Three Point Estimating

Three Point Estimating Training

Our one-day Three Point Estimating Course is aimed at estimating practitioners of all levels and senior stakeholders responsible for investment decision.  It provides a sound understanding of three point estimating concepts and the tools required to develop statistically valid cost-probability distributions for built environment applications.

A Three Point Estimates is an estimate of the range of possible outturns from a Plausible Minimum to a Plausible Maximum with the Most Likely outturn located between the two extremes.  In addition, a Three Point Estimate specifies the shape of the variation from Plausible Minimum, through Most Likely to Plausible Maximum. The figure below shows these key elements of a Three Point Estimate in comparison with a range estimating approach using a uniform distribution.

Frequency Distribution Curve

Three Point Estimates are produced for the various components of an estimate and combined using a statistical model to provide a realistic representation of the range of possible outturn costs.

The range between Minimum Plausible and Maximum Plausible represents the potential for variation in the quantum and price of all the items to deliver a project.  The confidence in the quantities allocated to the items will vary according to many factors, such as knowledge of the level of design completed and the amount of geotechnical and other survey work undertaken.  This confidence will increase as the project progresses through the project lifecycle and the range between Minimum Plausible and Maximum Plausible will narrow.

The range also reflects the potential variation in the item rates used for pricing. Where rates are based on historical cost data, consideration needs to be given to the suitability of the projects on which the item rates are based. It is important when using historical information, that the rates are resilient and have been verified against contemporary rates.  Where item rates have been built up from first principles then there will naturally be potentially variation in the method of calculation and the underlying production rates and resource rate data.