Muscle activity and archery scores: Suwarganda 2012

Three archers from the Malaysian national team, two men with Olympic experience and one woman, all with at least four years on the international circuit, shot between 29 and 36 arrows at 70 metres with surface electromyography on four muscles: the triceps of both arms, the left deltoid and the right trapezius. A radar gun behind the archer measured arrow speed and a former elite archer recorded the score arrow by arrow.

The mean per arrow was 8.93 for the first archer, 8.86 for the second and 9.14 for the third; mean speeds were 197.9, 219.4 and 220.3 km/h. The first was slower because he used a lighter draw weight.

The finding of the study is that no single muscle accounted for the score in all three. In the first archer, right trapezius activation was the only measure appreciably linked to score, and on its own it accounted for 22% of its variation; the regression pointed to right trapezius and left triceps together. In the second it was variability rather than level that mattered: that of the left deltoid and of the left triceps. In the third, the most accurate of the three, score was linked to the variability of the right triceps. For arrow speed the picture was just as individual: left triceps activation in the first, right triceps variability in the second, right trapezius and left triceps activation in the third.

The authors also note that the regression coefficients were smallest for the third archer (5.1 against 7.8 to 11.3 and 11.7 to 14.4), and read this as a sign that in the best of the group score and speed depended least on muscle activity.

The data were collected one archer at a time, on non-consecutive days, at the outdoor national training centre during regular sessions, in calm weather at about 28 degrees. After a warm-up the electrodes were fitted and the archer shot the series at 70 metres.

Twin electrodes were placed on the skin over the muscle belly and parallel to the fibres, on the right and left triceps, the left deltoid and the right trapezius. The signal was recorded with a Noraxon TeleMyo 2400T G2 system, rectified and filtered with a root mean square function over 100-millisecond intervals. The window analysed runs from the hand anchoring at the chin to the release.

How the values were expressed matters. The authors did not use maximal voluntary contraction: for each shot they divided the mean activity of that shot by the mean activity of all that archer's shots, and called the result standardised activation, where 1.0 means that shot matched that person's own average. Standardised variation is the standard deviation of activity within the shot, divided by the same average. Both measures therefore describe how far a given arrow departed from that archer's habit, not how hard the muscle worked in absolute terms.

Links between activation or variation on one side and score or speed on the other were explored in Excel, comparing the r squared of a straight line and of a second-order polynomial, then with multiple linear regressions in SPSS over all shots, one per archer, with score or speed as the dependent variable and significance set at p < 0.05; for curved relationships a hierarchical regression with the squared term was added. Some speed readings were lost to interference from team-mates shooting alongside: four for the first archer, six for the second, five for the third.

Three archers, one session each, a few dozen arrows: this is preliminary work and says so in its own title. At these numbers none of the associations stands on its own, and the authors say as much: more data are needed to know whether the links are repeatable, and it is not known how they change across the phases of a training cycle.

There is also a methodological limit that bears directly on the figures reported. By standardising each shot against that archer's own mean, the values end up calibrated around that person's average score, which here is always close to 9. The authors acknowledge that expressing everything as a percentage of maximal voluntary contraction would have been preferable, and that practical constraints and the squad's schedule ruled it out. Without that normalisation the three archers cannot be compared with one another, which is part of why the results come out differently from person to person.

Four muscles were recorded out of the many involved in the shot, chosen from among the proximal ones, and the r squared values are low even where the link is called significant: the highest reported for score is 0.32. Scores were read through binoculars from the shooting line rather than measured on the target face, and some of the speed readings were lost. There is no comparison with less experienced archers, nor with different sessions by the same archer.

One internal inconsistency is worth flagging: the results text says the mean per arrow ranged from 8.93 to 9.14, while the table gives 8.86 for the second archer, the lowest of the three. This record uses the table's figures.

The reason this study sits in the registry is not any of its correlations, which are too fragile, but its negative result: in three archers of the same level, training together and measured in the same way, the muscles that moved with the score were different for each of them. There was no muscle of the score.

The paper also introduces a distinction that is not usually made: how hard a muscle works is one thing, how much its work changes from one arrow to the next is another. In the two faster archers it was the second that tracked the score, in the slower one the first. The authors summarise it that way, but across three people it remains a question to pick up again rather than a result to lean on.

For an archer the practical sense lies mostly in what the study takes away: an electromyograph on one archer does not produce a profile that holds for anyone else, and a reference taken from a champion's shot does not transfer. Even the observation that the best of the three had the smallest coefficients is the authors' reading of a single case.

None of the figures in this work can be used to set anything on a bow or in training. What it gives is the knowledge that the link between muscle activity and score, at least in archers who already have their technique, has to be looked for one person at a time.