ABSTRACT
Introduction Up to this point task durations have been shown as a single time estimate and from that a project schedule and completion date were calculated. Sufficient background has now been developed to relax this simplifying assumption. It should now be obvious that project dynamics and the related complexity will make any work unit estimate somewhat uncertain. For this reason, it is time to explore how to handle this recognition. Think about the implications of this new view. If the plan says that a project is going to be completed on June 1, we now know that is highly unlikely to be the case. The question to ponder from this is, “what would be a more honest approach?” Would it not be more accurate to say that the most likely completion estimate is June 1, but there is a 50% chance that it could exceed August 1, or that there is a 90% chance of the completion being in the range of April 1 through August 1. It may well be true that stakeholders are more used to getting a single answer to this question, but from a maturity viewpoint, we need to infuse a more insightful (and accurate) approach to both scheduling and cost estimates. One way to accomplish this is by estimating task times using an assumed probability distribution with multiple estimated task duration values. The goal of this chapter is to describe a classic approach for dealing with this class of project schedule estimation.
