Formal Specification
Standing Seesaw Jumping — Formal Specification
North Korea · dexterity-throw · Confidence: high
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 2
# Neolttwigi — AI-optimized specification
## General idea
A two-agent cooperative-competitive jumping contest where agents stand at opposite ends of a balanced seesaw plank. By timing downward impacts on the plank, each agent converts kinetic momentum into vertical flight for their partner. Each cycle, agents select jumping power, landing balance adjustments, and mid-air acrobatic poses to maximize cumulative height and style points without misstepping or slipping off the board.
## Formal model
- **Abstraction taken:** Continuous physics of plank oscillation and aerial acrobatics is abstracted into alternating discrete flight cycles parameterized by impact timing error, vertical energy retention, and pose execution probability.
- **Player/agent count:** 2 agents (Agent A, Agent B) acting in alternating rhythmic sequence.
- **Turn model:** Alternating discrete half-cycles (Cycle = 1 jump/landing exchange).
## Formal rules
R1. Setup: Plank system initialized with base kinetic energy `E = 10 units`. Both agents start with balance index `100%`.
R2. On active half-cycle, descending Agent i commits an action pair: `{impact_force: (low, med, high), landing_adjustment: (heel, flat, toe), midair_pose: (none, twist, split, flip)}`.
R3. Landing accuracy is evaluated: `timing_error = |chosen_timing - optimal_cadence|`.
R4. Kinetic transfer: Upward energy imparted to partner Agent j: `E_next = E_current * transfer_efficiency(timing_error) + boost(impact_force)`.
R5. Jump height achieved: `H_j = f(E_next)`. Points awarded based on height and successfully executed midair pose difficulty.
R6. Balance degradation: High impact force or large timing errors reduce the agent's balance index: `Balance_i -= penalty(timing_error)`.
R7. [gap-fill] Fault threshold: If an agent's `Balance` drops below 0%, or `timing_error > max_tolerance`, the agent missteps off the board; the jump sequence ends immediately (Fault).
R8. [gap-fill] Competition format: The contest runs for a fixed 20 half-cycles (10 jumps per agent). If both agents complete the routine without faulting, the agent with the higher cumulative individual style/height score wins.
## Edge cases
- **Tie at the end condition:** Handled by comparing peak single-jump height achieved; if still tied, victory is shared.
- **A player with no legal move:** Not applicable; an impact commitment is always required every cycle.
- **Repetition / cycles:** Strictly bounded by the 20-cycle limit in R8.
- **Simultaneous claims on the same resource or position:** Not applicable; agents act strictly in alternating sequence.
- **Player elimination mid-round:** If either agent faults (R7), the match terminates immediately.
- **Resource exhaustion:** Energy and balance fluctuate based on execution precision.
- **Termination guarantee:** Guaranteed by finite 20-cycle cap and absorbing fault state.
## State representation
- **Persistent state:** `cycle_count (0..20)`, `cumulative_scores [score_A, score_B]`, `peak_heights [H_A, H_B]`.
- **Ephemeral state:** `current_energy E`, `active_agent`, `agent_balances [bal_A, bal_B]`.
- **Observable information:** Fully observable state.
## Action space
- `jump_and_land(force, timing_offset, pose_id)`: Select impact parameters and pose difficulty.
## Playability notes
- **Single-player-vs-AI:** Player controls one jumper trying to maintain a high-scoring cooperative rally with an AI partner, or competing against AI for highest personal score.
- **Physical-skill games:** Rhythm and balance are abstracted into timing-window tolerance and attribute checks.
## Implementation hints
Lightweight state machine. The primary engineering task is calibrating the energy transfer formula and timing window so that riskier poses require sharper timing precision.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Standing Seesaw Jumping (North Korea). (n.d.). Games of Countries. https://gamesofcountries.com/games/kp-north-korea/neolttwigi
@misc{gamesofcountries-kp-north-korea-neolttwigi,
title = {Standing Seesaw Jumping (North Korea)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/kp-north-korea/neolttwigi}
}