Technique of top recurve archers: Shafizadeh 2024 study

Thirty Olympic recurve archers, the top fifteen men and top fifteen women in the world rankings, observed in the public footage of the 2023 competitions: the World Cups in Berlin, Antalya and Hermosillo and the Asian Games. Eighteen shots each, taken from two different competitions and from the closing stages, semi-final, final or bronze match, individual or team. The shot cycle was broken into seven phases and each phase was classified by watching the footage.

In none of the seven phases is there a single form. Stance: eighteen archers square, twelve open, none closed. Bow raise: fourteen come straight up to the target line, sixteen go higher and then come down. Loading: twenty-three draw with a constant motion, seven with a variable motion, that is with a short pause. Full draw: nineteen bring the string along a straight path close to the jaw, eleven along a curved path that moves away from the face halfway. Anchor: twenty-one rest the string at the middle of the lips, nine more laterally, towards the corner of the mouth. Release: twenty-eight let go with relaxed fingers, two open them. String arm follow-through: in twenty-four the hand slides back towards the middle of the neck, in six towards the earlobe. Bow arm follow-through: thirteen move the whole arm, eight the hand alone, nine alternate between the two.

The frequency test reaches significance in four phases, loading, anchor, release and string arm follow-through, that is the ones in which one of the two forms is distinctly more widespread than the other.

A cluster analysis then divides the thirty into two profiles: twenty-three with a common style and seven with what the authors call a more complex style. The seven are set apart by three things, the extra movement in raising the bow, the change of pace with the short pause in loading, and the bow follow-through made with the whole arm. Those seven archers are all women, seven of the fifteen in the sample.

The authors conclude that the idea of a single model of elite technique does not hold, and that what is seen at the top are individual, stable solutions adapted to each archer's constraints.

The archers were chosen non-randomly from the world recurve rankings, consulted in May 2024, on the basis of overall performance in 2023, limited to the top fifteen of each sex. Approval from the university ethics committee. The footage comes from the World Archery channel on YouTube, public and requiring no permission, and was analysed without altering either the frame rate or the resolution.

The notational system was built in the NacSport software, with a panel of buttons defined on the phases of the shot cycle. The model used has seven phases and derives from a modification of the KSL method: stance, bow raise, loading, full draw, anchor, release, follow-through. The bow arm was observed only during the bow raise, the string arm during loading, full draw, anchor and release; the follow-through in both arms, and the stance only in the positioning of the feet.

Reliability was tested on the same observer: six archers chosen at random from those at the Berlin World Cup, three shots each, with the same procedure repeated twenty-four hours later, and a Kappa coefficient of 0.91. The authors state that, the study being exploratory, the tagging panel was revised whenever further action groups emerged during collection.

The analysis proceeds in three steps. First the types of action in each phase, with frequencies and percentages, compared with a one-way chi-squared test. Then a cluster analysis, to classify the archers by the styles observed and build a profile of them. Finally, to check the allocation to groups, two-way chi-squared tests on the same variables, giving 8 for the bow raise, 30 for loading, 5.28 for full draw, 3.91 for the anchor and 11.95 for the bow follow-through, all with p below 0.05. All in SPSS version 23, with 95 per cent intervals and two-tailed tests.

No score was collected: arrow values, match results and shot times do not enter the analysis. For the timings the authors state the reason, the quality of the public footage, which did not allow them to be measured.

Among the competitions analysed the article names the Asian Games in Thailand: the 2023 edition was held in Hangzhou, China, from the first to the seventh of October.

The test used does not answer the question it is cited for. The one-way chi-squared on a phase compares how evenly the two forms are spread, not whether archers differ from one another. Saying that the release reaches significance means only that twenty-eight against two is a long way from half and half; saying that the stance does not reach it means that eighteen against twelve is close to half and half. Neither statement concerns technique, and in the article's narrative the two words, significant and non-significant, end up sounding as though they did. The result that matters, that at least two forms coexist in every phase, is a description of frequencies and needs none of those tests.

The categories were born from looking at the data. The authors write that, the study being exploratory, the tagging panel was revised whenever further action groups emerged: the classes were therefore built during the coding, and then the frequencies of those same classes were put to a test. The reliability check, giving a Kappa of 0.91, was done on six archers and three shots each at one point in the work, and if the panel kept changing afterwards, that number does not cover the final version. It is moreover reliability of the same observer twenty-four hours apart, not between different observers: for a scheme made of judgements by eye on footage, agreement between two different people would count for more than the repeatability of one.

The two styles are then confirmed with chi-squared tests on the same variables the clusters were created from. Finding that the two groups differ on the variables that defined them is guaranteed by construction, so that step is not an independent check, and the article presents it as one.

The complex style is entirely superimposed on sex. The seven archers displaying it are seven of the fifteen women, and no men. With these data, two styles of Olympic recurve and a difference between men and women in three phases cannot be told apart, and the authors, who note the fact, explain it through morphology and strength without naming the simpler reading.

There is no score. Eighteen shots per archer taken from medal matches, where the value of every arrow is in the footage, and no value was recorded: nothing links a style to accuracy, to the result of the match or to consistency. So the study shows that the styles exist, not that one works better than another, nor indeed that they make no difference.

On the quality of the material the authors are honest by halves. They state that the public footage, unaltered in frame rate, cost them the timing analysis, and they do not extend the same concern to the categories they did code after all. Telling a string at the middle of the lips from a string moved towards the corner of the mouth, or fingers relaxing from fingers opening, is exactly what calls for a high frame rate and a suitable camera angle: the twenty-eight against two on the release may be saying how much the broadcast camera could see.

The eighteen shots also come from the closing stages, semi-finals, finals and bronze matches, that is from the heaviest shots of the tournament, and they mix individual with team events, where rhythm and time management differ. Neither point is discussed. And the sample is the absolute top, the first fifteen per sex chosen without any draw, which the authors do list among their limits: below that level the styles might be more numerous still.

A few incidental slips remain: the 2023 Asian Games, given as being in Thailand, were held in Hangzhou, and two works are cited in the text with a different year from the one they carry in the bibliography.

This is the study to hold up to anyone who says there is a right way to shoot. Thirty of the best archers in the world, watched in World Cup finals, and in all seven phases of the shot at least two forms coexist: those who raise the bow straight to the line and those who go up and come down, those who draw in a straight line close to the jaw and those along a curve that moves away, those who anchor at the middle of the lips and those towards the corner of the mouth, those whose bow arm follow-through moves the whole arm and those who move the hand alone. The stance splits eighteen to twelve between square and open, and nobody uses closed.

Two things, though, are nearly unanimous, and those do serve as a reference: twenty-eight out of thirty let the string go with relaxed fingers rather than opening them, and twenty-four out of thirty carry the hand back towards the middle of the neck in the follow-through. Where almost the whole of the top does the same thing, it is reasonable to treat it as the reference form; where it splits down the middle, it is not.

For a coach the working message is to be careful about what gets corrected. The bow raise, the path of the draw and the bow arm follow-through are, in these data, where the top of the sport divides: correcting them towards a single model means correcting towards one of the two halves, with nothing here saying that half shoots better. And on that the study is silent by construction, because no score was collected.

One thing should be read with care. The style the authors call complex, with the extra movement in raising the bow and the pause in loading, concerns seven archers who are all women. With seven people out of thirty and sex separating them exactly, it cannot be told whether this is two styles of recurve or a difference between men and women in three phases: in the registry's materials it should be cited as the second, not the first.

The registry keeps this work as the most representative description available of technical variety at the top, because it is made in competition rather than in a laboratory, and as evidence against the existence of a single elite technique; not as a measure of what produces a score.