Aiming and the final pull: Edelmann-Nusser 2006

The study looks at two things an archer does at the same moment: holding the aim steady and completing the final pull onto the clicker. Seven highly trained archers, all competitors at the Junior World Championships and the German National Championships, each shot 66 arrows at 30 metres indoors, with an optical system fixed to the riser recording the aim point and a strain gauge glued to the clicker to derive changes in draw length.

Aim point movement was summarised in a measure the authors call the hold box: a rectangle built from the standard deviation of the horizontal and vertical components of the trajectory, measured in square score points. Within each archer the hold box shrinks as the shot approaches: from the window 3 to 2 seconds before the shot, to the one 2 to 1 second before, to the final second, with a significant difference for all seven. The archer with the highest mean score goes from 1.793 to 0.475 to 0.315; the one with the lowest from 1.520 to 0.490 to 0.309. And for all seven, in the final second, the hold box on shots scoring 10 is significantly smaller than on shots scoring 8 or less.

Between archers, by contrast, there is no correlation between mean score and mean hold box size. The numbers show it: the best mean score is 9.31 with a hold box of 0.315, while 9.18 goes with 0.295, the smallest of all, and 8.72 goes with 0.309.

Another regularity holds for everyone: in the final second the horizontal width of the hold box is significantly greater than its vertical height. On average it runs from 0.595 to 0.913 score points horizontally against 0.412 to 0.647 vertically. The aim, in other words, swings more sideways than up and down.

On the final pull the results are qualitative. The archer with the highest score drew the arrow back steadily and continuously, without pauses; another systematically split the pull in two, a fast part and then a slow one, on all 66 arrows, and with that quirk had the second highest score. The two archers with the lowest scores showed the most obvious irregularities: one, on his three shots scoring 5, 6 and 6, stopped pulling when the clicker had nearly gone past and let the draw length come back, which archers call shrinking; the other, on his three shots scoring 6, paused without losing tension. For the first of the two, those three shots also had the largest hold boxes of the final second, and leaving out the shots with an irregular pull brought his mean down from 0.59 to about 0.45. For the second this did not happen.

There were seven archers, two men and five women aged 16 to 19, each using their own competition bow. They shot 11 ends of six arrows with no time limit at 30 metres indoors, on an 80 centimetre competition face divided into ten concentric zones, with instructions to shoot for maximum score. The score of each arrow was noted.

For the aim point, a system designed for rifle shooting was adapted: the Noptel ST-2000, normally mounted on a rifle barrel. It is an optical unit with an infrared transmitter and receiver that reads position through reflecting prisms fixed around the centre of the target, with the result sent to a computer at 100 scans per second. A dedicated fastener, designed with the Berlin institute for research and development of sports equipment, allows it to be mounted on the riser in place of the lower stabiliser. Since the system's axis does not point at the centre of the target, a visible laser is mounted on it: the archer shoots three or four arrows with the laser on, the prisms are fixed around the projection point of the laser beam, about 20 centimetres from the centre, the laser is then switched off, and the remaining positioning error is corrected in software by comparing real hits with computed ones. In archery there is no report, but the mechanical vibrations of the riser after the arrow leaves the string produce a similar reaction in the system, with a delay of 15 milliseconds.

The hold box is a rectangle built from the standard deviation of the horizontal and vertical components of the trajectory, and its unit is the square score point. Three hold boxes were computed for each arrow, one for the window 3 to 2 seconds before the shot, one for 2 to 1 second, and one for the final second. The within-archer comparison between the three windows, and the comparison between shots scoring 10 and shots scoring 8 or less, used the Mann-Whitney U test; the comparison between horizontal width and vertical height used the Wilcoxon test for paired samples. The threshold was five per cent.

For draw length the authors built a new system: a strain gauge glued to the clicker, connected to an amplifier by a thin four-metre cable. During aiming the clicker sits only a few millimetres behind the tip of the arrow, so every change in draw length alters the bending line of the blade because of the shape of the arrowhead, and from that bending, allowing for the shape, the course of the draw length can be computed. It is a lighter method than the one used in 1993 by Leroyer, the only earlier work on the subject, which fixed an electromechanical device to the grip and to the limb.

The section on aiming and the section on the final pull do not carry the same weight. The hold box is measured on all 462 arrows and compared with statistical tests; the final pull is described in words, on a handful of selected shots shown in figures, with no summary figures and no tests. The authors are explicit about why: they cannot pinpoint the instant the final pull begins, with a margin reaching 0.6 seconds, and on some shots they cannot distinguish fluctuations in draw length during aiming from the start of the pull. For this reason they decline to tabulate those instants and do not compute hold boxes within the final pull.

Their most quoted conclusion, that a regular and steady final pull is decisive for a high score, rests on that. The comparison is between one archer, the best, and the two weakest, out of seven; the shots shown for the last two were chosen because they scored low, so the coincidence of irregularity and low score is partly built by the selection; and the case of the archer who always split his pull in two and had the second highest score shows that one and the same irregularity, when repeated, can sit alongside a good result. The authors acknowledge this and explain it by training, which is an interpretation.

The sample is seven highly trained but very young people, aged 16 to 19, and five of the seven are women: not a group from which to generalise to archers scoring differently. There is a single distance, 30 metres indoors, far from the Olympic distance. The aim measurement system replaces the lower stabiliser with an optical unit, and the clicker with an instrumented one attached to a four-metre cable: two changes to the equipment whose effect on the shot is not checked. The sampling rate of the trajectory, 100 readings per second, is fine for aiming but coarse for what happens in the final tenth.

Finally the hold box describes the scatter of the aim point, not the point of impact: the link with score is established within each archer and absent between archers, and the authors say so plainly, that the archer with the smallest hold box is not the one with the highest score. A second paper by the same authors, devoted to draw length, uses the same arrows and is listed in the registry as a study with the same data.

The result that matters for an archer is the one within the individual, and it has to be read with its opposite beside it. In the same hands, in the final second, shots scoring 10 had a less mobile aim than shots scoring 8 or less: that held for all seven archers measured. Between archers, though, the same measure says nothing: the one with the steadiest aim was not the one scoring most. It is a caution about a way of watching a video or a screen: comparing your own steadiness with someone else's carries no information, comparing your own shots with each other does.

The second finding is the course over time. The amplitude of the aim point movement shrinks of its own accord as the release approaches, with a clear step between 3 to 2 seconds and the second that follows, and then a further reduction in the final second, even on shots where the final pull starts late. It is therefore not something the archer has to add: it is the way the movement draws in on itself as the three tasks (aiming, pulling through the clicker, holding the bow steady) overlap.

The third is the asymmetry. In all seven the aim swung more sideways than up and down, in the final second and, for five of them, in the one before as well. It is a feature of standing upright in front of a target, not a fault in any one person.

On the final pull the study describes two behaviours recognisable in the two lowest-scoring archers: draw length coming back when the clicker has nearly gone past, which is shrinking, and a pause at constant tension. On the shots shown, both went with scores between 5 and 6 and, for one of the two, with the largest hold boxes. It remains an observation on a few selected shots rather than a measurement, and the third behaviour observed, a pull always split in two by the archer with the second highest score, is a reminder that a repeated quirk is something other than an irregularity.