Formal Specification
Hopscotch (literally "hop paradise"; traditional Norwegian grid-hopping game) — Formal Specification
Norway · dexterity-small · Confidence: high
← Back to game pageImplementation Suitability
Suitability: engine-readyFidelity: probabilistic-proxyGap-fills: 6
# Hoppe paradis — AI-optimized specification
## General idea
Hoppe paradis ("Hopscotch" / Flyver'n) is a turn-based precision tossing and acrobatic hopping race through a numbered 10-zone geometric grid. Players sequentially progress a stone marker from zone 1 through 10 (Paradis), navigating alternating single-foot and straddle jumps while avoiding lines and retrieving the marker without losing balance.
## Formal model
- **Abstraction taken:** Physical stone-tossing accuracy and jumping balance are abstracted into stochastic execution checks conditioned on target zone distance and an agent skill vector . The geometric grid is formalized as a discrete 10-node progression graph with topology constraints (single vs double nodes).
- **Player/agent count:** 1 to 6 agents (standard model: 2 to 4 agents).
- **Turn model:** Sequential round-robin turns. An agent retains the turn upon a successful level completion or yields upon committing a foul [gap-fill: standard folk rule allowing continuous play while successful].
## Formal rules
R1. **Grid Topology:** The course consists of ordered zones where zones are `SINGLE`, are `DOUBLE_STRADDLE`, and zone 10 is `PARADIS` (rest and turn-around zone).
R2. **Level Progression State:** Each player tracks current target level , initialized to .
R3. **Toss Phase:** On player 's turn at level , the player throws the marker toward zone . Toss success probability is modeled as [gap-fill].
- If the toss fails (lands out of bounds or on a line), the turn ends immediately with a `TOSS_FOUL`. Player remains at level .
R4. **Hopping Execution:** If the toss succeeds, the agent executes the hop sequence traversing nodes , skipping zone .
- Hopping over single nodes requires 1 foot; double nodes require 2 simultaneous feet; zone 10 permits a 180-degree pivot.
- Each movement step incurs a small risk of balance failure: [gap-fill].
- At zone (or adjacent zone on return), the player executes `pickup_marker()`.
R5. **Foul Resolution:** If any balance fault or line touch occurs during the hop sequence, the turn ends with a `STEP_FOUL`. Player remains at level .
R6. **Level Clear & Continuation:** If the hop sequence and pickup are completed without fouls, . The player immediately proceeds to level on the same turn [gap-fill].
R7. **Victory Condition:** The first player whose target level exceeds 10 () wins the game immediately.
## Edge cases
- **Tie at the end condition:** Not applicable; turns are strictly sequential, and the first agent to clear Level 10 triggers an immediate game-ending victory.
- **A player with no legal move:** Not applicable; `toss_marker()` is always legal on an active agent's turn.
- **Repetition / cycles:** An agent failing attempts stays at . To prevent infinite loops among zero-skill bots, [gap-fill] a maximum turn limit rounds is enforced; if reached, the player with the highest current level wins.
- **Simultaneous claims on the same resource or position:** Not applicable; players execute hops sequentially on an isolated track.
- **Player elimination mid-round; turn order repair:** No player elimination; all agents cycle turns until victory.
- **Resource exhaustion:** Marker is indestructible and retrieved on each successful hop or foul reset.
- **Termination guarantee:** Guaranteed by monotonic progress toward Level 10 and the global match round cap [gap-fill].
## State representation
- **Persistent state:** `player_levels` (array with ), `active_player_idx` , `round_count` .
- **Per-turn state:** `marker_zone` , `turn_phase` .
- **Observable information:** Perfect public information. All player levels and toss/hop outcomes are visible to all agents.
## Action space
- `toss_marker(target_level, aim_offset)`: Throw action with parameterized aim precision.
- `hop_step(target_zone, foot_mode)`: Step execution declaring single vs straddle landing.
- `pickup_marker()`: Balance retrieval action from adjacent zone.
## Playability notes
- **Single-player-vs-AI:** Highly playable as an individual balance-and-accuracy solitaire challenge or race against AI ghosts.
- **Group games (n>4):** Scales up to 6 players naturally via round-robin turn taking. For a 2-agent duel, agents alternate turns directly on a shared or parallel grid.
- **Physical-skill games:** Physical hopping agility, single-leg isometric balance, and marker wrist snap are parameterized by scalar agent attributes.
- **What is lost in the adaptation:** Physical burning in the calf muscles, playground asphalt chalk dust, and joyous celebratory balance pivots.
## Implementation hints
- Model as an episodic discrete Markov Chain with deterministic transitions on success and resets on failure.
- Low state complexity; branching factor per step is minimal (1 or 2 discrete actions).
- Suggested first milestone: Implement deterministic level incrementing with synthetic 90% toss/hop success probabilities before integrating attribute-based sampling.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Hopscotch (literally "hop paradise"; traditional Norwegian grid-hopping game) (Norway). (n.d.). Games of Countries. https://gamesofcountries.com/games/no-norway/hoppe-paradis
@misc{gamesofcountries-no-norway-hoppe-paradis,
title = {Hopscotch (literally "hop paradise"; traditional Norwegian grid-hopping game) (Norway)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/no-norway/hoppe-paradis}
}