Policy Games for Strategic Management

Chapter 7 - Understanding the Policy Game Construct

by Richard D. Duke and Jac L.A. Geurts

7.3 Structure

Webster’s definition of structure, “the relationship or organization of the component parts of a work of art or literature” ... “to construct a systematic framework,” is useful for game design. Structure can be thought of as having three major components: the physical characteristics of the game (format, basic referent system); the procedures of play employed to ensure a logical progression of activities (policies, rules, scoring, steps of play); and the simulation (model, data/information, the accounting system, indicators).

7.3.1 The Physical Character of the Exercise

The game will be manifest through its physical character – the format and the referent system. The physical characteristics of policy games vary widely; familiarity with the literature and experience in the field will assist the designer in developing a range of technique from which to draw.

Format

Format is defined as the physical configuration (the documents, visuals and artifacts) of the game as well as the various processes that the participants encounter in the game; it is the environment the players will experience. It is important for the format to mimic the client’s environment as much as possible. The game is intended to be a playful, safe environment. The game format must encourage innovation and risk taking as a means of opening communication among players in the various teams as well as among the larger group during the critique and debriefing activities. As in any good negotiating situation, arguments can be transferred to the thing (format) rather than being focused on personalities. There are an infinite variety of formats that can emerge; the designer should avoid the mistake of copying an existing game format – let a new style emerge from the process! It should be clear, appropriate and powerful.

Frequently, a game combines role playing (based on the central figures from real life) with a physical setting that gives structure to the interaction (a board, map or flowchart, room arrangement, etc.). Participants are required to make decisions that are comparable to real-world decisions (e.g. run a hospital). As play begins, these activities are structured to simulate the present system; the players will “discover” that the system is sub-optimal. Success in the game will be indicated by their improved communication resulting in various alternatives being brought forth and evaluated; consequently better quality decisions will be achieved.

Game boards exist in three basic types: edge, grid, and patterned. One of the best-known “edge” boards is the commercial game Monopoly®. A second common board pattern is the “grid” board (a coordinate system that produces a series of cells that can be identified spatially by referencing the X and Y coordinates – this is well illustrated by a chessboard). The “patterned” board takes a great variety of shapes (e.g. the player progresses by moving symbols through an existing flowchart). In some instances, the players may be permitted to modify the flowchart as the game progresses (a more sophisticated example as used in a policy exercise might be the map of a region defined by its ecology; see Section 3.8).

Basic Referent Systems

The basic referent system is defined as one or more frames of reference that underlie and give order to the policy exercise. These reflect the intellectual discipline or set of substantive ideas that influence basic decisions about design. The selection of the basic referent system is extremely important because it must be in harmony with the objectives of the client and the characteristics of the participants. If these are properly matched, the task of getting the players quickly into the exercise will be much simpler than if they are forced into an alien referent system. A policy game requires a team of players, in an artificial environment and under severe time pressure, to deal with extensive substantive information in the context of a sophisticated conceptual referent system. There is a great deal for a player to assimilate in a few hours; it is a lot to squeeze into a game! It is not possible to give full treatment to this topic here; however, referent systems often reflect one or more of the following orientations: resource allocation, group dynamics, geography, demography, politics, business concerns, economics, sociology, psychology, anthropology, history, systems exposition, and others. Examples of basic referent systems reflected in the cases include Conrail (geographic/market place), Social Employment Program (decision tree), Employee Orientation Simulation (initially a silo, subsequently a matrix organizational structure) and the Rubber Windmill (the marketplace). The basic referent system is often a “board” (see the preceding paragraph).

7.3.2 The Procedures of Play

Procedures of play can be thought of as various facilitation mechanisms used to ensure a logical progression of activities; they can be classified as policies, rules and steps of play. There are a wide variety of procedures that may be presented as rules (not subject to change within the game) or as policy that is subject to player modification. The character and utility of a game are heavily influenced by the rules and policies employed. The game design may create a constrained or free environment as best meets the needs of the client specifications.

Policies

Policies can be defined as participant-imposed constraints controlling play. Players should be able to interact with the game in ways not initially perceived by the designer. When players are permitted to alter, amend or enrich procedures within the basic gaming structure (for example, moving from a nonexistent definition of acceptable player behavior to an advanced articulation in successive cycles of play), they can maximize learning without the unnecessary specification of an elaborate rule structure.

