Skip to main content
Formal Specification

Striking the Starling (Czech Tip-cat) — Formal Specification

Czechia · stick-team · Confidence: low

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: discretizedGap-fills: 1
# Tlučení špačka — AI-optimized specification ## General idea A turn-based bat-and-field game where batting agents execute a two-stage launch-and-strike sequence to send a tapered peg into the field. Fielders attempt aerial interception or return-throw base strikes to eliminate the batter, while surviving hits trigger an estimation and measurement betting mini-game in bat lengths. ## Formal model - **Abstraction taken:** Physical peg trajectories (launch angle, mid-air strike timing, field carry distance) are parameterized by a two-stage stochastic resolver `strike(power, angle)` yielding distance D in meters and catch probability Pcatch. The fielding return throw is modeled as a Bernoulli trial Phit(D). - **Player/agent count:** 2 teams of k agents (minimum 1 agent per team, standard 3 agents). - **Turn model:** Sequential within batting inning; fixed rotation order among batters. ## Formal rules R1. Setup: Total innings M (standard M=2) per team. Match state initializes with Team A batting, Team B fielding, inning = 1, current_batter = 1. R2. Inning structure: Each agent on the batting team takes one plate appearance per inning. R3. Batting phase: The active batter selects a launch profile (aim_angle,power∈[0,1]). The launch-and-strike resolver outputs: - *Whiff:* Peg not struck in 3 attempts; batter is OUT, score for turn = 0. - *Fair Hit:* Peg travels distance D>0 along vector θ. R4. Fielding phase: - *Catch check:* Given D and fielder coverage density, probability Pcatch resolves. If caught, batter is OUT, score = 0. - *Return throw check:* If uncaught, nearest fielder attempts return throw at the base. Probability of striking base is Pbase=max⁡(0.05,1.0−D/50.0). If base hit, batter is OUT, score = 0. R5. Estimation and scoring phase: If uncaught and base missed, true distance in bat lengths is Ltrue=⌊D/0.55⌋. - Batter declares estimate Lcalled∈ℕ+. - Fielding team chooses action: `ACCEPT` or `CHALLENGE`. - If `ACCEPT`: Batting team gains Lcalled points. - If `CHALLENGE`: If Lcalled≤Ltrue, batting team gains 2×Lcalled points. If Lcalled>Ltrue, batting team gains 0 points. R6. Inning transition: When all k batters have completed turns, roles swap (batters become fielders). When both teams complete inning m, inning advances to m+1. R7. Win condition: After M full innings, the team with the higher cumulative points wins. R8. [gap-fill] Tiebreaker: If scores are tied after M innings, each team nominates one representative for a single sudden-death batting turn; the higher single-turn score wins. ## Edge cases - Tie at the end condition: Resolved via R8 single-turn sudden-death playoff. - A player with no legal move: Not applicable; a batting agent always has the legal action space of aim/power/estimation, and fielding team always has accept/challenge. - Repetition / cycles: Not applicable; turn sequence is strictly finite (2×M×k total batting appearances). - Simultaneous claims on the same resource: Not applicable; turns and fielding resolutions are strictly sequential. - Player elimination mid-round: If an agent disconnects or forfeits, their turn is treated as an automatic OUT with 0 points; team rotation continues. - Resource exhaustion: Not applicable; no consumable tokens or pieces. - Termination guarantee: Guaranteed because innings and team rosters are finite, and each plate appearance concludes in exactly one resolution cycle. ## State representation - Persistent state: `inning: int`, `active_team: {0, 1}`, `scores: [int, int]`, `roster_index: int`, `completed_innings: int`. - Per-turn ephemeral state: `peg_distance: float`, `is_out: bool`, `called_estimate: int`, `challenger_response: {ACCEPT, CHALLENGE}`. - Observable information: Fully observable to all agents except during simultaneous decision points (none; estimate declaration precedes fielding response). ## Action space - `batter_swing(power: float, angle: float)`: Parameters ∈[0,1] and [−π/3,π/3]. - `declare_estimate(lengths: int)`: Integer estimate ≥1. - `fielding_response(decision: {ACCEPT, CHALLENGE})`: Discrete binary response. ## Playability notes - Single-player-vs-AI: Natural; 1 human vs. 1 AI agent across 2 innings with simulated fielding probabilities. - Group games (n>4): Generalizes naturally; each team fields k agents, taking individual batting turns while the AI fields. - Physical-skill games: Motor timing is converted into trade-offs between swing power (higher potential distance) and contact variance (higher miss/catch risk). - What is lost: The physical hand-eye precision of striking a spinning wooden peg mid-flight is abstracted into a parameterized probability distribution. ## Implementation hints Low complexity. The estimation sub-game introduces strategic bluffing (calling slightly below true distance to guarantee acceptance vs. pushing close to true distance to provoke a failed challenge). First milestone: implement R1–R7 with deterministic lookup tables before tuning strike variance.

Download as JSON

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

Download JSON
Cite this page

Striking the Starling (Czech Tip-cat) (Czechia). (n.d.). Games of Countries. https://gamesofcountries.com/games/cz-czechia/tluceni-spacka

@misc{gamesofcountries-cz-czechia-tluceni-spacka,
  title = {Striking the Starling (Czech Tip-cat) (Czechia)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/cz-czechia/tluceni-spacka}
}