Previous: A Strategic Metagame Player
for General Chess-Like Games
Up: A Strategic Metagame Player
for General Chess-Like Games
Next: Constructing a Metagame-player
Previous Page: A Strategic Metagame Player
for General Chess-Like Games
Next Page: Bias when using Existing Games
Virtually all past research in computer game-playing has attempted to develop computer programs which could play existing games at a reasonable standard. While some researchers consider the development of a game-specific expert for some game to be a sufficient end in itself, many scientists in AI are motivated by a desire for generality. Their emphasis is not on achieving strong performance on a particular game, but rather on understanding the general ability to produce such strength on a wider variety of games (or problems in general). Hence additional evaluation criteria are typically placed on the playing programs beyond mere performance in competition: criteria intended to ensure that methods used to achieve strength on a specific game will transfer also to new games. Such criteria include the use of learning and planning, and the ability to play more than one game.