Skip to main content
Formal Specification

Czech Dodgeball (Elimination Ball) — Formal Specification

Czechia · ball-team · Confidence: high

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: discretizedGap-fills: 2
# Vybíjená — AI-optimized specification ## General idea A two-zone tactical team elimination game where opposing squads maneuver on a partitioned 2D grid to target opposing agents with ballistic ball throws. Ball activation requires cross-field charging passes between court agents and an exiled captain zone. Eliminated agents move to the outer perimeter, creating crossfire angles and opportunities for self-redemption. ## Formal model - **Abstraction taken:** The 18×9 m court is discretized into a 1D/2D zonal graph: Zone A (Team A field, x∈[0,8]), Zone B (Team B field, x∈[9,17]), Outfield A (Team A exiled/captain zone behind Zone B, x∈[18,20]), Outfield B (Team B exiled/captain zone behind Zone A, x∈[−3,−1]). - **Player/agent count:** 2 teams of N agents (N≥4, default N=6: 1 Captain + 5 field players). - **Turn model:** Discrete time-ticked state machine with simultaneous movement and sequential ball throw resolutions. ## Formal rules R1. Setup: Agents initialize in their respective home zones. Captain of Team A starts in Outfield A; Captain of Team B starts in Outfield B. Ball initialized in Zone A. `is_charged = False`. R2. Ball Charging Rule: - An elimination throw is legal only when `is_charged == True`. - A pass between an agent in Zone A and an agent in Outfield A sets `is_charged = True`. - An interception or grounded ball resets `is_charged = False`. R3. Elimination Throw: - An agent possessing a charged ball in Zone A or Outfield A may execute `throw_at(target_agent_id)`. - Resolver evaluates: - *Catch:* With probability Pcatch(target_skill), target catches ball cleanly; no elimination occurs; target's team takes possession (`is_charged = False`). - *Direct Hit:* With probability Phit(thrower_skill)×(1−Pcatch), target is struck. - *Miss:* Ball lands on floor in target zone, resetting charge. R4. Elimination and Redemption: - If a field agent is struck: agent transitions from home field to team's outfield zone. - If an outfield agent successfully hits an opposing field agent: the thrower is redeemed and moves back into their home field zone. R5. Captain Entry and Lives: - When all field agents of a team are in the outfield, the Captain moves from the outfield into the home field zone. - The Captain possesses 3 lives (`captain_lives = 3`). - Each direct hit sustained by the active Captain decrements `captain_lives` by 1. R6. Win condition: The first team to reduce the opposing Captain's lives to 0 immediately wins the match. R7. [gap-fill] Turn cap / Stalling rule: If a team holds the ball for ≥3 consecutive ticks without making a pass or throw, possession is automatically transferred to the opposing team. R8. [gap-fill] Match time ceiling: If neither captain is eliminated within 200 simulation ticks, the team with the higher count of active field players (or greater remaining captain lives) wins. ## Edge cases - Tie at the end condition: Resolved via R8 (sum of active field players + captain lives). - A player with no legal move: Holding position or executing a safe pass is always available. - Repetition / cycles: Bounded by R7 anti-stalling rule and R8 200-tick termination cap. - Simultaneous claims on the same resource: Ball possession is binary; co-located agents resolve priority by proximity to ball impact coordinate. - Player elimination mid-round: Transition to outfield preserves agent active decision status without altering turn order. - Resource exhaustion: Ball and player counts invariant. - Termination guarantee: Monotonic life depletion tendencies bounded strictly by R8 tick cap. ## State representation - Persistent state: `team_scores: [int, int]`, `captain_lives: [int, int]`, `agent_locations: dict[int, Zone]`, `agent_status: dict[int, {FIELD, OUTFIELD, ELIMINATED}]`. - Ephemeral tick state: `ball_pos: tuple[float, float]`, `ball_possessor: int`, `is_charged: bool`, `holding_ticks: int`. - Observable information: Fully observable (all agent positions, lives, ball charge status, and possession are public). ## Action space - `move_to(target_pos: tuple[float, float])` - `pass_ball(target_teammate_id: int)` - `throw_at(target_opponent_id: int)`: Precondition `is_charged == True`. ## Playability notes - Single-player-vs-AI: Natural; 1 human commanding squad tactics or playing active captain against AI. - Group games (n>4): Natively designed for 6–12 agents per side. - Physical-skill games: Throw velocity and catch reactions are modeled via logistic skill functions P(hit)=σ(θthrow−θdodge). - What is lost: Physical dodging contortions, jumping over low balls, and gymnasium acoustics. ## Implementation hints Medium complexity. The game naturally splits into a discrete State Machine (Charging → Throwing → Hit/Catch → Transition). First milestone: 4v4 game with 1 life per player before integrating the 3-life Captain logic.

Download as JSON

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

Download JSON
Cite this page

Czech Dodgeball (Elimination Ball) (Czechia). (n.d.). Games of Countries. https://gamesofcountries.com/games/cz-czechia/vybijena

@misc{gamesofcountries-cz-czechia-vybijena,
  title = {Czech Dodgeball (Elimination Ball) (Czechia)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/cz-czechia/vybijena}
}