Kyudo: what Japanese research has measured

Why kyudo has a page of its own

Kyudo is Japanese archery: an asymmetric yumi about two metres twenty long, a target at 28 metres that is either hit or missed, and a shot codified in every step. Eighteen studies in the registry deal with it closely, from 1964 to 2025, mostly in Japanese university clubs. I give it a page because in kyudo release problems have names, and they have been counted and sorted by type as in no other shooting discipline. The engineering of the yumi and the practical question of the archer who feels the arrow go by itself deserve pages of their own, so this one stays with the practice and its numbers.

The four release problems

Tradition distinguishes four: hayake, the release that goes before the aim is complete; motare, the block that stops the arrow leaving when the archer wants it to; biku, a jerk of the arm while aiming; and yusuri, a tremor of the arm during the draw and the aim. In December 2018 a group of neurologists asked the members of the kyudo clubs at two medical schools, in Gifu and Kumamoto, whether they had ever had one. Sixty-five of seventy replied and 41 said yes, 63%, or 71% once the seven who could not say are left out: nearly two practitioners in three, counting episodes they have since got past.

The commonest is hayake, with 35 cases out of 41, and in 26 it is the only problem. Biku and yusuri appear in seven people each, motare in three and always on its own. Biku never appears on its own and, when hayake is there too, usually comes after it, which leads the authors to suggest that it grows out of the effort to hold back the early release.

Of the 35 with hayake, 12 say they have overcome it and 23 say they have not, so two in three still have it when they answer. For the other problems there are fewer than ten cases, so the claim that those who develop a problem often do not get past it rests mainly on hayake. Those who have had one have practised for five years on average, those who have not for two and a half, and it is the only factor that holds once the others are taken into account; but years of practice also measure how much time there has been to run into one. Once they enter the model, the links with injuries and with a heavier bow disappear.

Motare has been looked at more closely. In 2023 the same Gifu group examined nine practitioners with an interview and electromyography, three with the block, three with hayake and three with no problems. In the three with the block the problem was stereotyped and task-bound, eased with the instructor's touch in two cases, and in one the wrist flexors and extensors contracted together from about a second before kai. The authors liken this profile to a task-specific focal dystonia, and the registry keeps it as a described resemblance, based on six cases in two studies and measured with a rubber bow. Hayake is task-bound too, but shows none of the other signs.

In a high-school team many experienced members hold kai for two seconds or less, against the five the instructor asks for, and say that even when they try to hold, the arrow goes by itself. Outside kyudo there is a single count, a survey of British archers in which just under one in two say they have had target panic, without sorting it by type; how target panic shows itself in archers who shoot with a clicker is covered in the article on the subject.

The shot as measured

Kyudo has one of the first electromyographic recordings in archery, from 1971, and the measurements that followed, all of them small, track experience. Among fifteen high-school archers the position of the right elbow between two arrows shifted by less than a millimetre in the most consistent and by more than 36 in a first-year, and variability falls with years of practice. In nine men the muscle module of the second half of the shot relies on the deep abdominal muscles in the four experts and on the external oblique in the others. With an eye tracker, in four students with at least six years of practice the gaze leaves the target as the body takes up the load of the bow, while the beginner's stays put.

Then there is tenouchi, the way of holding the bow in which the left hand twists it around the grip. Nineteen members of a university club, measured with an instrumented bow, applied a torque of about 0.6 newton metres around the bow's axis. That the twist moves the arrow comes from a 1997 model and a 2008 test bench, where on a mechanical launcher the torque on the grip alone, between zero and about 14 newton metres, moved the point of impact at 28 metres from one side of the target to the other. Nobody has measured, in the same shot, the torque an archer applies and where that archer's arrow lands.

The mind and the heart

In 1964 fifteen students in Nagoya imagined their own shot while the electrical response of the skin was recorded: the four experts, from second to fourth dan, showed no clear changes, while most of the eleven less experienced did. In kendo the same study found the opposite link, so the finding that experts react less looks like a feature of kyudo rather than of experience.

In competition the heart rate climbs: in five archers from the Gifu club the highest heart rate of each shot rose above training in all of them, and in most reached around 150 beats per minute in the opening rounds of the match, well above the hundred beats usually recorded while shooting; yet in four of the five accuracy does not drop. It does change a great deal from day to day: in fifteen university archers the hit percentage went from 83.7% on the best day to 61.7% on the worst, and before the good mornings they had slept less, 287 minutes against 350. In thirty-six university archers working memory does not predict hits, and in a computer task archers and kyudo practitioners held back their blinking after a stimulus less than kendo practitioners, 10.6% against 18.2% of trials.

The body of the practitioner

In 1971 a questionnaire sent to 122 members of the Japanese federation, aged 18 to 78, found that about three in four reported at least one complaint, with the shoulder first (62 cases), followed by the lower back (59) and hand and fingers (53). It is one of the samples that put the shoulder at the top, and on the bow hand nearly everyone has a callus, from the base of the thumb to the web between thumb and index finger. Two single cases described by Japanese hospitals, a stress fracture of the first rib on the bow side and the rotated head of the shooting posture in a predisposed neck, show that something can happen, not how often. The broader picture is in the article on injuries.

Where the research is thin

Almost everything comes from university clubs, with small samples, histories rebuilt from memory and conference papers a few pages long. Nobody has followed a group of practitioners from their first year to see when hayake appears and who gets past it, and the same question has never been asked in recurve or compound, where the release is a different movement. Adult practitioners are missing too: the federation sample of 1971 had an average age of 43. For the explanations put forward for a blocked release I refer to the article on why target panic happens.

@misc{campagna2026_enresearchpopulationskyudo,
  author = {Campagna, Matteo},
  title = {Kyudo: what Japanese research has measured},
  year = {2026},
  howpublished = {Research, matteocampagna.com},
  url = {https://matteocampagna.com/en/research/populations/kyudo},
  note = {accessed 2026-10-16}
}