From the outside, an archer at full draw looks motionless. Underfoot, though, the centre of pressure never stops moving, and biomechanics now measures it with force plates, sensors on the bow and electromyography. The picture that emerges is less simple than the one taught on the shooting line. Most studies find that the less the body sways while aiming, the better the arrow tends to be, but how the body sways and when also matter: elite archers sway less and at the same time faster, and after the release the body moves more than twice as far as it does while aiming. Whether a steadier bow while aiming gives better shots, on the other hand, is contested, because each study measures something different and calls it bow stability.
Around these two points sit smaller findings, and each one is worth knowing on its own terms. Experienced archers prepare their balance before the release while beginners recover it afterwards, and in more experienced archers the trunk relies more on the deep abdominal muscles. Archers also stand more still than people who do not shoot, although single-leg balance tests have little to do with score. On training, the starting point is more cautious than people tend to say: core programmes lengthen trunk hold times, while on balance and on score the results disagree. At the end come three lesser-known cases: the finals platform, flat feet and the bite plate.
The centre of pressure and the score
The centre of pressure, abbreviated to COP in the research, is the point where the body's push on the ground is concentrated. It is measured with a force plate, and in an archer it sways continuously in two directions: along the line of the feet, which is towards the target, and between toes and heels, which is across the target. Studies describe how far it moves with different measures (the area it covers, the length of its path, the amplitude and the speed of the sway), and they do not always reach the same conclusions.
The general direction is clear: in the registry the claim that the less the body sways while aiming, the better the shot is reported, and two systematic reviews count the studies on the same side. The most direct comparison is within the same archer: in eight national-level recurve archers shooting six arrows each at 13 m, the path of the centre of pressure on the worst arrow was more than twice as long as on the best. Between archers of different levels, in 27 recurve archers measured on a force plate, the amplitude of the displacement was smaller in the elite than in the beginners, with the intermediates in between.
Three things, however, stop all this from turning into the advice to stand like a statue. The first is that how you move matters as well as how much: in a French thesis on 15 recurve archers, the elite showed smaller sway than the intermediates while aiming but also faster sway, and on their worst arrows the centre of pressure moved more quickly. The second is the time window, because in the 39 elite archers measured before a World Cup stage, what predicted the score of the arrow was the maximum sway speed, which always comes after the release, while the sway in the second before it did not enter the model. The third is that null results exist: in 33 national-level archers the displacement of the centre of pressure while aiming was not linked to level, and in eleven Russian archers who shot with their feet in a T stance, balance dropped by more than a third without the score falling.
One possible explanation comes from six elite Belgian archers, each shooting 90 arrows at 18 m. On the 9s and 10s the variability of the joints was organised to keep the direction of the arrow steady, without every movement being reduced to a minimum: the body moved, and the movements appeared to compensate for one another. The finding needs care, because the paper reports only an index and not the two measures it is built from, so greater compensation cannot be told apart from simply smaller disturbance. If the reading holds, standing still and controlling the shot are not the same thing.
Which direction dominates is also an open question, and the first problem is one of words. For an archer standing side-on to the target, "lateral" relative to the body means along the line of the feet, towards the target, while "lateral" relative to the target means between toes and heels. Once they are put on the same reference, the two older studies, with four elite archers each, find more movement across the target over the whole draw and aim; the three more recent ones find instead that the line of the target dominates close to the release, and in the 39 World Cup elite, in the final second, the centre of pressure moved 13.5 mm towards the target against 10.7 mm across it. This is why the claim that while aiming, the body sways more sideways across the target than back and forth towards it is contested in the registry. The most likely explanation is the time window, but nobody has yet compared the two windows in the same study, and on which direction weighs more on the score the data are few and do not all point the same way.
After the release the body moves more
The moment when the body moves most is not the aim. In the 39 World Cup elite archers, ten arrows at 70 m on a force plate, the centre of pressure moved on average 20 mm in the second before the release and 48 mm in the half-second after it: more than twice the distance in half the time. In 37 pre-elite recurve and compound archers the wrists went from 3-12 to 105-271 degrees per second, while the ankle stayed almost still, from under 2 to 2-4 degrees per second, so the jump concerns mainly the arms and the bow and much less the feet. In the registry the claim that after the release the body moves far more than it does while aiming is reported: three samples point the same way, but none of them compares the two phases with a statistical test and the measurement windows differ in length.
