Company Name

Distribution Game

The Distribution Simulation is a direct descendent of The Beer Game first introduced in the 60's by Jay Forrester and recently revitalized as a result of its appearance in Senge's widely read Fifth Discipline (1990). The Beer Game consists of four players who assume different roles in a beer distribution system: Retailer, Distributor, Wholesaler, and Factory. Each player receives orders from downstream that he/she tries to fill with current inventory and then sends his/her orders up stream to try and replenish what has been lost. Each excess case of beer in any player's inventory costs 50 cents a week and each case backordered (negative inventory) costs one dollar. The goal of the game is for a team of four players to try and minimize their total inventory costs over time. The team with the lowest costs wins. This kind of participatory model dramatically brings home the ways in which structure and well intended local actions can sum together to create unintended and undesirable system behavior. However, the simulation process associated with implementing the model has proven sufficiently cumbersome so as to present a real barrier to its successful implementation in classrooms. As part of the activity of this grant a hand-held based version for the TI-83 has been developed and linked to a STELLA model. A modification of the original game has been made so that each team of student does not start from scratch with the company but instead are "hired" as a consulting team to take control of a company after 11 weeks under a different management team. The students are given sets of graphs and data from the first eleven weeks.

Graph of typical "Consultant Team" Prediction

Subsequent Results from Playing the Consultant Version of the Distribution Simulation

They are then taken through a process of analyzing the current behavior, predicting where the company would end up if it continued under the current management strategy, coming up with a strategy for doing better than this prediction and then implementing that strategy with the goal of minimizing inventory costs from week 11 until week 22. At that point the progress of the team is compared with their predictions and the results from other consulting groups in the class. This game has been played for a number of iterations and has been mirrored with a STELLA version.

Uri Wilensky
Northwesten University
Evanston, IL

Walter M. Stroup
The University of Texas
Austin, Texas


Distribution Game

The Distribution Simulation is a direct descendent of The Beer Game first introduced in the 60's by Jay Forrester and recently revitalized as a result of its appearance in Senge's widely read Fifth Discipline (1990). The Beer Game consists of four players who assume different roles in a beer distribution system: Retailer, Distributor, Wholesaler, and Factory. Each player receives orders from downstream that he/she tries to fill with current inventory and then sends his/her orders up stream to try and replenish what has been lost. Each excess case of beer in any player's inventory costs 50 cents a week and each case backordered (negative inventory) costs one dollar. The goal of the game is for a team of four players to try and minimize their total inventory costs over time. The team with the lowest costs wins. This kind of participatory model dramatically brings home the ways in which structure and well intended local actions can sum together to create unintended and undesirable system behavior. However, the simulation process associated with implementing the model has proven sufficiently cumbersome so as to present a real barrier to its successful implementation in classrooms. As part of the activity of this grant a hand-held based version for the TI-83 has been developed and linked to a STELLA model. A modification of the original game has been made so that each team of student does not start from scratch with the company but instead are "hired" as a consulting team to take control of a company after 11 weeks under a different management team. The students are given sets of graphs and data from the first eleven weeks.

Graph of typical "Consultant Team" Prediction
Subsequent Results from Playing the Consultant Version of the Distribution Simulation
They are then taken through a process of analyzing the current behavior, predicting where the company would end up if it continued under the current management strategy, coming up with a strategy for doing better than this prediction and then implementing that strategy with the goal of minimizing inventory costs from week 11 until week 22. At that point the progress of the team is compared with their predictions and the results from other consulting groups in the class. This game has been played for a number of iterations and has been mirrored with a STELLA version.

Uri Wilensky Northwesten University Evanston, IL		Walter M. Stroup The University of Texas Austin, Texas