[MUD-Dev] The Best Guy on the Mud Thing

[J C Lawrence]

On Wed, 15 Sep 1999 01:49:52 -0500 
KiZurich  <Eli> wrote:

> I do not think that spare time alone should decide the "elite" of
> a mud.  However, I think it would be more productive to limit the
> effectiveness of the *playstyles* of powergamers, not to penalize
> players being online.  Just MHO.

I'd classify a finer distinction:

  What defines the most capable players in the game?

    Is it those players with the "best" stats?  (Prior advantage
    taken of the game world)

    Is it those players with the greatest (personal) knowledge of
    the game/world?  (prior OOC knowledge gained)

    Is it those players with the largest set of social ties to other
    players (currently playing, or perhaps just in total)?  (prior
    social activities)

   Or, is it some combination of the above, and if so, in what
   ratios?

Thinking about this briefly (its late and I'm tired) suggests that
this parallels the player suits in a rather interesting manner.  Its 
not really clubs, diamonds, hearts, and spades, but it also is 

> Ex: A player "hides" once in a room, moves and hides again
> quickly.  Hide has a frequency of one minute.  The second hide is
> not considered for improvement.  Two minutes after the first hide,
> the player hides again, four times fairly quickly.  Only the first
> of the four is considered for improvement, and it will be at least
> four more minutes before the player has a chance to improve it
> more.

I've been toying with the idea of attempting to detect character
stress and varying the skill improvement delta dependent on what the 
level of character stress is, and how significant the outcome of the 
skill application is.  There's a really softie area in here, but as
with the prior discussions, it seems that with just some basic
history keeping you can actually get a pretty decent idea of player
intent and from there derive a concept of risk and thus significance 
and stress.

Thus, Bubba hiding in the swamp while nobody is about does him
almost nothing.  Conversely, Bubba being chased by Tiamat and then
succssfully eluding Tiamat (high risk) by hiding in the swamp
(highyl significant outcome) would gain significantly more.  

One problem with this approach, tho it could be a benefit, is that
the concept of "practicing" goes away.  The only way to learn a
skill in this world is to go out and use it in real circumstances
where to some extent you put yourself on the line in doing so.

How basketweaving basketweaving or fishing (indirectly
survival-related skills) fits in here I'm not quite sure.  Perhaps
another pattern for those?

> Spamming a skill (or bot-like repetitive behavior) will have an
> effect opposite the desired.  If someone hides 60 times in a row,
> it will be at least an hour before they could hope to improve it
> again.  

This does not prevent the skill training from being scripted.  All
that is required is that the training script keep a tally of the
time since the last exercise of the skill and then runs it when
optimal.  In the "hide" case, you'd just find your character
seemingly randomly trying to HIDE every few minutes as the script
notices that the clock has wound down enough again.

Get enough of these scripts running of course and you could have a
character that just sits there, frantically excercising all his
skills in a carefully timed sequence as each one's clock winds out.

> The system would favor players who use a skill, then let it sit,
> and use it again later.  

Or who have a script that runs in the background that knows to try
and "XXX" every 90 seconds, and does it automagically.

> Of course, if players are told the details of the system, it could
> still be abused, but only once a minute.  :) 

Quite.  Given the reverse engineering efforts I've done and seen
done on various systems (the most recent example being the full stat 
progression charts empirically derived from UOL players), I'd take
it as a given that all players would know your mechanics.

That's the problem with mechanical systems.  They're mechanical.  As 
such they are inherently predictable and manipulable in determined
manners as machines.  Even if you add random factors the graphs in
almost all cases can still be approximated in a sufficiently useful
fashion by quite a simple curve.  

Heck, I don't care how many factors you have, almost any curve can
be approximated to a fairly useful degree by a couple straight
lines, a simple quadratic, a simple wave, or an asymptote.  There's
no magic there.  Jsut get enough data points and sketch something
roughly out and can have quite a significant player advantage.

> ...  But I have no idea how it might work out with real players.
> Ones who have the spare time to figure out how to exploit the
> system...  :)

That's the interesting bit: How to derive a systems which implements
non-deterministic feedback responses to each character
individually/dynamically without becoming so random as to lose all
sense of predictability and thus lose all sense of the possibility
of attaining a/the goal.

<ponder>

--
J C Lawrence      Life: http://www.kanga.nu/   Home: claw at kanga.nu
---------(*)                Work (Linux/IA64): claw at varesearch.com
 ... Beware of cromagnons wearing chewing gum and palm pilots ...


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