Abstract: Previous research shows that security incidents in an asymmetric conflict (counter-insurgency, war on terror) exhibit power-law behaviour in terms of casualties and incidents. This study, covering a significantly larger data set than the previous works, confirms the general findings of these earlier analyses. However, it appears that the daily incident counts obey power law distributions only for values greater than the value expected from corresponding monthly average count. The low counts seem to be driven by a completely different mechanism. The present paper proposes that this behaviour is consistent with a dynamical model of insurgency based on a time-dependent Hamiltonian system with a slow free energy (for the insurgency corresponding to their supplies and morale) build-up, followed by a rapid, avalanche-like release. Then the asymmetry in the behaviour for low and high counts can be accounted for as follows. The high counts redistribute the free energy in the form of avalanches of violent incidents (akin to explosive instabilities in many physical systems). On the other hand, the low counts correspond to introducing artificial energy sinks in the system (for example, by localized disruptions to the enemy supply chain or weather-related obstacles). The findings and the model seem to be in line with the studies suggesting that self-organized criticalities form a universality class, opening the door for modeling and possible predictive analysis.