Looking at each player type in more detail, then:

I) Achievers regard points-gathering and rising in levels as their main goal, and all is
ultimately subservient to this. Exploration is necessary only to find new sources of
treasure, or improved ways of wringing points from it. Socializing is a relaxing method
of discovering what other players know about the business of accumulating points, that
their knowledge can be applied to the task of gaining riches. Killing is only necessary to
eliminate rivals or people who get in the way, or to gain vast amounts of points (if points
are awarded for killing other players).

Achievers say things like:

"I'm busy."
"Sure, I'll help you. What do I get?"
"So how do YOU kill the dragon, then?"
"Only 4211 points to go!"

ii) Explorers delight in having the game expose its internal machinations to them. They
try progressively esoteric actions in wild, out-of-the-way places, looking for interesting
features (i.e. bugs) and figuring out how things work. Scoring points may be necessary
to enter some next phase of exploration, but it's tedious, and anyone with half a brain
can do it. Killing is quicker, and might be a constructive exercise in its own right, but it
causes too much hassle in the long run if the deceased return to seek retribution.
Socializing can be informative as a source of new ideas to try out, but most of what
people say is irrelevant or old hat. The real fun comes only from discovery, and making
the most complete set of maps in existence.

Explorers say things like:

"Hmm..."
"You mean you *don't know* the shortest route from <obscure
room 1> to <obscure room 2>?"
"I haven't tried that one, what's it do?"
"Why is it that if you carry the uranium you get radiation
sickness, and if you put it in a bag you still get it, but if
you put it in a bag and drop it then wait 20 seconds and pick it
up again, you don't?"

iii) Socializers are interested in people, and what they have to say. The game is merely
a backdrop, a common ground where things happen to players. Inter-player
relationships are important: empathizing with people, sympathizing, joking, entertaining,
listening; even merely observing people play can be rewarding - seeing them grow as
individuals, maturing over time. Some exploration may be necessary so as to
understand what everyone else is talking about, and points-scoring could be required to
gain access to neat communicative spells available only to higher levels (as well as to
obtain a certain status in the community). Killing, however, is something only ever to be
excused if it's a futile, impulsive act of revenge, perpetrated upon someone who has
caused intolerable pain to a dear friend. The only ultimately fulfilling thing is not how to
rise levels or kill hapless drips; it's getting to *know* people, to understand them, and to
form beautiful, lasting relationships.

Socializers say things like:

"Hi!"
"Yeah, well, I'm having trouble with my boyfriend."
"What happened? I missed it, I was talking."
"Really? Oh no! Gee, that's terrible! Are you sure? Awful, just
awful!"

iv) Killers get their kicks from imposing themselves on others. This may be "nice", i.e.
busybody do-gooding, but few people practice such an approach because the rewards
(a warm, cozy inner glow, apparently) aren't very substantial. Much more commonly,
people attack other players with a view to killing off their personae (hence the name for
this style of play). The more massive the distress caused, the greater the killer's joy at
having caused it. Normal points-scoring is usually required so as to become powerful
enough to begin causing havoc in earnest, and exploration of a kind is necessary to
discover new and ingenious ways to kill people. Even socializing is sometimes
worthwhile beyond taunting a recent victim, for example in finding out someone's playing
habits, or discussing tactics with fellow killers. They're all just means to an end, though;
only in the knowledge that a real person, somewhere, is very upset by what you've just
done, yet can themselves do nothing about it, is there any true adrenaline-shooting, juicy
fun.

Killers says things like:

"Ha!"
"Coward!"
"Die!"
"Die! Die! Die!"

(Killers are people of few words).

How many players typically fall within each area depends on the MUD. If, however, too
many gravitate to one particular style, the effect can be to cause players of other
persuasions to leave, which in turn may feed back and reduce the numbers in the first
category. For example, too many killers will drive away the achievers who form their
main prey; this in turn will mean that killers will stop playing, as they'll have no worthwhile
victims (players considered by killers to be explorers generally don't care about death,
and players considered to be socializers are too easy to pose much of a challenge).
These direct relationships are discussed in more detail towards the end of this paper.

