Quantifying war : from the Battle of Britain to terrorism

Conflicts are central events in history, defining eras and the lives of their peoples. In this thesis, we demonstrate the use of three quantitative methods to three case studies from historical and conflict modelling. We begin with the application of the bootstrap to the Battle of Britain. The bootstrap allows us to answer counterfactual questions about the Battle of Britain, including the importance of targeting and tactical decisions on the final outcome of the battle. We quantify the final outcome using theoretical prior distributions associated with historical viewpoints. We next conduct a changepoint analysis of historical battle deaths. This requires the adaptation of changepoint analysis methods to heavy-tailed data, for which we formulate an algorithm before applying the algorithm to the case study. We find evidence for changes in the distribution of battle deaths through time. We finish with a case study of coalescence and fragmentation modelling, which has been proposed for insurgent-counter-insurgent conflict. We demonstrate that gel-shatter cycles are a previously unrecognised yet ubiquitous feature of such systems and discuss the robustness of these systems to perturbations in the underlying rules. Together, these case studies demonstrate the ability of modern methods to refine and deepen our understanding of historic conflicts.