Game Development Reference
In-Depth Information
A
NALYZING A
S
INGLE
S
EGMENT
To analyze
only
the population of semi-logical guessers, we will isolate them from
the graph in Figure 11.2. If we look at the counts of the guesses of the people who
are in this category, we may see something like what is in Figure 11.3. Note that un-
like the random guessers, who were equally probably all across the range from 0 to
100, and unlike the targeted guessers who focused
only
on the numbers 33 and 22,
these guessers exhibit certain characteristics. Let's describe the observable features
of this group one by one.
Their guesses range from 0 to about 70.
The majority of people guessed between about 15 and 35.
There was a minority who guessed from 0 to about 15.
There was a minority who guessed from about 35 to about 60.
There were some rare “outliers� who guessed above 60.
Significantly more people guessed any given number in the majority range
(e.g., 27) than guessed any given number in the minority ranges (e.g., 50)—as
much as three or four times as many.
FIGURE 11.3
The guesses of the semi-logical guessers shown
are spread over the range from 0 to 70. The majority of the
guesses are clustered between approximately 15 and 35.
Even without the graph, from that list of descriptive characteristics, we can
begin to generalize what the population of semi-logical guessers looked like. We can
then start to construct an artificial curve using those characteristics that we could
use to represent the population. But what would we do with this information?