Player-generated policies appear quite frequently in virtually all policy games and are to be encouraged because their very formation implies a coherent consolidation of player purposes and objectives. They often deal with negotiations among players, teams, coalitions, and/or the game operator. They take an endless variety of forms, simple or complex, fixed or varying, from cycle to cycle. It is often necessary for the game operator to play the role of judge in overseeing these player-generated policies. Policies permit players to mold their simulated world into a more useful model during play. Negotiation should be used to address policy disputes; it is a good idea to require that policies be written down.

Rules

Rules are important tools that permit the designer to transform the client’s environment into the focused and abstracted world of the simulation. Rules are best defined as specific facilitator-imposed constraints that govern play. Players must obey rules; they cannot be changed by the participants. Rules must be stated clearly by the facilitator; they should be few in number, lucid, and essential to the logical progression of the game. These are not to be confused with rules for winning as defined in competitive games. Rather, they are a set of conventions that participants are asked to adopt for the duration of the game. Generally, rules as such should play a relatively minor part in a game. Nonetheless, some are inevitable and they must be addressed. They are of two types: those dealing with the accounting system and those controlling behavior during the course of the game. Rules dealing with accounting systems tend to be rigid to permit ease of quantification of results. Rules controlling player behavior have a greater degree of latitude than those governing the accounting system.

Games should be human driven rather than rule driven; the philosophy should be “it is your world, you solve it.” There has been undue emphasis on rules in gaming simulation, perhaps as a result of the strong heritage of rigid war games and game theory. The concept of rule use in the policy exercise is poorly defined in the literature; further, rules are often not explicitly stated in actual policy exercises. Because of game design considerations, certain conditions may be inviolable (for example, the requirement of iterative cycles and the calculations inherent to a particular model). A much more productive concept is “procedures,” intended as a flexible term to cover all mechanics of play, including any rule structure. A key skill of a talented game designer is the ability to create a game with rules that feel natural to the players and, at the same time, create a simulation that is doable, focused and structured.

Scoring

Scoring is defined as a system of penalties and rewards reflecting the results of decisions taken by the players; specific feedback must be developed for each role. Parlor games typically have a clear-cut scoring procedure. However, winning or losing is almost always an invalid concept when using the policy exercise; rather, the concept of win-win is appropriate. Win or lose scoring detracts from communication; the idea conveyed by scoring should be to illustrate what is possible to achieve in the client’s environment. For this reason, formal scoring mechanisms are typically not provided in an exercise; the “indicators” (see Section 7.3.3) are intended to serve this function. The indicators enable players to evaluate their own performance; specific feedback must be developed from the indicators to each participant. Indicators and scoring are closely linked concepts; a well-designed exercise lets the participant judge his or her own progress rather than have an artificial and meaningless number imposed as a score.

Steps of Play

Steps of play are defined as the sequence of activities that players must follow during each cycle of play. All games should consist of a series of cycles, which are iterative and which become more complex cycle by cycle. Within any given cycle, there is a standard set of steps that are confronted by the player. Steps of play are discrete activities that govern participant behavior; they must be quite simple, and each step should be a single action. This is essential if the participants are to understand what to do next and the facilitator is to control the timing of the game. Naturally, the first cycle is unfamiliar, and consequently takes longer than the following ones because the participants must be oriented to their environment. Steps of play will emerge during the construction of the game prototype. Figure 7.2 provides an idea of how activities might be structured in a typical policy exercise.

Image from book

Figure 7.2: Typical Steps of Play Sequence

The first step of play in a given cycle is the event; in turn, there is a discussion and interaction between the roles and the facilitator. Discussion should focus on reviewing the material (output) resulting from the start-up data or from the last cycle. Further discussion should be addressed to seeking information, forming coalitions, developing strategies and improving communication about the problem. In some instances, the game facilitator is obligated to initiate discussions. Examples include the formation of an ad hoc group to consider some problem, question, fact, or alternative at the request of the facilitator and to subsequently report to the assembled game participants.

Next in each cycle are the decisions which may be of both an inter-team and intra-team nature. The concept report will define the sequence in which these are to be made and give reasons for the sequencing. Generally, decisions should be made under reasonable time pressure.

