Systems analysis is applied in economics and engineering, biology and medicine, history and philology, politics and military affairs. It is used in theoretical and applied research, in selecting prototypes of new equipment and variants of capital investment, and so on. To present the material more clearly, let us first formulate the questions pertaining to systems analysis. The questions are posed in the form in which they arose during practical work in systems analysis.
It is well known that by defining a concept, we restrict, narrow, and impoverish it. Nevertheless, as a working tool, one of the possible definitions can be used as the principal one, by highlighting certain essential features of the phenomenon that distinguish it from others.
Systems analysis is a methodology for investigating properties and relations in objects that are difficult to observe and difficult to understand, by representing these objects as goal-directed systems and studying the properties of these systems and the interrelationships between goals and the means of their realization.
It is applied when the problem is posed in such a way that the information needed for its solution cannot be obtained by direct observation of the object. The object is then treated as a subsystem of some system, as a collection of subsystems interacting with other systems. This definition, which should be regarded as a working one, makes it possible to distinguish the methods of systems analysis from other methods of investigation and places it within a specific area of scientific knowledge.
Among the multitude of methods of investigation and analysis, the overwhelming majority are oriented toward direct observation of objects, taking into account their nature and specifics. It is always assumed that the object under study can be isolated from its environment, that it can be observed directly or by means of instruments. In contrast, the method of systems analysis, grounded in systems theory, takes into account the fundamental complexity of the object under study, its extensive and strong interconnections with the surrounding world, and the unobservability of a number of its properties.
Therefore, starting from a real phenomenon and from available factual data about its properties and connections with the surrounding world, the researcher translates them into the abstract categories of systems theory and, on the basis of known properties of systems, identifies new properties and new interrelationships. In most methods of investigation, the objects are precisely defined. A systems investigation usually includes, as one of its important stages, the very identification of the object — its discovery or construction. Almost all methods of investigation proceed from a clearly formulated problem. Systems analysis addresses the questions of how to correctly pose problems and which methods of investigation to employ.
The essence of systems analysis is how to transform the complex into the simple, how to turn a problem that is not only difficult to solve but difficult to understand into a clear series of tasks that have a method of solution.
Systems analysis is always concrete. It deals with a specific economic entity (even if initially vaguely defined) and with a specific economic problem (even if initially unclearly formulated).
Systems analysis is in no way opposed to other methods of analyzing problems and making decisions. What is new is the synthesis within a unified methodology of a certain interrelated set of concepts, methods, and techniques that were previously used in isolation for solving individual problems in science and engineering. Furthermore, this complex of systemic concepts and methods is extended to entirely new domains of activity where these concepts had not previously been used — the areas of planning and management.
The strength of the systemic method in analyzing complex problems lies in the fact that it allows, on the one hand, the decomposition of a problem that is too complex to solve into its components, down to the formulation of specific tasks for which well-established solution methods exist, while on the other hand holding these components together as a unified whole.
As the Soviet researcher S. P. Nikanorov put it very precisely, systems analysis makes it possible to represent the process of solving a problem as a process of designing, manufacturing, and using systems.
What is essential in systems analysis, what its attention is focused on, was formulated very precisely by the American researcher Quade: "Systems analysis is a way of looking at a problem. The mathematical formalism and use of computers may be necessary and even useful, but they may not be. Sometimes, careful reflection on the problem may suffice. But in any analysis related to decision preparation under conditions of uncertainty, regardless of its complexity, certain elements are present. These elements — a goal or goals, alternatives or means for achieving those goals, costs or everything that must be expended to achieve each alternative, a model or a description of the dependence between alternatives and what they accomplish and cost, and criteria according to which a preferred alternative is selected — are present in any analysis whose purpose is to influence the choice of a course of action."
It is perfectly obvious that the complex, specially developed, and rather cumbersome scientific formalism that systems analysis represents is worth applying only to sufficiently complex, large-scale problems involving the activity of many people and substantial material and other expenditures.
It is difficult to establish any classification of problems so as to say precisely where the systemic formalism can and where it cannot be applied. The origins of problems differ in terms of needs and opportunities. The first case is simpler: existing resources or working methods, or certain goods, have ceased to satisfy people, and the existing situation needs to be changed.
