Formal Specification
Hungarian Bat-and-Ball (Méta) — Formal Specification
Hungary · ball-team · Confidence: high
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: discretizedGap-fills: 5
# Méta — AI-optimized specification
## General idea
A traditional bat-and-ball team field contest where batters strike a ball into an open corridor, attempt to advance between two terminal safe zones, and score round trips while fielding agents retrieve the ball and attempt to tag runners out with thrown projectiles. The core strategic tension lies in calculating runner transit risk against ball recovery and throw velocity.
## Formal model
- **Abstraction taken:** The continuous 2D field (50m × 25m) is discretized into a 10×5 grid consisting of Home Base (Row 0), Active Field (Rows 1–8), and Far Base (Row 9). Batting outcomes and projectile throws are modeled via stochastic physics resolvers parameterized by agent attributes.
- **Player/agent count:** 2 teams of N agents (standard N = 8). At any time, 1 team is Batting and 1 team is Fielding.
- **Turn model:** Tick-based sequential phases within a pitch-and-hit cycle: Pitch Phase → Hit Resolution → Multi-Agent Simultaneous Movement & Fielding Resolution Tick (tick duration = 0.5s equivalent).
## Formal rules
R1. State Setup: Two teams (A and B). Home Base at `y = 0`, Far Base at `y = 9`. Fielders positioned across `y in [1..8]`. Batting lineup initialized in fixed rotation.
R2. Active batter executes `hit(angle, power)`. Hit outcome is resolved into landing coordinate `(x_ball, y_ball)` and flight duration `t_flight`.
R3. Catch Resolution: If any fielder is in range of `(x_ball, y_ball)` at `t_flight`, `P(air_catch)` is computed. If caught, immediate Role Reversal (R7) triggers.
R4. Movement Phase: While ball is active, runners at Home Base and Far Base choose `advance()`, `retreat()`, or `hold()`. Each movement tick moves agent 1 cell along `y`-axis.
R5. Tagging / Throw: Fielder with ball possession can `throw_tag(target_agent)`. Hit probability `P(tag_hit) = f(distance, target_speed, thrower_accuracy)`.
R6. [gap-fill] If a runner is hit by a throw while `y in [1..8]`, runner is tagged out, scoring zero, and a team Role Reversal is triggered.
R7. Role Reversal: Batting and fielding roles swap. [gap-fill] In digital formalization, the inning ends, accumulated points are committed, and the next half-inning begins with reversed roles.
R8. Scoring: When an agent starts at Home Base (`y=0`), reaches Far Base (`y=9`), and returns safely to Home Base (`y=0`), the batting team score increments by 1.
R9. [gap-fill] Game concludes after a fixed number of innings (e.g., 4 full innings per team). Highest cumulative score wins.
## Edge cases
- Tie at the end condition: If scores are equal after scheduled innings, play one additional sudden-death half-inning pair; if still tied, game ends in a draw.
- A player with no legal move: Not applicable — fielding agents can always move or hold; batters in queue wait their turn.
- Repetition / cycles: Bounded by max pitch count per inning (e.g., max 20 pitches or 3 outs per inning) [gap-fill].
- Simultaneous claims on the same resource or position: Grid cells in sanctuaries can hold multiple runners; in the active field, multiple runners occupying the same cell are resolved independently against area throws.
- Player elimination mid-round: If a runner is tagged out, the half-inning terminates immediately per R7; batting queue index resumes at next inning.
- Resource exhaustion (pieces, tokens, space, time): Single ball in play. Ball out-of-bounds resets to nearest boundary fielder with a 1-tick delay [gap-fill].
- Termination guarantee: Enforced by the bounded number of pitches/innings and deterministic tick limits per play cycle.
## State representation
- Persistent state: `inning_number`, `current_batting_team`, `scores[team_id]`, `batter_queue_index[team_id]`.
- Per-player state: `position (x,y)`, `state (in_sanctuary_home, running_forward, in_sanctuary_far, running_return, tagged_out)`, `skill_profile (batting_power, throw_acc, run_speed)`.
- Per-turn ephemeral state: `ball_pos (x, y, z)`, `ball_velocity`, `ball_carrier_id`, `active_pitch_tick`.
- Observable information: Fully observable 2D positional grid and score status to all agents.
## Action space
- Batters: `hit(angle_deg, power_ratio)`, `wait()`.
- Runners: `sprint_forward()`, `sprint_return()`, `hold_base()`.
- Fielders: `move_to(dx, dy)`, `intercept_ball()`, `throw_tag(runner_id)`, `pass_to(fielder_id)`.
## Playability notes
- Single-player-vs-AI: Highly viable — player controls active batter/runners or fielding captain with AI handling team coordination.
- Group games (n>4): Modeled as team multi-agent reinforcement learning (MARL) or tactical squad control.
- Physical-skill games: Physical batting/throwing execution converted to continuous parameter inputs passed through stochastic accuracy distributions.
- What is lost: The physical impact sensation of dodging a flying leather ball and real-time raw running athleticism.
## Implementation hints
Discretize into a discrete 2D spatial coordinate plane with a physics resolver subroutine. Suggested first milestone: implement a 1-batter vs 3-fielders prototype on a 10×5 grid to validate the hit-run-tag loop before scaling to full 8v8 squads.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Hungarian Bat-and-Ball (Méta) (Hungary). (n.d.). Games of Countries. https://gamesofcountries.com/games/hu-hungary/meta
@misc{gamesofcountries-hu-hungary-meta,
title = {Hungarian Bat-and-Ball (Méta) (Hungary)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/hu-hungary/meta}
}