distances (kilometres)
Figure 3. The scaling over logarithmic distance of Ambon’s Conflict measured from the Ambon City 1999-2004
Like wars in the world, the most frequent facts to be observed in a massive conflict
and civil wars are the death tolls. It is valid to simply said that the greater the casualties the
greater the size of the civil war. Of course, the geographically impact of the conflict can also
be seen as a measurement to the size of the war. Apparently, this is showed in figure [3]. It
depicted the scaling properties of the number of killings by logarithmically measure the
radius from the Ambon city to the places where the impact has taken other (and possibly
bigger) casualties. This fact is interesting as one of the discussions in the following section
later on the paper. Regarding to the figure [3], it is important to note that the displayed data
points of each figure represent the averages over non-overlapping interval bin on the rank
variable (x-axis) that is centered at the showing points. Here, the size of each bin is changing
on every step in such a way to have the constant value in the logarithmic scale. The aim is of
course to smooth the persisting fluctuations in the data in order to ease seeing the
emerging pattern.
Probably the most interesting feature we could see in the mass conflict and also the
main concern of our discussion is the presence of the power law distribution of the
casualties of the civil war, reminding us to the discovery of Richardson (1948). As it has been
discussed in Newman (2003), and with particular themes in Moura, et. al. (2006) as well as
in Situngkir (2007), we denote p(x) as the probability density function for the data or
statistical event of x and it follows power law of,
A
P (x) = -7, (1)
xα