Catherine Sibert, RPI Graduate Student

Catherine Sibert, RPI Graduate Student

The goal of most modeling work is to reproduce and predict behavior as closely as possible. For complex tasks like Tetris, where humans have to make rapid decisions for long term planning, the best models so far developed can only capture small aspects of the whole task. These models are far from perfect, but when applied as a measuring tool rather than as an explanatory model, they can provide insight into individual aspects of a more complex task. By creating a series of deliberately imperfect models, we can look at how speed and strategy contribute to expertise in Tetris.

 

Share|