting that small differences in start procedure and system configuration can shift a measured reaction time by tens of milliseconds. That is not a rounding error. It is larger than the gap between a legal start and a disqualification, which is precisely the gap the rule is being asked to adjudicate.

The most famous casualty of that gap remains Devon Allen, disqualified from the 110 metres hurdles final at the 2022 World Championships on a reading of 0.099. One thousandth of a second, on a measurement whose own literature cannot agree to within twenty thousandths.

None of this means the threshold is indefensible. A sport needs a line, lines have to sit somewhere, and a bright line that occasionally catches an innocent is a design most rulebooks accept. What it does mean is that anyone defending the number as physiological fact is defending something the physiologists have not agreed on.

Zero tolerance, and the two rules it replaced

The current rule is the third arrangement in about two decades, and each version failed for a different reason.

Before 2003: one per athlete. A sprinter who false started was charged with it and raced on. A second false start by the same sprinter meant disqualification. This is the version most people's instincts still run on, because it is how a normal person would design the rule: a warning, then a consequence, applied to the person who did the thing.

It broke down because it was gameable. In a field of eight, a race could absorb eight false starts before anybody was disqualified, and a nervous or tactical field could stretch a 100 metres final across several minutes. A sprinter who wanted to disrupt a rival's rhythm had one free attempt to do it.

From 2003: one per field. The first false start in a race warned everybody. The next one, by anybody, disqualified the athlete who committed it, whether or not they had anything to do with the first.

This closed the arithmetic loophole and opened a fairness one. A sprinter who had done nothing wrong could be disqualified for a first personal offence because somebody else in the field had already used the field's allowance. It was widely disliked for the obvious reason, and it produced the specific injustice of an athlete being punished for another athlete's nerves.

From 2010: none. One false start, disqualification, no warning, no allowance, applied to the athlete responsible.

The change was agreed in 2009 and took effect at the start of 2010. It is clean, it is easy to apply, and it removed both of the previous versions' defects at once. What it did not do is address the objection that the underlying detection threshold might be wrong, because that was never the problem it was designed to solve.

The real argument for zero tolerance is the schedule, and it should be made honestly

The fairness case for the current rule is genuine as far as it goes. Every athlete faces the same rule, nobody can absorb somebody else's mistake, and the sprinter who anticipates gains no advantage from having done so.

But the reason the rule moved when it did, and moved to zero rather than back to one per athlete, is the clock on the broadcast.

A championship sprint final occupies a slot measured in a few minutes of live coverage. Under the old rules, a final could consume that slot before anybody ran, with the field standing up, walking back, resetting, and the presenter filling. Televised athletics is a sequence of short events inside a long window, and the window does not move. Every false start pushes a field event's coverage out, delays the next race, and eats into a broadcast that has been sold on the basis of a running order.

This is not a cynical reading. It is a legitimate consideration for a sport that depends on being watchable, and the same pressure has reshaped presentation across the sport. What is objectionable is not that the schedule mattered. It is that the rule is defended almost exclusively on fairness grounds, when a rule made largely for the schedule should be argued for on those terms and weighed accordingly.

The cost of the design falls in one place, and it falls hard. A sprinter whose season, and in some cases whose career, has been built toward a single final can be removed from it before the race begins, in front of a stadium, by a machine reading, with no opportunity to demonstrate anything. There is no other event in athletics where a competitor is eliminated by an official measurement before performing. A wind reading can invalidate a performance after the fact, and a missed qualifying mark can keep you out of a championship entirely, but only the start removes you from a race you have already qualified for and travelled to.

The commands, the unsteady start and the recall

The procedure itself is more forgiving than the disqualification rule suggests, and this is the part casual viewers never see.

For races up to and including 400 metres, the starter calls "on your marks", waits, calls "set", waits, and fires. For 800 metres and longer there is no "set" command: the athletes go to their marks and the gun follows.

Two discretionary pauses are built into that sequence, and the starter controls both.

After "on your marks", athletes take up position and become still. The starter is not permitted to proceed until every athlete is settled, and there is no maximum time for this. After "set", athletes assume their final position, and again the starter holds until the field is steady. There is no prescribed hold time and no rhythm the athletes may rely on. A starter who used the same interval every time would be handing the field a metronome, so varying it is not merely permitted, it is the job.

If the field does not settle, or if an athlete moves in a way the starter judges to be accidental rather than an attempt to start, the start is declared unsteady. The athletes are stood up, the sequence begins again, and no warning is recorded against anybody.