Finally, the decisions must be processed through an accounting system, whether manual or computer. Regardless of the character of the processing system involved, it is extremely important for the game facilitator to review the decisions made by the players (before processing them) and to ensure that they represent the actual intention of the player. More importantly, the facilitator must check that they do not go beyond the capabilities of the game itself. It serves no purpose for a facilitator to process a decision knowing that the result will be an embarrassment to the player and an obstruction to progress in the game.

7.3.3 The Simulation

The simulation can be thought of as the quantitative components of the policy exercise; it consists of the model, a data/information system, the accounting system and the indicators. These will vary from highly simplified to rather complex components depending on client need.

Model

The model becomes explicit through the design sequence (in particular through the schematic – see Section 8.2.2, The Design Sequence, Step 7). It must correspond to the underlying systemic structure of the problem; it will mimic the client’s environment that is presented through the game. This simulated world will be used to process player decisions; it can be thought of as a formalized system for capturing the primary substantive focus of the exercise. At the start of the game, the model is typically presented to the players as an analysis of the situation in both its quantitative and qualitative respects. For example, in the Conrail Exercise (Section 3.4), a quantitative model was used to simulate the results of player decisions.

Games use different types of models: resource allocation (e.g. limited availability of municipal funds, competition by group members for budget, competitive bidding for land ownership), group dynamics (role-playing and/or interpersonal relationships), and system specification (the explicit expression of a complex system, its roles, components, and linkages). Many gaming/simulations reflect all three of these considerations but typically one theme will be dominant. This does not mean to suggest that other types of models cannot be applied to a broad selection of gaming/simulations.

The model that the game is intended to convey may be implicit (assumed to be known by the participants); explicitly presented to the participants in the game materials; or integral, becoming evident to the participants as they participate in the game. However presented, this model will be used to process player decisions. Depending on the purpose of the game, the model may be either descriptive or prescriptive. Most policy games will be based on a descriptive model (e.g. an estimation of probable voting behavior of various interest groups); however, this is often supplemented by a prescriptive model (e.g. budget model). If a descriptive model is used, the use of analogy (both physical and/or symbolic) is quite common. If a prescriptive model is used, it will be based on a structure of law, natural phenomena, and/or various man-made or scientific phenomena.

The model should represent an explicit expression of a complex system; it is useful to document the theoretical basis for the model in the concept report. If the theoretical basis is weak, this concern should be presented with equal candor so that those dealing with the problem will have some notion of where reality ends and fantasy begins. When dealing with the multidimensional world of gaming, the art of muddling through is essential. In gaming, data that is artificially generated can become a source of strength if it is presented with candor rather than through subterfuge.

Data and Information

The game is a device for exploring gestalt. To achieve this, players must be permitted to pursue any dimension of the subject from any perspective that seems relevant to them at an appropriate time. They must be provided with both a total systems overview and sufficient detail. For this reason, an important component of any game is a clearly referenced set of information at both general and detailed levels, as well as a carefully articulated information flow procedure. The information provided, in terms of quantity and depth, is dependent on the communication purpose, the sophistication of the players, and the conditions of play.

A significant factor in the development of a game is the loading (what, when and how to present data to participants) of data and information. During the first cycle of play, it is quite common to watch players grow increasingly apprehensive. This is inevitable because a great deal of information is being presented rapidly in an unfamiliar context. To counter this, players should be given no more information than is essential at any given moment. As a game progresses, each cycle becomes successively more involved and deals with an increased amount of information. As involvement and commitment increase, a self-generated need for information increases. Once a player is motivated to raise questions, he or she is able to assimilate surprisingly large quantities of information. As each cycle passes and the sophistication of the player increases, succeeding rounds become increasingly challenging.

The game should maintain a consistent level of abstraction, not only within a given role but also between the different roles and components of a game. When one role demands attention to detail and another role deals with questions of strategy and planning, communication between these players will be minimal. Equally important, the level of abstraction should permit the players to address the questions that are inherent to the policy objectives of the game. Games should not permit players to become too involved in detail –there will be no time to think about questions of strategy. If one is dealing with managers or other professionals, the level of abstraction in the game must permit analysis and synthesis of useful heuristics.

