Formal Specification
Tambourine Ball — Formal Specification
Italy · ball-team · Confidence: low
← Back to game pageImplementation Suitability
Suitability: engine-readyFidelity: discretizedGap-fills: 3
# Palla Tamburello — AI-optimized specification
## General idea
A turn-based team rally court sport played on an open boundary without a net. Teams of five specialized agents coordinate spatial zones to deliver, volley, or ground-return a high-speed rubber ball using circular tambourine-bats (*tamburelli*), scoring points along a 15-30-40-Game sequence to win 13 games.
## Formal model
- **Abstraction taken:** The continuous flight of a 150 km/h rubber ball over an 80m field is discretized into a rally exchange state machine. Each stroke assigns a target destination in the opposing half-court; the nearest defending agent attempts an interception based on velocity, reach, and positioning.
- **Player/agent count:** 2 teams of 5 agents (1 Server/Battitore, 1 Back/Spalla, 1 Mid/Mezzovolo, 2 Forwards/Terzini).
- **Turn model:** Alternating team rally phase; sequential individual agent hits.
## Formal rules
R1. Match setup: Target game threshold is games. Court is . Centerline is .
R2. Points progression inside game: .
R3. Service: Serving team's Battitore selects a serve vector landing in .
- [gap-fill] Service accuracy: .
R4. Rally mechanics: Defending team must intercept and return the ball before it bounces twice on the court ().
R5. Return stroke: The closest defending agent to the ball arrival point attempts a return stroke to the opposite half-court.
- [gap-fill] Return shot error: Landing position .
- If landing falls outside boundaries ( or ), the attacking team commits a fault.
R6. Point scoring: Faults (out of bounds, 2 bounces, body contact) award a point to the non-offending team.
- [gap-fill] At 40-40, the sudden-death "killer point" (*punto secco*) awards the game to the next point winner.
R7. Match termination: The first team to accumulate 13 games wins the match.
## Edge cases
- Tie at the end condition: Matches tied at 12-12 play a 25th deciding game.
- A player with no legal move: Not applicable; the assigned intercepting agent always chooses between legal target coordinates in the opponent's court.
- Repetition / cycles: Bounded because rally lengths are strictly stochastic with non-zero failure rate per stroke (), guaranteeing every rally terminates.
- Simultaneous claims on the same resource or position: When a ball lands in overlapping zones between Midfielder and Forward, the agent with smaller Euclidean arrival time executes the stroke.
- Player elimination mid-round; turn order repair: In team play, an injured/ejected player is replaced by a bench substitute; if bench is empty, remaining 4 agents expand coverage areas.
- Resource exhaustion (pieces, tokens, space, time): Single ball in active play; games accumulate monotonically to 13.
- Termination guarantee: Monotonic game score accumulation and positive point-ending probabilities per stroke guarantee finite termination.
## State representation
- Persistent state:
- `games_a`, `games_b`: integers .
- `points_a`, `points_b`: indices .
- `serving_team`: team ID.
- Per-rally state:
- `ball_pos`: .
- `ball_vel`: .
- `bounce_count`: integer .
- `active_team`: team ID.
- `agent_positions`: map of 10 agent coordinates on the court.
- Observable information: Fully observable positions, court markings, and score.
## Action space
- `serve(target_x: float, target_y: float, speed: float)`
- `return_ball(target_x: float, target_y: float, shot_type: str)`
## Playability notes
- Single-player-vs-AI: Engine-ready; play as team captain assigning tactical formations or controlling a specific player (e.g. Server or Forward).
- Group games (n>4): Can be reduced cleanly to 1v1 or 2v2 on a scaled court (e.g. ).
- Physical-skill games: Physical timing, forearm impact, and wind resistance are abstracted to Gaussian landing distributions.
- What is lost in the adaptation: The resonant booming echo of the stretched animal skin bat and high-speed reflexes.
## Implementation hints
Treat as an open-court tennis simulator with 5 cooperative agents per side.
First milestone: Build the 1v1 single-agent prototype on a simplified court with 15-30-40-Game logic, then scale to 5-agent cooperative zone partitioning.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Tambourine Ball (Italy). (n.d.). Games of Countries. https://gamesofcountries.com/games/it-italy/tamburello
@misc{gamesofcountries-it-italy-tamburello,
title = {Tambourine Ball (Italy)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/it-italy/tamburello}
}