Muscle coordination in archery: Edelmann-Nusser 1998

Nine recurve archers from the German national teams, with surface electromyography on eight muscles and an acceleration sensor fixed to the middle of the bow. Each shoots thirty-five arrows aiming for the centre and five more as extended shots, that is passing the clicker and then waiting more than a second before releasing, an exercise used in training so as not to shoot as an automatic reaction to the click.

The first result concerns the clicker. In each archer's electromyographic traces the normal shot and the extended shot share the same shape, the same moments and the same directions of the changes in activation, so no specific reaction to the click is visible. And the interval between the click and the release comes out shorter than the reaction time to a sound measured in the same archer: shorter in seven archers out of nine on the simple task, releasing a key at the signal, and in all nine on the task that adds holding a laser dot on the gold, with p below one per cent in each comparison. The authors conclude that the motor programme of the shot is already under way by the time the click is heard, hence open loop, and that the clicker's function is feedback after the event on the consistency of the draw: comparable to the sound a tennis player hears on striking the ball, from which they learn how the stroke went without being able to change it any more.

The second result concerns the way the release of the force balance is handled. A ratio between the activity before and after that moment is derived from the traces, and the nine archers give different values for it: no common technique is found, although all are of national and in some cases international standard. One thing alone is shared by all nine, the rising activity of the pectoralis major in anticipating that release. The traces of the body seen from above also show different configurations of shoulders and arms, and the authors put these down to differences in body measurements, in arrow length and to the redundancy of the degrees of freedom, which lead to different joint torques at the same force.

A technical detail useful to anyone studying the release: high-speed footage shows that the arrow leaves the string six to eight thousandths of a second before the first maximum of the bow's acceleration curve.

Proceedings of the sixteenth international symposium on biomechanics in sports, four pages, 153 to 156 of the second volume.

The eight electromyographic channels are, on the drawing side, the extensor carpi radialis brevis, the flexor carpi radialis, the deltoid, the trapezius in its descending and transverse part, the trapezius in its transverse part and the biceps; on the bow side, the pectoralis major and the deltoid. An acceleration sensor at the middle of the bow serves to synchronise three instants with the electrical signal: the fall of the clicker, the start of the arrow's acceleration by the string and the moment the arrow leaves the string. The article also describes the clicker for the non-archer reader, as a strip of spring steel five centimetres by five millimetres fixed to the middle of the bow.

Each archer shot thirty-five arrows in the normal condition, instructed to hit the gold, and five arrows in the extended condition, where after passing the clicker one waits more than a second before letting go of the string.

Reaction times were measured in two tasks, ten repetitions each. In the first a key is held down until an acoustic signal. In the second the same key is held down and in addition a laser dot is kept on the gold of the target as still as possible until the signal. In both the interval between the sound and the moment the finger leaves the key is measured.

To read the quality of the technique, a time-normalised ratio was built between two integrals of the electromyographic signal, computed over windows before and after the release of the force balance, with thirty thousandths of a second allowed for the electromechanical delay between activation and force. Each ratio is the mean of that archer's thirty-five arrows, and the chart gives one for each of the eight muscles and the nine archers. The authors note that values greater than one may indicate anticipation.

The last element of the protocol is a manipulation of the equipment: one arrow in five had been lengthened by five millimetres without the archers knowing.

The statistical comparisons are made archer by archer, with a stated threshold of p below one per cent for each.

The article does not contain a single value in thousandths of a second. The central result is that the interval between the fall of the clicker and the release is shorter than the reaction time to a sound, and neither the interval nor the reaction time is reported, for any archer or in either task. Anyone wanting to cite that comparison can cite its direction and not its size, and cannot set it against the times other studies report for the same interval. These are four pages of conference proceedings, with two charts and no table of results.

The comparison, moreover, is not between things of the same nature. On one side is the time between a sound and the moment a finger leaves a key, on the other the interval between the fall of the clicker and the opening of the hand holding the string at full draw. They are two different acts, with different effectors and different preparation, and the second reaction task adds aiming but still measures a key. That the reaction at the key is slower does not show that the release could not have been triggered by the click, because the two movements were never measured in the same way. The argument points in a direction; the article presents it as a conclusion.

The first of the two tasks, furthermore, does not hold in two archers out of nine, and nothing is said about the exception: neither which way it went, nor by how much. There is no overall analysis across the nine, no measure of effect, no distribution: there are nine separate comparisons against a threshold.

The other half of the clicker result is also judged by eye. That the traces of the normal shot and the extended shot share the same shape is a similarity read off the figures, with no quantitative comparison between the two forms. And the extended shot rests on five arrows per archer against thirty-five, a much narrower base.

Of the arrow lengthened by five millimetres nothing is known. The manipulation is described in the methods, one arrow in five and without the archers knowing, and no result concerns it anywhere in the article. Whatever it was meant to show is not reported.

On the second part of the work, the conclusion that no common technique exists comes from nine bar charts and nine traces looked at side by side, with no measure of similarity and no test. It is an observation that points the way of what one sees in a training hall, but an observation it remains. The same holds for the explanation the authors offer: body measurements were not taken and no relation between body dimensions and activation pattern was tested, so that part is a hypothesis.

The electromyographic ratios are not normalised to a maximum contraction, so they serve as a comparison within an archer and not between archers or between muscles. And it is the authors themselves who write that values greater than one may indicate anticipation, not that they do indicate it.

Finally there is no score. The archers were shooting for the centre, but where the arrows went is not reported, so nothing links the activation pattern, or the rising pectoralis activity, to accuracy. The bow side is read from two muscles alone.

This is the study behind one of the most useful things that can be said to an archer about the clicker: you do not shoot because you hear it. In nine national team archers the interval between the click and the release is shorter than the time those same people take to react to a sound, and the muscle traces of the normal shot are the same as those of the extended shot, where the clicker is passed and the archer deliberately waits. The authors read this as the movement already being under way by the time the clicker sounds, and the clicker serving as feedback on how consistent the draw has been rather than as a starting signal. The comparison they make is a good one: the sound of the ball on the racket tells the tennis player how the stroke went, it does not make them play it.

For a coach, this changes how the clicker is used in correction. Asking an archer to shoot faster after the click assumes the click is the start of the movement, and these data point the other way: if the programme is already running, the consistency of the draw and the moment of deciding to go come earlier, and the clicker reports on that.

The second thing to take away is that among nine archers of that standard no common technique is found, neither in muscle activation nor in the geometry of shoulders and arms seen from above, and that different geometry, at the same force on the string, produces different torques in the joints. One thing is common to all nine, though: the activity of the pectoralis major rising in anticipation of the moment the force balance is released.

What is not here is any number. No thousandths of a second, no score, and on the arrows lengthened by five millimetres no result. The registry keeps this work as the only study that addresses the clicker question directly, for the direction of its result rather than its size, and as one of the pieces of evidence that a single elite technique is not observed.