Skip to main content
Formal Specification

Jagara Boomerang Accuracy Contest — Formal Specification

Australia · dexterity-throw · Confidence: low

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 2
# Buran — AI-optimized specification ## General idea A boomerang accuracy contest scored on *return* rather than distance. Two features make it structurally unusual: the throw must pass beyond a far boundary to count, and the score is measured where the thrower catches it — so the catch is part of the score, not merely a flourish. ## Formal model - **Abstraction taken:** the flight becomes an outbound leg that must exceed 30 units and a return leg resolving to an interception point, drawn around the thrower's aim with spread widening in wind. Aerodynamics are not simulated. - **Player/agent count:** teams of three; players throw individually. - **Turn model:** sequential, five throws each, in rounds by boomerang hand. ## Formal rules R1. Concentric circles: a small circle with a centre marker, a 3-unit throwing circle, an 8-unit catching circle, and a 30-unit outbound boundary. R2. The thrower throws from inside the 3-unit circle. R3. The throw is void unless it travels beyond the 30-unit circle. R4. The thrower waits inside the 8-unit circle. A catch made outside it does not count. R5. The return resolves to an interception point drawn around the thrower's position, with spread increasing in wind strength. R6. On catching, the boomerang is placed on the ground directly below the catch point; that spot is the measured result. R7. Score is the distance from the centre marker to that spot — lower is better. R8. Each player has five throws. R9. Round 1 uses the right-hand *dunimgi*; round 2 the left-hand *watungi*. Boomerang size is chosen by wind: *dikir* in high wind, *mwoirnin* in light. R10. The team throwing closest to the centre wins. [gap-fill] How three players' results combine into a team score is not stated; the spec uses the team's single best result. R11. [gap-fill] A tie is re-thrown; no tie-break is attested. ## Edge cases - **Tie at the end condition:** possible on measured distance; R11 imposes a re-throw and flags it. - **Player with no legal move:** a thrower who has used five throws waits; a void throw under R3 or an uncaught return simply scores nothing. - **Repetition / cycles:** none. Throws strictly decrease and rounds are fixed by R9. - **Simultaneous claims:** throwers act in sequence and each catches their own boomerang (R7), so two players never contest one object. - **Mid-round elimination and turn-order repair:** nobody is eliminated; the rotation is fixed across both rounds. - **Resource exhaustion:** five throws per player per round; exhausting them ends that player's round. - **Termination guarantee:** guaranteed — a fixed team size, five throws each, two rounds. ## State representation Persistent: circle radii (3, 8, 30 units), marker position, wind strength, boomerang type in use. Per-player: throws remaining, best measured distance, skill parameter. Per-turn ephemeral: outbound distance achieved, return interception point, resulting measured spot. Information set: fully observable. ## Action space - `throw(aim, force, boomerang_type)` — precondition: actor inside the 3-unit circle, throws remain. - `catch()` — precondition: actor inside the 8-unit circle and the boomerang is returning to the actor's position. - `select_boomerang(type)` — precondition: before a round; constrained by wind under R9. ## Playability notes - **Single-player-vs-AI:** genuinely interesting, because R3 and R7 pull in opposite directions: the throw must be strong enough to clear 30 units, but a stronger throw returns less precisely. That trade-off is the game. - **Group games (n>2):** teams of three with no interaction; reduces to individuals without structural loss, though R10's team scoring then becomes vacuous. - **Physical-skill model:** return-point spread widening with force and wind; plus a catch probability, since an uncaught return scores nothing. - **What is lost:** the equipment knowledge. Four named boomerangs chosen by hand and wind condition is a body of expertise, and reducing it to a `boomerang_type` parameter records the taxonomy while discarding the judgement behind it. ## Implementation hints Implement R3's minimum outbound distance before anything else — without it the optimal throw is a feeble one that drops on the marker, and the game is trivial. With only two gap-fills and fully specified circle geometry, this is among the most completely documented games in the Australian set and a good candidate for a faithful implementation.

Download as JSON

Download the full game corpus entry for this game as structured JSON.

Download JSON
Cite this page

Jagara Boomerang Accuracy Contest (Australia). (n.d.). Games of Countries. https://gamesofcountries.com/games/au-australia/buran

@misc{gamesofcountries-au-australia-buran,
  title = {Jagara Boomerang Accuracy Contest (Australia)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/au-australia/buran}
}