The wind limit rounds against the sprinter, and the rule says how

The condition itself is familiar. For a performance made outdoors up to and including 200 m — except a 200 m on a 200 m standard oval track — wind velocity measured in the direction of running must be submitted, and if it averages more than 2 metres per second, the performance will not be ratified. The long jump and triple jump carry the same limit in the direction of jumping.

Reference The 2 metres per second limit for track events up to and including 200 m and for the horizontal jumps, and the exclusion of the 200 m on a 200 m standard oval track, from the record-ratification conditions in the Competition Rules.

What is less familiar is everything between the air and the number.

One point, one period, one direction

The gauge is not a field of sensors. It is a single instrument in a specified place.

For track events it is placed beside the straight, adjacent to lane 1, 50 m from the finish — 30 m for 50 m and 60 m races. It stands 1.22 m ± 0.05 m high and no more than 2 m from the track.

Reference The gauge’s position beside the straight adjacent to lane 1, the 50 m and 30 m distances, the 1.22 m ± 0.05 m height and the maximum 2 m from the track, from the Technical Rules on wind measurement.

And it runs for a fixed period, timed from the flash or smoke of the Starter’s gun. Ten seconds for the 100 m. Thirteen for the 100 m and 110 m hurdles. Five for the 50 m, 60 m and their hurdles. For 200 m races other than on a 200 m oval, the measurement is normally taken for ten seconds beginning when the first athlete enters the straight.

Reference The measuring periods for each event, timed from the flash or smoke of the Starter’s gun, and the 200 m provision beginning when the first athlete enters the straight, from the Technical Rules.

The numbers behind this chart

Row	Seconds of wind measurement, from the flash or smoke of the gun	Detail
50 m and 50 m hurdles	5	gauge 30 m from the finish
60 m and 60 m hurdles	5	gauge 30 m from the finish
100 m	10	gauge 50 m from the finish
100 m hurdles	13	gauge 50 m from the finish
110 m hurdles	13	gauge 50 m from the finish

Source World Athletics, Technical Rules, wind measurement periods.
Method Periods as tabulated in the rule. The 200 m is excluded from the chart because its period is defined by a different trigger — the first athlete entering the straight — rather than by the gun.

Three simplifications are built in before any arithmetic happens.

The wind is sampled at one place. An athlete in lane 8 is 8 or 9 metres from the instrument, has a different stand behind them and different turbulence around them, and is credited with a reading taken beside lane 1.

The wind is sampled over one interval. A ten-second average over the 100 m says nothing about whether the tailwind arrived during the drive phase or the last twenty metres, and those are not equivalent.

And only one component of the wind is measured: the velocity in the direction of running. A crosswind of any strength registers as nothing, which is why a mechanical gauge should have protection to reduce the impact of a crosswind, and why, where tubes are used, their length either side of the measuring device should be at least twice the tube’s diameter.

Reference The requirement to measure wind velocity in the running direction, the recommendation that mechanical gauges be protected against crosswind, and the tube-length guidance, from the Technical Rules.

None of this is a flaw. It is the necessary compromise of turning a moving fluid into one number that can appear beside a time. But it means the wind reading is a proxy, and the rule that follows treats it with more precision than a proxy can carry — deliberately.

The rounding rule

Here is the sentence that decides borderline records.

The gauge shall be read in metres per second, rounded to the next higher tenth of a metre per second, unless the second decimal is zero, in the positive direction. The rule gives its own worked examples: a reading of +2.03 m/s shall be recorded as +2.1, and a reading of −2.03 m/s shall be recorded as −2.0.

Reference The rounding rule, its “in the positive direction” qualification, the exception where the second decimal is zero, and both worked examples, from the Technical Rules.

Read those two examples together. The same three hundredths of a metre per second are rounded up when the wind helps and down in magnitude when it hinders. Rounding is not to the nearest tenth. It is always towards positive.

The consequence at the limit is exact. Any reading above +2.00 and below +2.10 rounds to +2.1, which exceeds 2 metres per second, and the performance is not ratified. So the effective limit is not “2.0 or less”; it is “at most 2.00 m/s to two decimal places”. A tailwind of +2.01 ends a world record.

And in the other direction the same rule reduces the recorded headwind. An athlete running into −2.03 is recorded as having run into −2.0, which is the smaller headwind. Where the number affects anything, it errs towards making the conditions look more favourable than they were.

That is a defensible choice, and its logic is conservative rather than punitive: when a measurement is uncertain, resolve the uncertainty against the record. A record should be beyond argument, so any doubt about whether the wind was legal is settled by assuming it was not. The rounding rule is that principle written as arithmetic.

The equipment the rule trusts

The rule is stricter about the instrument than about almost anything else in the measurement chain.

All wind-gauge equipment must be manufactured and calibrated according to international standards, and the accuracy of the equipment used in competition must have been verified by an appropriate organisation accredited by the national measurement authority. Non-mechanical gauges are required at the higher categories of competition and for any performance submitted for ratification as a world record.

Reference The requirement for manufacture and calibration to international standards, verification by an organisation accredited by the national measurement authority, and non-mechanical gauges at the top competition categories and for world records, from the Technical Rules.

The chain of trust ends at a national measurement authority, which is where legal metrology lives. It is the same institutional move a scale in a market makes: the reading is credible because a body outside the transaction certified the instrument.

The gauge may also be started and stopped automatically or remotely, with the information conveyed directly to the competition computer — and in practice for track events it is usually connected to the photo-finish system, so that a Wind Gauge Operator is often not required at all and their duties fall to the photo-finish team.

Reference That the gauge may be started and stopped automatically and/or remotely with data conveyed directly to the competition computer, and that in practice a Wind Gauge Operator is often not required because the gauge is connected to the photo-finish system, from the Competition Rules.

The limit itself

Nothing in the rules explains why the number is 2 and not 1.5 or 3, and it would be wrong to invent a reason. What the rules do establish is what kind of number it is.

It is not a threshold at which wind starts to matter — a tailwind of 1.9 m/s helps a sprinter, and is ratified without comment. It is a line drawn somewhere inside a continuous effect, chosen so that the great majority of performances are unaffected by it and only the clearly assisted ones are excluded.

That reading is supported by what the rule leaves alone. Altitude affects sprint times more than a legal tailwind does, and World Athletics ratifies records set at any altitude without adjustment. Wind is singled out not because it is the largest environmental variable but because it is the one that can be measured in real time, at the venue, with an instrument somebody can certify.

The rule, in the end, is about what can be witnessed. A tenth of a metre per second, rounded upwards, at one point beside lane one, for ten seconds — and everything else the air did that day goes unrecorded.