For the most part, though, the inter-relationships between the various playing styles are
more subtle: a sharp reduction in the number of explorers for whatever reason could
mean a gradual reduction in achievers, who get bored if they're not occasionally told of
different hoops they can jump through for points; this could affect the number of
socializers (the fewer players there are, the less there is to talk about), and it would
certainly lower the killer population (due to a general lack of suitable victims).

Making sure that a game doesn't veer off in the wrong direction and lose players can be
difficult; administrators need to maintain a balanced relationship between the different
types of player, so as to guarantee their MUD's "feel". Note that I am not advocating any
particular form of equilibrium: it is up to the game administrators themselves to decide
what atmosphere they want their MUD to have, and thus define the point at which it is
"balanced" (although the effort required to maintain this desired state could be
substantial). Later, this paper considers means by which a MUD can be pushed in
different directions, either to restore an earlier balance between the player types, to
define a new target set of relationships between the player types, or to cause the
interplay between the player types to break down entirely. However, first a means is
required of formally linking the four principal playing styles into aspects of a unified
whole; this helps account for different degrees of adherence to particular styles, and
aids visualization of what "altering the balance" of a MUD might actually *mean*.

INTEREST GRAPH

Consider the following abstract graph:

ACTING
Killers | Achievers
|
|
|
|
|
PLAYERS -------------------+------------------- WORLD
|
|
|
|
|
Socializers | Explorers
INTERACTING

The axes of the graph represent the source of players' interest in a MUD. The x-axis
goes from an emphasis on players (left) to an emphasis on the environment (right); the
y-axis goes from acting with (bottom) to acting on (top). The four extreme corners of the
graph show the four typical playing preferences associated with each quadrant. To see
how the graph works, it is appropriate to consider each of the four styles in detail:

I) Achievers are interested in doing things to the game, i.e. in ACTING on the WORLD.
It's the fact that the game environment is a fully-fledged world in which they can immerse
themselves that they find compelling; its being shared with other people merely adds a
little authenticity, and perhaps a competitive element. The point of playing is to master
the game, and make it do what you want it to do; there's nothing intrinsically worthwhile
in rooting out irrelevant details that will never be of use, or in idling away your life with
gossip.

Achievers are proud of their formal status in the game's built-in level hierarchy, and of
how short a time they took to reach it.

ii) Explorers are interested in having the game surprise them, i.e. in INTERACTING with
the WORLD. It's the sense of wonder which the virtual world imbues that they crave for;
other players add depth to the game, but they aren't essential components of it, except
perhaps as sources of new areas to visit. Scoring points all the time is a worthless
occupation, because it defies the very open-endedness that makes a world live and
breathe. Most accomplished explorers could easily rack up sufficient points to reach the
top, but such one-dimensional behavior is the sign of a limited intellect.

Explorers are proud of their knowledge of the game's finer points, especially if new
players treat them as founts of all knowledge.

iii) Socializers are interested in INTERACTING with other PLAYERS. This usually
means talking, but it can extend to more exotic behavior. Finding out about people and
getting to know them is far more worthy than treating them as fodder to be bossed
around. The game world is just a setting; it's the characters that make it so compelling.

Socializers are proud of their friendships, their contacts and their influence.

iv) Killers are interested in doing things to people, i.e. in ACTING on other PLAYERS.
Normally, this is not with the consent of these "other players" (even if, objectively, the
interference in their play might appear "helpful"), but killers don't care; they wish only to
demonstrate their superiority over fellow humans, preferably in a world which serves to
legitimize actions that could mean imprisonment in real life. Accumulated knowledge is
useless unless it can be applied; even when it is applied, there's no fun unless it can
affect a real person instead of an emotionless, computerized entity.

Killers are proud of their reputation and of their oft-practiced fighting skills.

