Formal Specification
Planting Buds and Flowers / Escalating High Jump — Formal Specification
Vietnam · dexterity-small · Confidence: medium
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 4
# Trồng nụ trồng hoa — AI-optimized specification
## General idea
Trồng nụ trồng hoa is a traditional Vietnamese jumping and body-hurdle agility game. Two seated "Planters" build an escalating living obstacle using their pressed soles, stacked feet, upright fists ("buds"), and open spread hands ("blooming flowers") across six rising tiers (15 cm to 80 cm). A queue of jumpers takes sequential running leaps over the human tower; touching any part of the barrier constitutes a fault, forcing a role swap with a seated planter.
## Formal model
- **Abstraction taken:** The continuous biomechanics of running approaches, vertical leap trajectory, and mid-air leg tucking are abstracted into a discrete stage-progression state machine where clearance success is sampled from parameterized jump probability distributions [gap-fill: physical athletic running approach, takeoff spring, and hurdle clearance kinematics are modeled as a stochastic stage-progression state machine with parameterized clearance probabilities].
- **Player/agent count:** 1 Jumper agent (in single-player challenge) or Jumper agents in a rotational queue ().
- **Turn model:** Sequential turn-taking queue advancing through 6 hurdle levels ().
## Formal rules
R1. Obstacle Levels and Setup:
- [gap-fill] Six standardized hurdle levels:
- Level 1 (*Gốc cây / Base*): Height , Base difficulty .
- Level 2 (*Chân đôi / Double Feet*): Height , Base difficulty .
- Level 3 (*Nụ một / First Bud*): Height , Base difficulty .
- Level 4 (*Hoa một / First Flower*): Height , Base difficulty .
- Level 5 (*Nụ hai / Second Bud*): Height , Base difficulty .
- Level 6 (*Hoa hai / Full Bloom*): Height , Base difficulty .
- Current obstacle level initialized at .
- Active jumper queue initialized as .
R2. Jump Action and Clearance Mechanics:
- Active jumper chooses a jump approach technique .
- [gap-fill] Clearance Probability Model:
where .
- A Bernoulli trial is sampled with parameter .
R3. Success and Advancement:
- If the jump is **CLEAN** ():
- Jumper clears the hurdle and joins the back of the queue.
- If all jumpers in the queue successfully clear Level :
- If , all jumpers achieve a Grand Round victory; level resets to (or advances to single-leg hopping challenge).
- Else, obstacle level increments: .
R4. Faults and Role Swap:
- If the jump is a **FAULT** ():
- Jumper touched the hurdle.
- Jumper is penalized: in queue mode, becomes a seated Planter (incurring +1 fault count), and the longest-seated Planter joins the jumping queue.
- The hurdle level for the new active jumper resets to .
R5. [gap-fill] Match Termination and Scoring:
- The match concludes after full level cycles, or when an individual jumper completes Level 6 cleanly without faults.
- Jumpers are scored by total clean leaps and highest level cleared.
## Edge cases
- Tie at the end condition: Jumpers with identical cleared levels are ranked by lowest fault counts, then by number of Level 6 clears.
- A player with no legal move: Active jumper must submit a jump technique action from the legal action set.
- Repetition / cycles: Bounded by R5 (hard cap of 5 level cycles).
- Simultaneous claims on the same resource or position: Jumpers execute attempts strictly one by one in queue order.
- Player elimination mid-round; turn order repair: A fault swaps the active jumper into the planter role; queue order advances smoothly to the next waiting jumper.
- Resource exhaustion: Hurdle levels 1..6 are persistent and reset cleanly upon faults or round clears.
- Termination guarantee: why the game provably ends: Guaranteed by R5 (finite cycle limit ).
## State representation
- Persistent state:
- `current_level`: integer .
- `jumper_queue`: ordered list of agent IDs.
- `jumper_faults`: array of fault counts per player.
- `cycle_count`: integer .
- Per-round ephemeral state:
- `active_jumper_id`: integer agent ID.
- `last_jump_result`: enum `{CLEAN, FAULT}`.
- Observable information:
- Fully observed by all agents: current obstacle level , hurdle height, queue order, and player skill profiles.
## Action space
- `execute_jump(technique)`: technique .
## Playability notes
- Single-player-vs-AI: Natural as an escalating height-clearance challenge (clear all 6 levels with minimum faults).
- Group games (n>4): The queue-and-swap mechanic naturally accommodates 3–8 children without losing the social rotation dynamic.
- Physical-skill games: Lower-body vertical leaping power and coordination are represented by the `skill` parameter and technique risk/reward tradeoffs.
- What is lost in the adaptation: The somatic thrill of vaulting over living friends' arms, wobbling hands, and the laughter when a toe brushes a finger.
## Implementation hints
- Implement as a sequential state machine with an array of 6 difficulty tiers.
- Suggested first milestone: Implement R1–R4 level transitions and verify that a jumper with skill=0.7 clears Levels 1–4 easily but faces a stiff challenge at Level 6.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Planting Buds and Flowers / Escalating High Jump (Vietnam). (n.d.). Games of Countries. https://gamesofcountries.com/games/vn-vietnam/trong-nu-trong-hoa
@misc{gamesofcountries-vn-vietnam-trong-nu-trong-hoa,
title = {Planting Buds and Flowers / Escalating High Jump (Vietnam)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/vn-vietnam/trong-nu-trong-hoa}
}