Formal Specification
Canarian Wrestling — Formal Specification
Spain · wrestling-combat · Confidence: low
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 2
# Lucha canaria — AI-optimized specification
## General idea
A two-agent stand-up grappling contest where agents start in a standardized mutual shorts-grip (*agarre*) and simultaneously select tactical throwing, unbalancing, or counter techniques (*mañas*). The bout operates under a strict sudden-death binary fall condition: the first agent to make any non-foot contact with the ground loses the round immediately.
## Formal model
- **Abstraction taken:** The continuous physical grappling exchange inside the circular ring (*terrero*) is abstracted into discrete tactical decision phases (Attack Technique vs. Defensive Counter). Outcome probabilities for ground fall or ring exit are calculated via a skill, mass, and technique matchup matrix.
- **Player/agent count:** 2 agents per individual bout.
- **Turn model:** Simultaneous action selection per exchange phase, resolved instantaneously.
## Formal rules
R1. Setup: Each bout begins at round score 0–0. Both agents start upright in the center of the circular *terrero* with the initial shorts grip (*agarre*) established.
R2. Round structure: A bout is contested as best of 3 falls (*tres mejores luchas*). First agent to 2 falls wins the bout.
R3. Phase execution: Each combat phase, both agents simultaneously select one action from the legal technique set .
R4. Matchup resolution: Technique choices interact through a rock-paper-scissors style advantage table modified by agent attributes (skill, balance, weight). The resolver yields one of three outcomes:
- Agent A falls (Agent B scores fall).
- Agent B falls (Agent A scores fall).
- Neutral stand / no fall (proceed to next exchange).
R5. Fall definition: A fall is registered the instant any body part other than the soles of the feet makes contact with the ground.
R6. [gap-fill] Ring boundary & stall penalty: If 5 consecutive neutral exchanges occur without a fall, both agents receive a stall warning; accumulating 3 warnings eliminates the passive agent.
R7. Simultaneous fall: If the resolver outputs a simultaneous ground contact (*revuelta*), neither agent scores and the round is restarted.
R8. Termination: The bout ends when an agent reaches 2 fall victories.
## Edge cases
- **Tie at the end condition:** Not applicable; rounds are re-fought on simultaneous falls until a decisive 2-fall winner emerges.
- **A player with no legal move:** Not applicable; defensive block or neutral balance posture is always available.
- **Repetition / cycles:** Bounded by R6 warning accumulation and non-zero fall probability per active attack ().
- **Simultaneous claims on the same resource or position:** Handled by R4 simultaneous resolution matrix.
- **Player elimination mid-round:** In team format (*lucha corrida*), an eliminated agent leaves the ring and the next roster agent immediately enters to face the surviving opponent.
- **Resource exhaustion:** Not applicable; stamina depletion can optionally degrade skill values over successive phases.
- **Termination guarantee:** Guaranteed by stall-warning elimination rules (R6) and strictly positive fall transition probabilities.
## State representation
- **Persistent state:** `falls_agent_A` (0..2), `falls_agent_B` (0..2), `warnings_A` (0..3), `warnings_B` (0..3), `active_round` (1..3).
- **Per-phase ephemeral state:** `chosen_action_A`, `chosen_action_B` (hidden until simultaneous reveal).
- **Observable information:** Fall count and warning tally are public; current-phase technique selection is private until resolution.
## Action space
- `execute_maña(technique_id)`: Select one specific technique from .
- `counter_stance(stance_type)`: Defensive posture targeting specific attack vectors (e.g. low base against leg trips, hip deflection against lifts).
## Playability notes
- **Single-player-vs-AI:** Natural; human inputs combat choices against an AI adversary operating with realistic risk/reward heuristics.
- **Group games (n>4):** 12v12 team matches are structured as a linear queue of 1v1 bouts, where individual wins carry team momentum.
- **Physical-skill games:** Modeled via: [gap-fill] .
- **What is lost in the adaptation:** Physical balance nuance, micro-shifts in grip tension, and sand friction dynamics are abstracted into the mathematical matchup table.
## Implementation hints
Low complexity. The core engine requires only a matrix of technique counters and a 3-state round tracker. First milestone: Implement 1v1 single-fall resolution with 4 basic techniques, then wrap in best-of-three round loop and warning system.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Canarian Wrestling (Spain). (n.d.). Games of Countries. https://gamesofcountries.com/games/es-spain/lucha-canaria
@misc{gamesofcountries-es-spain-lucha-canaria,
title = {Canarian Wrestling (Spain)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/es-spain/lucha-canaria}
}