Game Development Reference
In-Depth Information
Now we have taken the nine criteria that we started with and combined them
into more of a modular format. We are still using them all, but the relationships
between them have become more meaningful. Each time we combine factors, how-
ever, we will have to determine how to weight them and massage them to get the
effect we want.
Going back to our engagement decision example, even starting small, in con-
structing the risk to agent value, we need to determine how important the two
factors—agent's health and enemy's weapon—are, respectively, to that value. Are
they equal? Is the agent's health more important than the weapon? Do we calculate
it in terms of how long the agent can expect to live at a particular
rate
of damage?
Combining intermediate values can be a little more obscure. How do we process
total threat and perceived morale so that we are sure we are going to generate an ap-
propriate balance?
In large part, when dealing with complex, multilayered processes, the key is to
be confident with each step along the way. If you feel good about what went into
two smaller decisions, you can more accurately combine those two results into a
larger one. Likewise, when debugging or even tweaking a full decision, it is impor-
tant to work backward appropriately. If you are not satisfied with one layer, it may
not be the direct inputs to that layer that are the problem. You must question
everything that leads to that point. Referring again to Figure 3.8, if you are not
happy with the result you are getting from the total threat value, you have to ques-
tion not just how you are
combining
risk to agent and threat to enemy together, but
the factors that you used in arriving at those figures to begin with.
All of these approaches are the steps to a very complicated process of crafting
decisions. Once done, however, the resulting behaviors can be deep and powerful…
and yet easily changed and maintained.
That being said, don't expect “solutions� in this topic. Unlike an area such as
pathfinding, there isn't one
problem
here, much less a single
answer
to that prob-
lem. Using the metaphor from the beginning of this chapter, as you proceed
through the topic you can expect to find plenty of brushes, paints, palettes, and
other tools of the trade. You will learn some of the researched and studied history
of the art and how what those masters discovered can be applied in your own work.
You may find suggestions for what colors mix well to accomplish certain effects.
I will even invite you to look over my shoulder to see how I paint some sample
works. However, there is no way I can tell you how to do your own painting. There
are simply too many ways to do it.
All the artistry must be your own.
