“Systems analysis,” as a concept, can be difficult to define and pin down. For much of my life, I assumed it was some sort of generic back-office IT function (see, for instance, the hundreds of man-on-the-street “American Voices” interviews in The Onion, which describe respondents in equal measure as ‘unemployed’ or ‘systems analyst’). But given the complexities of almost, well, everything in the modern era, an understanding of the logical underpinnings of systems analysis is critical.

Essentially, single variables cannot be considered in isolation. A new weapons platform or technological development or re-basing movement must be thought of in the context of existing technology, logistics capacity, weather, enemy reaction, enabling capabilities, fixed facilities, power projection, and so on, down an otherwise infinite fractal list of factors.

Dr. Bernard Brodie, RAND Corporation (Wikimedia)

But all this is a long-winded introduction to Bernard Brodie’s hypothetical systems analysis example in Strategy in the Missile Age is one of the best,  most succinct ways of describing just how complex this interplay is. Brodie, of course, had a front-row seat to this effort, as the RAND Corporation was the earliest home to a methodological approach to the field. Beginning on page 381 in the 1967 edition:

Let us consider, for example, the problem of choosing between two kinds of strategic bombers. Each represents in its design an advanced “state of the art,” but each also represents a different concept. In one, which we shall call Bomber A, the designers have sought to maximize range. They have therefore settled for a subsonic top speed in a plane of fairly large size. The designers of Bomber B, on the contrary, have been more impressed with the need for a high dash speed during that part of the sortie which involves penetration of enemy territory, and have built a smaller, shorter-ranged plane capable of a Mach 2 dash for a portion of its flight. Let us assume also that the price of the smaller plane is about two-thirds that of the larger.

Perhaps we can take both types into our inventory, but even then we should have to compare them to determine which we should get in the larger numbers. Let us then pick a certain number of specific targets in enemy territory, perhaps three hundred, and specify the destruction of these targets as the job to be accomplished. Since we know that both types can accomplish this job with complete success if properly supported and handled, our question then becomes: which type can do it for the least money?

We do not ask at this stage which type can do it more reliably, because within limits we can buy reliability with dollars, usually by providing extra units. Some performance characteristics, to be sure, will not permit themselves to be thus translated into dollars-for example, one type of plane can arrive over target somewhat sooner than the other type, and it is not easy to price the value of this advantage but we shall postpone consideration of that and similar factors until later.

Let us assume that Bomber A has a cruising range of 6,000 miles, while Bomber B is capable of only 4,000 miles. This means that Bomber A has to be refueled only on its post-strike return journey, while Bomber B probably has to be refueled once in each direction. This at once tells us something about the number of “compatible” tankers that one has to buy for each type (“compatible” referring to the performance characteristics which enable it to operate smoothly with a particular type of bomber). Up to this point Bomber B has appeared the cheaper plane, at least in terms of initial purchase price, but its greater requirement in tankers actually makes it the more expensive having regard for the whole system. In comparing dollar costs, however, it is pointless to compare merely procurement prices for the two kinds of planes; one has to compare the complete systems, that is to say, the weapons, the vehicles, and the basing, protection, maintenance, and operating costs, and one must consider these costs for each system over a suitably long period of peacetime maintenance, say five years. These considerations involve us also in questions of manpower. We are in fact pricing, over some duration of time, the whole military structure required for each type of bomber.

A B-36 Peacemaker, B-52 Stratofortress, and B-58 Hustler from Carswell AFB, TX en route to the former’s retirement in 1958. The B-52 would long outlive the more advanced Hustler. (Wikimedia)

Now we have the problem of comparing through a process of “operations analysis,” how the two types fare in combat, especially the survival expectancy of each type of plane during penetration. In other words, we have to find out how much the greater speed (and perhaps higher altitude) of Bomber B is worth as protection. If the enemy depends mostly on interceptors, the bomber’s high speed and altitude may help a great deal; if he is depending mostly on guided missiles, they may help relatively little. Thus a great deal depends on how much we know about his present and projected defenses, including the performance characteristics of his major weapons.

If our Bomber A is relying mostly on a low altitude approach to target, which its longer range may just make possible (we are probably thinking in terms of special high efficiency fuels for wartime sorties), it may actually have a better survival expectation than its faster competitor. Also, we know that penetration capability is enhanced by increasing the numbers of bombers penetrating (again, a matter of money) or by sending decoys in lieu of extra bombers to help confuse the enemy’s radar and saturate his defenses. Perhaps we find that the faster plane would outrun the decoys, which again might tend to give it a lower penetration score than one would otherwise expect. But decoys are expensive too, in acquisition costs, basing, and maintenance, and involve additional operating problems. The faster plane may be less accurate in its bombing than the other, which again would involve a requirement for more aircraft and thus more money.

We have given just barely enough to indicate the nature of a typical though relatively simple problem in what has come to be known as “systems analysis.” The central idea is that no weapon can be considered independently of the other weapons and commodities that are used with it, that all endure through some period of time and require men to service them and to be trained in their use, that all these items involve costs, and that therefore relative costs of different systems, as considered against some common standard of function, are basic to the problem of choice between systems. Systems analysis, which brings what is modern to present-day strategic analysis, is mostly a post-World War II development.

The challenges herein are immense, which in part explains the explosion not only of defense research and development but also of the defense bureaucracy as a whole. It’s a sprawling, tangled mass that can in many ways only be understood in relation to itself. But systems analysis is at least an attempt to build that into other assumptions and considerations.

Using this technique is not only a way to compare technologies with like missions; it’s an excellent tool for use in wargame design. This too is in fact an iterative process, as the insights from a wargame itself might reveal further interrelationships, which might then be used to craft a more complex operating environment (or refine the mechanics used to select force lists), and so on ad infinitum.

Practicality aside, Brodie’s writing serves as an excellent primer to what systems analysis entails, and more broadly, to the change in strategic thought and analysis since the end of World War II.