An adventure game can ask you to combine objects in an unexpected way. The difficulty becomes frustrating when the game gives you no useful reason to try that combination. “Moon logic” is a criticism of this gap between the solution the designer knows and the clues the player receives.
Strangeness alone is not the problem. A fantasy world can establish rules that make an impossible action understandable. Conversely, an ordinary-looking room can conceal a puzzle so poorly explained that the player can only try every item on every object. The criticism concerns the route to the answer, not whether the answer could happen in real life.
A contemporary design argument
Ron Gilbert addressed arbitrary puzzles in “Why Adventure Games Suck”, written in 1989 and summarised in Computer Gaming World in April 1990. He argued that a difficult solution should make sense once discovered, rather than depend on luck or exhaustive trial and error. He also acknowledged faults in his own Maniac Mansion designs.
That essay provides a useful design test; it does not establish who coined the phrase “moon logic”. Nor does it say that every demanding puzzle is unfair.
Gilbert separately discussed recognising the player’s intended action. A player might have the right solution but be unable to express it through the exact word a text parser expects. That is a different failure from an arbitrary solution, even though both can leave someone repeating actions without knowing what is wrong.
The cat-hair moustache
Gabriel Knight 3 supplies a concrete example of the argument. In a November 2003 interview, The Inventory asked Jane Jensen about the puzzle in which Gabriel uses cat fur as a moustache while disguising himself as Mosely. Her reply was brief: “Yeah, yeah. It made sense at the time.”
The same interview covered Le Serpent Rouge, a much more elaborate puzzle involving research into a French mystery. Jensen described adapting material until it became something players could solve. Together, the two answers show why “complex” and “arbitrary” are not synonyms. A lengthy chain of reasoning can give the player something to work with; a short item combination can leave them guessing.
Judging a puzzle
For a particular puzzle, useful questions are whether the goal is clear, whether relevant clues are available, and whether the game recognises a reasonable attempt to act on them. Difficulty and fairness need separate judgements. Calling a puzzle moon logic should identify what the player could not reasonably infer, rather than serve as a general complaint that an adventure was hard.
Sources
- Ron Gilbert, “Why Adventure Games Suck”, written 1989, republished 12 May 2004: firsthand design essay, especially “Arbitrary puzzles” and “Reward Intent”.
- Computer Gaming World 70, April 1990, p. 42: contemporary summary of Gilbert’s essay from the Journal of Computer Game Design.
- Jane Jensen interview, The Inventory, November 2003, preserved transcription, “The Past”: cat-hair disguise and Le Serpent Rouge. Jensen’s own account of her design work.