Skip to main content
Formal Specification

Kick Ball / Kick Shuttlecock — Formal Specification

Philippines · dexterity-small · Confidence: low

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: probabilistic-proxyGap-fills: 3
# Sipa — AI-optimized specification ## General idea A physical dexterity kick-juggling game modeled as a sequential Markov decision process. On each strike, the active agent selects a body contact technique (e.g., inside foot, knee, heel, outside flick), balancing technique difficulty, control stability, and score multipliers against a stochastic failure risk (ground contact or illegal hand touch). The goal is maximizing total consecutive airborne strikes over R=3 attempts. ## Formal model - **Abstraction taken:** The continuous 3D physics of kicking a lead/plastic shuttlecock is abstracted into a discrete technique choice with an evolving stability/fatigue state S∈[0.0,1.0]. Each strike samples a success probability P(success)=f(technique_difficulty,S,skill). - **Player/agent count:** 1 to N agents competing sequentially. - **Turn model:** Sequential attempt rounds (3 attempts per agent). ## Formal rules R1. Setup: - Each agent begins with `score = 0`, `attempt = 1`, and initial control stability S=1.0. R2. Strike Action & State Transition: - On each airborne step, the agent chooses an action from the strike dictionary: 1. `instep_kick`: base difficulty D=0.15, stability delta ΔS=−0.02, score +1. 2. `knee_bump`: base difficulty D=0.25, stability delta ΔS=−0.04, score +2. 3. `outside_flick`: base difficulty D=0.40, stability delta ΔS=−0.06, score +3. 4. `heel_strike`: base difficulty D=0.60, stability delta ΔS=−0.10, score +5 [gap-fill]. - Strike outcome is sampled: P(success)=1.0−(D×(1.5−S)×(1.0−skill)). R3. Turn Continuation or Drop: - If success = `true`: - `current_streak` increments by +1, `round_score` increments by the technique value. - S←max⁡(0.1,S+ΔS). - Agent continues to next strike. - If success = `false` (drop / foul): - `attempt_score[attempt] = round_score`. - `attempt` increments by +1, S resets to 1.0. - If `attempt > 3`, the agent's turn concludes. R4. Winner Determination: - An agent's total score is ∑r=13attempt_score[r] (or maximum single attempt [gap-fill]). The agent with the highest score wins. ## Edge cases - Tie at the end condition: In the event of equal top scores, a 1-round sudden death attempt is executed between tied agents [gap-fill]. - A player with no legal move: Strike actions are always valid while S>0. - Repetition / cycles: Bounded because stability S monotonically degrades with each consecutive strike, driving P(success)→0, which guarantees stochastic termination. - Simultaneous claims on the same resource or position: Not applicable under sequential solo play. - Player elimination mid-round: In circle/elimination mode, an agent who reaches 3 failed volleys is removed; remaining players adjust rotation order. - Resource exhaustion: Not applicable; attempts are fixed at R=3. - Termination guarantee: Enforced by the finite attempt limit (R=3) and the strictly diminishing success probability curve. ## State representation - Persistent state: - `agent_scores`: array of size N. - `current_agent`: integer 0..N−1. - `current_attempt`: integer 1..3. - `control_stability`: float 0.1..1.0. - `current_streak`: integer. - Observable information: Fully observable public state. ## Action space - `strike(technique_id)`: integer 0..3 representing `[INSTEP, KNEE, OUTSIDE, HEEL]`. ## Playability notes - Single-player-vs-AI: Highly playable as a classic push-your-luck arcade scoring game. - Group games (n>4): Generalizes naturally to leaderboard ranking across any number of agents. - Physical-skill games: Physical foot-eye coordination is mapped directly to the risk-reward payoff curve of technique difficulties. - What is lost in the adaptation: Wind drift, foot orientation angle, and physical leg flexibility are abstracted into stochastic multipliers. ## Implementation hints Extremely lightweight state representation (O(1) memory). Can be evaluated via simple Monte Carlo rollouts or value iteration to find optimal risk-seeking policies. First milestone: Implement single-attempt loop with fixed skill parameter to verify streak distribution.

Download as JSON

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

Download JSON
Cite this page

Kick Ball / Kick Shuttlecock (Philippines). (n.d.). Games of Countries. https://gamesofcountries.com/games/ph-philippines/sipa

@misc{gamesofcountries-ph-philippines-sipa,
  title = {Kick Ball / Kick Shuttlecock (Philippines)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/ph-philippines/sipa}
}