Javelin

Every rule below is quoted from the World Athletics Book of Rules, Book C – C2.1, Competition Rules & Technical Rules, the English edition effective 1 July 2026. That is the edition this game was built against.

The throw

• TR 38.1 — the javelin is held at the grip with one hand only and thrown over the shoulder or upper part of the throwing arm; it may not be slung or hurled, and non-orthodox styles are not permitted.
• TR 38.2 — “A throw shall be valid only if the metal head strikes the ground before any other part of the javelin.” The current edition prints a note saying that references to the “tip” have been removed and replaced by the head, because head shapes vary so much; flat and tail-first landings are still red-flagged, and how small an angle counts is left to the Judges.
• TR 38.3 — until the javelin has been thrown the athlete may not turn completely around so that their back is towards the throwing arc.
• TR 38.4 — a javelin that breaks during the throw or in the air is not a failure; the athlete gets a replacement trial.
• TR 32.14.4 — touching the lines that mark the runway, or the ground outside, is a failure. Only the feet matter.
• TR 32.17.1 — leaving the runway before the implement has landed is a failure. TR 32.17.3 — the first contact on leaving must be completely behind the arc, or behind the lines drawn from its ends; once the implement has landed, being behind the 4 m marks and inside the runway also counts as having left correctly.
• TR 25.17 — one minute for a trial in a throwing event, two minutes for consecutive trials.

The facility

• The runway is at least 30 m long (33.50 m in the top competition tiers, 36.50 m where conditions permit), marked by two parallel white lines 50 mm wide and 4 m apart.
• The throw is made from behind an arc of a circle of radius 8 m, a strip at least 70 mm wide, flush with the ground, with lines at right angles at its ends at least 0.75 m long.
• TR 32.12.2 — the sector lines pass through the two points where the inner edges of the arc meet the runway lines and meet at the centre of the circle: “The sector angle is thus 28.96°”. This game derives that angle from the 8 m radius and the 4 m width and gets °.
• TR 32.16 — a failure if the head, in first contacting the ground, touches the sector line or the ground outside it.
• TR 32.20.2 — the distance is measured from the head’s mark to the inside edge of the arc, along a line to the centre of the circle — so a throw off the axis loses a little. TR 32.19 — recorded to the nearest 0.01 m below.

The competition

• TR 25.6 — eight athletes or fewer, six trials each; more than eight, three trials and then three more for the eight best valid performances. Athletes with no valid trial in the first three rounds throw first in the later rounds. TR 25.6.1 — the later rounds run in reverse ranking order. Note (iii) — a governing body may set the number of trials, provided it is no more than six.
• TR 25.21 — each athlete is credited with the best of all their trials. TR 25.22 — a tie on the best mark is decided on the second best, then the third, and so on. TR 25.23 — if they are still equal, the tie remains: this game shares the place rather than inventing a decider.

TR 38.10 — “Minimum weight and diameter limits for admission to competition and acceptance of a Record (inclusive of the cord grip)”. Every figure in this table is read out of the engine’s own constants, so it cannot drift from the physics.

TR 38.12 — the angle of the point is not more than 40°; the diameter 0.15 m from the tip is at most 80 % of the maximum shaft diameter, and at the midpoint between the centre of gravity and the tip at most 90 %.

TR 38.13 — the tail: the diameter at the midpoint between the centre of gravity and the tail is not less than 90 % of the maximum shaft diameter, at 0.15 m from the tail not less than 40 %, and at the very end not less than 3.5 mm. Those minimums are what the 1986 change added.

TR 38.8 — the cord grip must cover the centre of gravity and may not exceed the average shaft diameter by more than 8 mm.

What this game builds

The profile between the rule’s checkpoints, the mass distribution and the pre-1986 tail are reconstructed: the rules fix the total mass, the centre of gravity, the length and the diameter at a handful of stations, and leave the rest to the manufacturer. Everything reconstructed is listed under About & credits.

