Formal Specification
Taiwanese Dragon Boat Racing (Flag Grabbing) — Formal Specification
Taiwan · water · Confidence: low
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: discretizedGap-fills: 3
# Longzhou Duobiao — AI-optimized specification
## General idea
A traditional aquatic sprint regatta and flag-snatching contest based on Taiwanese Dragon Boat Racing (*Lóngzhōu jìngdù duóbiāo*). Crews coordinate paddling cadence, drummer stroke timing, and rudder steering over a 400-meter straight river course, culminating in a high-risk flag grab (*duóbiāo*) where the bow grabber lunges past the dragon head to pluck the floating finish flag.
## Formal model
- **Abstraction taken:** Hydrodynamic fluid turbulence, 18 individual paddler biomechanics, and dragon hull hull-wave interactions are abstracted into a lump-parameter boat dynamics model driven by stroke rate (SPM), crew stamina, and a terminal reach-and-lunge flag snatch resolver.
- **Player/agent count:** 2 to 4 competing crews (each commanded by 1 AI/human agent managing pace and grab timing).
- **Turn model:** Real-time ticked simulation () over a 400-meter linear aquatic lane.
## Formal rules
R1. **Course & Boat Dynamics:**
- Course length in parallel lanes of width .
- Floating finish buoy with target flag is stationed at , lane center.
- Boat state: position , velocity , heading , crew stamina .
R2. **Crew Command Inputs (Per Tick):**
The crew commander selects:
- `drum_cadence`: Stroke rate strokes/minute.
- `stroke_effort`: (scales thrust and stamina expenditure).
- `rudder_angle`: .
- `lunge_trigger`: boolean (orders the bow grabber to lunge forward).
R3. **Acceleration & Stamina Model:**
- Forward thrust: .
- Water drag: .
- Velocity update: .
- Stamina drain: .
R4. [gap-fill] **Flag Snatching Action (*Duóbiāo*):**
- When the boat bow is within the capture window :
- If `lunge_trigger == True`, the grabber extends forward with reach distance .
- Lateral offset error: .
- Grab success probability:
- If the grab succeeds, the flag is secured and the boat logs its final official finish time.
- If `lunge_trigger` is initiated prematurely () or fails, the grabber misses the flag.
R5. **Victory Condition:**
- The winning boat is the first crew whose flag grabber successfully secures the finish flag.
- Crossing the line without successfully grabbing the flag results in an unranked/disqualified run (*No Flag = No Finish*).
R6. [gap-fill] **False Start & Lane Violation:**
- Starting before the signal incurs an immediate 5.0-second delay penalty.
- Deviating outside lane boundary () incurs disqualification.
R7. [gap-fill] **Race Timeout:** Any boat failing to reach the finish line within seconds is logged as DNF.
## Edge cases
- **Tie at the end condition:** If two boats secure flags within the same tick (), rank is decided by high-speed photo-finish timestamp; if still tied, a 200m re-row heat is held.
- **A player with no legal move:** Control inputs are continuous bounded vectors available throughout the race.
- **Repetition / cycles:** Water current and continuous paddling ensure monotonic progress along the 400m course toward .
- **Simultaneous claims on the same resource or position:** Each boat has an independent lane and separate target flag buoy at the finish line.
- **Player elimination mid-round:** If a grabber falls into the water or misses the flag completely, that boat is logged as DNF while remaining lanes continue racing.
- **Resource exhaustion:** Stamina decreases monotonically; course distance is finite.
- **Termination guarantee:** Guaranteed by finite course length and maximum race clock .
## State representation
- **Persistent state:**
- `crews`: array of `{id, lane, boat_mass, crew_stamina_base, skill_rating}`.
- `finish_times`: mapping of `crew_id` float.
- `race_status`: enum `[Ready, Racing, Finished]`.
- **Per-tick ephemeral state:**
- `elapsed_time`: float.
- `crew_telemetry`: array of `{x, y, v, heading, stamina, flag_secured, grabber_lunged}`.
- **Observable information:** Fully observable boat positions, current velocities, relative distance gaps, and lane flags.
## Action space
- `set_pacing(spm, effort, rudder_angle)`:
- *Parameters:* .
- `trigger_flag_grab()`:
- *Preconditions:* Boat bow , executed once per race.
## Playability notes
- **Single-player-vs-AI:** Highly playable; an AI agent optimizes pacing strategy (energy pacing curve / burst sprint vs. aerobic conservation) and precise millisecond timing for the flag lunge.
- **Group games (n>4):** Scales across 2 to 4 parallel racing lanes simultaneously.
- **Physical-skill games:** 20-person synchronized exertion and lunge dexterity are formalized into coupled ODE dynamics and a timing-accuracy probability function.
- **What is lost in the adaptation:** Deafening resonance of wooden war drums, water spray in the face, and the visceral danger of hanging inverted over churning river water.
## Implementation hints
- Simple 1D/2D kinematic simulation with quadratic drag.
- Optimization problem is well-suited for dynamic programming or reinforcement learning.
- Suggested first milestone: Implement single-boat energy-pacing simulation with flag-grab window detection and success scoring.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Taiwanese Dragon Boat Racing (Flag Grabbing) (Taiwan). (n.d.). Games of Countries. https://gamesofcountries.com/games/tw-taiwan/longzhou-duobiao
@misc{gamesofcountries-tw-taiwan-longzhou-duobiao,
title = {Taiwanese Dragon Boat Racing (Flag Grabbing) (Taiwan)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/tw-taiwan/longzhou-duobiao}
}