Skip to main content
Formal Specification

Chiba (Mock-Combat Tournament, Clump Point) — Formal Specification

Australia · stick-team · Confidence: low

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: discretizedGap-fills: 7
# Chiba — AI-optimized specification ## General idea A continuous two-sided missile exchange with no possession phase: both teams throw and retrieve at once, so the state is a constantly shifting pool of live missiles rather than a turn-based contest. ## Formal model - **Abstraction taken:** the court becomes a bounded zone split into two team territories; each thrown missile becomes a projectile with a flight delay and a hit-probability check against its target, rather than simulated physics. - **Player/agent count:** 2 teams; team size undocumented, [gap-fill] assume 6-10 per side by analogy with prun/puloga. - **Turn model:** ticked and simultaneous — all agents on both teams may throw or move each tick. ## Formal rules R1. The court is divided into two team territories; out-of-bounds zones (service/front-court lines) are marked and off-limits while play continues. R2. Each tick, an agent holding a missile may `throw(target_direction)` toward the opposing territory, or `move(direction)` to reposition or retrieve. R3. A thrown missile arrives after a flight delay and hits its target agent with a probability that decreases with distance and increases with the thrower's skill. [gap-fill] no accuracy model is given in the source. R4. A missile landing without hitting anyone becomes retrievable by any agent who reaches it; [gap-fill] first agent to occupy its landing cell picks it up. R5. [gap-fill] A hit agent is marked contacted for scoring purposes (R6) but is not removed from play, since the source describes continuous play rather than elimination. R6. Score by one of the agreed variant methods: total contacts made (Variant A) or missiles remaining in the opposing territory after a fixed time (Variant B). [gap-fill] the base card does not state which scoring method is the default; both are explicit variants in the source. R7. The team with the higher score at the fixed time limit wins. ## Edge cases - **Tie at the end condition:** equal scores at time-up are possible with either scoring method; [gap-fill] resolved by sudden-death overtime, since the source gives no tie-break. - **Player with no legal move:** an agent holding no missile and out of retrieval range has only `move()` available — legal and expected. - **Repetition / cycles:** none structurally — missiles are continuously exchanged, not cycled through a fixed sequence. - **Simultaneous claims:** two agents reaching a landed missile on the same tick — [gap-fill] resolved by nearest-agent priority, since the source does not address simultaneous retrieval. - **Mid-round elimination and turn-order repair:** no elimination under R5 — hits are scored, not removed, so no repair is needed. - **Resource exhaustion:** the pool of missiles in play is fixed at round start; none are created or destroyed, only relocated by throws and retrievals. - **Termination guarantee:** guaranteed by the fixed time limit (R7). ## State representation Persistent: court bounds, out-of-bounds zones, time limit, scoring method in use. Per-agent: position, territory, missiles held. Per-turn ephemeral: in-flight missiles with target and arrival tick, contact/score tally. Information set: fully observable — an open exchange with no hidden information. ## Action space - `throw(target_direction)` — precondition: actor holds a missile. - `move(direction)` — precondition: none, always legal. - `retrieve()` — precondition: actor occupies a cell with a landed missile. ## Playability notes - **Single-player-vs-AI:** workable as one agent within a team of AI teammates against an AI-controlled opposing team. - **Group games (n>4):** scales directly with the gap-filled team size; the continuous-exchange structure does not require any reduction. - **Physical-skill model:** throw accuracy as a function of distance and thrower skill (R3) stands in for real throwing proficiency. - **What is lost:** the war cries, chants, taunts and dances the source records as part of the contest, and — because team size and traditional missiles are undocumented for this specific card — a fully attested traditional baseline to calibrate the model against; this spec leans more than usual on the gap-filled defaults. ## Implementation hints Because team size and scoring default are both undocumented here (unlike the better-specified prun/puloga cards), implement against prun's numbers first and treat chiba as a naming variant rather than a separately tuned ruleset. The interesting mechanic is the flight-delay/accuracy trade-off in R3 — get that right before scoring.

Download as JSON

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

Download JSON
Cite this page

Chiba (Mock-Combat Tournament, Clump Point) (Australia). (n.d.). Games of Countries. https://gamesofcountries.com/games/au-australia/chiba

@misc{gamesofcountries-au-australia-chiba,
  title = {Chiba (Mock-Combat Tournament, Clump Point) (Australia)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/au-australia/chiba}
}