Searching the game tree quickly and correctly is a crucial part of designing intelligent AI (Artificial Intelligence). The algorithm to search game tree determines the response time and shrewdness of the AI player's moves. Alpha-Beta pruning is an algorithm which can be used to search the game tree efficiently. We define the nature of the game and provide the naïve search algorithm. MinMax algorithm is used as the naïve search algorithm. We examine drawbacks and the areas of improvement in MinMax algorithm and show how Alpha-Beta algorithm exploits them to boast the performance. We then analyze the performance of Alpha-Beta algorithm and the conditions under which it can produce optimum results. We present a variation of the Alpha-Beta technique which relies on hardware to allow the search on independent branches of game tree to be done in parallel.
