Decision Analysis for Management Judgment, 6th Edition

8Applying simulation to decision problems

Introduction

When the payoff of a decision depends upon a large number of factors, estimating a probability distribution for the possible values of this payoff can be a difficult task. Consider, for example, the problem of estimating a probability distribution for the return that might be generated by a new product. The return on the investment will depend upon factors such as the size of the market, the market share which the product will achieve, the costs of launching the product, manufacturing and distribution costs, and the life of the product. We could, of course, ask the decision-maker to estimate the probability distribution directly (e.g., we might ask questions such as: ‘What is the probability that the investment will achieve a return of over 10% per annum?’). However, it is likely that many people would have difficulty in making this sort of judgment since all the factors which might influence the return on the investment, and the large number of ways in which they could interrelate, would have to be considered at the same time.

The decision-analysis approach to this problem is to help the decision-maker by initially dividing the probability assessment task into smaller parts (a process sometimes referred to as ‘credence decomposition’). Thus, we might ask the decision-maker to estimate individual probability distributions for the size of the market, the market share that will be achieved, the launch costs, and so ...