Matteo Campagna
ITEN
Menu
Menu

The movement parameters measured during the shot account for only part of an arrow's score

9 studies · 154 archers measured

ForContextAgainst20102015202020252022, context, n = 412017, for, n = 392017, context, n = 322026, context, n = 152021, for, n = 112008, context, n = 82025, context, n = 82016, for, n = 42012, for, n = 32007, for, n = 1
Each circle is a study: its position gives the year, its row whether it is for, against or context, its size how many participants it measured. Light circles are studies on other populations or without archers.

How much of an arrow's score the movement parameters account for, model by model

Each row is a model, and the dot is the share of the variation in the arrow's score that it accounts for; whatever is missing to reach a hundred stays outside the model. Only Spratford and Campbell put several archers into the same calculation, 39 archers with 10 arrows each at 70 metres on a force platform: the other eight rows are one archer each. Yi 2007 is a regression over 60 arrows with five shoulder and trunk parameters; with step-by-step selection from eight parameters two remain, both of the shoulder, accounting for 15.6%. For Suwarganda the row is the strongest relationship between score and a single muscle, with the R squared value the authors print above their charts: linear in the first archer, curved in the other two, significant in the second and not in the third, while for the first the study does not say. In the two curved cases the authors put score on the horizontal axis and the muscle on the vertical one, so their R squared shows how closely the curve follows the points and stands for a share of the score only roughly. Quan and Lee build one model for each archer, over 29 to 36 arrows at 30 metres, choosing the variables step by step from nine sensors; adjusted for the number of variables, the same models fall to 49.66, 20.96, 53.91 and 35.54%. Their model on the mean data of the four archers, with R squared close to 1, is left out because it cannot be compared with the others. None of the four studies gives an interval for these shares, and the chart draws none. Data from: Yi et al. 2007, Suwarganda et al. 2012, Quan and Lee 2016 and Spratford and Campbell 2017.
The numbers in the figure
Study or groupValue (% of the variation in score accounted for (R squared))IntervalnStance
Yi 2007, world number two, five parameters17.91for
Suwarganda 2012, archer 1, right trapezius221for
Suwarganda 2012, archer 2, variability of the left deltoid321for
Suwarganda 2012, archer 3, variability of the left triceps8.41for
Quan 2016, archer 1, 10 variables64.041for
Quan 2016, archer 2, 3 variables27.931for
Quan 2016, archer 3, 11 variables72.021for
Quan 2016, archer 4, 5 variables45.621for
Spratford 2017, three variables4239for

Every study that has tried to put a figure on it finds the same thing: even when several measures are combined, the share of an arrow's score that movement accounts for stays a minority, or varies widely from one archer to the next. The strongest sample, 39 elite recurve archers, reaches 42% with three variables, namely clicker reaction time, draw weight and sway speed after release; in the world number two archer five shoulder and trunk parameters, over 60 arrows, account for 17.9%; in 3 Malaysian national archers muscle activity does not exceed a third. The picture has two limits. These are associations from a single session, on small samples and often in regional journals; and the share accounted for depends on how the sum is done, because in 4 young female archers models built archer by archer, with up to eleven variables, reach between 28 and 72%, and in 11 Japanese archers bow tremor tracks each archer's total score (R = −0.74) but not the score of the single arrow. The most recent study, of 8 elite French archers, does not calculate a share but shows what is left out: shoulder and elbow angles and chest movement do not separate arrows scoring 8, 9 and 10 in any archer, whereas some muscles, the centre of pressure and clicker reaction time do.

Also phrased as: "biomechanics explains the score"; "sensors tell you why the arrow went out"; "how much the measured technique matters for the score"; "shot measurements predict the arrow"

In practice

With sensors, force platforms and cameras ever more common, archers hear that if the shot is measured well enough it becomes possible to see why one arrow goes into the ten and another does not, and so to correct it.

Evidence

  • 2007
    For Archers Published later

    One recurve archer ranked number two in the world, 60 arrows at 70 m: five shoulder and trunk kinematic parameters in a multiple regression account for 17.9% of the arrow score, with individual contributions of 7.7, 0.1, 8.5, 0.7 and 0.9%. Trunk angle during back tension does not predict the score (β = 0.092, p = 0.520; on its own R squared 0.1%, p = 0.81), and the consistency of the vertical shoulder angle does not follow the score (coefficient of variation 7.34% on 8s, 8.91% on 10s, 15.27% on 9s).

