Smart terrain causality chains for adventure-game puzzle generation

Isaac Dart, Mark J. Nelson (2012). Smart terrain causality chains for adventure-game puzzle generation. In Proceedings of the IEEE Conference on Computational Intelligence and Games, pp. 328–334.

Abstract

Adventure videogames have the player assume the role of protagonist in an interactive story, which is primarily driven by exploration and puzzle-solving. A major drawback with this genre is minimal replayability, since the player has already seen what there is to explore, and knows how to solve the puzzles. We propose a technique to generate variations on puzzles that fit in the same location in the original story, and therefore don't require fully procedural story generation. We keep a database of smart terrain items, which can have effects on other items. Puzzles are generated by taking advantage of a duality between puzzle-solving and generation. Once we build smart terrain causality chains (STCCs) of puzzle solutions, a puzzle known to be solvable can be generated by simply inserting the items contained in a causality chain into the environment. We demonstrate this technique in an experimental videogame, Space Dust which shows that even a very short adventure game can produce multiple interesting playthroughs when STCC-based puzzle generation is added.

Supplemental links


Back to publications.