Systems analysis is applied when a very complex problem arises. The second case is in principle more complex: a new opportunity appears, such as space flight or the use of thermonuclear energy. In each case, the development of such opportunities is associated with the creation of entirely new branches of the national economy. In this case, the application of systems analysis proves not only useful but absolutely necessary, since the problem itself — being complex in character — still needs to be formulated, because the realization of new opportunities affects the most diverse aspects of human activity, of the life of society and the state. One can also note even more complex cases, where new opportunities arise everywhere and almost hourly: experimental chemistry, for example, synthesizes up to three hundred thousand new compounds per year with the most diverse properties, which can be used in many, if not all, branches of production.
Human activity can be conventionally divided into two areas: the area of routine activity — that is, regular, everyday tasks — and the area of solving new, previously unencountered tasks. In the first area, the methods of solving tasks are usually well established, and there is no ground for systems analysis, although the very existence of routine in some cases constitutes a problem in itself. Thus, the existence of a multimillion-strong economic management apparatus itself creates difficulties in management and generates problems; the tendency toward constant growth in the number of management personnel constitutes a major problem. But in the sphere of human activity associated with solving new, previously unknown tasks (for example, in long-range planning, in science, in design development), the methods of systems analysis are applicable almost universally, and in some cases are indispensable.
In the method of systems analysis, it is also customary to distinguish problems by the degree of their structuredness — that is, by the clarity and awareness of how they are posed, by the degree of detail and specificity in the understanding of their components and interrelationships, and finally by the ratio of quantitative to qualitative factors noted in the problem formulation. With this in mind, three classes of problems are distinguished: well-structured, quantitatively formulated problems; ill-structured, or mixed, problems that contain both quantitative and qualitative assessments; and unstructured, or qualitative, problems.
For the first class of problems, the methods of systems analysis are not needed, since a well-developed and powerful formalism of mathematical modeling and rigorous quantitative solution methods already exists. The second class of problems — ill-structured, with mixed quantitative and qualitative assessments — is identified as the primary domain of application for systems analysis methods. It is considered that unstructured problems are not solved by systems analysis methods; for their solution, so-called heuristic methods are applied. Of course, it is difficult to draw any clear boundaries between the three classes of problems, but the essence of the matter lies elsewhere.
The method of systems analysis is precisely a method of structuring, of ordering problems. Systems analysis is applied in order to at least weakly structure an initially unstructured, vaguely defined problem, and then to gather new additional information about it, to establish the interrelationships of its components, to give, wherever possible, quantitative assessments (even if subjective, expert-based), and to move the problem into the category of structured problems, to which the formalism of mathematical modeling and the selection of optimal solutions can already be applied.
In a number of cases, attempts are made to define systems analysis through the spheres of human activity in which it finds predominant application — as a methodology for solving research, military, political, and economic problems.
The first condition for the success of systems analysis is its application where it is truly needed. The very nature of a systems investigation, devoted to complex questions concerning complex objects, requires a high level of knowledge and the involvement, at one stage or another of the analysis, of highly qualified specialists. Meanwhile, the fashion for systems analysis has prompted numerous researchers, graduate students, and even undergraduate thesis students to take it up. Naturally, in such cases, a "manageable" and entirely surveyable object is chosen, or some hypothetical model with which any operations can be performed, or else particular problems of cost, labor productivity, product quality, and so on. A systems investigation cannot succeed if the management of the commissioning organization or the researchers themselves approach it in bad faith.
The conditions for the success of systems analysis are considered to be the presence of three elements: (1) a clearly understood need, goal, or purpose; (2) a source of ideas, accumulated information, experience, and understanding of the subject; (3) resources — experienced specialists, as well as equipment, materials, and financial means.
Systems analysis is itself a tool for ensuring the presence of all three of these elements: identifying and elaborating goals, the corresponding information and resources, and linking them to the goals. Thus, specialists in systems analysis must know the conditions for the success of their work and undertake it only when those conditions are met.