Wind assistance rules in athletics, and the +2.0 line

Every sprint result in athletics comes with a second number, printed small, immediately after the time. Wind assistance rules in athletics turn that small number into the difference between a performance that exists and one that formally does not. Above +2.0 metres per second, the time stays on the scoreboard, the winner keeps the medal, and the mark is struck out of the record book before anybody has finished celebrating.

The number is not a suggestion. It is a hard edge, applied by arithmetic, with no discretion available to any official in the stadium.

What makes it interesting is not the threshold. It is everything the threshold does not cover: the events with no gauge at all, the half of the 200m nobody measures, the rounding convention that quietly moves the limit, and the thin air at altitude that does the same job as a tailwind and has never been regulated at all.

• 2Tailwind limit for a valid record, in m/s
• 10Measurement window in the 100m, in seconds
• 13Measurement window in the sprint hurdles, in seconds
• 1.22Height of the gauge's measuring plane, in metres

All four are set by the World Athletics technical rules and do not vary by competition.

The limit is +2.0 m/s, and it covers a shorter list of events than most people think

Six events at senior level carry a wind limit: the 100m, the 200m, the 100m hurdles, the 110m hurdles, the long jump and the triple jump. Junior and masters programmes add shorter sprints and hurdles on the same principle.

That is the whole list. Nothing else in athletics has a wind rule.

The 400m has no gauge. Neither does the 800m, the 4x400m relay, the steeplechase or anything else that goes round a bend and comes back. The reason is straightforward once you picture it: an athlete running a full lap faces every wind direction in turn. A tailwind down the back straight becomes a headwind down the home straight, and in between it is a crosswind pushing sideways against a body already leaning into a curve. There is no single reading that would describe what any of that did, and a reading taken at one point on the track would be actively misleading about the rest of it.

The vertical jumps have no limit either. High jump and pole vault are affected by wind, sometimes badly, but the effect is not a simple assist in one direction. A tailwind on a pole vault runway helps the run-up and then has to be dealt with at the top of the vault, where a vaulter is trying to hold a position that a gust will happily rearrange.

Then there are the throws, which is where the absence of a rule gets genuinely strange. A javelin is an aerofoil. It generates lift, it holds an angle of attack, and a headwind of the right strength makes it fly further rather than shorter, because the extra airspeed over the implement buys more lift than it costs in drag. Discus behaves the same way and is even more sensitive. Nobody measures the wind for either, and nobody caps it. A thrower who catches the right conditions has caught the right conditions, and the mark goes into the book without an asterisk.

So the wind rule is not a general principle about fair conditions. It is a specific rule about a specific kind of help: a following wind, on a straight line, in an event short enough that the athlete never turns round.

The gauge is a set of coordinates before it is an instrument

Most explanations of the wind rule describe the threshold and stop. The rules spend far more of their length on where the gauge goes, because that is what determines whether the number means anything.

For track events, the gauge is placed beside the straight, adjacent to lane one, fifty metres from the finish line. Its measuring plane sits 1.22 metres above the ground, give or take five centimetres, and no more than two metres from the edge of the track.

Those coordinates encode several decisions. Fifty metres from the finish puts the gauge at the midpoint of the last half of a 100m race, in the region where athletes are at or near top speed and where aerodynamic drag is doing most of its work. The 1.22 metre height is roughly the mid-torso of a sprinter in full flight, which is where the frontal area that meets the air is largest. It is also, not coincidentally, exactly four feet, a legacy imperial figure that survives all over the sport's dimensions, including the 1.22 metre lane width that the rules on the start turn out to depend on as well.

The two metre limit on distance from the track is the one that gets overlooked. A gauge parked further out is reading the wind in a different place, and near a stadium wall or a stand that difference is not small. Wind inside a bowl does not behave like wind in open ground. It separates, recirculates and forms a low pressure region downwind of the structure, so a reading taken five metres off the track edge can be measuring an eddy that no athlete ever ran through.

For the horizontal jumps the gauge is placed twenty metres from the take-off line, beside the runway, at the same height and the same maximum offset.

