"Not Another Z Piece!": Adaptive Difficulty in TETRIS

Difficulty in TETRIS is adjusted by adapting the speed with which blocks fall. In this contribution, we describe results of an exploratory study in which we investigated relationships between players performance and their subjective assessment of difficulty and fun. We tested five different algorithms that, instead of adjusting game speed, adjust difficulty by choosing blocks based on the current game state. With our results, we establish pile height and bumpiness as parameters that indicate the performance of a player during a live game, discuss the inherent difficulty of different block choosing algorithms and show how the relationship between fun and perceived difficulty varies for distinct player groups. With regard to adapting difficulty, we argue that one can still teach an old dog such a TETRIS a lot of new tricks.