Lewis Fry Richardson (1881–1953)

It's hard to get hold of this book, but having read a few analyses (and even some CAS experiments verifying its findings), it can be summarized as follows: The distribution of armed conflict outbreaks mirrors that of a Poisson distribution, the same distribution used to model chaotic processes such as radioactive decay or DNA mutations. In other words, when observed in aggregate, the outbreak of war is not statistically differentiable from any other form of random catastrophe such as show more hurricanes or earthquakes.

Note that this is just the "10,000 foot view" of the matter when taken in a purely statistical sense based on the number of conflicts over time. It does not mean that individual wars happen randomly like natural catastrophes, for each conflict has its own set of causes and drivers (people still have agency). It's just that as with hurricanes or earthquakes, it's difficult to predict how many conflicts going to happen and when, at least currently. show less