I’ve spent so much time working out details that I didn’t realize that I am now at something of a plateau. I have the basic conversational system in place, utilizing the verbs greet, what’s new?, gossip, no news, I see, request deal, what is your goal?, agree, reject, goal aura count, tell auragon, threaten, defy, and submit. That’s 14 verbs in a nice little cluster, and they all work together well. The time has come to move to the next big chunk of work. But what should that chunk be?
The most obvious factor to add is mood changes and their effects on behavior. The current system takes into account only the long-term personality traits. Inasmuch as moods and their manipulation should be an important part of the interpersonal strategy, I really ought to factor in mood changes, as well as their effects on behavior.
However, there are a few catches to impede me. The first is that there’s no mood for the honesty trait. I could alter the engine to take that into account. Currently there are four mood traits: Disgusted_Aroused, Sad_Happy, Fearful_Angry, and Tired_Energetic. These are the general-purpose moods inherited from SWAT.
I can easily dispense with Disgusted_Aroused; there’s nothing in the game that will arouse either of these extremes of the same mood. Dumping Tired_Energetic isn’t so obvious a deletion, but I think that I can do without it. Sad_Happy and Fearful_Angry are just as obviously important to retain. So the question is, should I add a mood corresponding to the Honest trait? Can an actor become temporarily more suspicious? Yes, I think so.
So I have some engine work to do: delete two moods and add a new one. The next problem will be deploying the moods. They must change in response to events. Those changes should be fairly easy to implement. And their effects on decisions should also be fairly simple to add to the existing option desirabilities.
Adding lies
But there’s another area that deserves my efforts: lies. An actor can lie about two things: auragon counts (as a result of a deal) or gossip. (I have decided to eliminate betrayals, the deliberate reneging on a deal.) The decision to lie, however, is an extremely complicated one. The benefit of the lie must be calculated in terms of its effect on the listener — that gets us into some hairy stuff. The costs of the lie will be based on the probability of being caught, the likelihood that it will lead to suspicion, and the effect on the liar’s reputation. This is messy stuff; I think that I shall put it off until after I’ve done with the mood work.