This is the release valve, and it matters more than its obscurity suggests. A twitch after "set" is not automatically a false start. The starter has to form a view about whether it was a start attempt or a wobble, and where it was a wobble the correct response is to stand the field up rather than to disqualify. The zero-tolerance rule sits on top of a procedure that still contains a human judgement about what happened.

A recall is the related mechanism after the gun. If the start was unsatisfactory for any reason, a recall gun stops the race and the field returns. Where the cause was an athlete, that athlete is disqualified. Where the cause was not an athlete, and this includes equipment failures, external noise and a starter's own error, no warning is recorded and the field simply restarts.

The starter is the last human left in an automated procedure

It is tempting to read the modern start as a machine that disqualifies people while an official stands nearby holding a pistol. That is wrong, and the remaining discretion is larger than the rulebook makes it look.

The starter decides when the field is settled, and that decision has no appeal and no criteria beyond the starter's own judgement. A field that is not settled cannot legally be started, so a starter who holds and holds is not being awkward, and a starter who fires into a restless field has made an error the athletes will pay for.

The starter decides whether a movement after "set" was a start attempt or an accident, which determines whether the next thing that happens is a disqualification or a stand-up. The electronic system offers no view on this at all, because the system only speaks after the gun.

The starter can abort before firing for any reason, including a noise in the stand, a photographer moving, a camera crane, or a competitor in another event crossing behind the blocks. Nothing is recorded and nobody is warned.

And the starter is not alone. A championship start is officiated by a team: the starter, and a set of recallers or start judges positioned to see the field from angles the starter cannot cover. Each carries a recall gun. Any of them can stop the race, and their brief is to catch what the equipment cannot, which is movement that does not register as force.

That is the part most people miss. An athlete can be disqualified for a false start with no electronic evidence at all. A hand lifting off the line before the gun, a knee coming up, a shoulder rolling forward: none of these necessarily change the load on the footplates, and all of them are false starts if the officials judge them to be start attempts. The start information system is a supplement to human observation, not a replacement for it, and the technical rules have never said otherwise.

The reverse also holds. Where the system flags a reading below the threshold and the officials saw nothing, the reading governs, because the whole point of the threshold is to catch what the eye cannot.

1. 1"On your marks"Athletes move into the blocks and settle. The starter may not proceed until the field is still, and there is no maximum time for this.
2. 2The starter waits, or stands the field upIf the field does not settle, or an athlete is delaying, the athletes are stood up. Nothing is recorded and the sequence begins again.
3. 3"Set"Athletes take their final position. The starter holds for an interval of their own choosing, deliberately varied so that no rhythm can be learned.
4. 4An accidental movement, or a start attemptThis is the starter's judgement and the equipment has no view on it. An accidental movement makes the start unsteady, the field is stood up, and no warning follows.
5. 5The gun firesThe signal reaches the timing system, the speakers behind every set of blocks and the start information system at the same instant.
6. 6Recallers and the start information system both reportOfficials watch for movement the sensors cannot see. The system reports any reaction time under 0.100 seconds. Either route can stop the race.
7. 7The recall gun sounds, or the race continuesA recall returns the field to the blocks. Whether anybody is charged depends on what caused it.
8. 8The card goes upGreen if no athlete was at fault. Yellow for a first offence in combined events. Red and black, divided diagonally, for the disqualification.

Only races up to 400m use the "set" command. The two decision points that belong entirely to the starter are marked by what does not happen: a start that is aborted or stood up produces no record against anybody.

Above 400 metres there are no blocks, and the rule changes shape

Races longer than 400 metres use a standing start, no blocks and no "set" command. The starter calls the athletes to their marks and fires when the field is still.

Everything electronic drops out at that point. There is no footplate, no force trace and no reaction time, so a false start in a middle-distance race is whatever the officials say it is, judged entirely by eye against a field that is standing upright and moving slightly by definition.

In practice the standard applied is different in substance as well as in method. A sprinter is in a held position that they are trying to explode out of, so the line between settled and going is razor thin. A runner standing at the line is not straining against anything, and a false start there means a plain, visible break for the track before the gun. It happens, mostly in tactical championship finals where the field is packed and jostling for position on the first bend, and it is rare enough that most viewers have never seen one called.

The rule that applies is still zero tolerance. The evidence available to enforce it is a hundred years older than the rule.

Green, yellow, red: reading the cards

The cards are the sport telling the stadium what has been decided, and they are worth knowing because they are usually the first honest information available.

A green card, shown to the whole field, means the aborted or recalled start was nobody's fault. No warnings, no records, restart. If you see green, an athlete has been cleared rather than let off.

A yellow card is a warning, and in the context of the start it belongs almost entirely to combined events, where it marks an athlete's first false start.

A red and black card, divided diagonally, is the disqualification, raised against the athlete responsible.

