Heartbeat and release in archery: Keast 1990 study

Four high-level Australian archers, two men and two women, shot 240 arrows on a force platform, wired for an electrocardiogram, with a high-speed camera on their fingers. Each arrow was classified good, average or bad by combining the score with the archer's own judgement and the coach's: 164 good, 45 average, 31 bad.

The firmest result concerns when the release falls within the cardiac cycle. All four, regardless of sex, age and bow type, started the first finger movement of their good arrows in the ST phase or in the mid-cycle phase of the electrocardiogram, that is, near maximum contraction or in the pause between beats; with average and bad arrows the movement was scattered across the whole cycle. The length of the beat points the same way: before good arrows the cardiac cycle lengthened measurably, and the lengthening began at least six beats before the release. With average and bad arrows this did not happen, and the cycle sometimes shortened as the release approached.

Postural sway goes in the expected direction but less tidily: the area covered by the centre of pressure tended to grow as arrow quality fell, from 75 square millimetres for a good arrow to 3,326 for a bad one, but arrows of every quality fell right across that range. With bad arrows, sway often exceeded what was measured in the same archer standing relaxed in front of the target with the hands at the sides. The movement contributing most to sway was the lateral one, across the target face. Mean heart rate, by contrast, did not change with arrow quality: it rose with the act of shooting, by 3 beats per minute in the compound archer and by 6 to 25 in the recurve archers, in proportion to the draw weight held at full draw.

Aiming times for good arrows ran between 2.12 and 5.82 seconds for the recurve archers and reached 8.40 for the compound archer; there is no common direction, since in two cases aiming time grew as the arrow got worse and in three it shrank.

The archers were a man shooting compound (22 years old, peak draw weight 60 pounds dropping to 30 at full draw), a man shooting recurve (17, 36 pounds), a woman shooting recurve (21, 35 pounds) and a woman of 51 who shot both recurve with sight and stabilisers (30 pounds) and barebow recurve (34 pounds), so five configurations across four people. All had represented Australia nationally or internationally.

Shooting took place from inside the university biomechanics laboratory, out through an opening, at a target 30 metres away with a 60 cm ten-ring face, so that wind did not enter into the sway. The archer stood barefoot on a force platform, with foot position marked and kept for the whole session. Natural sway was first recorded over ten periods of 30 seconds, standing relaxed in front of the target with the hands at the sides and 30 seconds of rest between periods. Then up to two ends of six arrows to set the sight, and finally 24 consecutive scored arrows in ends of six.

The release was detected by a light clip on the string that closed an electrical circuit as it left; the electrocardiogram came from three chest electrodes. Data were recorded from the moment the archer settled into the final aim until just after the release, over intervals of 4 to 25 seconds, 15 on average. Finger movement was filmed separately, with a camera at 400 frames per second and an electronic timer in frame, and from that film the time of the first discernible finger movement was obtained and carried back onto the cardiac trace. The position of the release was then assigned to one of the five classical phases of the electrocardiogram and compared, by chi-squared, with a random distribution of the same number of arrows. Each archer came back several times: the data reported come from the first and third sessions, between 32 and 100 days apart.

Four archers, one of whom counts twice because she shot in two styles, and 240 arrows: consistency over time is the strength of this work, the size of the sample its most obvious limit. Nothing here says how common the phenomenon is among archers in general.

The arrow classification is partly circular. A good arrow has to score 8 to 10 and carry two favourable judgements; a bad one scores 5 to 7 with two unfavourable judgements. The score ranges overlap across the three categories, and the archer's and the coach's judgements were given knowing where the arrow had landed. What is compared, then, is not a good shot against a poor one by an independent criterion, but a category built partly on the outcome.

The statistical comparison of cardiac phases has a design problem. The five phases of the electrocardiogram are not of equal length: the mid-cycle phase is long, the QRS complex very short. Comparing the number of releases in each phase with a uniform distribution makes the longer phases look crowded by construction, and two of the four phases in the count are precisely the long ones. The result stays interesting because the scatter appears only in average and bad arrows, but the significance level quoted should be read with that reservation.

The timing of the first finger movement was not measured alongside the electrocardiogram: the film was shot on a separate occasion and the figure carried back, that is, with an average correction applied to the traces. Postural sway varies as much between sessions in the same archer as between arrow qualities, and for one archer the area for good arrows in the first session was eight times that in the third. Shooting took place in a laboratory, barefoot, wired up, at a single distance. The authors close by saying that biofeedback could be developed as a training technique: that is a proposal, not something they tested.

This is the study underneath one of the most repeated ideas in precision shooting: that the best shots leave at a consistent point in the heartbeat. In the four archers measured, the finding is there, and it holds three months apart, with good arrows leaving during full contraction or in the pause between beats and the worse ones scattered.

What it does not say is just as useful. The beat is not something to go looking for: none of these archers knew they were shooting at a particular phase, and the phenomenon was found in the traces, not taught. Training to synchronise with the heartbeat was not shown to raise the score, and mean heart rate, the figure a watch shows, did not separate good arrows from bad ones: it rose only in proportion to the poundage held at full draw.

On sway the message is more measured than the way it usually gets reported. Moving more goes with worse arrows, but with a wide overlap, and a good arrow can leave with more sway than the archer shows standing relaxed. The movement that weighs is the lateral one, along the face of the target.