CTArcade : Learning Computational Thinking While Training Virtual Characters Through Game Play
Web Service 09.2011
ABSTRACT. CTArcade is a web application framework that can engage users through game play and result in learning computational thinking (CT) skills. Our formative study indicates that CT skills are employed when children are asked to define strategies of common games such as Connect Four. In CTArcade, users can train their own virtual characters while playing games with it. Trained characters then play matches against other virtual characters. Based on reviewing the matches played, users can both improve their game character and their own CT skills.

>> extended abstract (PDF)
Figure 1. Trainer mode. (Left) An ordered list of rules that user's character currently knows. Rules can be moved up & down to re-prioritize. (Center) Tic-Tac-Toe board where user can play with their own character. (Right) Console showing rules that have been applied to the last move made. In this example, user's latest move (middle tile in the upper row) matches with TAKE RANDOM and TAKE ANY SIDE rules.
Figure 2. Custom Rule Creator. A powerful, but more advanced method is to create custom rules. First, the user needs to define a base pattern on the board (by assigning one of five cases for some board tiles) and then generalize the pattern by applying various transforming operations.
Figure 3. Match Review using graph visualization (top) and three other visualization methods (below); List view simply shows all board states from left to right. Step-wise animation allows user to replay all the matches in forward/backward direction. Group by winner shows winning and losing games separately to find a pattern for each result easily. Game tree graph compresses similar board states into a graph node and connects them with accumulative edges whose thickness represents how often the transition occurred.
(c) 2011 Tak Yeon Lee