Information will generally be in one of three categories; reference materials, output from the accounting system, and information from participants. Reference information will include any relevant material dealing with the subject matter. Standard reference materials should be roughly the same ones the participants would choose if they were to make their own library search for supporting material. It is a mistake to generate greater detail than the average player can readily assimilate; an over-saturation of information may make the game harder to understand than reality. In an on-the-job situation, one at least has the ability to make on-the-spot judgments as to what information to retain and what to avoid. In a game where the setting is controlled by the facilitator, the player will assume that it is appropriate to attempt to comprehend and deal with any information presented during the exercise; an overload or poor selection by the developing team can create problems.

A glossary is a useful item as it sets forth the definition of terms employed. Another value of the glossary is that it serves as an indication of the degree of complexity of the game. Other reference materials that can be displayed visually should be provided. In many games, a variety of wall charts need to be posted showing the progress of different variables, cycle by cycle, as they are computed by the accounting system. Any reference material that will facilitate communication among players should be provided.

The most important use of information in any game is during the discussions (both inter-team and intra-team) and during the critique at the end of each cycle. The values of these discussions cannot be overemphasized, for it is through them that much of the information to which the player has been exposed is synthesized. Specific data that documents themes, issues and/or alternatives must be made available to players as they make their decisions.

Obtaining a thorough and valid empirical database on which the game message is formulated can be a costly and time-consuming process. There is a serious obligation on the part of the designer to document the source and validity of the data in the concept report. Parsimony is central when developing a database for the participants – too much detail will drive the game lower on the cone of abstraction; as a consequence, players may miss the big picture and spend their time fruitlessly debating detail. Tests for data include the relevance to the model, an acquisition plan, a careful file structure and storage system and a coherent and readily available list of variables that are used.

Accounting System

The accounting system is the process by which player decisions are captured and recorded, processed to ensure accountability of the roles, and the results reported back to the player to engender discussion. There are an infinite variety of accounting systems which may be invisible to the players. The accounting system may be formal (embedded in a computer model) or some less rigorous approach. A central question to be addressed by the design specifications is whether a computerized accounting system is desirable, necessary, and/or practical. Transparency of the model is essential for a successful game; an elaborate computer model can cloud the picture. This is especially true in those cases where complexity, time, and lack of proven science prevent predictive models. If a computer is used, care must be given to facilitator training and providing access to appropriate technology during the use of the exercise.

The accounting system consists of a system of accounts and underlying model(s). Systems of accounts become fixed procedural agreements, whether or not known to the players, by which decisions of the players are processed and forwarded to another component of the game. Design specifications should determine what information flows are to be provided during the exercise, and which are to be monitored, recorded, and preserved beyond the exercise for purposes of evaluation.

A well-designed game obtains the commitment of the players by requiring explicit decisions that are processed through the accounting system. Even the casual observer can see the anxiety that is generated when players must first commit their position to paper. At this point, the players become involved and want to know whether their decision is valid. The results of player decisions become one of the most important sources of information and dynamics during the game. Players will be frustrated by the failure of the game mechanics to give an accurate and rapid response. Most policy games require a relatively formalized accounting system that deals consistently with player decisions. The EDF exercise required that players have access to a wide variety of factual information; the accounting system made extensive use of graphic displays, indexed notebooks, and computerized database capabilities (see Section 3.3).

An important element in game construction is the establishment of the order of processing through the accounting system. Most games deal with systems that are complex and nonlinear; in reality, the problem being addressed entails many simultaneous activities. However, in a game, the accounting system is inevitably rigidly sequential because of mechanical constraints. In some cases, many simple accounting components are linked together into a totality that is quite complex. The order of processing of various models, components, or decisions will inevitably be artificial in some sense.

After decisions are processed (by hand or by computer), the results will be of two basic kinds: player specific and general information. Player-specific information is delivered to individuals; general information should be posted for all players. The accounting system should continuously illustrate relationships among the actors. It requires sophisticated judgment by the designer to ensure that the participant experiences the results of the accounting system as an overview rather than as disconnected fragments of information.

Indicators

To the extent that a formal accounting system is used, it is necessary to focus the results of the decisions on the central aspects of the model under investigation; indicators are useful in this regard. Indicators are defined as a few specific outputs of the accounting system (presented through graphs, charts, etc.) that are available to the players as feedback of the results of play. These indicators should focus discussion on the accounting system results that address the most important aspects of the problem.