Game Development Reference
In-Depth Information
Evil Genius
M/G
10.8
0.71
4.18
11.7
Baddie 3
Shotgun
11.2
0.69
4.87
11.4
Baddie 3
M/G
12.6
0.61
5.48
10.0
Baddie 2
R/L
13.1
0.59
6.07
9.7
Examining the weights of the options, we can see how the weights are propor-
tional to how often we would like them to occur. For example, the last option (Baddie
2 with the rocket launcher) would occur about 60% as often as our top choice.
The last two columns give us another look at our results. The fifth column,
Edge, shows us what the edges of the buckets would be if we put all our weighted
options end to end. The final column shows the percentage chance of each option
occurring based on the weights. If we had picked randomly, each option would
have had a 12.5% chance of being selected. Because of the application of our
weights, those percentages now range from 16.5% for our most preferable selection
to 9.7% for our eighth-place option.
The Haves and Have Nots
While it is noticeable, there is not much of a spread between the percentages of
those eight options. We can see a more distinct difference when we use data that is
more disparate—such as the modified data (gray bars) in Figure 16.1.
Name
Weapon
Score
Weight
Edge
%
Evil Genius
R/L
7.7
1.00
1.00
25.8
Boss Man
R/L
8.5
0.91
1.91
23.5
Baddie 3
R/L
18.1
0.43
2.34
11.1
Evilmeister
R/L
19.3
0.40
2.74
10.3
Evil Genius
M/G
22.6
0.34
3.08
8.8
Baddie 3
Shotgun
26.9
0.29
3.37
7.5
Baddie 3
M/G
27.4
0.28
3.65
7.2
Baddie 2
R/L
33.1
0.23
3.88
5.9
Because the scores for this batch of data are more widely spread, the resulting
occurrence percentages are significantly different as well. We can see that the two
options with the relatively low scores of 7.7 and 8.5 will occur 25.8% and 23.5% of
the time, respectively. That is almost 50% of the time between the two of them. We
are not precluding the other six options from selection, however. They will occur,
but at a reduced frequency that is on par with their proportional score.
