Experimentation

Mechanism · IX – Learning & Adaptation immediate / delayed · reversible / conditional · concealed / dynamic · solo

Experimentation discovers unknown rules, relations or possibilities through systematic trial: a hypothesis becomes a discovery or a failure via the attempt.

Design question – How do agents in your game discover the unknown through trial – and how do you reward it?

Reference example – Baba Is You: push rule words around and see what happens – “Wall is Win”? Zelda: Breath of the Wild: fire in tall grass creates an updraft – discovered by trying it out. Mao: the rules are secret; you only work them out by trial and penalty card. Minecraft (Crafting Discovery) · Subnautica (Exploration) · Sandbox Spiele (Emergent Strategy)

Structural pattern

Input -> transformation -> output: an unknown relation (hypothesis) becomes a discovery or a failure (insight) via an attempt (trial).

Regulated in permitting mode (attempts by combining, acting or exploring), constraining mode (what remains unknown; attempts have a cost or are counted) and conditioning mode (an attempt leads to a discovery, a failure or partial information). In a game, costs and number of attempts become specifications.

Design parameters

• Knowledge space – finite (all recipes findable) ↔ infinite (procedural)
• Attempt cost – free ↔ risky
• Result transparency – immediately recognizable ↔ ambiguous
• Repeatability – identically repeatable ↔ changed by the attempt
• Shareability – exclusive (only the discoverer knows) ↔ public
• Reward – knowledge advantage ↔ new options

Related Mechanisms

• Revelation (MEC-18) – Frequently combined: reveal the result.
• Matching (MEC-31) – Frequently combined: test combinations.
• Production (MEC-01) – Frequently combined: discovery -> new recipe.
• Progression (MEC-15) – Frequently combined: discovery unlocks.

Ludic Organisational Structures

Evolution / Adaptation · Exploration System · Crafting / Recipe System · Commons (dynamic)