Thomas F. Pray

Historical review of algorithm development for computerized business simulations

This article presents a 25-year historical review of algorithm development for computerized business simulations. The articles selected were drawn extensively from the annual proceedings of the Association for Business Simulation and Experiential Learning (ABSEL) conferences. The authors have categorized the historical review into two broad areas: functional business topics and special topics of interest. Both narrative and summary tables address marketing, accounting, finance, human resource, and production/ operations areas. The special topics of interest reviewed include the algorithms for quality and the time-based modeling concerns. The authors conclude the review with issues future modelers of business games should consider.

The Production Frontier: Modeling Production in Computerized Business Simulations

The production function defines the technology of the business and the relationship between the inputs used by the business and the quantity of goods or services provided. The specification of this relationship should play an important role in business or management simulations. The costs of doing business and profits are directly related to the production function. The production function will influence the firm’s optimal mix of inputs, inventory policy, capital utilization, and capital investment. Business and management simulations are modeled to represent the &dquo;real world&dquo; firm. Participants in a simulation should gain insights into the workings of the real world through involvement with the simulation. As a result, it is necessary that the functions and algorithms within the simulation be consistent with the economic theory of production. Although the economic theory of production is well known, the task of incorporating the theoretical properties into a computerized business simulation is not straightforward, as Goosen (1986) stated:

Changing Customer Preferences and Product Characteristics in the Design of Demand Functions

A mathematical approach is developed to model product characteristics in demand functions that considers the complexities of time, technological change, and consumer preferences. Changing product characteristics are represented by assigning a new set of attributes desired by the consumer. A demand function, which allows for intertemporal changes in product attributes and consumer preferences, is simulated. Incorporating new product development in contemporary business simulations allows a number of enrichments. These include direct market influences from technology and intertemporal changes in consumer tastes, the creation of market segment diversity in a single product simulation, creating a structure permitting firms to target specific markets with different characteristics, including fact-based data for use in customer satisfaction surveys, and the creation of a modeling approach that introduces quality factors as attributes in a simulated demand function.

The use of the gamma probability distribution: a critique of Carvalho's demand simulator

The article by Carvalho (1995, this issue) advocates the use of a stochastic demand simulator based on the use of the gamma probability distribution. He argues that this simulator is an improvement over existing demand models because it is consistent with the theoretical properties of demand and is considerably more flexible than those currently available. Our critique focuses on four major points: (a) Carvalhos assumptions are too restrictive, (b) the parameters of the Carvalho model are difficult to set and thus lack flexibility in classroom use, (c) the demand function may be unstable when used in a dynamic simulation environment, and (d) the model does not have the unique property of independence between consumers and suppliers claimed by the author

Integrating Visualization into the Modeling of Business Simulations

This article demonstrates the advantages of using visualization as part of the modeling process. Several examples are given to show how visualization can help developers to more completely understand the range of behaviors for their algorithms. Specifically, the Cobb Douglas function and Gold and Pray demand system are examined using a tool that combines mathematical modeling with visualization capabilities.