Matteo Campagna
ITEN
Menu
Menu

A round of thirty-six arrows measures an archer's level reliably enough

4 studies · 41 archers measured

ForContextAgainst20202022202420262022, context, n = 202024, context, n = 202023, context, n = 12020, context
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.

There is little on this question, and what there is comes second-hand. Two Indonesian training studies, from 2022 and 2024, use a round of 36 arrows at 30 m, out of a maximum of 360, as their accuracy test and credit it with the same reliability, 0.825: the first calls it test-retest, the second Cronbach's alpha, which measures the internal consistency of the test rather than its stability from one day to the next. In neither is the coefficient calculated on the study's own sample, twenty young or club archers in both cases: the first does not say where it comes from, and it appeared in a journal with questionable editorial standards; the second cites it from earlier work by the same group, where that figure does not appear. Even taken at face value, a coefficient of 0.825 leaves a margin of error that neither study expresses in points. Two studies that do not measure repeatability across archers give the scale of that margin. A 2020 model, for a top compound archer at fifty metres with the sight set correctly, gives 36 arrows a standard deviation of 2.7 points from one round to the next, enough to mask small improvements. And a 2023 laboratory study shows how wide that margin is for a single archer: for an advanced barebow archer at 50 m, the authors' simulation gives a 30-arrow round a standard deviation of about six points, and a stable average would take between 900 and 1,050 arrows. A round of thirty-six arrows may therefore be enough to separate archers of very different levels, but nobody has measured how repeatable it is for reading differences of a few points. A coefficient that is cited rather than calculated on a study's own sample is not a measurement, and so the status is not studied.

Also phrased as: "36-arrow test"; "30 m accuracy test"; "one round is enough to judge an archer"; "reliability of a shooting test"; "test score and actual level"

In practice

Clubs and selection panels often use a 36-arrow round to decide who is in form and who is not, as if that day's score were a snapshot of the archer's level.

Evidence

  • 2020
    Context Model Published later

    A Monte Carlo simulation of a top-group compound archer at 50 m on an 80 cm face, with the sight set correctly and never moved: over 200,000 sets of 36 arrows the mean score is 350.74 with a standard deviation of 2.70, and over 72 arrows 701.48 with a standard deviation of 3.81. The author concludes that the normal variation from round to round is enough to mask small improvements or degradations. It is a model, not a repeated test on archers.

    The impact of off-centred groups on an archer’s score · James L Park

  • 2022
    Context Archers Published later

    Twenty Indonesian archers drawn from a population of 58, split into two groups matched on the pretest to compare fixed-distance and changing-distance training. The accuracy test is a round of 36 arrows at 30 m; the methods credit it with content validity and a test-retest reliability of 0.825, without saying which sample this was calculated on. Journal with questionable editorial standards.

    The Effect of Fixed and Changing Distance Drilling Exercise Methods on Archery Accuracy Questionable venue · Muhammad Wahyu D. Saing, Suharjana, Ahmad Nasrulloh, Betrix Teofa Perkasa Wibafied Billy Yachsie, Afeb Chesa Arianto

  • 2023
    Context Archers Published later

    One advanced barebow archer, 30 arrows at 50 m on a 122 cm face with each of eleven weight configurations: scores from 228 to 262 out of 300. Using a simulation based on Park's model, the authors estimate a mean of 256 and a standard deviation of 6.15 for a 30-arrow round, and conclude that a stable average would take 30 to 35 ends of 30 arrows, that is 900 to 1,050 arrows per configuration.

    Analysis of preferred weight configuration for barebow shooting style · Francesco Pietra, Philip Loveday, Robert Reid

  • 2024
    Context Archers Published later

    Twenty Indonesian competitive archers aged 15 to 20, eighteen sessions over six weeks of fixed-distance or changing-distance training. The accuracy test is a round of 36 arrows at 30 m, maximum 360, with content validity of 0.935 and Cronbach's alpha of 0.825, figures cited from earlier work by the same group. That work, however, reports a validity of 0.895 and a reliability of 0.944 for its 40 m test. Just before, the methods call the same test a 40 m test.

    Distancia de perforación fija versus cambiante: ¿qué ejercicio es mejor para el tiro con arco de precisión? (Drilling distance fixed vs changing: which exercise is better for accuracy archery?) · Betrix Teofa Perkasa Wibafied Billy Yachsie, Doni Pranata, Yudik Prasetyo, Siis Suhasto, Putri Famelia, Amri Hartanto

Status history

  1. 2020 Not studied Before any check on archers, a model gives the scale of the problem: for a top compound archer at fifty metres, with the sight set correctly, 36 arrows vary from one round to the next with a standard deviation of 2.7 points, enough to mask small improvements or degradations. It says how much noise there is in a round, not how repeatable the test is when archers shoot it, and nobody has measured that yet: the question stays open.
  2. 2022 Not studied The first indication comes from an Indonesian training study of twenty archers: the accuracy test, 36 arrows at 30 m, is presented with content validity and a test-retest reliability of 0.825. The methods do not say which sample the coefficient was calculated on, though, and the journal has questionable editorial standards. A coefficient cited with no source does not measure the repeatability of the test: it is context, and the question is still unmeasured.
  3. 2023 Not studied A laboratory study of barebow weights shows the limit of a single round. An advanced archer shot 30 arrows at 50 m with each of eleven configurations, scoring from 228 to 262 out of 300, and the authors' simulation gives a 30-arrow round a standard deviation of about six points: a stable average would take between 900 and 1,050 arrows per configuration. It does not measure the repeatability of the test across archers, so it is context and the status stays at not studied.
  4. 2024 Not studied A second Indonesian study, of twenty competitive archers aged 15 to 20 and in a journal with no flag against it, uses the same test of 36 arrows at 30 m and credits it with a content validity of 0.935 and a Cronbach's alpha of 0.825. The coefficient is cited from earlier work by the same group, which in fact reports different values for its own test, and it matches the 2022 figure; here, moreover, it is a measure of internal consistency, not of stability over time. It is not a measurement on the study's own sample: it is context, and the status stays at not studied.

Related

Related articles

Research lines

Cite
APA

Campagna, M. (2026). A round of thirty-six arrows measures an archer's level reliably enough (Version 3). In Evidence Registry. Retrieved 16 October 2026, from https://matteocampagna.com/en/registry/competition/36-arrow-test-reliability

BibTeX
@misc{campagna2026_test36frecceaffidabile,
  author = {Campagna, Matteo},
  title = {A round of thirty-six arrows measures an archer's level reliably enough},
  year = {2026},
  howpublished = {Evidence Registry, matteocampagna.com},
  url = {https://matteocampagna.com/en/registry/competition/36-arrow-test-reliability},
  note = {Version 3; accessed 2026-10-16}
}

Other claims in this domain

See all claims on competition and scoring