Skip to main content
Formal Specification

Malaysian Five-Stone Jacks — Formal Specification

Malaysia · dexterity-small · Confidence: low

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: probabilistic-proxyGap-fills: 2
# Batu Seremban — AI-optimized specification ## General idea A turn-based multi-stage dexterity progression game. Agents attempt a sequence of 5 distinct grabbing challenges and a culminating hand-flip weighing maneuver. Success at each stage depends on hand coordination parameters, spatial stone scatter geometry, and risk-management decisions when choosing the mother stone and pickup paths. ## Formal model - **Abstraction taken:** The physical mid-air toss, hand sweep, and catch are modeled as an execution probability P(success|stage,distance,skill) where difficulty increases non-linearly with the number of stones scooped simultaneously and their spatial spread on the 2D surface. - **Player/agent count:** 2 to 5 agents, taking sequential individual turns. - **Turn model:** Sequential individual attempts. An agent continues advancing through stages until a fault occurs, at which point the turn passes to the next player. ## Formal rules R1. State tracking per agent i: Current stage Si∈[1..5], current score Ci∈ℕ. R2. Stage definition: - Stage 1: 4 separate pickups of 1 stone each. - Stage 2: 2 pickups of 2 stones each. - Stage 3: 1 pickup of 1 stone, 1 pickup of 3 stones. - Stage 4: 1 pickup of 4 stones simultaneously. - Stage 5 (Timbang): Back-of-hand toss-and-catch scoring. R3. Stage execution: At stage s∈[1..4], stones are scattered on a 2D plane: coordinates (xk,yk) for k∈[1..5]. - The agent chooses one stone as *ibu*, and selects the pickup grouping sequence. - For each pickup group G, success probability is: Ppickup=Skilldexterity×exp⁡(−λ⋅spread(G))×γ|G|−1 where spread(G) is the maximum pairwise distance among stones in group G. R4. Resolution: - If all pickups for stage s succeed, the agent advances to Si←s+1 and immediately plays the next stage. - If any pickup fails (or an unselected stone is touched), the turn ends immediately; Si remains at s, and turn passes to agent (i+1)modN. R5. Timbang resolution (Stage 5): - When an agent reaches Stage 5, they execute the *timbang* maneuver: - Number of stones caught on back of hand: k1∼Binomial(5,pback). - Number of those caught in palm: k2∼Binomial(k1,ppalm). - Score awarded: Ci←Ci+k2. - Reset stage: Si←1. Turn passes to next agent. R6. [gap-fill] Loop breaking & win condition: The match terminates immediately when any agent reaches Ci≥TargetScore (default 50 points). R7. [gap-fill] Target score tie-breaker: If multiple agents cross the target in the same round, the agent with the highest total score wins. ## Edge cases - Tie at the end condition: Highest score above target wins; if equal, a sudden-death single *timbang* roll breaks the tie. - A player with no legal move: Not applicable; an agent always has valid stone-grouping choices. - Repetition / cycles: Since pback>0 and ppalm>0, positive points accumulate monotonically, guaranteeing eventual termination. - Simultaneous claims on the same resource: Turns are strictly sequential; each agent plays on their own fresh scatter. - Player elimination mid-round: Not applicable; all players remain active until a winner emerges. - Resource exhaustion: Fixed 5 stones. - Termination guarantee: Monotonically increasing score totals with non-zero probability of advancement at every turn. ## State representation - Persistent state: `scores[N]`, `stage[N]`, `active_agent_id` (0..N−1), `target_score` (integer). - Per-turn ephemeral state: 2D scatter coordinates of 5 stones, selected *ibu* index, chosen grouping plan. - Observable information: Fully observable by all agents. ## Action space - `attempt_stage(ibu_index, grouping_plan)`: Where ibu_index∈[1..5] and grouping_plan is a partition of the remaining 4 stones matching stage requirements. ## Playability notes - Single-player-vs-AI: Natural; player attempts stages against AI agents with parameterized dexterity models. - Group games (n>4): Clean round-robin rotation without coalition interference. - Physical-skill games: Complex hand-sweep timing is abstracted into spatial distance metrics and skill parameters. - What is lost: The micro-tactile sensation of fabric weight and precise hand-eye reaction speed is mapped to probabilistic outcomes. ## Implementation hints Extremely lightweight state machine. Can be easily implemented as a discrete Markov process. First milestone: build a simulation loop verifying convergence to 50 points across varying skill values.

Download as JSON

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

Download JSON
Cite this page

Malaysian Five-Stone Jacks (Malaysia). (n.d.). Games of Countries. https://gamesofcountries.com/games/my-malaysia/batu-seremban

@misc{gamesofcountries-my-malaysia-batu-seremban,
  title = {Malaysian Five-Stone Jacks (Malaysia)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/my-malaysia/batu-seremban}
}