Moving a lot after the release is neither a fault in itself nor a virtue. Across levels, the French elite archers moved more than the intermediates after the release on every centre-of-pressure measure; within the same level, by contrast, the worst arrows came with a higher peak speed. In the 39 World Cup elite, between the stillest and the most mobile archer after the release, the predicted score per arrow went from 8.1 to 5.4, and that peak speed was already signalled by the speed at the moment of release. For the archer the consequence is practical: the release and the follow-through are as much a part of stability as the aim.
A steady bow while aiming: a contested finding
On the shooting line archers are taught to look for the steadiest possible aim and not to let the shot go while the sight is wandering. It is reasonable advice, but the claim that the less the bow moves while aiming, the better the shot is contested in the registry. The supporting results are the more numerous: in seven international-level junior recurve archers, the window covered by the aiming point in the final second was smaller on the 10s than on arrows scoring 8 or less, in all seven; in the eight national archers mentioned above, the bow travelled 28.8 cm on the worst arrow against 17.8 cm on the best; in eleven Japanese university archers, an accelerometer on the bow predicted the score of the end.
Null results, however, appear at both levels. In four elite archers the amplitude of the aiming sway was the same on the 10s and on the 8s; in the same seven juniors, those who held the bow steadier did not have the higher average; in 32 young Malaysian archers and in 27 regional compound archers, bow movement did not separate the stronger from the weaker. None of these results has been explained, and they cannot be added to the others because each study measures something different: the area covered by the aiming point, the amplitude of the sway, the acceleration of the bow.
There are, on the other hand, two situations in which movement while aiming increases. The first is fatigue: in twelve Italian archers a fatigue protocol increased the lateral sway of the bow in all twelve, by 20% on average, while the centre of pressure underfoot did not shift, and the claim that muscle fatigue increases bow sway while aiming is reported. The second is a long hold: in eight recurve archers from the British national programme, with the aim stretched to twice and three times their usual time, the wrist of the bow hand moved more and the average score per arrow fell from 8.60 to 8.13 and then 7.75, with a significant difference only between the usual hold and the triple one; the longest hold, however, always came last, and the bow was not tracked. I cover bow-arm tremor and aiming time in two separate articles.
Before the release: balance is prepared
When the body knows a push is coming, it anticipates it, and the postural muscles switch on a moment early to counter it. In the shot that push comes with the release, and the question is whether experienced archers anticipate it. The only direct comparison dates from 2024: nine Korean archers with sixteen and a half years of practice on average and nine novice university students simulated the shot on a force plate, without arrows, with a laser pointer on a target at 10 m. In the experts the load on the feet was already shifting before the release, in the direction opposite to the push of the bow, and in five of the eight leg muscles the peak fell in the window before the release, whereas in the novices the work came afterwards. After the release the experts regained their balance sooner, despite a similar maximum excursion.
In the registry the claim that experienced archers prepare their balance before the release, while beginners correct it afterwards is reported: it rests on a single cross-sectional study of eighteen men, with two trials each, without arrows or score. The other study on the subject, on nine elite Korean women archers, confirms that experts anticipate, mostly in the last fifty milliseconds before the release, but finds that a larger adjustment does not go with a better aim: sideways there is no link, and in the fore-aft direction, when the archer chose the moment of release herself, a smaller adjustment went with a more precise aim.
The drawing path and the type of bow
The way you reach the aim also leaves a mark on balance, but the data here are very thin. One experienced recurve archer shot three arrows with each of three drawing paths, from above (his usual one), direct and from below, at 12 m with one foot on each of two force plates. In the second before the release the direct draw gave wider sway in the shooting direction, 1.0 mm against 0.6 mm for the draw from above, and shifted the mean fore-aft position of the centre of pressure by about ten millimetres; in the half-second after the release nothing changed. The claim that the path by which the bow is raised and drawn changes balance while aiming is reported, with limits that matter: one archer, three arrows per technique and a significance threshold widened to p below 0.10. All nine arrows went into the centre, and the comparison may reward habit more than the drawing path itself.
The comparison between bows comes from eleven qualified Russian archers aged 17-18, six compound and five recurve, on two stabilometric platforms. In the two seconds before the release a composite index of balance quality was 69.9% in the compound archers and 40.3% in the recurve archers, and when the archers shot with their feet in a T stance the drop hit the recurve archers much harder than the compound archers. In the registry the claim that compound archers hold their balance better than recurve archers in the seconds before the release is reported and no more: the paper does not test the two groups against each other, the journal is regional, and with five and six archers the bow cannot be separated from the people shooting it.
