Randomness usually belongs in presentation rather than the core decision space

What it is

Cosmetic randomness places unpredictability in visual treatment, sound, narrative phrasing, environmental detail, or equivalent reward variants, while keeping the player's consequential decision space relatively stable. A tactical objective may arrive with different decoration, dialogue, or enemy appearance, while resource trade-offs, available strategies, and key success conditions remain learnable. It is one way to locate randomness between novelty and control.

Why it happens

Presentation variation breaks repetition without erasing the causal relationships players learn from. Players can therefore have both "this run is different" and "I know how to handle it": attention is refreshed by a new combination, while strategic knowledge transfers. When randomness changes whether a core option exists, whether a developed strategy suddenly fails, or whether a win condition can be met, it rewrites rather than enriches the decision space. Players must first re-establish whether the system is trustworthy before making a plan. Not every play variable must be fixed, but the nearer it is to consequential commitment and skill evaluation, the greater its explanation cost.

Where it stops holding

Roguelikes, deckbuilders, and tactical endgames draw pleasure precisely from changing core choices per run. They should not read stable decision space as a fixed deck or route; what they need is stable meta-rules. Players must understand the generation pool, appearance constraints, and response tools to strategize around each run's different option set. Conversely, variation limited entirely to appearance can feel cheap when players expect exploration or combinatorial depth. The test is whether a variable asks players to relearn an invisible rule, or merely to apply judgement anew within a known rule.

Applying it

• Classify random variables as changing only feeling, changing in-run tactics, or changing long-term commitments; progressively strengthen disclosed constraints, compensation, and experimentation for the latter two.
• Give equivalent rewards and repeat events several recognizable treatments, so variation serves affect and world texture without secretly changing value.
• If available strategies must be randomized, show generation rules, substitute routes, and obtainable in-run recovery tools.
• Keep core inputs, adjudication, and win conditions stable across updates and events; do not manufacture novelty through surprise rule changes.
• Verification: have players compare two runs and identify which differences alter atmosphere, which demand a new strategy, and why. Inability to separate them means presentation and rules were not clearly divided.

Related

• Same group: W1.05.1 Randomness gives repeat play unpredictable novelty · W1.05.2 Pure luck deciding outcomes weakens control over one's actions · W1.05.4 Too much luck makes high-level skill meaningless
• Nearby: W4.03 Reward structures · W5.02 Interface and immersion trade-offs · W10.02 Loot boxes and probability disclosure
• Search terms: cosmetic randomness · decision space · procedural variation · systemic design