Equipment specification follows the same pattern of care. Non-mechanical gauges, which in practice means ultrasonic units that time a pulse travelling between transducers, are compulsory at the top tier of international competition and for any performance submitted for record ratification. A mechanical gauge, which is a propeller in a tube, is allowed lower down, and the rules require it to be shielded against crosswind, with the tubes extending on each side of the measuring device by at least twice their own diameter. The shielding exists because a propeller does not know which way the wind is coming from. It spins in response to whatever crosses it, and a strong crosswind can drive a reading that has nothing to do with the wind running along the track.

1. 1The gun firesThe gauge is triggered on the flash or smoke of the starter's pistol, not on a signal from the timing system and not by an official pressing a button when they notice the race has begun.
2. 2The gauge samples continuouslyIt records wind velocity along the axis of the track, positive for a following wind and negative for a headwind, for a window fixed by the event rather than by how long the race lasted.
3. 3The window closesTen seconds for the 100m, thirteen for both sprint hurdles, ten for the 200m starting from the moment the first athlete enters the home straight.
4. 4The samples are averagedOne number for the whole window. Gusts inside the window are flattened into it, so a violent two second gust and a steady breeze can produce identical readings.
5. 5The average is roundedRounded to the next higher tenth in the positive direction unless the second decimal is already zero. A reading of +2.03 becomes +2.1. A reading of -2.03 becomes -2.0.
6. 6The reading is attached to the raceEvery athlete in that heat receives the same figure, regardless of lane, regardless of finishing position, regardless of the fact that the leader finished several tenths before the last athlete.
7. 7The comparison is madeAt or below +2.0 the mark is legal for every purpose. Above it the result stands but the mark is wind assisted, and the record book and the entry standard lists both refuse it.

Only the averaging window and the rounding step are fixed by rule. Everything else is the physical situation the rule is trying to describe.

The measurement windows are set by the clock, not by the race

The rules give a fixed number of seconds per event, measured from the flash or smoke of the gun. Ten seconds for the 100m. Thirteen for the 100m hurdles and thirteen for the 110m hurdles. For the 200m, ten seconds beginning when the first athlete enters the straight.

Look at what that produces in practice.

A world class men's 100m final is over in under ten seconds. The gauge is still averaging after the race has finished, and the last fraction of the window is measuring wind that blew across an empty track. In a slower heat, the window ends before the back markers have crossed the line, so the tail of their race goes unmeasured. The window is a fixed rectangle laid over races of varying length, and it fits none of them exactly.

The hurdles get thirteen seconds because the events take longer, and the extra three seconds is a reasonable compromise. It also means a hurdles reading is an average over a longer period, which makes it slightly more stable and slightly less representative of any particular moment.

100m10s

200m10s

100m hurdles13s

110m hurdles13s

Long jump5s

Triple jump5s

Periods set by the World Athletics technical rules. The sprint windows run from the gun; the 200m window starts when the first athlete enters the home straight; the jump window starts when the athlete passes a mark on the runway.

Show the numbers

How long the gauge averages for, by event

Item	Value
100m	10s
200m	10s
100m hurdles	13s
110m hurdles	13s
Long jump	5s
Triple jump	5s

The 200m reading describes the part of the race it did not measure

This is the sharpest problem in the whole system, and it is written into the rule in plain sight.

In the 200m, the gauge does not start when the gun goes. It starts when the first athlete enters the home straight. Everything before that, which is the entire bend, roughly half the race, is unmeasured. Nobody knows what the wind was doing there, because no instrument was pointed at it.

That is not a small omission. On the bend an athlete's heading rotates through about ninety degrees. A wind that is a pure tailwind on the home straight was a crosswind at the top of the bend and something close to a headwind at the start. Or the reverse. The reading you eventually get tells you about the last hundred metres and says nothing at all about the first.

The published modelling of wind effects in the sprints makes exactly this point: because the gauge is only operated once the leader has come off the curve, the conditions through the curve are unknowable, and any attempt to correct a 200m time back to still air inherits that ignorance. A correction built on a home straight reading can be substantially wrong in either direction depending on what the wind was actually doing where it was not being watched.