    Evaluation on Motion Features of the World's Second Archer during Back-Tension in Archery · J.-H. Yi, C.-K. Hah, J.-S. Ryu, K.-C. Kim

  • 2008
    Context Archers Published later

    8 Korean university recurve archers, 3D kinematics at 200 Hz, 12 arrows each: the coefficient of variation of the bow's vertical displacement in the follow-through (0.05-0.14) is not linked to mean score (8.58-9.50), at either p < 0.05 or p < 0.10. The same paper finds that mean shoulder speed, calculated by score level from 7 to 10, lines up with score with R squared 0.874 at p < 0.10: it is a line through four means, not a measure of how much the shoulder accounts for the single arrow.

    Evaluation on Kinematic Factors Affecting Scores of Olympic Round Game during the Follow Through in Archery · Chong-Ku Hah, Jae-Hun Yi

  • 2012
    For Archers Published later

    3 Malaysian national archers, 29-36 arrows each at 70 m, electromyography and radar on the arrow: the best fits between muscle activity and score have R squared 0.22 (right trapezius), 0.32 (variability of the left deltoid) and 0.084 (variability of the left triceps), and 0.19-0.35 with arrow speed. In the slowest archer (198 km/h) the score follows the level of activity, in the two fastest (219 and 220 km/h) its variability: the share accounted for does not exceed a third and changes in kind from one archer to the next.

    Influence of Muscle Activity on Shooting Performance in Archery: Preliminary Findings · Edin Suwarganda, Ruhil Razali, Barry Wilson, Ahmad Pharmy

  • 2016
    For Archers Published later

    4 middle-school female archers aged 14-16, accelerometers on 9 segments, 133 shots out of 144 at 30 m: the movements of the two seconds before release account for 64.04, 27.93, 72.02 and 45.62% of each one's score (significant models, p between 0.0012 and 0.016), with 10, 3, 11 and 5 variables entering the model. Higher shares than elsewhere, but obtained with step-wise regressions archer by archer on thirty or so shots, and different from one to the next.

    Relationship between Aiming Patterns and Scores in Archery Shooting · ChengHao Quan, Sangmin Lee

  • 2017
    For Archers Published later

    39 elite recurve archers from four countries, 10 arrows at 70 m on a force platform: clicker reaction time, draw weight and peak sway speed after release together account for 42% of the shot score (r squared 0.42, F = 8.53, p < 0.001), with the equation score = 8.772 − 15.955 × clicker time + 0.098 × draw weight − 0.005 × peak sway speed. Sway varies by 51% between archers and is worth 1.20 points across the observed range, against 0.54 for the clicker and 0.50 for draw weight. Most of what moves the score of a single arrow stays outside the model.

    Postural stability, clicker reaction time and bow draw force predict performance in elite recurve archery · Wayne Spratford, Rhiannon Campbell

  • 2017
    Context Archers Published later

    32 young Malaysian archers aged 13 to 20, six arrows at 50 m, sensors on the pelvis, bow and forearms: a cluster analysis splits them into 20 higher scorers and 12 lower scorers, and a discriminant analysis recognises the group in 83.38% of cases with three variables and 90.63% with two. It is not a share of the arrow's score: the groups come from the same measures, score included, that then recognise them, and the count concerns archers, not individual arrows.

    Analysis of biological and mechanical related performance parameters of Malaysian senior youth archers · Zahari Taha, Mainul Haque, Rabiu Muazu Musa, Mohamad Razali Abdullah, Ahmad Bisyri Husin Musawi Maliki, Siti Musliha Mat-Rashid, Norlaila Azura Kosni, Aleesha Adnan

  • 2021
    For Archers Published later

    11 Japanese university recurve archers, accelerometer on the bow, 36 arrows at 50 m: tremor while aiming estimated from the accelerometer tracks each archer's total score (R = −0.74, p < 0.01; line score = −498 × tremor + 320), but with the score of the single arrow there is no significant correlation. Even at zero tremor the line stops at 320 points out of 360, about 8.9 per arrow, and the authors list among the missing factors forearm control at release, the passage through the clicker, concentration in the follow-through, arrow straightness and the wind.

    Archery Skill Assessment Using an Acceleration Sensor · Takayuki Ogasawara, Hanako Fukamachi, Kenryu Aoyagi, Shiro Kumano, Hiroyoshi Togo, Koichiro Oka

  • 2022
    Context Review Published later

    A systematic review of 41 studies on performance assessment in archery: study design is the weakest quality category, many papers have fewer than 5 participants, and keeping only the studies with a high quality score leaves the proportion between those finding a parameter important and those not finding it essentially unchanged. It does not calculate the share accounted for, but it shows what the others' calculations rest on.

