Bow sway on three axes in junior archers: Talayero 2026

Eight young recurve archers from the Madrid federation, five boys and three girls aged between 14 and 21, shot twelve arrows at 18 metres in a laboratory with eight infrared cameras. Three markers on the bow, at the tip of the long rod, on the grip and at the tip of the lower limb, made it possible to track the midpoint between them on three axes: vertical, lateral and longitudinal, which means towards the target. The amplitude of displacement was calculated over a fixed five-second window before the release.

On average the bow moved 8.56 millimetres vertically, 6.65 longitudinally and 5.30 sideways, with standard deviations almost as large as the means. One archer weighs very heavily on these values: the one with the lowest score (6.00 points per arrow) and the widest displacements on all three axes, between 11 and 21 millimetres. Leaving that archer out, as the authors had planned to do for anyone well outside the distribution, the means drop to 6.83 millimetres vertically, 5.24 longitudinally and 4.42 sideways. The order stays the same, and archer by archer the vertical movement exceeds the lateral one in six cases out of eight.

The mean score was 8.70 points per arrow, 9.08 without the outlying archer. The only association that survives correction for multiple comparisons is the one between score and group area, measured as the smallest circle containing the six arrows of an end (rho -0.96). Score goes with longitudinal displacement in the expected direction, rho -0.68, and -0.52 without the outlier, but the confidence interval reaches +0.04 and the correction leaves it well short of the threshold. With vertical and lateral displacement there is no stable association: the sign changes when a single archer is removed. Grip strength, 30.5 kilograms on average, bears no relation to the score (rho -0.19), while the draw weight measured at full draw, from 31.5 to 47.8 pounds, points towards a higher score (rho 0.36, and 0.54 across seven) without reaching significance.

The study is cross-sectional, with a single session of about half an hour per archer in the laboratory of the Universidad Europea de Madrid. Entry criteria required at least four years of structured training and competition at national or international level, at least six hours of training a week and no injury in the previous six months; each archer shot with their own competition equipment, unchanged. One archer in the sample is left-handed.

Grip strength was measured with a Takei dynamometer, two attempts per hand, keeping the higher value. Draw weight was measured with a digital dynamometer by drawing the bow to the archer's own draw length, three times, keeping the mean. After a warm-up each archer shot two ends of six arrows at 18 metres, with a few minutes' rest between them.

Movement was recorded at 120 frames per second by an eight-camera Vicon system, with fifteen markers on the body and three on the bow. To locate the phase before the release, the authors followed the distance between the wrist of the drawing hand and the grip: the analysis window is a fixed five-second interval within the stretch where that distance changes little, after the draw and before the abrupt change of the release. The authors make clear that the five seconds are not the length of the aim but a window of the same length for everyone. For each archer the mean of the twelve arrows is kept. Group size came from a photograph of the target after each end, using the smallest circle that contains the six impacts. The associations are Spearman correlations with confidence intervals, corrected with the Benjamini and Hochberg procedure and repeated without the outlying archer.

There are eight archers and twelve arrows each, in a single morning, indoors. The authors themselves write that their figures should be taken as estimates for planning larger studies, not as demonstrated associations: the confidence intervals are very wide, and several coefficients change sign when a single archer is removed. The strongest association, between score and group area, measures the same thing twice, because arrows grouped more tightly around the centre score more.

The instrument measures how far the bow moves, not where it points. The midpoint of three markers can move a few millimetres vertically without the aim changing much, and the aim can move a good deal across the target with a small rotation of the bow that barely shifts that point. The same applies to the longitudinal axis: a bow moving a few millimetres towards the target does not shift the aim, and over the window analysed the distance between the drawing hand and the grip kept increasing, so that displacement may also contain the push of the final draw. It is not clear how amplitude was calculated, whether as maximum excursion or otherwise, and there is no test comparing the three axes with one another.

Some figures do not match between the text and the tables: the group area of the outlying archer is 384.1 square centimetres in the text and 348.1 in the table; in the analysis on seven archers, for longitudinal displacement, the corrected p value is lower than the raw one, which the procedure used cannot produce, and the confidence interval is identical to that of another variable. The last part of the paper recommends working on longitudinal stability and moving up in draw weight, but neither suggestion rests on a result that survives the correction.

The study is the first three-dimensional measurement of the bow with an optical system on young archers at this level, and the most useful thing it leaves behind is a sense of scale. Over the five seconds before the release, a bow held by a good young archer moves a few millimetres on each axis, typically between three and nine; anyone going beyond fifteen millimetres is also the one with the worst scores and the widest groups. In this sample the largest displacement was the vertical one, not the lateral one, but this is the bow's movement in space, which is not the same as the sight's movement across the target.

On the physical side, the data do not suggest that a stronger grip helps: among archers who are already trained, grip strength does not separate those who score more from those who score less. Higher draw weight goes with a higher score, but in eight young people of different ages, where draw weight also rises with age and experience, and without the association passing the statistical test. Before moving up in pounds to shoot better, a study following the same archers over time would be needed.