There is a second, quieter problem. The gauge sits next to lane one. In an eight lane race the athlete in lane eight is more than eight metres away from it, on the other side of a track, potentially in a different flow regime, and quite possibly in the partial shelter or partial funnel created by whatever the stadium has built on that side. Everyone gets lane one's wind.

The rounding rule is not neutral, and it moves the threshold

The rounding convention is written into the technical rules and it is worth reading slowly. A reading is rounded to the next higher tenth of a metre per second in the positive direction, unless the second decimal is zero. So +2.03 is recorded as +2.1. And −2.03 is recorded as −2.0.

Both of those move the number in the same direction: towards the positive, which is to say towards more apparent assistance. A tailwind is rounded up, so a marginal tailwind becomes a bigger tailwind. A headwind is rounded towards zero, so a marginal headwind becomes a smaller headwind.

The practical effect is that the true threshold is 2.00 metres per second, not 2.049. Any measured value above two point zero zero, by any margin the instrument can see, gets rounded to +2.1 and the mark is gone. There is no rounding down into legality.

That is a deliberate choice and a defensible one. If you are going to have a hard limit on assistance, you want the ambiguity resolved against the assisted mark rather than in favour of it. The record book is a place where a false positive costs much more than a false negative. But it does mean that the widely quoted "two metres per second" limit is, in the way the sport actually operates, a limit of exactly two point zero zero, enforced against a measurement whose own accuracy is a question the rule does not discuss.

Raw reading	Recorded as	Status
+1.97	+2.0	Legal
+2.00	+2.0	Legal
+2.01	+2.1	Wind assisted
+2.03	+2.1	Wind assisted
-2.03	-2.0	Legal, and running into it

The jumps measure five seconds of a run-up, and each attempt gets its own number

The horizontal jumps work on a completely different principle from the track, and the difference is more interesting than it first appears.

The gauge starts when the athlete passes a mark placed beside the runway, forty metres from the take-off line for the long jump and thirty-five metres for the triple jump. It runs for five seconds. If an athlete's approach is shorter than that distance, measurement begins when they start running.

Five seconds at approach speed covers most of the second half of a long jump run-up and finishes around the take-off. It measures the wind acting on the athlete while they are building speed, which is exactly the right thing to measure, because the overwhelming determinant of a horizontal jump is how fast the jumper is travelling when their foot hits the board. The flight itself is short, the athlete is a poor aerofoil, and the wind's contribution during flight is small next to what it did to the approach.

The difference from the track is that every attempt gets its own reading. In a race, one gauge reading covers eight athletes. In a long jump competition, a jumper takes six attempts and receives six separate wind figures, one per jump.

That creates a situation with no analogue on the track. A jumper can produce a series in which the best mark is wind assisted and the second best is legal, so the same competition yields both an official personal best and a longer distance that officially never happened. The competition result uses the longer jump. The record book, the qualifying lists and the athlete's own legal best all use the shorter one. Two truths, one afternoon, and the reason they diverge is a reading taken over five seconds while the jumper was still accelerating.

Combined events use an average, and the algebra is the whole rule

Three events in the decathlon carry a wind reading: the 100m, the long jump and the 110m hurdles. Three in the heptathlon: the 100m hurdles, the long jump and the 200m.

For the total score to be valid as a record, the rule does not look at those readings individually. It takes the algebraic sum of them, divides by the number of wind-measured events, and requires the average not to exceed plus 2.0 metres per second.

Algebraic sum is the load-bearing phrase. Negative readings count as negative. A headwind in one event genuinely cancels a tailwind in another, which means a decathlete can have a single event blown along at well over the individual limit and still finish with a legal total, provided the day balanced out.

• 3.5m/s100m reading
• 1.4m/sLong jump reading
• -0.7m/s110m hurdles reading
• 1.4m/sAverage of the three

Invented readings used to show the arithmetic, not a record of any competition. The algebraic sum of +3.5, +1.4 and -0.7 is +4.2, and dividing by three gives +1.4. That sits inside the +2.0 limit, so a total scored on this series would be valid despite one event running well over the individual sprint limit.

