Chess
Introduction
There is nothing magic about computer playing chess, it's just alpha-beta search. Theorectically, computer can check all possible move sequences, and know the game result even before the game starts. Practically, the problem is that the search space is too large. Suppose on average each game has 50 steps, and each step has 20 possibilities, then the computer has to search 20^50 combinations. Of course, pruning can terminate some branches early, but it does not change the fact that the search space is exponential.
In order to have a reasonable response time, this implementation of compuater AI only looks 3 steps ahead (computer move, your possible response, and computer conter-response). Therefore, if you can think more than 3 steps forward, you will definitely win the game!
Client Requirements
This chess game is implemented as a Microsoft .NET Windows form user control. It requires both Internet Explorer and .NET Runtime 1.1 or later.