The "interest graph" is a representational structure which can chart what players find of
interest in a MUD. The axes can be assigned a relative scale reflecting the ratio of an
individual's interest between the two extremes that it admits. Thus, for example,
someone who thinks that the people who are in the world are maybe twice as important
as the the world itself would lie on a vertical line intersecting the x-axis at a point 1/6 of
the distance from the origin to the left edge; if they had little interest in bending the game
to their will, preferring their actions to have some give and take, then they would also lie
on a horizontal line at the bottom of the y-axis. The intersection of the two lines would
put them in the socializer quadrant, with leanings to explorer.

It is, of course, possible to analyze the behavior of individual players quantitatively by
processing transcripts of their games. Unfortunately, this is very difficult to do except for
very limited domains (e.g. forms of communication (Cherny, 1995a; Cherny, 1995b)). An
alternative approach might simply be to ask the players what they themselves like about
a particular MUD: even a short questionnaire, completed anonymously, can give a fair
indication of what players find enjoyable (Emert, 1993). Such information can then be
used to determine the make-up of the MUD's player base, so that in times of falling
player numbers the current composition could be compared against some earlier ideal,
and remedial action taken to redress the imbalance. This "ideal" configuration would,
however, be specific to that particular MUD, and its precise form is therefore not
addressed here. Instead, the more general issue of how to alter the balance between
player types is considered, along with the gross effects that can be expected to follow
from having done so.

CHANGING THE PLAYER TYPE
BALANCE

A stable MUD is one in which the four principal styles of player are in equilibrium. This
doesn't imply that there are the same number of players exhibiting each style; rather, it
means that over time the proportion of players for each style remains roughly constant,
so that the balance between the the various types remains the same. Other factors *are*
important, to do with the rate at which new players arrive and overall player numbers, but
their consideration is not within the brief of this paper; the interaction between players of
different types *is* within its brief, however, and is discussed in some detail later.

The actual point of balance (i.e. whereabouts in the interest graph the center of gravity of
the individual players' points lies) can vary quite enormously; it is up to individual
administrators to determine where they want it to lie, and to make any programming or
design changes necessary to ensure that this is where it actually does. What kind of
strategies, though, can be employed to achieve this task?

In order to answer this question, consider the interest graph. If it is regarded as a plane
in equilibrium, it can be tilted in a number of ways to favor different areas. Usually, this
will be at the expense of some other (opposite) area, but not necessarily. Although tilting
can in theory occur along any line in the plane, it makes sense (at least initially) to look
at what happens when the tilt lines coincide with the x and y axes if the graph.

What follows, then, is a brief examination of means by which a MUD can be adjusted so
as to favor the various extremes of the interest graph, and what would happen if each
approach were taken to the limit:

PLAYERS

Putting the emphasis on players rather than the game is easy - you just provide the
system with lots of communication commands and precious little else. The more the
scales are tipped towards players, though, the less of a MUD you have and the more of
a CB-style chatline. Beyond a certain point, the game can't provide a context for
communication, and it ceases to be a viable virtual world: it's just a comms channel for
the real world. At this stage, when all sense of elsewhere-presence is lost, you no longer
have a MUD.

WORLD

Tilting the game towards the world rather than its inhabitants is also easy: you simply
make it so big and awkward to traverse that no-one ever meets anyone in it;
alternatively, you can ensure that if they do meet up, then there are very few ways in
which they an interact. Although this can result in some nice simulations, there's a loss of
motivation implicit within it: anyone can rack up points given time, but there's not the
same sense of achievement as when it's done under pressure from competing players.
And what use is creating beautifully-crafted areas anyway, if you can't show them to
people? Perhaps if computer-run personae had more AI a MUD could go further in this
direction (Mauldin, 1994), but it couldn't (yet) go all the way (as authors of single-player
games have found (Caspian-Kaufman, 1995)). Sometimes, you just *do* want to tell
people real-world things - you have a new baby, or a new job, or your cat has died. If
there's no-one to tell, or no way to tell them, you don't have a MUD.