Formal Specification
Historic Florentine Football — Formal Specification
Italy · ball-team · Confidence: low
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: discretizedGap-fills: 3
# Calcio Storico Fiorentino — AI-optimized specification
## General idea
A high-dimensional team invasion game combining territory progression, ball-carrying mechanics, and 1v1 combat pin mechanics on a discretized 2D grid. Teams score full points by carrying/throwing the ball into the entire opposing baseline or concede half-points on errant high shots, with combat encounters deciding tactical lane control.
## Formal model
- **Abstraction taken:** The 50x100m sand pitch is discretized into a grid. Continuous physical brawling is abstracted into 1v1 combat state locks (where two adjacent agents can become "pinned" in grappling, removing them from movement phases until broken).
- **Player/agent count:** 2 teams of 27 agents (or reduced to 5 agents per team in small-scale simulations: 1 Goalkeeper, 1 Defender, 1 Blocker, 2 Runners).
- **Turn model:** Real-time ticked simulation (e.g. 1 tick = 1 second) or simultaneous round-based commitment phase where all non-pinned agents select actions simultaneously.
## Formal rules
R1. State setup: Pitch is grid . Team A defends ; Team B defends . Ball starts at center . Match duration is ticks (50 minutes).
R2. In each tick, all active agents independently choose one action: `move(dx, dy)`, `tackle(target_agent)`, `pass(target_coord)`, `carry()`, or `shoot(baseline)`.
R3. Combat resolution: When Agent A executes `tackle(Agent B)` within distance 1:
- [gap-fill] A stochastic contest resolver compares combat attributes: .
- If pinned, both agents enter `grappling` status and cannot move or handle the ball until a breakout action succeeds.
R4. Illegal combat: [gap-fill] If a third agent attacks an already grappling pair (2-on-1 infraction), the offending agent is permanently ejected from the match (`ejected` status), leaving their team undermanned.
R5. Scoring mechanics:
- If a ball carrier enters the opponent's baseline row or passes accurately into the baseline goal zone, Team scores *caccia*.
- If a shot on the baseline is attempted but misses wide/high (sampled via accuracy variance), the defending team receives *caccia*.
R6. Post-score reset: Following a scored caccia, teams swap defending baselines (), and the ball is reset to the center spot.
R7. Win condition: The team with the higher score total at tick wins. [gap-fill] If tied at tick , sudden-death extra time continues until the first caccia is scored.
## Edge cases
- Tie at the end condition: Resolved via sudden-death extra time (R7) until the next whole or half caccia is registered.
- A player with no legal move: Pinned agents have only the `attempt_breakout()` action; non-pinned agents can always `hold_position()`.
- Repetition / cycles: Bounded strictly by the absolute 50-minute time clock ( ticks).
- Simultaneous claims on the same resource or position: If two opposing agents enter the ball cell simultaneously, a 50/50 loose-ball contest resolves possession.
- Player elimination mid-round; turn order repair: Ejected or injured players are permanently removed from the active agent roster; no replacements are permitted.
- Resource exhaustion (pieces, tokens, space, time): Match concludes unconditionally when the simulation tick timer reaches zero.
- Termination guarantee: Clock monotonicity ensures guaranteed termination in exactly 50 minutes of game-time (or bounded sudden death).
## State representation
- Persistent state:
- `score_a`, `score_b`: floats (supporting half-cacce).
- `time_remaining`: integer .
- `active_agents`: list of 54 agent structs with positions , status (`free`, `grappling`, `ejected`), team ID, and role.
- Per-tick state:
- `ball_position`: coordinate or `held_by(agent_id)`.
- `pending_actions`: dictionary mapping agent IDs to chosen actions.
- Observable information: Fully observable spatial grid to all agents.
## Action space
- `move(dx, dy)`: .
- `tackle(target_id: int)`: adjacent opponent.
- `pass(dest_x: int, dest_y: int)`: ball holder only.
- `shoot()`: ball holder targeting opponent baseline.
- `attempt_breakout()`: grappling agent only.
## Playability notes
- Single-player-vs-AI: Needs adaptation; control an individual forward or play as tactical team captain giving high-level lane commands.
- Group games (n>4): Full 27v27 can be reduced to 5v5 without losing the core dual-track mechanic of combat lane-pinning versus ball progression.
- Physical-skill games: Combat execution is parameterized by strength and wrestling attributes; ball throwing by distance accuracy kernels.
- What is lost in the adaptation: The intense atmosphere, historical costume pageantry, and visceral physical endurance of sand fighting.
## Implementation hints
Treat as a hybrid between cellular automata tactical invasion and simultaneous grappling state machine.
First milestone: Build the 5v5 discretized invasion engine with simple ball passing and goal line scoring, then add the 1v1 combat-pin state lock.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Historic Florentine Football (Italy). (n.d.). Games of Countries. https://gamesofcountries.com/games/it-italy/calcio-storico-fiorentino
@misc{gamesofcountries-it-italy-calcio-storico-fiorentino,
title = {Historic Florentine Football (Italy)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/it-italy/calcio-storico-fiorentino}
}