Simulated archery release and posture: Song 2023

The study compares two ways of initiating a release, one chosen by the archer and one in response to an external signal, and looks at what changes in standing posture. The action is simulated: eight healthy men with no shooting experience held an instrumented bow frame in both hands, standing on a force platform, taking the draw to 40 per cent of their maximum finger force and holding it steady for five seconds before letting go.

With the self-chosen release the virtual hits were tighter: the precision index, which measures the scatter of impact points around their own centre, came to 0.017 against 0.023 in the external cue condition, with p of 0.01 and a large effect. In the two seconds before release, sway of the centre of pressure was smaller with the self-triggered release, but only in the fore and aft direction, and the difference concerned the component the authors call trembling, the one tied to peripheral mechanisms, rather than the slow drift component.

Postural muscle activity did not change between the conditions, while the way the muscles were coupled did. With the external cue, co-contraction rose between tibialis anterior and soleus on the same side, on both legs, and between the two soleus muscles, meaning the legs stiffened. The decomposition of the signals points the same way: three muscle groups accounted for about seventy per cent of the signal in both conditions, but only with the external cue did all four leg muscles fall into the same group, in seven of eight participants, whereas with the self-triggered release no common pattern emerged.

The result the authors treat as central concerns coordination between those groups. The synergy index, which measures how far variation in one group offsets variation in the others so as to hold the position of the centre of pressure steady from trial to trial, was higher with the self-triggered release, and only in that condition did it exceed the threshold above which such compensation can be said to occur. Lastly, sway velocity correlated with the scatter of hits only in the self-triggered condition, in the fore and aft direction, with coefficients of 0.80 for the centre of pressure, 0.79 for the slow drift and 0.66 for the trembling.

The eight participants averaged 30.5 years of age, 1.71 metres and 72.1 kilograms, were right-handed and right-leg dominant, were not archers and had never competed in archery. The bow frame was built in aluminium and carried five six-axis force transducers, one for the bow hand and four for the fingers; it weighed 1.53 kilograms, less than a bow, a choice stated as a way of limiting fatigue over repetitions. A seventy-five inch monitor three metres away showed finger force in real time.

Before the trials the maximum voluntary force of the four fingers of the right hand was measured by pressing the transducers in the same posture as the task, and the draw target was set at 40 per cent of that. Participants stood barefoot on the force platform, feet shoulder width apart and parallel, their positions marked so they could be repeated; the left elbow was extended as with a bow and the right hand reached the centre of the chest. The orientation of the frame was constrained to zero degrees, and the virtual arrow was projected to the participant's left.

The platform sampled at one thousand times a second, and eight wireless electrodes recorded at two thousand times a second from tibialis anterior, soleus, external oblique abdominis and erector spinae on both sides, placed according to the usual guidelines. Each participant performed twenty trials per condition, in randomised order, with fifteen seconds of rest between trials and five minutes between conditions; the whole session lasted sixty to seventy minutes and nobody reported fatigue. At least twenty minutes of practice preceded the trials.

In the self-triggered condition the release came at a moment of the participant's choosing, in the external cue condition as soon as possible after a random signal appeared on the monitor; the mean time from cue to the onset of release was 318 milliseconds, in line with a visual finger reaction time. The analysis covers the two seconds before the onset of release. The centre of pressure was decomposed into two components, a slow drift derived from the instantaneous equilibrium points and a trembling obtained by subtraction, and for each of them the root mean square, mean velocity and sway area were computed. The muscle groups were extracted with a principal component analysis, chosen instead of non-negative matrix factorisation because orthogonal groups were needed, and then used within the uncontrolled manifold framework, which splits the variability between trials into a part that does not move the centre of pressure and a part that does.

None of the participants was an archer, and the study involved no bow, no arrow and no target. The frame weighs 1.53 kilograms and has an inertia unlike that of a bow, as the authors acknowledge; the draw is at 40 per cent of finger force rather than the load of a bow at full draw; and the hits are computed from finger forces and the orientation of the frame rather than observed on a target. The precision index, moreover, measures scatter around that person's own mean impact point, not the distance from the centre of the target: it is a measure of consistency rather than of accuracy.

The sample is eight men, and the authors write that the number makes their conclusions tentative and that replication with a larger cohort is needed; they announce a follow-up study on high-level archers with real equipment.

Translation to actual shooting is the most fragile part, and deserves care. In the external cue condition the go signal is a random visual cue on a screen, whereas an archer's clicker is produced by the movement of that archer's own draw, is predictable, and works through a different channel. The authors' conclusion about the clicker is therefore an extension of their result rather than something they measured: their external cue is less predictable than a clicker, and the difference they find may rest on exactly that unpredictability.

Some uncertainties remain inside the result. The effect on sway appears only in the fore and aft direction and only in one of the two components; the recorded muscles are four per side and all belong to the trunk and the leg, so no arms and no shoulders; and the correlation between sway and hit scatter is computed over eight points, that is eight people, where a single participant can move the coefficient a long way. Finally the self-triggered condition allowed the moment to be chosen, so a participant who was steadier could also wait for the most favourable instant: the design does not separate the direction of that link.

The study deals with something an archer recognises: the difference between letting go when the shot is ready and letting go because a signal has arrived. In the simulated task, those who decided for themselves had tighter hits, swayed less fore and aft, and coordinated their legs so as to offset their own sway; those reacting to a cue stiffened their ankles, with the same amount of muscle activity but distributed so as to lock rather than adjust.

The passage about the clicker needs reading with care. The authors offer it as an interpretation of their data, but a clicker is not an external cue in the sense of their experiment: the archer produces it with their own draw, and feels it coming. What the study measures is the difference between an action prepared in advance and a reaction to an unpredictable cue, and that is a distinction about how a release is arrived at rather than about equipment.

Nothing here bears on points in competition. Eight people with no shooting experience, no arrows, no target, and a scatter measure computed on virtual hits: the study describes a postural mechanism in a constructed task, and the authors themselves defer to later work on real archers. What remains for someone who shoots is the observation that the legs are not spectators of the release, and that stiffening them is not the same thing as holding the position steady.