Loki-1 was a intermediate level poker player as shown by the experimental
results
in [19]. Also, Loki-1 demonstrated the benefits of using
opponent modeling. However, weaknesses in its
betting strategy were hampering the overall performance of the program
and the Opponent Modeler required improvement. These issues led to a
redesign of Loki's architecture to facilitate the addition of new
components to the program: probability triples and selective sampling
simulations.