Game Development Reference
In-Depth Information
amount of animation, we show a power bar. It's both more efficient (requiring fewer
visual assets) and more effective (the player can read it instantly). Besides, the simu-
lation of the fighter's health isn't accurate anyway, because in the game the fighter
fights at full strength until the last moment. The game, a stylized simulation of
fighting, focuses on the most interesting aspects of the system it represents and shows
these aspects with much more clarity.
Considering this difference between games and simulations, it is curious that game
developers spend so much effort on making games more realistic. Realistic games are
like iconic simulations: They try to create mechanics that resemble the mechanisms of
the real thing they represent as closely as possible. Although realism and iconic simu-
lation in games is not a bad thing, it's generally a mistake to concentrate on realism
in games at the expense of enjoyable gameplay or to assume additional realism will
lead to more fun. Games for entertainment should concentrate on communicating
their ideas through other, noniconic forms of simulation instead. Later in this chap-
ter, we will explore the notions of
analogous
and
symbolic
simulation
in more detail.
aBsTracTiOn
In either a scientific or a game simulation, we have to build mechanisms that are
simpler than the mechanisms in real life. This is necessary because otherwise we
would build a replica of the original system, which would run at the same speed and
operate on the same scale. We wouldn't be able to use a replica to fast-forward time
or test ideas in a safe environment. Because a simulation must be simpler than the
system it represents, the simulation designer makes the decision to leave out certain
details. This process is called
abstraction
.
There are two kinds of abstraction:
elimination
and
simplification.
In general, you
can safely eliminate factors from your simulation that have little or no effect on the
operation of the mechanics. In simulating the aerodynamics of an automobile, it
simply may not be worth going to the trouble of including the windshield wipers or
the radio antenna; their influence is too small to bother with. And of course some
details, such as the interior décor, are completely irrelevant.
When we abstract through simplification, we look for features of a simulation that
contribute to its overall mechanics but whose inner workings don't really matter.
Then we model those features in a very simple way, without including those inner
details. An example will show what we mean. Suppose you are trying to model the
effects of military vehicle failure on military readiness on a grand scale—all the
vehicles in an entire nation's armed forces. Suppose that you also know from collected
statistics that one in every 10,000 aircraft landings puts the airplane out of commis-
sion because of damage to the landing gear. Your job is not to actually figure out what's
wrong with the landing gear but simply to include this factor in your model of over-
all military readiness. Instead of modeling the landing gear machinery in detail, you
just build in a random 1-in-10,000 loss factor for landing gear damage. You have
abstracted the landing gear problem to a simple random factor. When you run the
