The game is all but finished but I am still torn over the artificial personality and its implications for gameplay. As I noted in the previous design essay, the indirect statements are all but useless, and decisions based on them are accordingly useless. This strongly suggests that I should dump indirect statements completely, but as I noted previously, this would make it impossible to use much guile in talking to people; they already know how you feel!
In the current design, you are given perfect indirect knowledge in the 4-player game, and less in each of the higher games.
Here are some solutions:
Everybody gets complete knowledge of all affinities at the outset of the game, but not as they change. If you tell ‘em what they want to hear, then the penalty for lying should be equal to the benefit for matching their feelings, in which case, there’s no way to win.
Another way is to change the nature of indirect gossip: you are supposed to gossip about what somebody else actually said. You can still lie, but people keep track of what was said, not the pAffinities. That way they can accuse you of lying. Thus, on the first turn, you’re not allowed to make an indirect statement. After that, you can, but people will compare histories and you can get clobbered retroactively.
Does this mean that I should have a phone call in which the caller accuses the callee of lying? That would certainly be dramatic, but what good would it do the caller? I see no reason to indulge in that anger: it won’t make you popular with the callee and it probably won’t do any good for any of your other relationships, except as a way of distancing you from an outcast. Besides, we’re talking about a completely different turn structure – out of the question at this late date.
So we revert to the idea that indirect relationships are only stated as actual facts. You don’t relate what you think pAffinity is, you relate what you were actually told. The listener compares that with the history he knows, looking for a match. Then he reacts accordingly.
Reactions must take place in two dimensions: similarity of affinity and falsehood. But now falsehood is a binary quantity.
You start off knowing only your own feelings about others. On the first turn, you discuss those. After that, you can discuss indirect affinities, but to make that effective, we’d have to expand the statement of indirect affinity to permit all such statements (that is, A tells B that C feels x about D, as distinguished from the current version: A tells B that C feels x about C).
This returns me to an old question: should I once again permit 4-person statements of indirect affinity? I set that up initially, but then removed it on the grounds that it was too complicated. Should I bring it back?