Modelling the mobile land battle: Combat degradation and criteria for defeat

This article reviews some of the evidence concerning what is known about the degradation of combat skills in battle. It puts forward a scheme for the representation of this effect in battle models, and then links this to the odds of ‘victory’ or ‘defeat’ in mobile land warfare. The historical evidence suggests that, in mortal combat, only a modest proportion of weapon crews can be relied on to make a fully active contribution to the battle. Of the remainder some will make only an intermittent contribution, and some no contribution at all. It appears that there are relatively stable differences in these proportions from one army to another. The evidence also suggests that the contribution of the less effective is likely to be somewhat more in the attack than in the defence. The article shows how this phenomenon could affect the form of mathematical models and predictions commonly used to represent combat attrition. Historical analysis also suggests that, at the tactical level, successful resistance to attack depends less on attrition than it does on maintaining the spatial integrity of the defence. Clearly, this integrity is more likely to be compromised as the proportion of non-contributing defenders increases. A simple modelling scheme is therefore proposed. That sector in which the attacker intends to break through is designated as the ‘critical point’. If, when the attacker reaches this ‘critical point’, the number of his survivors equals or exceeds a pre-determined multiple of the active surviving defenders in this sector, then the attack will be deemed to have ‘succeeded’. Although simplistic, and obviously in need of refinement, this scheme does provide a plausible explanation for some observed operational relationships: that armies which characteristically impose low casualty rates on their attackers tend to surrender when their own casualty rates are low, and also tend to retreat at a faster rate as a function of local force ratio.