Continuous Flow Model of a Historical Battle: A Fresh Look at Pickett's charge

TL;DR

This paper introduces a continuous flow model for simulating infantry behavior in historical battles, applied to Pickett's charge, demonstrating improved realism over agent-based models and analyzing terrain effects and outcome probabilities.

Contribution

The paper presents a novel continuous flow modeling approach for battlefield simulation, capturing terrain and troop interactions more accurately than traditional agent-based models.

Findings

Flow model better captures terrain and force interactions.

Terrain effects slow battle pace and favor defenders.

Only 6% of randomized scenarios resulted in Confederate victory.

Abstract

A continuous flow model of infantry behavior, based on conservation of individuals and tracking of subunit identity, has been developed in sufficient detail that it can now be applied to a realistic simulation of a historical battle. Pickett's charge during the 1863 Battle of Gettysburg, Pennsylvania in the U.S. Civil War was chosen as an initial application of the model. This scenario is a good test of the current mathematical model because many modern military tactics were employed, in a context where the action took place on foot or horseback, and the historical map and troop numbers are available. Compared to a discrete agent model, the flow model was found to better capture the interaction of the forces with the terrain and each other. A brigade-level simulation, faithful to the details of the historical events, was performed. The main source of asymmetry in the numbers of casualties was found to be the inability of the Confederate forces to use effective ranged fire while they were moving. Comparison of simulations with and without terrain effects showed that they slow the pace of battle and favor the defenders, exposing the attackers to heavy ranged fire for an extended period. A statistical analysis of possible outcomes for an ensemble of 1000 randomized perturbations of the baseline brigade-level scenario was carried out. Consistent with historical events, it was found that only 6 percent of the scenarios resulted in an outcome that could be considered a Confederate victory.