Anders L. Madsen and Lars M. Nielsen and Finn V. Jensen
We present an application designated to the calculation of probabilities in the card game bridge. The application can be used by bridge players to improve their decisions in the game. We describe in detail how a Bayesian network is dynamically generated depending on the case presented to the applications. We also present a communication language which makes the specification of all kinds of relevant evidence simple. The communication language also makes it easy to specify requests for probabilities of complex properties of the distribution of cards. We explain how the probability of a strategy for declarer’s play being successful is calculated by transforming the corresponding condition to disjunctive normal form. Finally, we give examples showing the use of ProbSy.