Underneath the record rule sits a second, separate provision that catches people out. For the narrow purpose of deciding whether an athlete has achieved the entry standard for a combined events competition, World Athletics allows either of two conditions to be satisfied: no individual event over plus 4.0 metres per second, or the average not over plus 2.0. Either one is enough. That is a genuinely more permissive test than the record rule, and it exists because a decathlon takes two days, the athlete cannot choose the weather, and a system that voided entry standards on one gusty morning would be punishing something nobody controls.

The third layer is the one that trips up statisticians. An individual mark inside a combined event is still judged on its own reading when anybody wants to use it as a standalone performance. A long jump of a given distance inside a decathlon, measured at +2.6, is a wind assisted long jump. It scores its full points towards the decathlon total, and that total may still be perfectly legal on the average. But nobody may list the jump itself as a legal mark, because for that purpose the individual reading governs and the average is irrelevant. How those points are converted into a score is a separate mechanism entirely, and one worth understanding on its own terms.

Altitude does the same job as a tailwind, and nothing in the rules stops it

Here is the asymmetry at the centre of the subject.

A sprinter is fighting air. The drag force on a body moving through a fluid scales with the density of that fluid, the frontal area presented to it, a drag coefficient describing the shape, and the square of the speed relative to the air. Written out, the relative velocity term is the athlete's ground speed minus the wind speed, which is precisely why a following wind helps: it reduces the speed of the athlete relative to the air without reducing their speed relative to the track.

A tailwind attacks the velocity term. Altitude attacks the density term. Both reduce drag. Only one of them is regulated.

Air density falls as you climb, and it falls fast enough to matter over the range of altitudes where athletics stadiums are built.

Calculated from the International Standard Atmosphere, which fixes sea level density at 1.225 kg per cubic metre and a temperature lapse rate of 6.5 degrees per kilometre. These are standard-atmosphere values, not measurements taken at any particular venue, where temperature and humidity shift the number further.

Show the numbers

Air density falls with altitude, and drag falls with it

Item	Air density
0 m	1.23kg/m3
500 m	1.17kg/m3
1000 m	1.11kg/m3
1500 m	1.06kg/m3
2000 m	1.01kg/m3
2400 m	0.97kg/m3

At two thousand metres the air is around eighteen per cent thinner than at sea level. Drag scales directly with density, so at the same running speed a sprinter at that altitude meets roughly eighteen per cent less air resistance than the same sprinter at sea level. Aerodynamic drag accounts for a substantial share of the resistance a sprinter overcomes at top speed, so removing a fifth of it is not a rounding error.

World Athletics does not correct for this, and does not limit it. A mark set at altitude is a mark. Statistical practice is to flag performances achieved above a thousand metres as altitude aided, which is a labelling convention rather than a rule, and the flag has no effect on validity. The record book takes them.

So a following wind of +2.1 metres per second, which is a breeze you would barely notice walking, voids a performance permanently. Thin air worth as much or more, sustained for the whole race, over every race in the meeting, does not.

There are four defences of that position, and they are not all equally good.

Altitude is known in advance and shared by everybody. Every athlete in that stadium runs in the same air, on the same day and in every other race of the season held there. Wind is momentary and unequal: the heat before yours can get +1.8 and yours can get −1.2, and nothing about the schedule was in anybody's control. That is a real distinction. The rule against wind is partly a rule against luck.

Altitude is not a pure gift. Thin air helps a sprinter and punishes anybody running long, because the same reduction in density means less oxygen per breath. It also hurts the aerodynamic throws: a discus and a javelin both need air to generate lift, and at altitude there is less of it to work with. Venue altitude is therefore not a single lever labelled "faster". It is a set of trade-offs that fall differently on different events, which makes a blanket correction much harder to justify than it looks.

Correction requires a model, and models can be argued with. To adjust a time for altitude you need a physiological and aerodynamic model of that athlete, and any number you produce is the output of assumptions rather than a reading from an instrument. The wind rule survives because it is enforced against a measurement, however imperfect. An altitude rule would be enforced against a calculation, and every affected athlete would be entitled to dispute the inputs.

Too much of the record book is already there. Whatever the merits, retroactive application would rewrite a great deal of history, and a governing body that has decided to leave the record book alone in one respect finds it awkward to disturb it in another.

