The game has come together and everything is working fine. Of course, it’s not fun yet – that gels only in the final adjustments, which I’m now struggling with. I’m worried about the role of indirect statements of affinity. There’s really not much strategy to use with them. I had previously thought that they’d be good for breaking up friendships: you could tell Ella that Owen hates her, making Ella like him less. But that entails a lie that will surely cost you in her estimation. It therefore seems a pointless strategy. I need some way to make it worthwhile, or to remove it.
Worse, the information I get from other characters about me is useless; it’s always dated because characters don’t get much opportunity to talk about me. The odds of character A talking to character B about character C on any given turn are only 0.05; hence the likelihood that they can give me up-to-date information is insignificant. So I don’t need this information, and my ability to use it to influence others is limited. Finally, the information I already have when I start, based on the symmetry of relationships, is better than I can get from others.
The implication is that I should completely dump indirect statements. This would leave a simpler, faster game. But would it be interesting? It would be even simpler than the original Gossip game. The problem with this approach is that it permits a simple “tell ‘em what they want to hear” strategy that is optimal. There can’t be any penalty for lying because people don’t share your conversations with others. The only way to get proper balance against the simple lock on victory is to have checks for what others have said. But that in turn requires that people have more indirect conversations. Perhaps I should expand the indirect statements to include every other person; this was the original arrangement. When A and B discuss C, they swap what they know about C’s feelings for everybody else. That would greatly slow down the game but it would permit richer game play.
I don’t know; this is a difficult decision.