[MUD-Dev] Geometric content generation

Dave Rickey daver at mythicentertainment.com
Wed Oct 3 09:28:56 New Zealand Daylight Time 2001

-----Original Message-----
From: Ola Fosheim Grøstad <olag at ifi.uio.no>
>"Koster, Raph" wrote:

[the behaviour of rats are representative of humans]

>> Well, if you believe the behavioral psych people, there is no
>> difference, right? :)

> If you believe a carefully selected subset of behaviourists,
> maybe. Hmm, how to put this.  Yes, behaviourist thinking can
> probably be valuable for training people with braindamages, maybe
> the only way. Does that mean that we should treat most people as
> if they are braindamaged!?  YES, the gamedesigners shout, we are
> catering for the masses! :^)))

I have yet to find an example of mass game behaviour in these games
that can't be accounted for by either behaviourist theory or game
theory (which is related).  Make any game design assumptions that do
not carry the implicit assumption that the players will pursue their
own interests and rewards above all else, and you're guaranteed to
end up disappointed.

> Oh, wait, no feedback is never good because then you don't know
> what you are doing.  However, they will pursue the activities if
> there is no explicit reward IF that activity is _meaningful_ to
> them (or compelling).  The reward does not make the activity
> itself meaningful though (unless it is strongly related to the
> activity).  However, sometimes you need a carrot to get people
> started, but when they are IMMERSED into the subjectively
> MEANINGFUL activity then they don't need the carrot anymore.  If
> you have to PAY your players to participate in activities, then
> there is bound to be something boooring and unsatisfactory about
> those activities.  People go biking and skiing without getting
> paid, don't they?  They also ROLEACT without getting paid.  Those
> activities are inherently meaningful to those that participate in
> them.  I believe that they partially are so because the player
> himself can increase the level of challenge and change the focus
> and variation, but partially also because they have a very high
> potential for being "compelling" along multiple dimension.

Not from what I've seen.  People roleplay when roleplay is a means
to a desired goal, period.  Oh, you always get abberations, but when
the system does not reward them, that's all they are.  In a small
group with a dedicated GM, it is possible to make RP a rewarded
activity, and you can get a lot more of it.

Most of the problem comes from recognizing the behaviour, a computer
can only reward what it can measure.  It's feasible (albeit somewhat
challenging in practice) to measure and reward the traditional
XP-hunting PvE activities in a way that cannot be manipulated (or
more precisely, can only be manipulated in desired ways).  It's much
more difficult for an automated system to measure and reward social
activity, or building activity, or other internalized activity, in a
way that is not subject to manipulation that is more focused on
tricking the measuring system than actually pursuing the activity.
So our systems *don't* for the most part.  AC's allegiance system
was a primitive attempt to measure and reward social activity, The
Sims Online's "Relationship Web" looks to be another angle on the
same thing.

EQ, and to an extent Camelot, take a different tack, by trying to
present the players with challenges within the traditional PvE
reward context that cannot be met without the players first solving
problems within the social context (like how to reliably organize
40+ coordinated players of the right classes to go on a Dragon-Raid
class encounter in EQ).

--Dave Rickey

MUD-Dev mailing list
MUD-Dev at kanga.nu

More information about the MUD-Dev mailing list