On 1 April 1986 the men’s 800 g javelin was re-specified, and the record book was reset. The reason usually given — Uwe Hohn’s 104.80 m of 20 July 1984 — is not the reason the governing body gives.

“The IAAF Technical Committee decided to change the rules for javelin construction because of the increasingly frequent flat landings and the resulting discussions and protests because of attempts declared valid or invalid by competition judges. … In demanding a change of the construction rules however, our primary goal was to achieve an exactly measurable landing of the javelin so that it was no longer completely up to the discretion of the judge on the infield… ever since then the media have incorrectly reported that this throw was the cause for the change in the rules.” Erich Bremicker, member of the IAAF Technical Commission, in IAAF New Studies in Athletics 15:3/4 (2000), 29–31

World Athletics says the same thing in its own 2022 feature: “contrary to popular belief, those record-breaking distances were not the reason”.

What actually changed

• The centre of gravity moved forward. Bremicker: trials in 1982–83 moved it by 2, 3, 4 or 5 cm; 3 cm “was sufficient to guarantee the javelin landing point first”, but the Committee proposed 4 cm to avoid having to change the rules again. World Athletics’ 2022 feature says three centimetres. The two governing-body sources disagree, so this game carries both and reports both.
• A minimum tail diameter was introduced, so that “other adjustments could not be made that would render the shift in the centre of gravity useless”.

What this engine gives

Press the button to run the comparison in your browser: it finds each specification’s own best legal throw at the same release speed.

Run the 1986 comparison

And a second change, far more recent: the 700 g javelin’s specification was amended by Council on 14 August 2023 with effect from 1 April 2025 — overall length 2300–2400 mm became 2400–2500 mm and tip-to-centre-of- gravity 860–1000 mm became 850–990 mm. That is the implement Hatton and Parkes complained about in Athletics Weekly in 2005 as “the 700 has yet to be modified … about half of all throws flat landing even as far out as 60 m”.

A javelin is a slender lifting body whose angle of attack is a state variable. The relationship between the centre of gravity and the centre of pressure decides whether the nose comes down, and the whole coupled state — position, velocity, the body axis, its transverse angular velocity, the axial spin and the roll phase, fourteen numbers — is integrated together at one order with a fixed timestep against an absolute clock.

The aerodynamics

The load on a body of revolution at incidence is split the way Allen and Perkins split it: a potential term distributed as the rate of change of cross-section, and a viscous crossflow term distributed as the local diameter. Both are integrated over the same profile the specification produced, so the centre of gravity and the tail shape enter the pitching moment through the geometry rather than through a fitted constant. The crossflow term is evaluated with each station’s local transverse velocity, which includes the rotation, so the aerodynamic pitch damping arrives with no extra constant at all.

The five remaining constants are calibrated once, against magnetic-suspension wind-tunnel measurements on a full-size 600 g implement (Seo, Okuizumi, Konishi, Kobayashi, Hasegawa and Obayashi, Scientific Reports 13:391, 2023). The men’s implement, in either era, is then a prediction.

That paper’s own headline is worth knowing: the pitching moment of a real modern javelin is positive — nose-up — below about 12° of incidence, so the 1986 design intention of a moment that is nose-down everywhere is not actually met.

What a thrower controls

Release speed falls as the release angle rises, because a steeper delivery shortens the pull: about 0.13 m/s per degree (Hatton, IMA Sport 2007, agreeing with Red and Zogaib, J. Biomechanics 1977). With that trade-off in, the best release angle comes out near the 32–38° real throwers use, not the 45° of a drag-free projectile.

What this is

Javelin is an independent reimplementation of the athletics event, not a port of any existing game. There is no single original author or publisher to credit: the javelin throw is a track and field event, contested at the Olympic Games for men since 1908 (London) and for women since 1932 (Los Angeles), and governed today by World Athletics (founded 1912 as the International Amateur Athletic Federation, the IAAF, and renamed World Athletics in 2019). The modern hollow, aerodynamic javelin was invented by the American thrower Bud Held and developed and manufactured by his brother Dick Held in the 1950s.

