Attrition rates and maneuver in agent-based simulation models

摘要

Agent-based military combat simulation models make use of a variety of methods to determine individual engagement and casualty effects. The agent-based approach differs to the deterministic approach of Lanchester and Osipov, whose combat attrition calculations pre-dated computationally intensive physics-based simulation models. Deterministic approaches to analyze force attrition rates and the outcomes of combat have been noted by many authors to lack realism and to ignore many fundamental aspects of modern combat However, there is little available evidence to suggest that modern simulations provide a more realistic attrition rate, despite their complexity. This paper seeks to explain the importance of understanding the implications of the tactical aspects of multiple simulation packages and to ensure these are considered within the underlying combat model. The principle factor this paper will discuss is attrition and its relationship to maneuver. This work analyzes the macro-attrition rates emergent within a number of unclassified agent-based simulations using the Helmbold space. The results demonstrate insufficient sensitivity to the influence of maneuver and the inconsistent influence of force ratios on combat outcomes. Simulation supported by well-documented testing and validation is necessary for analysts to support the military training imperative and provide realistic experimental engagement results.