Re: Goals and motivations

I think MadKeith is right on this one, at some point you have to start the story moving. I think this would give you more authorial control, btw, then evolving goals from the ground up. Evolving might work, but the recent stuff on learning game AI I''ve seen doesn''t fill me with a lot of hope.

Now, I think you could change motivations along the way (in fact, good strategy AI does this... is it time to build or fight?) and like DungeonMaster said this would make for logical and cohesive plot. For instance, V doesn''t go after the king anymore if the player does something to make this impossible.

But as far as motivations go, you have to start somewhere. (Although, there might be something to randomly generating a character''s history like in the old role-playing book central casting... but I still think you''d have to be sure to generate from certain tables, like a ''villain'' table, and a ''supporting character'' table. Otherwise, how can you be sure what you create will have conflict?)

--------------------
Just waiting for the mothership...

quote: Original post by Wavinator
I''d favor this approach because, in terms of AI, it can be systematized. For instance, you could make loyalty a stat and have the king do a search on the most loyal character (heh, which could be the player too ). With this sort of key and lock approach, the king AI might be able to look at X in the same way a player looks at heroes in an empire game (as a modifier for armies or cities or some other important post). On a basic level, the king really seeing solutions that will fill a puzzle (which we''d hardcode as his motivation, too keep the kingdom running).

Of course, how loyal a character seems might have nothing to do with how loyal he/she intends to be.

-Jussi

Interesting read guys.
Seems like you guys want to create a never-ending game huh? Cool!

I agree with the idea of generating different *class* of NPCs. Of course this is just a start. During game different stats could change according to the level of interaction with the game world.

A question. What if the NPC does over-thrown the king. Say V sucessed. What then? Does he become the kind? Inherit previous king''s goals? He will have his own goals based on his on state and interaction with the game world. If V character is evil, he''s probably going to power trip.

What if the player over-throws the king? The game becomes a ruling game?

Although *I* wouldn''t mind if the game became a ruling game, I think most players would have a problem with that. So we need to have a normal mechanism for some NPC becoming king in the event of the king''s death.

And I think it''d be interesting if V did become king. Then the player would have a whole new set of problems to deal with (hah! forget about boss monsters, pal! How about the entire Royal Guard!)

''Course, I could see this branching out in so many different ways. Perhaps the player has to pretend to support the new, evil king V? Or go into hiding and gather strength? Maybe some other disenchanted NPCs join him, and we play out a sort of Man in the Iron Mask tale? Rich with possibilites!!!

--------------------
Just waiting for the mothership...

Ahh you are starting to see just how rich this could be
And why wouldn''t the player want to become king? Could be fun! Half the other kingdoms trying to assassinate you or bribe you, wars to run, taxes to collect, ministers to keep happy. You''ll have your work cut out for you

quote:
> SIR! The peasants are revolting!
> Yes, they''ve always BEEN revolting, but now they are rebelling too.

People might not remember what you said, or what you did, but they will always remember how you made them feel.
~ (V)^|) |<é!t|-| ~

Hey, I''m must be mad. Stayed up all night reading posts on the fourm, especially here in game design.

Too bad I only have one class later today...NOT!!!

Anyone down to code some concepts? I will tried to squeeze it in bettween school and my current project.

I''m already looking around a bit to see if I can find an easy way to implement the basics of this ( i.e. a small 2D tile based framework for studying the emergent behaviours )


People might not remember what you said, or what you did, but they will always remember how you made them feel.
~ (V)^|) |<é!t|-| ~

Actually, I think we are going a bit too far off tracks here.
THe initial reason I started this thread was to understand how we could change the classic :
go to L1 to trigger switch S1, then come back to talk to C1. If S1 is true, then you get the reward, otherwise you get a message that you have to finsih the quest.
What I''d like to see is a new way to create quests. By taking into accounts the characters involved. Just making them actually move when they say "I''ll meet you there", so that if they don''t meet you there, they must have had a problem, or you''ll simply camp until they arrive, or you can''t wait but you can''t afford going on without the support, so do you split your party ?
I think what it come down to is a better way to create quests (or whatever we might call them .. maybe game goals ?), because if you look at it, I think you could probably make a quest as deep as in most RPG with the current entity scripting that FPS games offer.
So the question is , how can we model a quest in a formal way ? What kind of parameters ? shoud we use scripts, should we leave room for emergence jsut like you all described ? How do we give the impression of freedom (that would imply emergence) without losing the ability to shape the story as we want (as designers/writers).

I think your answer is that folks have now designed away the traditional ''quest'' in favor of something with far more depth.

NPCs will have desires. One way to accomplish the goals is to have a hero do the task. The nice thing about the more in-depth NPCs is that it does away with frivolous tasks that it would be easier for the NPC to accomplish on their own.