Game Development Reference
In-Depth Information
the numbers, the odds of the two potential outcomes would change as well. The
histogram is representative of the odds. As such, the shape of the curve that it ex-
hibits is an important tool in determining what outcomes we can draw from its data.
Throughout this topic, I will use the word
curve
often in regard to graphs of
data. While the line implied by the data may or may not be curved, I will use
curve
regardless. Please do not get out a straight edge to determine if a line that
looks straight to you actually has a curve to it simply because I referred to it as
such.
S
EGMENTING THE
P
OPULATION
In the Guess Two-Thirds Game in Chapter 6, we identified some general categories
of people. There were the random guessers who we termed Index 0. We saw a spike
in the number of people who guessed 33 (Index 1) and a spike in the number who
guessed 22 (Index 2). Those were three distinct categories for which we could make
a reasonable case as to
why
they selected what they did. However, a significant
majority of the population guessed plenty of answers that were neither mindlessly
random nor calculatingly centered on the magical numbers of 33 and 22. Why did
those people guess that way? The distribution of guesses showed that it wasn't com-
pletely random, but it wasn't completely rational either. So how do we get into
those people's heads?
The truth is that we don't really need to get inside their heads as to exactly why
they chose the way they did. If we were to ask 100 members of that large majority
of people what their rationale was in choosing their number, chances are we would
get 100 different answers. Everyone has a different thought process that may or may
not be based on logical premises. It may even be based
partially
on a logical premise
but has been obscured or adulterated by an interloping,
non
-logical thought. In
Chapter 6, we covered many potential pitfalls that, if they don't completely ensnare
a person's train of thought, can at least make it stumble somewhat. Each of the in-
dividual person's perceptions and thought processes may be composed of a differ-
ent cocktail of truth, falsehood, error, and subjective coloring. Trying to model all
of those individual rationales is not only prohibitive but usually largely irrelevant.
We must ask ourselves, what problem are we trying to solve? If we were (for
some obscure reason) trying to simulate people trying to play the Guess Two-Thirds
Game, we would have to include three different approaches (Figure 11.2). First, we
would have to model the intelligent calculators. This is actually a fairly straightfor-
ward process of determining what percentage of people picked those two numbers
(33 and 22).