The trunk and the deep abdominals
That the trunk should be held steady from the inside, by engaging the deep core rather than bracing the abdominals, is one of the most common cues in archers' conditioning. Direct measurements are few, but they point the same way. In kyudo, among nine practitioners measured with electromyography of eight trunk muscles, in the second half of the shot the experts relied on the internal oblique and transversus abdominis, the less experienced on the external oblique. In 28 Indian archers at three levels, the transversus abdominis worked in the elite at 22.9% of maximal contraction at anchor and 32.9% at release, against 7.4% and 12.2% in the beginners, and level affected the transversus far more than the multifidus. This is why the claim that in more experienced archers the transversus abdominis plays a larger part in stabilising the trunk is reported in the registry.
These are comparisons between groups, with no link to score, and where score is measured the picture becomes more complicated. In eight elite French archers, 549 arrows at 70 m, the multifidus on the string side was more active while aiming on the 10s than on the 9s and on arrows scoring 8 or less; in the follow-through of the arrows scoring 8 or less, however, the transversus abdominis was more active, not less, than on the 9s and 10s, so a more active transversus is not enough on its own to make a better arrow.
Then there is the question of how the core is measured in the field. Hold tests, the side plank and trunk extension, are often used on the assumption that whoever holds longer controls the deep muscles better. A 2025 study set them beside ultrasound in fifteen young elite archers: only the side plank on the non-dominant side tracked how much the back muscles thickened at the level of the fifth lumbar vertebra, while the side plank on the dominant side and trunk extension were linked to nothing, and none of the three tests said anything about the deep abdominals. The claim that of the clinical trunk tests, only the side plank is linked to how well an archer activates the deep back muscles is reported, and it calls for caution, because 24 correlations were calculated on fifteen people without correction and only two came out significant.
Archers compared with people who do not shoot
Archers stand more still than people who do not shoot. Ten Bulgarian Olympic recurve archers with more than five years of practice, compared with fifteen untrained peers, swayed less on the force plate in almost every condition, eyes open and closed, on a firm surface and on foam; in a Japanese study, 24 archers swayed less than 34 non-athletes when the dotted box surrounding them started to move, although the difference disappeared by the third trial. The most interesting detail is the form the advantage takes: closing their eyes disturbed the Bulgarian archers very little, while the foam removed almost all of their advantage in sway speed, as if they relied more on their feet and body than on their vision.
In the registry the claim that archers stand more still than people who do not shoot is reported. These are cross-sectional comparisons, and they do not say whether shooting makes people steadier or whether the steadier people take up archery and stay with it. They also show something else: standing still on two feet is not the same as balancing on one leg, and on that test the archers' advantage is far less clear.
Static balance and dynamic balance
The two balance tests archers are given most often measure different things. Static balance is usually measured with the stork test, the number of seconds you can stand on one leg; dynamic balance with the Y-Balance Test, the distance you can reach with the other leg without losing your footing. One test gives a time and the other a distance, so the results cannot be put on the same scale, and in the Turkish study that measured both on the same archers the two went together, but only the second was linked to score.
For static balance the picture is unfavourable. Nine samples of archers have set a single-leg test beside score, and seven found nothing. The most striking result, a correlation of 0.906 in 31 young archers, is explained by the fact that in that sample balance, strength and endurance were almost the same measure, and as soon as they entered the model together balance stopped counting. The claim that single-leg static balance is linked to score is therefore contested, with one exception to keep an eye on: in 30 professional compound archers, after twelve weeks of BOSU work, those who had improved most on the stork test had also gained the most points, but in a single group with no control.
For dynamic balance the picture is a little more favourable. In eighteen recurve archers from a Turkish club, those who reached further on the Y-Balance scored more at 50 m, but once leg length was taken into account the link held only on the left; in the 30 compound archers, those who improved the dynamic balance of the left leg most also gained the most points, while on the right the link was not significant. The claim that dynamic balance is linked to the score is reported: these are associations in small samples, each on one leg only, and they say that the two tend to go together, not that training dynamic balance raises the score.
How stability is assessed in practice
In my work a postural assessment runs on three levels. The structural level covers how the body is set up at rest, how much mobility the joints have and whether there are asymmetries to look into; the functional level covers tests of strength, holding and movement control under load; the kinematic level covers centre-of-pressure measurement, video analysis of the shot and, where needed, electromyography. The first two need little equipment and a trained eye, the third needs instruments or a properly equipped centre, and where specific instruments and expertise are required the assessment belongs to someone trained in sport science, physiotherapy or biomechanics.
