All versions of the Bayesian Poker Player to date have used a succession of four distinct Bayesian networks, one per round. In a DDN model for poker, each round in the poker betting would correspond to a single time slice in the DDN, which would allow explicit representation of the interrelation between rounds of play.

1. How would the network structure within a single time slice differ from the DN shown in Figure 5.14?

2. What would the inter-slice connections be?

3. What information and precedence links should be included?