[MUD-Dev] Evolutionary Design

Sasha Hart Sasha.Hart at directory.reed.edu
Wed Jun 12 04:34:41 New Zealand Standard Time 2002

[Daniel Cook's article]

> I used a crude yet remarkably effective method of observing the
> game.$A0 While playing the game, I watched my fellow players.$A0
> Whenever there was frustration or boredom, I asked why.$A0 I also
> asked people to make suggestions if they noticed something that
> felt odd or could be improved.



With some games, I don't really know if they even bothered to play
them first. Maybe that has something to do with the very rigid
top-down design philosophy.

> The use of web forums, open and closed betas can glean valuable
> knowledge from outside testers.

I would personally trust web forums about as far as I can throw 'em.
*I* have posted stuff that was basically nonsense in such forums ;)

Did I say that out loud?

> The amount of psychological pleasure, or value, of rewards is
> dependent on a wide variety of factors, but there are several key
> ones.

It's simply untrue that things which cost more are more
pleasurable. Would you rather pay $1 for a great slice of pie, or
$500? Would you rather work 40 hours a week for your paycheck, or

The econ 101 exception is certain luxury goods. I think a pretty
good case could be made that this works because the price is an
indicator of what you can get out of it (if nothing else, in terms
of the reactions of others).

> I find that staggered reward schedules tied to the player's
> investment work best.$A0 Imagine a set of rewards given
> approximately every thirty seconds, every minute, every 5 minutes,
> every ten minutes, and every 20 minutes.$A0$A0 The result is that
> the player is constantly bombarded by positive reinforcement.

Intermittency of rein. can be interesting, especially if the basic
task isn't so interesting, but more important here, I think, is the
contingency - the bit where the rewards I get depend on what I do,
at least enough to trick me. I have a hunch that this is
particularly important in games, where the "rewards" ultimately
aren't valuable in and of themselves, but perhaps as feedback about
performance per se. For me anyway, I at least need an illusion that
the game is about me, my actions, and doesn't just impinge on
me. Some behavior really does depend on the schedule matching up to
what has been done prior, not just the timing of
rewards. Schedule-induced behavior is worth knowing a little about,
but isn't usually so interesting to study or engage in. Unless you
get a big kick out of timing or counting.

> However, the 20-minute reward is inherently more valuable than the
> 30 second award because the player must invest playing time in
> order to achieve it.$A0

Again, maybe valuable in the sense that it cost more, or was rarer;
not valuable in the sense that the player gives it more value,
e.g. prefers it.

> Many players will become numb to the pleasures associated with
> 30-second rewards, but will still play for several minutes longer
> to reach the 'larger' reward. Sid Meier used this technique to
> great effect in Civilization.$A0 How many players were bored with
> killing a single unit, but kept playing to finish construction of
> the Wonder they started 20 minutes ago?

Wonders weren't valuable just because they took forever to make
(otherwise you could have stupid units in civ that took 100 turns to
make and would therefore be well loved??!?) but because, and this is
anecdotal, they showed you a graphic, etc. and did amazing things
for your empire. Also, of course, because they weren't beaten to
death so much that I ignored them (habituation; that set in when I
started cheating).

> Imagine a hill.$A0 The high points on the hill are areas of great
> player enjoyment.$A0 The low points make players miserable.$A0 The
> evolutionary process is often called a hill-climbing algorithm.$A0
> Your initial game idea is a point on the side of the hill.$A0 Each
> iteration of the design process, you stop and "Ask which direction
> should I move to increase player enjoyment?"$A0 Slowly, but surely
> your game will climb the hill. The equilibrium point of
> evolutionary design corresponds to the peak of the player
> enjoyment.

As noted elsewhere, hill climbing needs smooth hills to climb, and
even then gets stuck on local maxima. The way to fix either is to
take differently sized steps up the hills. In other words, to sample
the space more widely! Am I saying that this evolutionary concept is
a bad idea? NO.  I couldn't write code, work out an experimental
procedure or post to the list if I didn't have the ability to revise
and retry repeatedly. But the more angles I really try (e.g. go
through the iterative process enough to get a hint about the part of
the space I'm in), the more likely I am to hit at least a higher
local maximum rather than a dud.

With the preface that one's trash is another's treasure, I will talk
about Counterstrike. Before it became the juggernaut it is now, it
was a bit of a shot in the dark. I remember being utterly fascinated
and frustrated by the "unfairness" of it. Racking up score would
give me weapons which help me rack up score even more. I couldn't
stop playing it. Several releases later, it was definitely a more
mature game. After that it was mature enough to secure its author a
job working on his baby; by all accounts, a tremendous success. It
climbed the hill, right?

The problem is that you not only have problems within the
fitness-scape of the game you are working on, you have another
problem: WHICH fitness-scape you are hill-climbing in! I lost
interest in counterstrike at about the third or fourth release and
never went back for more than a few minutes. I wouldn't buy a
counterstrike product. My pet theory for why this is the case has
always been that the designers of the game paid so much attention to
their feedback that they lost the aspects of the early game that I
thought were really new and hot. Player demand for conservative
"balance" upset a game which I thought was brilliantly unbalanced
which made it unique. Subjective? Yes. So is any feedback you will
receive. I don't know if I feel that players of a game are also good
direct analysts of games. I don't question the need for testing, but
players' opinions need to be taken with a big FAT grain of
salt. Some forms of bitching are, I believe, a natural byproduct of
a game which has reached top shape. ;)

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

More information about the MUD-Dev mailing list