War & statistics

[huy.duong@ctsu.ox.ac.uk]

Dear anh Tuan,

What I meant was that although wars might be related, eg all wars by Napoleon,
because:

1) for example Napleon might not start his wars in one year,
2) for most of the period 1500-1931, ie before the days of global
connectedness, the number of wars that are related to each other is small
compared to all the wars in the world in that period,

it is resonable to see all the wars started in any one year as independent.

Random or otherwise is a property that depends on the level that we observe
and what we can observe and calculate (perhaps quantum effects and only quatum 
effects are fundamentally random? Any expert opinions here?). At the level
we are looking at, I think it is reasonable to say that the number of wars that 
starts in one year is random. This is not to say that wars start at random.
Eg given that Saddam invaded Kuwait, we would like to think that the decision
to start the Gulf war was not made at random.

I wonder if the rate of war is actually approximately constant for 1500-1931.  
Have you got the rate for each of the century or 50 year period in question?
I don't think that because the fit is good you can say that the rate is
constant.

Huy