On Sun, 2018-09-30 at 22:26 -0600, Reuben Staley wrote: > That's actually a really cool idea. One thing, though, I think a week is > too short. That would be as few as 0 and as many as 2, but generally 1 > report per instance. Unless we decided to get rid of all weekly > Cartographor reports and instead do a new report for each instance, > which is just messy.
I guess my idea was that the game would have few inputs from the Agoran gamestate (although maybe it could depend on things that happened in previous weeks), just output to it in a way that didn't really affect future instances of the game. So if you cared about "this week's subgame" you could find all the information you needed easily. > Also, that short of games would require constant monitoring of your > opponents' actions, and I'll tell you what, I sure can't focus that much > time on watching this game. Usually, I have less than an hour a day I > can focus on Agoran business. Hmm, so maybe we want some sort of game where everyone's actions are secret from everyone else's? Like, you have one week to submit your strategy for that week to an officer, then the officer publishes a report saying what the results were. (Presumably we need to give the officer in question a fairly large salary to compensate for their inability to participate in the game normally.) This implies that maybe it should be a mixed cooperative/competitive game (a simple start would be: you get a bonus for correctly predicting other players' actions, and a penalty for having your actions correctly predicted; collusion would be possible here but zero-sum, so it'd probably lead a lot of interesting deals under the table). > That's why I said about a month. It's long enough that you can fit > several reports in there, and you don't have to be constantly monitoring > your opponents' actions. > > Of course, I could just be looking at this from the wrong angle and > misunderstanding your point, so feel free to correct. Well, my idea was that participating in the subgame would only be necessary if you either wanted to a) grind points, or b) go for some economic goal (pending proposals, expunging, etc.). So you could drop in and out at will. That means that the cycles would need to be very short by Agoran standards. Perhaps we could reinforce this by giving players an advantage in the subgame based on how long they hadn't participated in it. (A side effect of this is that new players would pretty much always win the subgame round just after they registered, as they hadn't participated in it before, helping to fix the "new player problem".) That way, not participating (because you're busy or whatever) arguably wouldn't even hurt your economic standing. > One thing we should definitely keep from this message is the amendments > only effect the next round and are easier to do idea. If the goal is > repeated iterations, that's a good way to do it. Right. I got the idea from this by observing a long-defunct (but apparently successful) nomic which was set up for the purpose of trying to build a game, and basically had separate "building" and "playtesting" parts. -- ais523
