Formal Specification
Inuit Football — Formal Specification
Greenland · ball-team · Confidence: high
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: discretizedGap-fills: 5
# Arssaq — AI-optimized specification
## General idea
A traditional Greenlandic Inuit football game played across open snow fields, sea ice, or tundra. Two large teams contest possession of a heavy stuffed sealskin ball (*arssaq*), advancing it across slippery arctic terrain via foot kicking and body checking to cross the opposing settlement or landmark goal line.
## Formal model
- **Abstraction taken:** The continuous vast open-ice terrain is discretized into an grid (e.g. cells). [gap-fill] Player locomotion on snow/ice incorporates traction and momentum checks, with foot-kicking and physical tackling resolved via stochastic success probabilities.
- **Player/agent count:** 2 teams of agents each ().
- **Turn model:** Simultaneous-action simulation ticks (each tick represents 1 discrete tactical phase: move, kick, tackle).
## Formal rules
R1. Setup: A grid of cells (length , width ). Team A defends goal line ; Team B defends goal line . The heavy sealskin ball begins at center coordinate .
R2. On each tick, all agents simultaneously submit an action:
- `move(direction)`: Advance 1 cell orthogonally or diagonally. On ice, moving onto a new cell has a small slipping probability [gap-fill] (, causing the agent to lose movement for that tick).
- `kick_ball(power, direction)`: If the agent occupies the same cell as the ball, kick the ball in a specified direction with displacement cells. Hand touching is strictly prohibited.
- `tackle(target_agent_id)`: If adjacent to an opposing agent possessing the ball, attempt a physical shoulder charge to dislodge the ball.
R3. Tackle resolution: If a tackle is declared on the ball carrier, success is evaluated:
If successful, the ball becomes loose on an adjacent cell; if failed, the ball carrier retains possession.
R4. Goal scoring: If the ball crosses completely over Team B's goal line (), Team A scores 1 goal. If the ball crosses over Team A's goal line (), Team B scores 1 goal.
R5. Reset: Following a goal, the ball is reset to center , all agents reset to their defensive halves, and play resumes.
R6. [gap-fill] Time limit / match duration: A match is played for ticks, or until one team reaches goals.
R7. Win condition: The team with the most goals at (or first to reach ) wins. [gap-fill] If tied at , sudden-death extra time (first golden goal) determines the winner.
## Edge cases
- Tie at the end condition: Resolved by the golden-goal sudden-death period in R7.
- A player with no legal move: Not applicable — moving, kicking (if adjacent), or tackling is always an available legal action.
- Repetition / cycles: Bounded by the fixed tick horizon .
- Simultaneous claims on the same resource (loose ball): If two opposing agents move onto the loose ball cell on the same tick, possession is resolved by a simultaneous 50/50 contested scramble check weighted by agility.
- Player elimination mid-round; turn order repair: No player elimination; injured/fatigued agents can rest on sidelines without disrupting team tick updates.
- Resource exhaustion: Ball and field cells are permanent and non-depleting.
- Termination guarantee: Provably terminates because tick count strictly increments to .
## State representation
- Persistent state: Grid dimensions , goal target , tick limit , current score per team.
- Per-tick state: Ball position , ball carrier ID (or `null` if loose), agent positions array , agent stamina pools.
- Observable information: Fully observable — all player coordinates, ball trajectory, and scores are public.
## Action space
- `move(dx, dy)`: Move adjacent cell.
- `kick(direction, power)`: Propel ball forward.
- `shoulder_charge(target_id)`: Attempt dispossessing tackle.
- `hold_position()`: Guard zone.
## Playability notes
- Single-player-vs-AI: Needs adaptation — 1 human player controls the primary ball handler while team AI scripts guide defensive spacing and offensive runs.
- Group games (n>4): Naturally 2 teams of 4–12 agents; [gap-fill] reduction to 2 agents (1v1 soccer duel) preserves the full footwork and goal-scoring loop.
- Physical-skill games: Snow-running balance and kicking force are parameterized by agent physical attributes.
- What is lost: The freezing arctic wind, heavy sealskin ball texture, and cultural community celebration honoring the Northern Lights.
## Implementation hints
Moderate complexity. Standard 2D grid-iron team sports model with simultaneous movement and collision resolution. First milestone: implement a 1v1 grid game with loose ball physics and kicking to verify goal detection and reset mechanics.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Inuit Football (Greenland). (n.d.). Games of Countries. https://gamesofcountries.com/games/gl-greenland/arssaq
@misc{gamesofcountries-gl-greenland-arssaq,
title = {Inuit Football (Greenland)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/gl-greenland/arssaq}
}