Skip to main content
Formal Specification

Ball with Armband — Formal Specification

Italy · ball-team · Confidence: low

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: discretizedGap-fills: 3
# Pallone col Bracciale — AI-optimized specification ## General idea A continuous high-velocity team court rally simulation in an asymmetric 2.5D walled arena (*sferisterio*). Teams of three specialized agents (Battitore, Spalla, Terzino) alternate striking a heavy leather ball using spiked wooden arm gauntlets, leveraging wall rebounds and spatial coverage to win points on a 15-30-40-game tennis-like ladder. ## Formal model - **Abstraction taken:** The 80x16m sferisterio is partitioned into discrete defensive zones (Service Ramp, Deep Backcourt, Midcourt, Net/Centerline, Wall Rebound Zone). Continuous 130 km/h ball flight and wall angular reflection are abstracted into discrete rally exchanges resolved via spatial reachability and skill-based return distributions. - **Player/agent count:** 2 teams of 3 agents (Battitore, Spalla, Terzino). - **Turn model:** Alternating team rally phase; sequential individual agent returns based on ball intercept trajectories. ## Formal rules R1. Match setup: Target game threshold is G=12 games. Score inside a game progresses: 0→15→30→40→Game. R2. Service phase: Serving team's Battitore executes service from the ramp zone. - [gap-fill] Service success probability: P(valid serve)=σ(skillserve−difficulty). - A valid serve lands in the receiver's half-court (Y∈[40,80]), optionally glancing off the high lateral wall (X=0). R3. Rally exchange: The defending team must intercept the ball before it bounces twice on the floor (Nbounce≤1). Wall impacts reset horizontal trajectory without incrementing the floor bounce counter. R4. Return selection: Defending agent within spatial intercept radius executes a return shot, specifying direction (Deep, Cross-court, High-Wall Rebound, Low-Skimmer) and power. - [gap-fill] Return shot error model: P(return in)=f(agent_stamina,bracciale_skill,ball_incoming_speed). R5. Point resolution (*Quindici*): - A team scores a point if the opponent fails to return the ball across the centerline, allows 2 floor bounces, hits out of bounds, or suffers a body fault. - Scoring progression: 0→15→30→40→Game. [gap-fill] At 40-40 (deuce), the next point wins the game (*punto secco / vantaggio immediato*). R6. Side swap: Teams swap serving/defending court ends every 3 games. R7. Match termination: The first team to win 12 games wins the match. ## Edge cases - Tie at the end condition: Games tied at 11-11 proceed to a decisive 23rd game; no endless tiebreak loop. - A player with no legal move: Not applicable; within an active rally, the assigned intercepting agent always chooses among the 4 return shot directions. - Repetition / cycles: Bounded because rally lengths are stochastic with P(point scored per stroke)≥ϵ>0, ensuring every point terminates in a finite number of strokes. - Simultaneous claims on the same resource or position: When a ball lands in overlapping zones between Spalla and Terzino, the agent with higher intercept priority (closer Euclidean distance) executes the stroke. - Player elimination mid-round; turn order repair: In traditional play, an injured player is replaced by a benched substitute; if no substitute exists, remaining 2 agents expand coverage zones. - Resource exhaustion (pieces, tokens, space, time): Single ball in active play; game score increments monotonically to 12. - Termination guarantee: Strict score monotonicity and strictly positive point-ending probabilities per stroke guarantee that matches terminate in finite time. ## State representation - Persistent state: - `games_team_a`, `games_team_b`: integers ∈[0,12]. - `points_team_a`, `points_team_b`: indices ∈{0,1,2,3} (0,15,30,40). - `serving_team`: team ID. - Per-rally state: - `ball_state`: position (x,y,z), velocity vector (vx,vy,vz), and `bounces_on_floor` ∈{0,1,2}. - `active_receiver_id`: agent ID responsible for next hit. - `rally_length`: integer. - Observable information: Ball position, wall contact state, court markings, and opponent positioning are fully observable. ## Action space - `serve(power: float, target_zone: int)`: Battitore only. - `return_stroke(shot_type: str, wall_rebound: bool, target_x: float, target_y: float)`: Active defender. ## Playability notes - Single-player-vs-AI: Needs adaptation; human can control the team's Battitore or manage team tactical zone assignments. - Group games (n>4): Naturally models 3v3 team dynamics; can be reduced to 1v1 singles on a smaller court layout. - Physical-skill games: Physical impact of 2 kg bracciale against 350 g leather ball is abstracted to stamina consumption and velocity transfer curves. - What is lost in the adaptation: The immense physical shock, resonant wooden cracking sound, and raw athletic sprinting down the springboard ramp. ## Implementation hints Treat as an advanced 3D trajectory-intercept state machine with 1-wall bounce physics. First milestone: Build the point-to-game scoring machine (15-30-40-Game) with a Markov chain rally outcome generator before adding spatial agent positioning.

Download as JSON

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

Download JSON
Cite this page

Ball with Armband (Italy). (n.d.). Games of Countries. https://gamesofcountries.com/games/it-italy/pallone-col-bracciale

@misc{gamesofcountries-it-italy-pallone-col-bracciale,
  title = {Ball with Armband (Italy)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/it-italy/pallone-col-bracciale}
}