[MUD-Dev] Re: Quest engines

[ApplePiMan at aol.com]

At 10/3/98 12:11 PM Raph Koster (rkoster at origin.ea.com) altered the 
fabric of reality by uttering:

>> -----Original Message-----
>> From: James Wilson [mailto:jwilson at rochester.rr.com]
>> Sent: Saturday, October 03, 1998 12:30 PM
>> To: mud-dev at kanga.nu; Koster, Raph
>> Subject: Re: [MUD-Dev] Re: Storytelling vs. Simulationist 
>> (Was Re: Room
>> de scriptions)
>> 
>> 
>> 
>> On Sat, 03 Oct 1998, Koster, Raph wrote:
>> 
>> >Missed this the first time around--but in any case--there's 
>> no question
>> >that you can get by with less admins. There are MANY systems 
>> out there
>> >for story generation and quest management... just damn few 
>> are in muds.
>> >Plenty of "quest engines" and "storytelling engines" have been
>> >designed--heck, I've designed three myself.
>> 
>> can you give some details?
>
>Uh, sure. :) I guess the basics are something that can easily be
>discussed without violating confidentiality. I've moved this thread to a
>new title to reflect the change in topic.
>
>Most quest engines rely on basic plot elements that are combined and
>recombined in interesting ways. The strength of them is that you can
>build an easily extendable database to generate a TON of unique quests.
>The weakness--quests can seem rote or repetitious or simplistic, and the
>work involved in making the database large enough to prevent this
>problem is very large.

Keeping plots fresh and pertinent, and taking into account world history, 
are indeed the problems that can't adequately be addressed 
programmatically (as far as I can see) by a system with this sort of 
combinatory engine at its core. Your database will require a steady flood 
of new plot elements to keep it vital.

<snipped long example of combinatory quest generator>

>The hard parts:
>
>- Error handling--multiple people participating; someone may find the
>guy in the alley BEFORE the girl crying, for example; or they may get
>the fourth amulet for the mage not knowing that there's a guy out there
>with the third one; or they may kill the spider queen outright instead
>of asking for her help. You can handle these sorts of problems in
>various ways.
>- Making it fictionally compelling. If you make them go after things
>that do not seem to have significance, it'll feel rote and stupid. It
>may still feel rote and stupid anyway, if you don't have enough
>variations.

And another hard one for "extra credit" =) (though this could and should 
actually be subsumed by the "making it fictionally compelling" point): 
determining which pieces to put together in what order to produce the 
storyline with the most (literary) conflict. Pure random selection can 
also lead to rote and stupid stories, or contradictory ones, even with 
individual pieces that are worthy individually.

It's in deciding at what scale to implement the stories, tweaking the 
basic engine and working out solutions to the "hard parts" that 
proprietary systems are born. =)

-Rick.


---------------------------------------------------------
Rick Buck, President and CEO  <mailto:rlb at big-i.com>
Beyond Infinity Games, Inc.
See you in The Metaverse! <http://www.big-i.com>