One solution to the problem of getting expert knowledge into expert systems would be to endow the systems with powerful learning procedures that could discover appropriate behaviors by observing an expert in action. A promising source of such learning procedures can be found in recent work on connectionist networks, which are massively parallel networks of simple processing elements. This report discusses RAMBOT, a connectionist expert system that learns to play a simple video game by observing a human player. The game, Robots, is played on a two-dimensional board containing the player and a number of computer-controlled robots. The object of the game is for the player to move around the board in a manner that will force all of the robots to collide with one another before any robot is able to catch the player. The connectionist system learns to associate observed situations on the board with observed moves, and is capable not only of replicating the performance of the human player, but also of learning generalizations that apply to novel situations. Diagrams illustrate the Robots game, game strategies, and RAMBOT's performance. Listings of references, ICS Technical Reports, and earlier reports from the Cognitive Science Laboratory are provided. (KM)