Looking at the Range of Outcomes
Our next step is to develop a cumulative probability graph (sometimes called an s-curve) for our decision. This will show us the full range of outcomes given our assessed inputs. No single outcome on the tree is more or less likely than any other point—my friend David Skinner expresses this as, “You’ve bought the whole curve.” Comparing the curves for each alternative usually gives us insight as to which alternative best fits our appetite for risk.
There are two ways to develop the cumulative probability graph. We can
• run each case in our decision tree after we have assessments for p10/50/90s for each uncertainty and then solve (roll back) the tree or
• run a Monte Carlo simulation. ...