What it is faithful to

• The World Athletics Book of Rules, Book C – C2.1, English edition effective 1 July 2026: TR 25.6, 25.6.1, 25.6.2, 25.7, 25.17, 25.19, 25.21, 25.22, 25.23; TR 32 (the runway and the arc), 32.12.2, 32.13, 32.14.4, 32.15, 32.16, 32.17.1, 32.17.3, 32.19, 32.20.2; TR 38.1 to 38.13 and Figure TR38.
• The implement specification for both eras: TR 38.10 as it stands, and the pre-1986 800 g window derived from the sourced 4 cm shift; the 600 g pre-1999 window from the sourced “from 95 to 92 cm”; and the 700 g column as it stood before 1 April 2025, read out of the previous edition.
• The 28.96° sector, derived from the 8 m arc radius and the 4 m runway width rather than copied.

What is reconstructed, and said so

• The diameter profile between the rule’s checkpoints — taken piecewise-linear, which TR 38.9’s “straight or slightly convex” allows, and at the limiting value at each checkpoint.
• The pre-1986 tail, where no minimum diameter applied: a straight taper from the rear of the grip to the 3.5 mm the rules still allow at the very end. With that shape the centre of pressure sits within a few millimetres of the centre of gravity, which is the one independent statement available about the old implement.
• The mass distribution — a metal head of uniform linear density plus a thin-walled shaft — solved to reproduce the sourced mass and centre of gravity. The transverse moment of inertia and the axial one follow from it; the rules fix neither.
• The aerodynamic constants, calibrated to one published wind-tunnel data set at up to 18° of incidence. Beyond 18° the model extrapolates.
• The thrower levels, the failure rates and the run-up braking deceleration. World Athletics publishes no table of how far a club thrower throws or how often anyone steps over the arc.
• The rate at which a javelin breaks under TR 38.4. The rule exists, so the game makes it reachable; the frequency is invented.

How it differs from a real competition

• A run-up is a timing bar, not a run. Six athletes, one round of the competition, no qualification round, no protests, no wind measurement (there is none in the javelin).
• The judge’s discretion under the note to TR 38.2 — “there must be some angle on landing, however small” — is taken as strictly “the nose is below the horizontal”.
• The implement is drawn thicker than it is, or it would be one pixel wide at 80 m.
• The model does not reproduce actual flat landings. It gets the direction right — the pre-1986 implement lands far shallower — but not the 18 % of flat landings the governing body was counting in the 1980s.

Sources

• World Athletics, Book of Rules, Book C – C2.1, Competition Rules & Technical Rules, English edition effective 1 July 2026, and the previous edition (amendments in force from 1 November 2023).
• E. Bremicker, “Why did the senior javelin specification have to be changed?”, IAAF New Studies in Athletics 15:3/4 (2000) 29–31.
• World Athletics, “From Held to Hohn and beyond – the evolution of the javelin”, feature, 8 October 2022.
• K. Seo, H. Okuizumi, Y. Konishi, T. Kobayashi, H. Hasegawa and S. Obayashi, “Measurement of aerodynamic force and moment acting on a javelin using a magnetic suspension and balance system”, Scientific Reports 13:391 (2023).
• L. Hatton, “Optimising the javelin throw in the presence of prevailing winds”, IMA Sport 2007 proceedings; L. Hatton and B. Parkes, “Javelin throwing – the appliance of science”, Athletics Weekly, 30 November 2005.
• H. J. Allen and E. W. Perkins, “A study of effects of viscosity on flow over slender inclined bodies of revolution”, NACA Report 1048 (1951) — the load decomposition the aerodynamics uses.
• The men’s world-record progression and the yearly world-leading marks, from the English Wikipedia article “Javelin throw”.

No trade mark of World Athletics or of any implement manufacturer is used or implied. Released under the MIT licence — see LICENSE.txt and CREDITS.txt.