Value of Perfect Information

Another way to determine if you should expend resources to refine your model is to use the Value of Perfect Information tool. This tool calculates the value of knowing today what will happen in a future event.

Suppose someone knew for certain if your rezone request was going to be approved. What would you be willing to pay for this information?

To find out, point to Decision Tree in the Tools menu and click Value of Perfect Information. Next, enter the name of the root node in your model and the node you want to value.

DecisionPro will display the result as follows:

What does this result mean?  Seldom will you run into a situation where someone can sell you definitive information about what will happen in the future. However, you can generally buy research that will help you come closer to knowing what will happen. Calculating the value of knowing precisely what will happen sets an upper limit on the value of that research. This value helps you determine if you should invest in additional research and the maximum amount you should be willing to pay for it.

How It Works

The Value of Perfect Information tool calculates a result by internally restructuring your tree to reflect a situation where you have perfect information. Then, it compares this to the initial problem.

If you could buy information about whether or not the rezone request will be approved, the decision tree would look something like this:

You don't know at the time you buy the information what it will tell you. But, you already know there is a 60% chance the information will tell you that the rezone will be approved and there is a 40% chance it will be denied.

The value of this tree is $27,000, or $12,800 more than the $14,200 value in the original tree. Therefore, the perfect information is worth $12,800.