Waterman [30] and Smith [25] used poker as a testbed for automatic learning techniques. Both of them worked on the problem of acquiring problem-solving heuristics through experience.
Waterman worked in two areas: 1) the representation of heuristics as production rules to facilitate their dynamic manipulation, and 2) the automatic modification and creation of these heuristics by a learning program on the basis of information obtained during training. During his research, five computer players were created, differing in the number and source of the heuristics initially provided to the program. The performance of his best program was evaluated to be the same degree of skill as a ``nonprofesional but experienced human player''. His programs played a two-player standard version of five-card draw poker. He stated that by choosing poker, the representation and generalization techniques he developed were shown to be an effective approach to implementing decision making and learning in an imperfect information environment.
Smith proposed an alternative method for dynamically learning heuristics by using adaptive search (genetic algorithms). Poker was used as a testbed for this technique to provide a basis for comparison with Waterman's work.