From the observables experienced, the player can make educated guesses about the locations of certain other things (i.e. the wumpus, pits and bats), and use this information to help win the game. This can be implemented in EDEN as a series of functions such as "likelihood of wumpus in room 14" which depend on observables such as "can the player smell a wumpus from a room near 14?" or "has the player entered room 14 without being eaten by the wumpus?"
The DoNaLD and Scout notations can be used to display a pictorial representation of the rooms of the cave and their interconnections, as well as a "player's view" showing the destination room numbers of the tunnels leading out of the player's room. Most implementations of Hunt the Wumpus do not provide the player with a graphical map of the caves, nor do they keep the room numbering scheme consistent. Therefore, it should be possible to hide the map (or only show those parts of the map that have been visited) in order to model a player who has no map, and is relying only on the map of the parts of the cave already seen and the information in the "player's view".
In the model, the user could try out various scenarios, as with the Pjawns model. For example, the model could be used to explore questions relating to the positions of pieces in the game, such as "what might the game's outcome have been had the wumpus been in room 2 instead of room 1?" or questions that explore different versions of the game, for example, "what might happen if I can smell the wumpus from three rooms away, instead of only two?" This is intended to encourage the user to become familiar with the game and its dynamics through experiment.