The honest summary is that the sport regulates the assistance it can measure cheaply at the moment it occurs, and tolerates the assistance that is baked into a venue. That is a practical distinction rather than a principled one, and the argument that altitude aided sprint marks deserve their own category has been made seriously in the academic literature for decades without ever winning.

A wind assisted mark is not deleted, it is discounted

The most common misunderstanding about wind assistance rules in athletics is that an illegal reading erases the performance. It does not. Different parts of the sport treat the same run in three different ways, and they are all correct at once.

The competition keeps it. The result stands. Places are as run, medals are awarded, prize money is paid, and the athlete who crossed the line first has won the race. A wind reading has never changed a finishing order.

The record book refuses it, and so do entry standards. The world record rules require wind information to be submitted for outdoor performances up to and including 200m, and for the long jump and triple jump, and refuse ratification when the average exceeds two metres per second. Championship entry standards go a step further: a performance that is wind assisted, or for which no wind reading is available at all, is not accepted as a standard. That last clause is why a meeting with a broken gauge can be a disaster for athletes chasing a qualifying mark, and it feeds directly into the route to a major championship, where a missed standard means depending on ranking position instead.

The rankings accept it and apply a correction. This is the part almost nobody knows. The World Athletics rankings include wind aided results rather than discarding them, and adjust the result score to account for the conditions. There is no adjustment at all between zero and +2.0, so ordinary legal conditions are treated as neutral. Above +2.1 points are deducted on a linear scale, and because adverse conditions are equally real, a headwind earns points back on the same scale. A result with no wind information available takes a fixed deduction of its own. The five wind-measured disciplines that carry a ranking adjustment are the 100m, the 200m, both sprint hurdles, and the long and triple jump.

Two systems, two verdicts, one race. The record book says the mark never happened. The ranking says it happened, here is what the wind was worth, and here is the score with that subtracted. Which of those is the better answer depends on what you want the number for, and the fact that the ranking system has to make this call at all is a good illustration of how much heavier a job it is doing than a record list.

Why the rule exists at all, and what it costs

The wind limit is old, and it was written for a reason that has not changed: without it, the sprint record book would be a list of the windiest afternoons in the history of the sport.

A following wind is worth a great deal in the 100m. The exact amount is a modelling question rather than a measured constant, and estimates depend on the athlete's speed and shape, but it is comfortably large enough that a limit-legal tailwind is a bigger effect than the margin separating a good performance from a historic one. Without a cap, the incentive structure would be obvious: schedule races on windy days, in stadiums with an open end, facing the right way. That is not a hypothetical. Meeting organisers have always known which of their venues run fast.

The cost is borne by the athletes who fall on the wrong side of a number that describes something they did not choose. There is no appeal against a wind reading. There is no mechanism for saying that the gust arrived after you had finished, or that lane eight was in shelter, or that the gauge was reading a recirculation off the back straight stand. The reading is the reading, it is applied to everyone in the race, and it is final.

It is also, in the strict sense, an approximation being enforced to the second decimal place. A ten second average of a turbulent flow, measured at one point beside lane one, at one height, is a genuinely reasonable summary of the conditions and nothing like a complete description of them. The rule works because it is consistent, not because it is exact. That is true of a surprising amount of officiating in track and field, and it is the same bargain the sport strikes with a start information system that judges anticipation from a force trace, or with the rules that decide when technology in footwear stops being a shoe.

What to check on any wind assisted performance

Four questions will tell you almost everything about a mark with a wind figure attached to it.

Which event was it, and does that event even have a gauge? If it is longer than 200m, or a throw, or a vertical jump, there is no reading and no limit, and any claim about the conditions is a description rather than a ruling.

Was it the 200m? If so, the reading covers the home straight only, and tells you nothing about the bend, which is half the race.

Was it a jump, and what did the rest of the series read? Horizontal jumps get one reading per attempt, so the athlete's best legal mark from that competition may be a different jump entirely.

Was it inside a combined event? Then the individual reading governs the standalone mark and the three-event average governs the total, and the two can easily disagree.

And one more, for the marks that carry no wind figure and never will: check the altitude of the stadium. The rule that sends a +2.1 sprint into the footnotes has nothing to say about a venue where the air itself is doing the same work, all day, for everyone.