The card also tells you what the officials believe happened. Green after a stopped start means the mechanism failed or the environment intervened. No card and a stand-up means the starter judged the field unsteady, which is a judgement about the athletes without being a charge against them.

Combined events kept the older rule, and the reason is instructive

In the heptathlon and decathlon, an athlete who false starts is warned, and only a second false start disqualifies. The old per-athlete rule survives in exactly one corner of the sport.

The justification is proportionality. Disqualifying a decathlete from the 100 metres does not remove them from one event; it removes them from a competition spread over two days that they have prepared for as a single unit. A rule whose severity is tolerable when the cost is one race becomes grotesque when the cost is ten of them.

That reasoning is sound, and it is also an admission. The sport has accepted, in writing, that automatic disqualification is too harsh when the consequences are large enough. The obvious question is why the consequences of removing a sprinter from a world final are treated as small enough to bear, and the honest answer is that the sprint final is the event the schedule cannot absorb a delay in.

Combined events also run with a starter who knows the athletes are not specialists, which changes the practical texture of the whole procedure without changing a word of the rule.

What you can appeal, and what you cannot

An athlete who believes the call was wrong may ask the referee to be allowed to compete under protest. The referee decides on the spot, the race is run, and the result stands or falls once the protest is resolved.

The guidance to referees is restrictive, and deliberately so. Where a start information system that appears to be working properly has detected the false start, or where the offence was plainly visible, the referee should refuse the protest. In practice that closes off nearly every case, because those two categories cover nearly every disqualification.

What is left is narrow: a system malfunction, an external disturbance the starter should have accounted for, or a procedural error. An appeal against the referee's decision goes to the jury of appeal, and the jury is deciding whether the process was properly applied rather than whether the athlete is quick.

There is no route by which an athlete can argue that they genuinely reacted in 0.098 seconds. The rule does not contain that possibility, so no tribunal can find it. That is the difference between a threshold and a test.

How sprinters actually train the start under this rule

The rule has reshaped the technique it polices, and this is the most underrated part of the subject.

The first adaptation is that elite sprinters do not chase the fastest possible reaction. They chase the fastest reaction they can produce reliably, which is a different target, and they leave a margin above the threshold because the downside is unbounded. Losing two hundredths at the gun costs a fraction of a metre. Being disqualified costs everything. Any rational athlete sits above the line with room to spare, and the sport as a whole therefore starts slower than it could.

The second is that the response is trained as a response, not as a prediction. Athletes rehearse with randomised hold times so that no rhythm can be learned, because a rhythm is exactly what produces a sub-threshold reading. The starter varies the hold; the training has to mirror it.

The third is that block setup is treated as a reaction-time variable rather than only a power variable. The pressure a sprinter holds against the plates in the set position, the block spacing and the angle of the front plate all affect how quickly force rises after the command arrives, which is the quantity the system is actually measuring. Getting into a position from which force can build fast is part of the same job as getting into a position from which the first stride is powerful, and the mechanics of how a sprinter leaves the blocks is where the two goals meet and sometimes conflict.

The fourth is psychological and the least discussed. Under a one-per-athlete rule a sprinter could risk an aggressive start knowing a mistake was survivable. Under zero tolerance the same sprinter is managing a small, permanent probability of catastrophe on every start of their career. That changes what happens in the set position, and everyone who has stood in blocks under both rules says the same thing about it.

Relay teams live with an adjacent version of the problem, since the lead-off runner faces the identical rule while the rest of the team's risk sits in the changeover zones instead. One squad, two entirely different ways to be disqualified before the finish.

What to watch for at the start of a race

Watch the starter's hold, not the athletes. The interval between "set" and the gun is the only variable the starter controls and it is the one thing the field cannot prepare for. A long hold is a legitimate tool and it produces most of the twitches.

Watch for the stand-up with no card. That is an unsteady start, the mechanism that keeps honest jitters out of the disqualification column, and it happens more often than the disqualifications that get replayed.

Watch which card comes out. Green clears the field. Red and black removes somebody. The absence of a card after a stopped start is itself information.

Watch the reading, then remember what it is. The number beside a sprinter's name is the moment force crossed a threshold on a particular manufacturer's equipment. It is precise. Whether it is accurate in the way the rule assumes is the question the published research has been arguing about for years without settling.

The rest of the archive on how track and field is measured and governed sits in the athletics section, and the same tension appears everywhere in it: a sport that has instrumented itself to the millisecond still has to decide what the instruments mean. The rankings that determine who gets into a championship are built on the same instinct, which is that a published number is fairer than an official's opinion. Usually it is. At the start line, once a season or so, it very publicly is not.