shihonage said:
Simulation > economy. Economy is merely a subset of the infinite amount of things that could be simulated.
A game's goal is to establish the mechanics of the world which interact with each other. Those mechanics may be "economy", or something else entirely, like need/emotion management. They're what makes the world become emergent and alive.
<Snip>
A large roaming world and proper aggro management are a start. A required base for other mechanics that can be built upon it.
It's what I was talking about back on p.1, sorry I got sidetracked with economic simulation. What I was saying is that faction management is just another kind of economic sim, because just like the economic sim, actor interaction is based on transactions.
Monsters eat people because they operate with a different sort of economy than people do. To people, people aren't a natural resource. To monsters, they are.
Guards fight monsters because to them, monsters are a resource. Guards are paid to fight them.
The cool part of basing it on simulated transactions, is that it enables actors to reconsider the value of whatever they're up to, on the fly. If an overlord doesn't pay the guards well enough, guards might just run away & take up banditry instead of fighting slavering beasts for pennies. Or the monsters might reconsider attacking armed guards. Or go fetch some friends and eat everyone in the city. Or if the protection offered by the guards is worth the protection money the citizens pay the overlord, something else might happen.
The same goes for actor emotions. It's just another kind of transaction. Work, kids and a nagging wife might deplete a resource only offered by bars. Bars require money. Money requires work, and eventually children, which requires at least one husband or wife, and so on...
Throw in a few fixed aptitudes & attitudes, and you suddenly have a hugely dynamic sandbox where both NPCs, monsters and factions will fuck with you, themselves and each other, and where all of it is wide open to manipulation by the player.
It quickly becomes unmanageable, but it doesn't have to be taken to the extreme. Just have enough simulations to account for what's actually in the game. If there's a bar, guard, farmer, overlord or monster, justify their existence within the context of the other crap in the game. I'm not saying there has to be lumber mills if there's trees in the game, but if there's lumber mills, there better be people cutting up trees and generating wealth. Whether it's simply hauling off lumber to the harbour and selling it to a nebulous elsewhere doesn't matter, what matters is whether and to what extent the NPCs and business involved can mess with/be messed with the player and the rest of the stuff in the game.
hakuroshi said:
A good sandbox game should provide a not-too-easily dispellable illusion of life going on around the player, not nessesary simulate it actually. While realisation of ecology/economic as a core element of a game could be a great foundation for sandbox, I think that efforts required could be better applied elswhere. Majority of players who enjoy sandbox games would not spend times tracing crop circulation or following random NPC to verify that he actually doing something meaningful.
The depth of the illusion doesn't matter, what matters is whether it's playable. A game doesn't become a sandbox game because it, like Oblivion, offers the illusion of being more than a corridor with random spawns. It becomes a sandbox game when the gameplay incorporates whatever's in the sandbox.
Whether you want to tail random farmers is really besides the point. The point is that the farmer has dependants and dependencies, properties that can be manipulated by the player (and actors in the game) and have consequences in the gameworld. That you can tail a farmer to see what he does if the game's realised in that much detail, is just added benefit for the weird.