An Engineer's View of Game Design

I once attended a lecture by Doug Church at the Game Developers
Conference in 1997. Doug influenced some of the best hard-core design
(Thief, System Shock, Deus Ex, etc). He tried to put a rational
perspective on the "art" of game design. I summarize his views (A.
immersion and B. economy), which explain his games quite well. I
suggested an additional point, based on the adaptation cycle. These
ideas do not apply just to games, but to any kind of entertainment.
Our discussion never got written down, but I'm appending my notes at
the end of this e-mail.

Anyway, if you look at the reward section, you will notice that
different people prefer different kinds of reward. This explains why
there are different genres of games. Some prefer cognitive rewards
(puzzle solving), others prefer dramatic rewards (adventure games),
yet others prefer kinesthetic rewards (arcade games), or glandular
rewards (...). A good game is able to provide a set of rewards. Also,
not all games do the full cycle of challenge-learning-reward.
Non-interactive "games" only have challenge-reward, because you
observe a character that does the learning for you. Some games have
trivial challenges, meaning that it takes little planning, everything
is focused on the cycle between learning/acting and reward/punishment.
On the other hand, some games, especially turn-based strategy, are
predominantly challenge-learning/acting, because the rewards and
punishments are very far off.

I know I'm reinventing the wheel, and that I'm outdated, but
nonetheless this might incite some discussion.

Best regards,
    Aleks Jakulin


DOUG CHURCH'S MODEL:

===
* A: game as a model of the world:

1. Immersion is the aim.
2. Immersion requires involvement (the player must care).
3. Involvement requires:
      - Power of expression: an orthogonal, consistent, simple,
reasonable way of interacting with the environment
      - Control over destiny: the player is solely responsible for his
destiny, he cannot blame others, he can only blame himself, therefore
replaying the game, the player must feel he can achieve the goal
      - Facilitated planning: the player must be allowed to plan, the
player must know what's the aim of the game

* B: economics of game development:

1. An ideal game would take an infinite amount of time.
2. How to reduce the development time for the same gameplay effect:
      - Reuse the world: keep the player interested when he has to
re-explore the world for several times: different tasks, modifications
to the environment, requires attention to different elements
      - Weak task info: let the player fill in the details
      - Nonlinearity: let the player choose the order in which
problems are solved


ALEKS' MODEL EXTENSIONS
===

* C: Adaptation cycle:
---
There are three mechanical psychological sources of player's joy when
playing the game, challenge->learning/acting->reward/punishment. The
role of immersion (A.) is to free the player from all distractions
while running in this circle. A good game should let this cycle spin
and spin and spin. It should be smoothly started (so that the player
doesn't drop out), varied in speed (so that it is not boring), and
stopped in one of the following ways: slowly (relaxing the player,
comforting him), or abruptly (leaving the player confused, puzzled and
(hopefully) hoping for a sequel).

CHALLENGE:
      - Goal
      - Problem

Examples of implicit challenges:
      - learn more stuff
      - finish the game
      - get more power
      - get to a new level
      - see more graphics
      - create a superman out of your character
      - beat a friend
      - exploring: new graphics, story, music, enemies


LEARNING-ACTING:

It's possible to reduce most of the satisfaction with the explorative
elements of the game to *learning*. About learning:
      - The player must be learning all the time, this is motivating
the player to continue
      - Learning gives the player a sensation of progress, and a
continuous inflow of satisfaction
      - If the player learns something useful in the domain outside of
the game itself, the game is no more a "waste of time." This implies
even more satisfaction.
      - Learning how to deal with the enemies
      - Learn by experimentation

What isn't learning:
      - Tediousness (no goal, just routine, fighting against the
enemies you have already mastered)
      - Boring stuff blocking the player from continuing towards the
goal
      - Problems that were solved before
      - Lack of new information
      - Repetition
      - Irrelevance (figuring out alien lettering, what's that useful
for if you're not a puzzle junky?)
      - Limitedness
      - Anticipated predictability

REWARD:

There are many kinds of "reward":
- EMOTIONAL/DRAMATIC, achieved through self-identification and
catharsis. Read Aristotle's Poetics, especially his discussion of
tragedy. That's how the story should be designed in order to get
maximum impact on the beholder. Also, both punishment and reward is
provided through association, once the self-identification has been
achieved. Keywords: relevance, self-identification, catharsis.

- SOCIAL, achieved through achieving status among other (human!)
players. This explains why multiplayer games are popular, and why
people discuss about their successes in playing single-player games.
Keywords: success, prestige, ranking, score, level

- COGNITIVE, achieved through player learning useful skills, seeing
new stuff, learning the rules of the (real!) world and solve problems.

- KINESTHETIC, achieving proficiency in controlling the avatar,
pressing the keys, driving the car, etc.

- GLANDULAR: induced excitement (achievement, beauty, fear, sex,
taste) and the atmosphere.


Managing reward:

The excitement should appear in a rhythm:  Continuous extreme
excitement can cause the player to get killed, or not to notice the
excitement any more. Let there be several peaks of increased
excitement, with plateaus between them, letting the adrenaline levels
in blood to drop down, so that they can burst again later. The player
enjoys to relax after moments of intense action. Anticipate this
relaxation, and facilitate it.

Read (not watch!) Macbeth, or watch a movie (thrillers are very
appropriate) to see how to manage several peaks of excitement.

References:
http://www.theinspiracy.com/400_project.htm
http://www.gamedev.net/reference/list.asp?categoryid=23
http://gamestudies.cdis.org/~rocketship/ionstormterms.htm
http://www.gamasutra.com/features/20011012/garneau_01.htm
http://www.planetdeusex.com/witchboy/articles/thefuture.shtml
http://www.gamasutra.com/features/19990716/design_tools_01.htm