Designing a new simulation involves assembling a range of simulation models (algorithms), testing them and adjusting how they transform decisions into results. This is a time consuming and risky process that can be shortened and the quality ensured if you can call on a library of pre-designed and tested models. The table below lists some of the models that I have in my library that can be used to develop a business simulation for you. Below this I describe some simulation model design problems.

Business Simulation Model Problems

Over the years I have made a lot of mistakes when designing business simulation models, learnt from them and happily ensured that they never got into the final design.

Accounting Errors

A common problem is where the Balance Sheet does not balance because costs and cash expenditures conflict. For example, sometime ago I was asked to update a business simulation for a client. Unfortunately, the cost models and cash expenditure models had been designed by two dofferent people and as a result the Balance Sheet did not balance - yet the business simulation had been used for several years (by school children rather than accountants). When designing my Banking Challenge simulation because of the way a bank handles impaired loans getting the Balance Sheet to balance too a considerable time but this was primarily because I could not use my accounting models out-of-the-box!

I build into my simulations reports that reconcile accounting calculations. Not only do these reports ensure that there are no accounting errors but during use the tutor can access the reports answer learner questions. However, incorporating reconciliation reports increase reporting package significanttly.

Extreme Decisions

There is an apocryphal story about how a learner entered a price of many million pounds (or dollars or euros) and sold a single unit making a huge profit and as a consequence won the business game. I remember, when piloting my INTEX business simulation one team increased advertising a hundredfold. This irrational decision led to huge losses but did not destroy the simulation - it may be of interest that the team leader worked for an advertising agency and, I think, learnt a useful lesson.

How you handle extreme decisions is difficult - you have to allow the learners freedom to make unusual, perhaps, radical decisions but not allow them to make extreme decisions that would destroy the learning experience. Handling this involves two things - the design of the algorithms and screening decisions, commenting if they are unusual and rejecting them if they are illegally large or small. An example of an unusual decision that one company's learners made frequently was to zero advertising in one market sector so that they withdraw from that market and resulted in the business shrinking to an unviable level..However, building in a decision screen adds to the size of the model especially where decisions are progressively entered as the simulated time passes.

Overly Complex

To learn, the learners must be able to identify how their decisions impact results. And, if the learners cannot identify cause and effect they see the activity as a waste of time and become disengaged. An example of a complex simulation where it is difficult to identify cause and effect is my SMITE - Sales Management simulation. Here (and in general) the problem can be alleviated by providing comments and using the Tutor's Audit to help the trainer proactively managing the learning.

Dynamics and Stability

A business simulation involves a feedback process with decisions impacting results for several periods. This was illustrated my CISCO simulation that replicated a business that bidded for design and build contracts. Like the real world managing such a company is difficult because a characteristic is that the company swings from feast to famine. In other words, there were periods of too much business (with missed deadlines and  penalties) followed by periods where there was too little work (and idle staff). We managed to deal with this to an extent by adjusting the flow of contracts as the simulation ran. But despite this being real, as it was difficult for the learners to be successful, commonly they became disengaged.

Other Problems

The section on Software Soundness in my book on simulation design discusses other elements that can impact design quality (see LOVEANODDFAILURE model) - page 30 Chapter 8 - Design for Quality)

© 2011 Jeremy J. S. B. Hall

Most recent update 28/03/12
Hall Marketing, Studio 11, Colman's Wharf, 45 Morris Road, London E14 6PA, ENGLAND
Phone +44 (0)20 7537 2982 E-mail jeremyhall@simulations.co.uk