Confirming the Lanchestrian linear-logarithmic model of attrition

Technical Report · Sat Dec 01 00:00:00 EST 1990

DOI:https://doi.org/10.2172/6077918· OSTI ID:6077918

Hartley, III, D S

This paper is the fourth in a series of reports on the breakthrough research in historical validation of attrition in conflict. Significant defense policy decisions, including weapons acquisition and arms reduction, are based in part on models of conflict. Most of these models are driven by their attrition algorithms, usually forms of the Lanchester square and linear laws. None of these algorithms have been validated. The results of this paper confirm the results of earlier papers, using a large database of historical results. The homogeneous linear-logarithmic Lanchestrian attrition model is validated to the extent possible with current initial and final force size data and is consistent with the Iwo Jima data. A particular differential linear-logarithmic model is described that fits the data very well. A version of Helmbold's victory predicting parameter is also confirmed, with an associated probability function. 37 refs., 73 figs., 68 tabs.

• Research Organization: Oak Ridge K-25 Site, TN (USA)
• Sponsoring Organization: USDOD
• DOE Contract Number: AC05-84OT21400
• OSTI ID: 6077918
• Report Number(s): K/DSRD-263; ON: DE91006904
• Country of Publication: United States
• Language: English

Related Subjects

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEFENSE
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
WARFARE
MATHEMATICAL MODELS
ALGORITHMS
COMPUTERIZED SIMULATION
DATA ANALYSIS
HISTORICAL ASPECTS
MILITARY STRATEGY
REGRESSION ANALYSIS
STATISTICAL MODELS
VALIDATION
VARIATIONS
MATHEMATICAL LOGIC
MATHEMATICS
SIMULATION
STATISTICS
TESTING
450000* - Military Technology
Weaponry
& National Defense
990200 - Mathematics & Computers