Muscles in aiming and follow-through: Watanabe 1993

A short paper, in Japanese, on a question nobody had yet measured in the early 1990s: what the muscles do after the arrow has gone. Ten members of a university club shot three arrows each at eight metres on a fifteen-centimetre face, with electromyography at six sites, the deltoid and the ascending and descending parts of the trapezius on both sides. The authors compared the integrated activity over one second of aiming, the second immediately before the clicker fell, with one second of follow-through after the release.

The most active muscle was the deltoid, particularly on the bow arm, with a mean value of 987.0 microvolts per second. Going from aiming to follow-through, almost every site measured fell, settling at around 60% of the aiming value; the exception is the two ascending parts of the trapezius, which rose, to 118.3% on the bow side and 147.5% on the drawing side.

The finding the paper rests on is a comparison between people. In the five archers with the best scores, follow-through activity held at around 80% of the aiming value, while in the others it dropped much further. Relating that ratio, computed over the three sites on the drawing side, to each archer's best single-round score, the correlation was 0.791 across the eight archers who had competition experience, and 0.949 across the six who had passed 1,100 points. In the four above 1,200 points the ratio sat between 79.4% and 117.9% across all six sites and on the drawing side.

The ten participants were members of an archery club and used their own bows, so as to express their habitual technique. The participants' table reports sex, years of practice, height and weight along with best single-round and half-round scores: on average 166.1 centimetres, 56.4 kilograms, 1,109.8 points on the single round and 594.9 on the half. The tests took place indoors, to keep environmental conditions constant, at eight metres from a target fifteen centimetres in diameter, the one used in indoor shooting. After a check of the bow, each archer shot three practice arrows and then three recorded arrows, and all the values reported are the mean of those three.

The electromyography is surface, with bipolar electrodes three centimetres apart, inter-electrode resistance below 10 kΩ and a time constant of 0.03 seconds. The clicker falling was detected with a small metal plate fixed to the side of the grip, closing an electrical contact. Muscle signals and the clicker signal were recorded on a tape recorder, then digitised at 1 kHz and analysed on a personal computer.

The two analysis windows are defined relative to the clicker. Aiming is the second immediately preceding the signal. The follow-through runs from 0.5 to 1.5 seconds after the signal, a choice made to allow for the reaction time between the clicker falling and the arrow being released. For each window and each site the authors compute the integrated electromyogram in microvolts per second, and then the ratio of the follow-through value to the aiming value, expressed as a percentage. The comparison with performance uses that ratio, computed over the six sites together, over the three on the bow side and over the three on the drawing side, related to each archer's best single-round score by Pearson's correlation coefficient.

This is a small study, and the record stays short for that reason. Ten archers, three recorded arrows each, a single club: thirty shots in all carry the whole analysis, and each archer's values are means of three arrows, with a variability between people the authors themselves describe as large and without a common trend.

The link with performance is not a link with those arrows. The recorded arrows are shot at eight metres on a fifteen-centimetre face, and their score is not reported; what is correlated is the best single-round score each archer had achieved in their career. It is therefore a correlation between people, between a laboratory measure taken over a few shots and a historical performance, across eight archers and then six. The correlation of 0.949 concerns six people, and those six are the subgroup selected by score after looking at the data: on a number that small, the value says almost nothing about the strength of the relationship, and the band of uncertainty around it is very wide.

The electromyographic signal is not expressed as a percentage of a maximal contraction, so the absolute values cannot be compared between archers or between muscles; the ratio between the two phases partly gets round the problem, but it remains sensitive to how the electrodes sat on each person. Neither the draw weight of the bows nor any strength measure is reported, so the difference between those who hold tension and those who lose it might also reflect how heavy the bow is for the person drawing it.

The direction of the relationship stays open. The study observes that the stronger archers hold tension after the release: it does not show that holding it raises the score, and no training was tested. The coaching suggestions the authors offer in the discussion, such as watching that the bow hand does not drop and that the tension does not dissolve immediately after the release, are their reading rather than a measured result. The paper, finally, is from 1993, is written in Japanese with the abstract and tables also in English, and appeared in a national physical education journal.

This paper's contribution is to have put a number on something usually taught in words. In the four archers with the highest scores in the group, muscle activity in the second after the release sat between 79.4% and 117.9% of that in the second of aiming; in the others it fell towards 60%. The follow-through, in those archers, is not a pose held out of habit: the muscles of the shoulders, neck and back are still working when the arrow is already in flight.

The interesting detail concerns which part of the trapezius. Between aiming and follow-through almost every site measured falls, while the ascending part of the trapezius rises on both sides, and more so on the drawing side. It is not a muscle that relaxes after the shot, in this sample.

What the record cannot say matters just as much. The link is between archers, not within the same archer: those who shoot better hold more tension, but nobody tried making an archer hold more tension to see whether the score rose. And the three recorded shots are at eight metres on a small face, with the score of those arrows never reported, so there is no connection between what the electromyography shows and where the arrow landed.