In considering them and Alex's sandbox Traveller game, I came to the conclusion that I had made a choice without realizing it. I had been running an open world in the Morrowind style - things mostly static except where the PCs acted and set off chains of events. The advantage of this idiom is that the blame for everything almost always falls back on the PCs. This is good, because it places the focus on the consequences of the actions of the PCs; player agency is emphasized. It is eminently suited to classically tragic play; mistakes born in hubris come back and lay the mighty low. Finally, this style works well for post-apocalypses where things have stabilized and everyone is afraid to leave their homes, or for wildernesses where a ecological equilibrium can be assumed.
It is, however, no less fundamentally exceptionalist than the 1st-to-20th adventure path railroad. The PCs are necessarily unlike NPCs, in that they are (the only) causal agents in the world. What's more, this makes for a pretty lousy simulation of settings where NPCs have any degree of power, as well as settings where the world is undergoing substantial changes. But I am a lazy DM, and like to roll with material my players give me.
The pure-open world also has the disadvantage of leaving the PCs sans hooks or direction unless provided by themselves. Reacting to external threats is easy; picking from the infinite menu is difficult. Threat also unifies, while selection divides parties.
If I were to run more games, I think I would desire more world-engines; systems which drive the world in patterns, assist me with adjudication, and remind me that the Mongols are due to invade this winter. One of the places ACKS shows great promise is in the noble court; we know how many vassals and henchmen everyone has, and those henchmen's henchmen, and so on down the chain. This cast of underlings could be a rich source for plotting, conspiracy, roleplaying, and other such shenanigans. Part of the fun of the low-level ACKS game was the treachery; the competition for the allegiance of henchmen, the bribery and charm personing, and the never-quite-actual backstabbing were all quite entertaining for me as a DM, and it sounded like the crew I was playing with at the time enjoyed it too. I suspect that an intriguey game could be fun for even my non-treacherous / Lawful Good players, provided that the party remained unified and plotted against NPCs. But managing the sort of personality traits and relationships between NPCs that I'd need is at least an O(n^2) task, and beyond my brain - hence the necessity of tools. Renegade Crowns' personality and relationship material was a good start, but I think I could do better. VBAM's AIX stats are also moderately inspiring, while Pendragon's virtues might also come pretty close to what I'm looking for if I added a relationship score (persistent modifier to reaction rolls) on top of it. I dunno; it's a hard problem. People are complicated. But if I did it right, it could become an endless font of procedural rumors, plots, allies, and enemies.
Posts potentially in the pipeline:
- Barbarians are still coming (strongly related to living worlds / part of the impetus for this post)
- Domains at War - Battles sub-platoon scale? Six-man mercenary squads in a standard dungeon 10x10 square (er, hex)?
- Domains at War - Putting together a sample campaign? I think I want to set it in Midnight, in a section of the Ice Forests. The orcs have a line of forts guarding a supply line to the south; the rebels have a set of villages out in the sticks. The orcs have a stronger supply situation and a larger force, but lack intel and need to balance striking at the enemy with guarding their supply train, while the rebels have better intel but need to feed off of orcish supplies and rely on attrition from small skirmishes in order to survive.
- The time may have come for me to turn the ACKStools into a webapp. I think I am going to use Flask; we're using it at work, and might benefit from some extra familiarity with it. It is not going to be pretty, but it will work.