An Adaptive Hint System for Puzzle Games: A Multimodal-based Approach

For a puzzle game, the embedded hint system is a vital medium to improve the quality of the gameplay. A well-designed hint system can help balance the difficulty of the puzzle game and offer the player a better game experience. However, according to the questionnaire-based survey and the industrial investigation we made for puzzle games, two significant problems with the current hint systems are discovered: a) Most hint systems are rule-based. Many players consider that the timing of the hints cannot match the actual states of their game flows. b) In addition, the presentation modes of the hints are sometimes too abstract or too straightforward to follow and even leak the upcoming puzzles. Hence, to improve the game experience for the players when involved in puzzle games, we design and develop a multimodal-based adaptive hint system to help players get through the puzzles more smoothly with timely hints that are presented in proper modes. To test the system's effectiveness and usefulness, we conducted a control experiment based on three puzzle games, with a control group playing the original games and an experimental group playing the games with the multimodal-based adaptive hint system. The experimental results showed that our system has a significant effect in helping players improve their game experience and balance the difficulty of the game.