The measurements above help in choosing the tests. The stork test on its own says little about shooting, plank hold time says little about the deep abdominals, and a dynamic balance test is linked to score a little more often than a static one; the most informative thing is still to watch the body while it shoots, during the aim and in the half-second after the release. The mistake I see most often is jumping straight to the kinematic level: centre of pressure and electromyography are measured on an athlete whose structural limits have never been identified, and the data can be read in ten different ways. That is why I work in the opposite order, from structural to kinematic, even though it takes longer.
How stability is trained
On training, the data are more cautious than what you hear on the range. That core exercises improve an archer's balance is contested: two studies out of three point that way, with figures that do not fully add up, and the third points the other way. That balance training improves shooting accuracy is contested as well. The strongest result comes from 30 national-level Indian archers assigned at random: with six weeks of balance work alone the score over twelve arrows went from 108.5 to 111.4, in the group that added nothing from 107.6 to 106.2, and the difference does not survive the statistical test. In beginners the picture is more favourable, but more fragile.
The reading for the core is similar. The claim that core training improves the score is reported, but the chain from exercise to arrow is closed by a single study, and poorly: in the same 30 national-level archers only the group doing Pilates plus balance pulled away from the control group, but it had started twelve points below the others, and compared with balance work alone the difference did not hold. Among the associations, in 23 archers from a university club the plank hold explained about 7% of the variability in score at 18 m. The trunk does respond to training: in 38 competitive archers assigned at random, a twelve-week circuit took the plank from 78.5 to 102.3 seconds, but the study does not link that gain to score, and the circuit included shooting too.
More broadly, the claim that physical conditioning improves shooting accuracy is reported: a dozen studies on archers find higher scores after the programme and none finds a decline, although many have no control group and shooting alone also raises the score. As for holding, the idea that for archers, isometric endurance matters more than maximal strength is contradicted: the two studies that come closest to the question point the other way, and in the little there is, holding does not turn out to matter more than strength.
Within these limits, I keep to a few criteria with my athletes. Exercises reflect the posture and load of the shot, because the transfer from the gym to the shot is the least documented part. Strength and holding are trained together, without sacrificing one to the other. Balance work is done little and often, for example ten minutes a day rather than an hour once a week, and that is a practical guideline, not a measured finding. For balance, core and proprioception programmes and their figures, there is a dedicated article.
Three particular cases
That leaves three questions often heard on the range, each with at least one measurement behind it. The first concerns the raised finals platform. A poster prepared with elite Australian archers ahead of the Rio 2016 Games compared twelve arrows at 70 m shot from the ground with twelve shot from a platform built to reproduce the Olympic one: on the platform the left foot swayed more, both sideways and fore-aft, and the overall centre of pressure swayed more fore-aft, while the right foot did not change. The claim that shooting from the raised finals platform increases body sway is contested, because the same study points in two directions, and it carries little weight: the poster does not say how many archers it measured and does not record the score. Nobody has yet shown, then, that the extra sway costs points.
The other two are documented curiosities rather than guidance. At a university in Jakarta, ten archers, ten footballers, ten futsal players and ten pencak silat practitioners left a wet footprint: flat feet affected three archers in ten against one in ten in football and futsal, but four in ten in pencak silat. The claim that flat feet are more common among archers than in running sports such as football is reported, but with no statistical test, and with ten people per group the difference may be chance. In 30 Indian archers measured with a different instrument the most common foot type was normal, and foot posture was not linked to body sway.
The last concerns the bite plate. In 1992 a group of dentists measured a 22-year-old archer on a stabilometric platform in the aiming position, after establishing his normal variability over three days, and had him wear three resin bite plates that raised the occlusion by 2, 5 and 10 mm. Only the thickest changed anything: the area and length of the centre-of-pressure path fell below his normal variability as early as five minutes after it was fitted, with the largest reduction at thirty minutes, while with 2 and 5 mm nothing changed up to sixty minutes. The claim that a bite plate that raises the occlusion reduces body sway while aiming is contested: one bite plate in three had an effect, in a single archer, with no sham plate as a control and no score.
Stability and the rest of the shot
Stability on its own is not enough. A rock-steady archer whose eyes do not know where to settle is still limited, and a stable archer with good quiet eye but haphazard periodisation will arrive at competitions in form more by luck than by design. What the measurements add is that the stability that matters is not the same as stillness: elite archers sway little but quickly, experienced archers, in the only direct comparison, prepare their balance before the string leaves, and after the release elite archers move even more than intermediates.
In an initial assessment, all the same, posture is almost always the first thing I look at, because it shapes everything else rather than because it is the most important thing in absolute terms. If you are an intermediate archer who has been stuck for months, before changing bow, arrows or anchor, look at what your body does in the seconds before the release and in the half-second after it.