WEIGHTLIFTING
An independent simulation of Olympic weightlifting: the snatch and the clean & jerk.

WHAT THIS IS
This is not a replica of a video game. The thing being replicated is a SPORT, and the document that
defines it is the International Weightlifting Federation's Technical and Competition Rules &
Regulations 2025 (89 pages, with the Annex of modifications as of 5 November 2025). Every clause
number in this app is that document. The lifter is a planar multibody model written for this app; the
physics is checked against published biomechanics measurements, and the places it disagrees with them
are printed on the page rather than tuned away.

It is not affiliated with, endorsed by or connected to the IWF, the IOC, or any federation or athlete.
The seven rival lifters are invented: their names and bests belong to no real athlete. Their
simulated success rates are checked against published results.

NO THIRD-PARTY CODE
The multibody dynamics, the mesh generator, the WebGL renderer, the rules implementation, the
independent rules oracle and both harnesses were written for this app. There is no library, no CDN,
no external script, stylesheet, font or image. Everything ships in the bundle.

THE DOCUMENTS THIS WAS BUILT FROM
  IWF Technical and Competition Rules & Regulations 2025, and its Annex of 5 November 2025.
  D. A. Winter, Biomechanics and Motor Control of Human Movement, 4th ed., Fig. 4.1 and Table 4.1
    (after Dempster, and after Drillis and Contini).
  Gordon, C. C. et al., 1988 Anthropometric Survey of U.S. Army Personnel (ANSUR).
  Cunanan, A. J. et al. (2020), Sports 8(9):118 - 319 snatches, Vorobyev trajectory types with
    Hiskia's fourth.
  Harbili, E. (2012), J. Sports Sci. Med. 11:162; Harbili & Harbili (2014) on juniors.
  Liu, G. et al. (2018), Heliyon 4:e00658 - bar height normalised by stature.
  Safrushahar, Y., Norhaslinda, R. & Wilson, B. (2002), ISBS - every landmark as a percentage of the
    lifter's own stature.
  Xu, Y. et al. (2024), Life 14(9):1086 - 46 competition jerks, and the bar is not ballistic.
  Wang, Y. & Liu, Y., ISBS - one elite snatch timed event by event, with the arm force in flight.
  Deming, W., Kangwei, S. & Yunde, L. (1993), ISBS - the centre of gravity driven down faster than g.
  Jon, M. H. & Rim, U.-R. (2024), Engineering Reports, doi 10.1002/eng2.13042 - bar oscillation.
  Genakos, C. & Pagliero, M. (2011), Collegio Carlo Alberto WP 196 - 41,550 attempts.
  Flores, F. J. & Redondo, J. C. (2020), Int. J. Advanced Research 8(04):901 - the snatch:C&J ratio.
  Arauz, P. et al. (2026), J. Biomechanics 195:113106 - snatch and clean grip widths.
  Olympedia results 2013218 and 2013263, and the Paris 2024 attempt tables.

PROVENANCE
Every constant and claim is tagged DOCUMENTED / DERIVED / MEASURED / CALIBRATED / RECONSTRUCTED and
listed in js/provenance.js, which the page reads at run time and the page harness recounts from the
shipped file. The tally is 83 documented, 17 derived, 11 measured, 6 calibrated and 12 reconstructed
of 129 items - 64.3 per cent documented, 9.3 per cent reconstructed. One item is flagged as general
reference rather than a document opened for this build.

WHAT THIS MODEL GETS WRONG, STATED PLAINLY
  The barbell does not loop back toward the lifter. Published measurements put the backward loop at
    3.6 to 7 cm; this model's is zero and all its horizontal travel is forward.
  The arm force on the bar in flight is about 700 N against a published 150, and the peak ground
    reaction is 2.5 times system weight against a published 1.4 to 1.8. Both follow from PULL_ASSIST.
  The snatch comes out at 0.95 of the clean & jerk instead of the published 0.821, because in this
    model the clean's ceiling is set by a planar lifter losing balance rather than by the barbell.
  The jerk's split catch cannot be solved by a left-right symmetric model: solving it gives 98.2 per
    cent of stature against a published 87.7, so the published figure is used and the gap reported.
  The catch is adjudicated from the state rather than integrated through.

THE ONE CALIBRATION
PULL_ASSIST = 4.0, applied to the leg joints only. It is the factor by which this model must exceed
published peak joint moments to reach published competition loads, and it is a measurement of the
model's own shortfall in the third pull rather than a property of an athlete.

LICENCE
See LICENSE.txt.