Skip to main content
Formal Specification

Quail Fighting — Formal Specification

Afghanistan · animal-contest · Confidence: low

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 5
# Bōdana-jangī — AI-optimized specification ## General idea An avian territory-contest duel between two male quails placed in a bounded circular ring. Handlers select tactical urging stances (whisper, snap, clap) that modify their bird's aggression and stamina expenditure, while an underlying stochastic combat resolver computes peck damage, pushback distance, and morale checks until one bird flees, submits, or is knocked out of the ring. ## Formal model - **Abstraction taken:** The continuous physical pecking, flapping, and territorial posturing of live quails is replaced with discrete rounds of simultaneous tactical stance selection followed by a stochastic duel resolver that updates bird stamina, ring position, and morale state. - **Player/agent count:** 2 handler agents (Player 1 and Player 2), each controlling 1 quail entity. - **Turn model:** Discrete simultaneous rounds. Each round represents approximately 5–10 seconds of active pecking and sparring. ## Formal rules R1. **Setup:** Each agent enters a bird entity with initial attributes: `stamina = 100`, `morale = 100`, and `position = 0` (ring center on a 1D axis from `-5` [Agent 1 boundary] to `+5` [Agent 2 boundary]). [gap-fill: numerical range and starting attributes are formalized for simulation]. R2. **Round Action:** Each round, both agents simultaneously select one urge command: `AGGRESSIVE_SURGE` (+attack power, -stamina), `STEADY_PRESSURE` (balanced attack/defense), or `DEFENSIVE_PERCH` (+morale recovery, -pushback resistance). R3. **Combat Resolution:** The engine computes the net force delta `ΔF = (Bird1_Power - Bird2_Power) + ε`, where `ε ~ Normal(0, σ)`. - If `ΔF > 0`, Bird 2 takes `|ΔF|` damage to stamina and is pushed toward its boundary by `ceil(ΔF / 10)` units. - If `ΔF < 0`, Bird 1 takes `|ΔF|` damage to stamina and is pushed toward its boundary by `ceil(|ΔF| / 10)` units. - [gap-fill: formula and pushback scaling are formalized]. R4. **Morale & Refusal Check:** If an agent's bird stamina drops below 30, it performs a morale check `P(submit) = 1.0 - (morale / 100)`. If it fails, the bird enters `taslim` (submission chirp). R5. **Refusal Counter:** If both birds choose defensive stances and zero net movement occurs for 3 consecutive rounds, a refusal counter increments. Upon 3 consecutive refusals, the bout resets to center with a stamina penalty. [gap-fill: refusal threshold and reset penalty]. R6. **Victory Conditions:** A bout ends immediately and Agent A wins if: - Agent B's bird crosses its rear boundary limit (`position > +5` or `position < -5`, representing flight/fleeing `farār`). - Agent B's bird fails its morale check (`taslim` submission). - Agent B's bird reaches `stamina == 0` (physical exhaustion). - Agent B's bird refuses engagement 3 times consecutively while Agent A's bird remains aggressive. R7. **Tie Resolution:** [gap-fill: If both birds simultaneously reach `stamina == 0` or cross boundary limits on the same tick, the bird with higher remaining morale wins; if morale is equal, the match is declared a draw]. ## Edge cases - Tie at the end condition: Resolved via R7 by comparing remaining morale score; if precisely equal, the match resolves as a formal draw. - A player with no legal move: Not applicable; all three urge commands (`AGGRESSIVE_SURGE`, `STEADY_PRESSURE`, `DEFENSIVE_PERCH`) are always legal every round. - Repetition / cycles: Bounded because stamina strictly decreases each round for both birds due to exertion; exhaustion forces a morale check and terminal loss within a finite number of rounds. - Simultaneous claims on the same resource or position: Actions are committed simultaneously and resolved together via the algebraic difference formula in R3; position updates symmetrically along the 1D ring axis. - Player elimination mid-round; turn order repair: Not applicable; the game is strictly a 2-agent duel terminating instantly upon the first victory condition triggered. - Resource exhaustion (pieces, tokens, space, time): Stamina is the sole depleting resource. Total round cap is set to `50` rounds; if no bird has fled or submitted by round 50, the bird with higher remaining stamina wins. [gap-fill: 50-round safety cap]. - Termination guarantee: Guaranteed because non-zero stamina depletion per round ensures stamina reaches 0 or triggers a terminal morale failure within at most 50 rounds. ## State representation - **Persistent state:** - `round_number`: integer (1..50) - `agent1_bird`: `{stamina: 0..100, morale: 0..100, position: -5..+5, consecutive_refusals: 0..3}` - `agent2_bird`: `{stamina: 0..100, morale: 0..100, position: -5..+5, consecutive_refusals: 0..3}` - `match_status`: `ONGOING | P1_WON | P2_WON | DRAW` - **Per-turn ephemeral state:** - `agent1_action`: selected urge command - `agent2_action`: selected urge command - `resolved_delta`: net combat force delta - **Observable information:** Fully observable. Both agents observe all bird positions, approximate stamina bars, and past round history. ## Action space - `select_command(command)`: - `command`: enum `[AGGRESSIVE_SURGE, STEADY_PRESSURE, DEFENSIVE_PERCH]` - Preconditions: `match_status == ONGOING`. ## Playability notes - Single-player-vs-AI: Natural fit. A human agent commands one bird against an AI handler utilizing heuristic state evaluation (e.g., switching to defense when stamina is low, surging when opponent is near boundary). - Group games (n>4): Not applicable as a single bout is 1v1. Tournaments reduce to standard single-elimination brackets among multiple agents. - Physical-skill games: Bird physical conditioning and aggressiveness are modeled as static initial weight/power attributes modified dynamically by handler commands. - What is lost in the adaptation: The auditory stimuli (chirping cues), handler tactile familiarity, live bird psychology, and spectator ambiance are abstracted into discrete stochastic state transitions. ## Implementation hints Low complexity. State space consists of two scalar vectors and a position integer. Branching factor is tiny (3 choices per agent = 9 joint action pairs per round). Suggested first milestone: implement the deterministic state updater with fixed stamina loss, then integrate the stochastic normal-distribution force resolver and test termination under random agent policies.

Download as JSON

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

Download JSON
Cite this page

Quail Fighting (Afghanistan). (n.d.). Games of Countries. https://gamesofcountries.com/games/af-afghanistan/bodana-jangi

@misc{gamesofcountries-af-afghanistan-bodana-jangi,
  title = {Quail Fighting (Afghanistan)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/af-afghanistan/bodana-jangi}
}