Procedural content and emergent gameplay, a brainstorming

[cmagoun]

I think emergent properties can come from higher level abstractions.

It is number crunching like a 4X game that AIs can do but it is playing around with itself, creating an enviroment that the player can alter.

This is a good idea.

I was thinking about this thread over the weekend and I came to a similar conclusion: We already have numerous examples of computer "AI" that can play a strategy game reasonably well. Why not use a similar concept for an emergent RPG? So, instead of sweating over the AI of the 100s of (mostly) meaningless NPCs in the game world, focus on the interactions of faction-level AI on a strategic gameboard and then create a component that "translates" the action of the factions into situations with which the PC can interact.

For instance, imagine a generic fantasy RPG that takes place in a single town. The Thieves' Guild "scores" in the strategy game by increasing its wealth. So, it plans robberies, sets up businesses, tries to capture other factions' caravans, etc. The Guard faction scores by lowering the crime rate in the city. They send out patrols to guard caravans, scour the streets for thieves and raid their safehouses and front businesses. The Merchants just want to get their caravans through the wilderness and keep their businesses safe and profitable.

The PC interacts in this game through rumors, random events, and quests provided by the various actors in the strategic game. The PC might stumble onto a robbery, or walk into a shop to find a member of the Brotherhood of Bad Lighting extorting a shop owner. The leader of the guard might get fed up if he starts losing the game and hire the PC to raid Guild hideouts. If the Guild is losing, they might set up an assassination attempt against the leader of the Guard. The PC could get wind of that attempt and choose to stop it, or aid it, or ignore it.

This type of system would be much easier to bring to fuitition than one that relies on more individual AI interaction. I think it would also be more likely to bring about "meaningful" events in which the player can participate. Unfortunately, to do this, it takes a step away from true emergent gameplay because you have to "translate" the events on the gameboard into RPG events which means... scripting again.

Just a thought brought about by Human Shield's post,

 

[DarkSign]

Well that sounds well and good,but youre still dealing in abstraction. The reason why parts of the discussion are NPC based is because usually its interactions with individual NPCs that make up quests.

Put another way, creating large processes in the background are great, but if it doesnt turn into something concrete that translates into gameplay for the human user, there's no point.

Your comment sort of goes into specifics with the actions carried out towrds the metagoal, but until you really design the game you wont know if those procedures are meaningful for the player.