A PLANNING APPROACH TO DECLARER PLAY IN CONTRACT BRIDGE

• DOI: 10.1111/j.1467-8640.1996.tb00255.x
• Publication Date: 2007-04-02T22:56:38Z
• Authors: S. J. J. Smith; D. S. Nau; T. A. Throop
• Keywords: Bridge (graph theory); Perfect information; Tree (set theory); Game tree

Although game‐tree search works well in perfect‐information games, it is less suitable for imperfect‐information games such as contract bridge. The lack of knowledge about the opponents’ possible moves gives game tree a very large branching factor, making impossible to significant portion this reasonable amount time. This paper describes our approach overcoming problem. We represent information bridge task network extended multi‐agency and uncertainty. Our game‐playing procedure uses generate trees which set alternative choices determined not by actions, but available tactical strategic schemes. have tested on declarer play bridge, an implementation called Tignum 2. On 5000 randomly generated notrump deals, 2 beat strongest commercially program 1394 1302, with 2304 ties. These results are statistically at α= 0.05 level. searched average only 8745.6 per deal time 27.5 seconds Sun SPARCstation 10. Further enhancements currently underway.