This study introduces a blazingly fast, randomized, no-loss expert system for tic-tac-toe using decision trees called T3DT, which tries to emulate human gameplay as closely as possible and is over 22k times faster than non-randomized minimax based approaches.