    Performance assessment in archery: a systematic review · Eleonora Vendrame, Valeria Belluscio, Luigi Truppa, Lorenzo Rum, Aldo Lazich, Elena Bergamini, Andrea Mannini

  • 2025
    Context Archers Published later

    8 high-level French recurve archers, 549 arrows at 70 m split into arrows scoring 8 or less, 9 and 10: drawing-shoulder abduction, drawing-elbow flexion and range of chest movement show no significant difference between the three levels, either in the group or in any of the 8 archers, while aiming or after release (Kruskal-Wallis, threshold p < 0.05). At group level, by contrast, seven muscles, three centre-of-pressure parameters and clicker reaction time separate the arrows while aiming, and four muscles and three centre-of-pressure parameters in the follow-through. The study does not calculate how much of the score they account for.

    A multi-parameter investigation of elite archery: distinguishing individual and shared performance strategies during the aiming and follow-through phases · Raphaël Jacquot, Khaireddine Ben Mansour, Kévin Bouillet, Jean-Philippe Jehl, Gérôme Gauchard

  • 2026
    Context Archers Published later

    A doctoral thesis gathering the same 8 elite archers and the same 549 arrows as the 2025 paper: no kinematic variable separates the arrows in either phase, in the group or for any individual archer, whereas some muscles, the centre of pressure and clicker reaction time do so at group level. It does not calculate a share accounted for. It counts as the same study as in 2025.

    Analyse multidimensionnelle de la performance en tir à l'arc · Raphaël Jacquot

Status history

  1. 2007 Reported The first calculation concerns a single archer, the world number two: five shoulder and trunk kinematic parameters, combined in a regression over 60 arrows at 70 metres, account for 17.9% of the variation in score, and trunk angle on its own does not predict it at all. It points the claim's way, but it is a single case in a regional journal: reported.
  2. 2008 Reported In the same journal, in 8 Korean university archers, the variability of the bow's vertical movement in the follow-through is not linked to mean score. The same paper finds shoulder speed lining up with score, but on four means by score level and at p < 0.10, which says nothing about how much of the single arrow it accounts for. Context, and the status stays reported.
  3. 2012 Reported Three Malaysian national archers, measured arrow by arrow with electromyography: muscle activity accounts for at most a third of the variation in score (R squared between 0.08 and 0.32), and the muscle that matters changes from one archer to the next. A second small case pointing the same way: reported.
  4. 2016 Reported In four female archers aged 14-16, with accelerometers on nine segments, the movements of the two seconds before release account for 64, 28, 72 and 46% of each one's score. These shares are higher, but they come from models built archer by archer, with up to eleven variables over thirty or so arrows, and they vary widely from one to the next. Four cases in a regional journal do not move the status.
  5. 2017 Reported The first large sample: in 39 elite recurve archers, clicker reaction time, draw weight and sway speed after release together account for 42% of the score of the single arrow, and of the three it is sway that varies most between archers and is worth most points. The picture firms up, but it remains a cross-sectional association from a single session: reported. In the same year, in 32 young Malaysian archers, a discriminant analysis recognises the higher and lower scorers in eight or nine cases out of ten, but the groups are built from the same measures that recognise them and the count does not concern the arrow: it is context and moves nothing.
  6. 2021 Reported An accelerometer on the bows of 11 Japanese university archers separates levels well: tremor while aiming tracks each archer's total score (R = −0.74), but with the single arrow there is no significant correlation, and even at zero tremor the line stops at 320 points out of 360. It is another association, and the status stays reported.
  7. 2022 Reported A systematic review of 41 studies on performance in archery shows what these calculations rest on: design is the weakest part, many papers have fewer than five participants, and keeping only the best studies leaves the proportion between those finding a parameter important and those not finding it unchanged. It adds no sample, and the status stays reported.
  8. 2025 Reported From 8 elite French archers, 549 arrows at 70 metres, comes a study that does not calculate a share but looks at which measures separate arrows scoring 8, 9 and 10: drawing-arm shoulder and elbow angles and chest movement do not, either in the group or in any archer, whereas some muscles, the centre of pressure and clicker reaction time do. It shows what is left out, not how much it weighs: it is context, and the status stays reported.

Related

Related articles

Research lines

Cite
APA

Campagna, M. (2026). The movement parameters measured during the shot account for only part of an arrow's score (Version 2). In Evidence Registry. Retrieved 16 October 2026, from https://matteocampagna.com/en/registry/technique/movement-parameters-and-score

BibTeX
@misc{campagna2026_parametrimovimentopunteggio,
  author = {Campagna, Matteo},
  title = {The movement parameters measured during the shot account for only part of an arrow's score},
  year = {2026},
  howpublished = {Evidence Registry, matteocampagna.com},
  url = {https://matteocampagna.com/en/registry/technique/movement-parameters-and-score},
  note = {Version 2; accessed 2026-10-16}
}

Other claims in this domain

See all claims on technique and biomechanics