lost and the damage caused to the targets using the respective mathematical models taken from IABG’s air war model.4 Never have I forgotten the disputes between players and OR analysts. The blue players considered the target damage calculated as too low and the loss of their air sorties as too high. In the midst of the game they suggested that the analysts manipulate the “critical” inputs of the assessment models so that the outputs would be closer to their judgment. Not surprisingly, the analysts rejected the suggestion arguing that rather than manipulating the game halfway, to improve its results for Blue, it would make more sense to end the game and, thereafter, revisit its data and the assumptions underlying the assessment models. That is what we did.
On the basis of the air war model the Air OR group then developed, together with the military advisory group associated with IABG’s Study Division, an interactive wargame that was successfully tested within a high‐level planning exercise of the GAF in 1970. In 1972 I became head of IABG’s System Studies Division that included OR/SA support for all three service branches of the Bundeswehr. In a discussion with STC’s Andreas Mortensen5, about the upcoming issue of modeling in support of overall force capability assessment, we agreed that the most difficult and least documented aspect of military OR/SA seems to be the land war. Thus we felt that a scientific conference on land battle systems modeling would not only contribute to a better understanding of the implications of different modeling approaches, but also help preclude undesirable redundancy through better familiarity with models available elsewhere.
The Special Program Panel of Systems Science of the NATO Scientific Affairs Division6 accepted and funded, together with the German Ministry of Defense, my proposal to organize, together with my co‐chairmen, Lynn F. Jones of the UK’s Royal Armaments Research and Development Establishment and Egil Reine of the Norwegian Defence Research Establishment, the scientific conference on “Modeling Land Battle Systems for Military Planning” held at the War Gaming Center of IABG7 in Ottobrunn, Germany, 26–30 August 1974. The keynote address of the Deputy Under Secretary of the US Army for Operations Research Dr. Wilbur Payne began with the following statement:
As we are less and less able to rely on historical European combat data and as we see more and more the necessity of evaluating issues in large contexts, gaming and simulation emerge perhaps as the only tools able to organize large quantities of information and discipline our thinking and communication about them.8
While not having strong opinions on some methodological aspects of the problem of modeling land battle systems, Dr. Payne expressed strong opinions on certain problems that require more attention than we have given in the past. From the point of view of a senior member of the profession9, and a bureaucrat involved in trying to use the results of research to generate defense programs and convince others that the programs are worthy of support, he pointed out some of these problems such as (1) identifying radical changes in the general structures of combat; (2) interactions between weapon system development and tactics development; (3) the issue of “quality versus quantity” in the weapon systems design and selection, and (4) developing estimates of combat losses. He believed that all of these imply basic and extensive improvements in both modeling techniques and how models are used. Thus, he proposed to discuss these key problems “to get a better idea of the direction we should take to improve the ability of our models to handle these four problems.”
In the context of the problems listed by Payne, a constructive assessment technique in form of force on force models would be appropriate such as, for example, changing the structure and tactics of defense forces to improve their deterrent capability. This was exactly the idea of a group of German political scientists and retired military officers who proposed, in the late 1970s and early 1980s, that Germany’s all‐active armored defense forces be supplemented by reactive forces thus improving the chance that an eventual attack by WP forces could be stopped at the demarcation line, between East and West, without NATO having to employ nuclear weapons.10 The sometimes bitter debates between the retired military members of the group and their active peers in the Ministry of Defense, on the pros and cons of the group’s proposals, were characterized by arguments based mostly on military judgement rather than analysis.
Therefore, together with my colleague Prof. Hans Hofmann of the Institute of Applied Systems and Operations Research (IASFOR) at the Bundeswehr University in Munich11, we initiated a project to take a look at the arguments of both sides on the basis of the outcomes of battle simulations involving 12 reactive defense options of four categories12 using the Monte Carlo‐type model BASIS. In cooperation with the military authors of the options, this model was developed by Hofmann and his research assistants over a period of three years, accounting, in great detail, all essential interactions affecting the dynamics and outcome of ground battles, for the simulation of battles between battalion‐sized German ground forces defending against a sequence of regimental‐sized Soviet attacking forces supported by organic and higher level fire support on both sides. More than 500 battle simulation experiments were conducted in different type of terrain and visibility to generate sufficient data for a detailed analysis of each of the 12 reactive options.13
The results were discussed at a Workshop with international experts organized by the German Strategy Forum on “Long‐Term Development of NATO’s Forward Defense,” held 2–4 December, 1984, in Bad Godesberg/Bonn. The overall conclusion of the analysis suggested that properly equipped and trained reactive defense forces being available on short notice might be an effective and efficient tool to absorb the initial attack by fighting, at the demarcation line, an attrition‐oriented delaying battle thus providing the time for the active defenses to deploy at the points of the enemy’s main thrusts and for counterattacks into the enemy’s exposed flanks. The main reason why the Bundeswehr and its NATO partners did not consider following up the options investigated by IASFOR was that restructuring the all‐active forces, deployed at the time, in the Central Region close to the demarcation line, would involve some time of conventional weakness and strategic risk considering the strategic situation in the 1970s and 1980s. However, given today’s strategic situation between NATO and Russia, it seems that NATO partners in the East might well revisit some of the reactive options investigated by IASFOR for their territorial defense forces.
Comparing the proceedings14 of the follow‐on conference held in Brussels under the aegis of Panel 7 of the Defense Research Group (DRG)15 – eight years later – with the proceedings of the conference in Ottobrunn (see footnote 8), both published by Plenum Press, the number of papers that addressed inter‐active and computer simulation models had increased by 36% and their average length of papers by 60% suggesting that battle simulation modeling had expanded and intensified significantly in the eight years between the two conferences.16 And this expansion went on with the reorganization of NATO in 1998 when the former DRG panel 7 established, parallel to its System Analysis and Studies Group (SAS), the Modeling and Simulation Group (MSG).17 While the mandate of SAS did not change, MSG’s primary mission areas include standardization of modeling and simulation (M&S), and education and associated science and technology “to promote co‐operation among Alliance Bodies, NATO Member and Partner Nations