Skip to main content
Formal Specification

Five Stones / Knucklebones (Tiraka) — Formal Specification

Montenegro · dexterity-small · Confidence: high

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: probabilistic-proxyGap-fills: 2
# Piljci — AI-optimized specification ## General idea A traditional five-stones dexterity game where agents progress sequentially through a standardized progression of catching and manipulation stages (*jedinci*, *dvojci*, *trojci*, *četvorci*, *kapija*). On each stage, the agent tosses the lead stone (*mati*) into the air, scoops up the required subset of stones from the ground, and catches the falling stone without dropping or nudging adjacent stones. ## Formal model - **Abstraction taken:** Hand dexterity, aerial toss timing, and spatial stone spread are modeled via a discrete stage counter `stage in [1..5]` and sub-step counter `step`, with success probabilities `P(success) = f(stage_difficulty, stone_clustering, agent_dexterity)`. - **Player/agent count:** 1 to 4 agents. - **Turn model:** Sequential; an agent continues playing until a fault occurs, whereupon the turn passes to the next player. ## Formal rules R1. Setup: Agents track `current_stage[agent] in [1..5]` and `current_step[agent] in [1..4]`, all starting at stage 1, step 1. R2. Stages: - Stage 1 (*Jedinci*): 4 sub-steps, picking 1 stone per toss. - Stage 2 (*Dvojci*): 2 sub-steps, picking 2 stones per toss. - Stage 3 (*Trojci*): 2 sub-steps, picking 3 stones, then 1 stone. - Stage 4 (*Četvorci*): 1 step, picking all 4 stones in one sweep. - Stage 5 (*Kapija*): 4 sub-steps, flicking stones through a hand gate. R3. Turn action: The active agent submits `toss_and_gather(risk_mode: {CAUTIOUS, BALANCED, AGGRESSIVE})`. R4. Resolution: - [gap-fill] Success probability is evaluated: `P_succ = base_prob[stage] * risk_multiplier * dexterity_skill`. - If successful: Sub-step advances. If all sub-steps in a stage are completed, agent advances to the next stage. - If failed (foul): Turn ends immediately. State is preserved at the current stage/step; turn passes to the next agent in rotation. R5. Game end: The first agent to successfully complete Stage 5 and the final palm catch is declared the winner. R6. [gap-fill] Cycle bound: If 50 full rotations pass without a winner, the agent who has advanced furthest in stage/step progression wins. ## Edge cases - Tie at the end condition: Turns are sequential; the first agent to complete the final stage triggers immediate game victory. - A player with no legal move: Not applicable; `toss_and_gather` is always available on an active turn. - Repetition / cycles: Bounded by R6 and monotonic progress retention (agents never lose completed stages on a fault). - Simultaneous claims on the same resource or position: Not applicable; turns are strictly sequential. - Player elimination mid-round; turn order repair: An agent who forfeits is dropped from rotation; turn order skips directly to the next active agent. - Resource exhaustion: The set of 5 stones is permanent and never depleted. - Termination guarantee: Progress is monotonic across turns (successful stages are permanently retained); eventual random success guarantees termination. ## State representation - Persistent state: `progress: map`, `active_agent_idx: int`, `round_rotation_count: int`. - Per-turn state: `toss_outcome: {SUCCESS, FAULT}`, `stones_gathered: int`. - Observable information: Fully observable public game state. ## Action space - `toss_and_gather(risk_mode: {CAUTIOUS, BALANCED, AGGRESSIVE})`: Parameter choice adjusts the success probability against execution speed. ## Playability notes - Single-player-vs-AI: Natural; race against AI profiles with varying motor-dexterity parameters. - Group games (n>4): Sequential rotation scales smoothly to any reasonable party size. - Physical-skill games: Motor dexterity abstracted into stage-specific success distributions. - What is lost: The tactile delight of smooth river pebbles clicking against palm skin and the tension of holding multiple pebbles in one hand. ## Implementation hints Discrete progression state machine. Milestone 1: Implement the 5-stage transition ladder and foul pass-over mechanics. Milestone 2: Implement probability tuning curves for each stage.

Download as JSON

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

Download JSON
Cite this page

Five Stones / Knucklebones (Tiraka) (Montenegro). (n.d.). Games of Countries. https://gamesofcountries.com/games/me-montenegro/piljci

@misc{gamesofcountries-me-montenegro-piljci,
  title = {Five Stones / Knucklebones (Tiraka) (Montenegro)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/me-montenegro/piljci}
}