Archery shooting dynamics: Erkek 2010 study

Four Turkish women archers of international level shot 30 arrows each at 18 metres indoors, with electromyography on the drawing arm, two force plates, a system recording the movement of the aim on the target and a chronometer linked to the clicker. All the instruments were synchronised to the clicker's snap. Of the 120 arrows, 108 were analysed, those that scored 8, 9 or 10: thirty-four eights, thirty-five nines, thirty-nine tens.

Across all the arrows, two things change with the score. The activity of the finger flexors in the second before the clicker snaps is lower on the tens: 80.0 per cent of maximal contraction, against 84.4 on the nines and 85.6 on the eights (F 5.4, correlation with score of minus 0.29). And the draw time, measured from the first activation of the trapezius to the clicker, is shorter on the tens: 4.8 seconds, against 5.0 and 5.6 (F 5.0, correlation of minus 0.28). Deltoid activity in the second after the clicker also falls as the score rises (31.6, 30.6 and 29.7 per cent), though with no significant difference between the groups.

The reaction time to the clicker, which the abstract presents as shorter on the better performances, does not move in the data: 143.4 milliseconds on the eights, 142.5 on the nines, 142.3 on the tens, with an F of 0.05 and a correlation of minus 0.03. Nor do the finger extensor, the trapezius, the aiming time on the target (3.6, 3.7 and 3.7 seconds) or the extent of the aim's wander on the target, about 13 centimetres horizontally and 12 vertically at every score. Lateral postural sway comes out correlated with the score, but the means do not follow any order (31.7 mm on the eights, 12.5 on the nines, 16.7 on the tens) and the comparison between the groups is not significant.

Looked at archer by archer, the picture breaks up. In the first, the reaction time to the clicker falls from 159.8 milliseconds on the eights to 140.4 on the tens; in the second, flexors, trapezius and extensor all fall as the score rises, and the draw time drops from 6.6 to 5.3 seconds; in the third the extensor is more active on the better shots, which is the opposite; in the fourth the flexors and the lateral sway fall. The authors conclude that each archer develops a muscular and postural strategy of her own.

The four participants had represented or were representing Turkey at national or world events, all right-handed and shooting recurve. The characteristics reported are: mean age 25.7 years, weight 54.0 kg, height 168 cm and draw weight 16.9 kg, with very small standard deviations, notably 0.1 kg on the draw weight.

Surface electromyography recorded four muscles of the drawing arm at 600 Hz: flexor digitorum superficialis, extensor digitorum, posterior deltoid and trapezius. Before the shooting, each muscle's maximal voluntary contraction was measured, and the amplitudes are expressed as a percentage of that maximum. Since the trapezius is the first to contract, the authors used its first activation to define the total draw time, which ends at the clicker's snap, and the point at which it reaches its maximum to define the full draw time.

The force plates recorded centre of pressure displacement at the same rate, with the axis parallel to the target face for lateral sway and the perpendicular one for fore and aft sway, along with the weight taken off the legs after the snap. The feet were marked on the floor and the position held throughout the session. For aiming, an electronic home trainer was used with a transmitter mounted on the bow, which starts recording when the aim stays on the target for more than a tenth of a second. Reaction time to the clicker was taken with a shooting chronometer with a resolution of one millisecond and a stated error of 0.01 per cent.

The analysis used correlations, analysis of variance with Bonferroni correction and t-tests, with a threshold of 0.05. Arrows, not archers, are the unit of the main comparison: the three score groups contain 34, 35 and 39 arrows coming from the same four people.

Four people, a single session, 108 arrows treated as independent observations when they come from only four archers: one archer who shoots many tens and who habitually draws quickly shifts the mean of the ten group on her own. The authors in effect acknowledge this, by showing how much the picture changes when they look at them one at a time. Arrows scoring below 8 were discarded without the number being stated (it is twelve, by subtraction).

Abstract and conclusions do not match the tables. The text states that the reaction time to the clicker is shorter on the better performances, while the three values are 143.4, 142.5 and 142.3 milliseconds with an F of 0.05: that is no difference at all. The authors themselves, in the discussion, write that they formed that hypothesis after looking at the results of two archers out of four, and the introduction cites an earlier study that found no relation between clicker reaction and score in sixteen archers.

The signs of the correlations in the individual table do not match the direction of the means: for the second archer the flexors fall from 87.0 to 80.2 per cent going from eights to tens, and the correlation reported is positive, 0.557. The same happens for that archer's trapezius, extensor and draw time. On lateral sway too the text says performance improves as sway falls, while the correlation reported is positive and the means are not in order.

The table of physical characteristics invites doubt: four archers with draw weights varying by a hundred grams. The same table, with the same numbers, appears in a second paper from 2011 under Tinazci's name alone, where the same four people become men.

Finally, the study measures the score of the single arrow, not the group nor the outcome of a competition, and everything takes place at 18 metres indoors, with no wind and no competitive pressure.

The finding that holds up best is the one on the finger flexors: in the second before the clicker, on the ten-scoring arrows the string hand works less, 80 per cent of maximal contraction against 86 on the eights. That fits the idea, common among coaches, that the string hand should stay as inactive as possible until the release, but here it is measured on four people in one session and it does not say which way the causation runs: a relaxed hand may produce the ten, or a shot that is already well aligned may allow the hand to stay relaxed.

The shorter draw time on the better arrows, 4.8 against 5.6 seconds, is the second significant result. It counts as an observation about the group, not as a prescription: one of the four archers shows the opposite pattern on other measures, and nobody tried changing the draw time to see what happens to the score.

What this study does not show, though it is often cited for it, is that reacting faster to the clicker produces higher scores. Here the three mean times are practically the same number. Anyone citing this work for the opposite is citing the abstract rather than the data.

One point of method remains: with four archers, group means conceal opposite strategies. The same table shows one archer who shoots better with the extensor more active and another who shoots better with every muscle quieter. Taking the mean of a small group of elite archers and drawing a rule from it leads nowhere good.