[MUD-Dev] Thoughts about smarter Sims

[Justin Rogers]

> > Jon A Lambert wrote:
> > This sort of AI is quite common to other games.  Not so in most muds.
> > The key here is not just designing goals that "win" games for NPCs since
> > in a mud there is usually no "game-over" position for an NPC other than
> > death.  But goals that attempt to maintain and extend a winning position
> for
> > a given NPC.

    The question is whether or not every NPC has to "win"...  In every social
order there exists a slightly different set of winning conditions for every
person, but on top of this there also exists ambition to change a winning
condition to a better winning condition.
    The idea is to make a class of AI that applies to a general group of
NPC's (villagers)...  Let the computer decide an initial Magnitude of what
the NPC has to do to "win", (sell 3 daggers/day or 30 daggers/day).  Ideally
you can reduce any simple AI down to a mathematical equation that is quickly
rendered by computer so training of the Range of AI will be necessary, but
you basically start off with a medium winning condition and pass events through
your net until your villagers raise and lower their expectations until you have
a fairly decent curve on low winning conditions versus high winning conditions.
Then you have a series of functions that can be passed a variable (1..10)
depending
on the magnitude (level) of the NPC, and the appropriate winning condition will
be tested.  An NPC who achieves their "winnage' for a long period of time may
advance in level slowly and become more ambitious.  And those that don't can
lower
their ambitions.  What happens IRL is the best model of what happens in the
game.

    - Justin Rogers, CEO DigiTec Web Consultants
    [COOP]DigiTec - Half-Life, Q3A, Unreal Tournament




_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev