In his preface, Dr David Hillson describes David Hulett’s book as “how to actually apply the statistical techniques of Monte Carlo analysis in a way that combines intellectual rigour with practical realism”. This book is all that, and more. It provides a structured development of ideas for modelling all random and systematic effects on the schedule.
Practical Schedule Risk Analysis describes the inherent difficulty in scheduling, and observes that project scheduling is unevenly practiced in delivery organisations. It discusses the varying attitudes taken by senior management to the information provided by the schedule; unrealistic constraints and completion dates, the assumption that the estimating exercise provides definitive answers, the power imbalance between management and schedulers, and the resultant fudging to create artificial schedules that meet demanded dates.
The complete build-up of the statistical analysis is taken step-by-step, the first consideration being a probability distribution for the estimates of a single task – in essence, if a large number of similar project teams undertook the same task, how would the durations taken compare? This is developed into a simple project comprising four sequential activities, showing how statistics combine across a project. This model is used to introduce the S-curve, to show that there is increasing probability of delivery with increasing duration. This vital step takes us away from considering the truth to be represented by the single-valued, deterministic schedule normally produced by a scheduler with one of the common tools available, to a model that quantifies the uncertainty of delivery.
These ideas are extended into more complex schedules with parallel paths and merge points common in most schedules, e.g. when the products of sub-projects are brought together for integration into a system. The model is used to elaborate further on the strengths of Monte Carlo analysis and to point to the shortcomings of the traditional Critical Path Method in complex schedules; an area that is developed further in an appendix.
Analysis includes Risk Management, where some activities within a plan may only be executed in response to a set of conditions, as is the case with contingency plans. The effects on the schedule of risks affecting many activities are also discussed, with explanations for how these are modelled.
The logical and structured development of the ideas is slightly impeded by the uneven editing of this book. Although all chapters finish with a summary, not all have an introduction. The chapter addressing the fundamentals of good scheduling, which would have been useful at the beginning, does not appear until part way through the development of the approach to analysis. The reasons for using large numbers of iterations in a Monte Carlo model are not well presented, and the development of ideas on correlation seems complicated for a practical guide.
In summary, this is an extremely important book, which presents a depth of understanding of estimating and scheduling rarely seen in industry. It explains the techniques used in a number of schedule analysis software packages, and encourages a more mature approach to understanding the information presented by a project schedule.
It should be required reading for all project managers, schedulers, and risk practitioners; and also for programme managers, sponsors, and all staff involved in contracting for project delivery.
ABOUT THE REVIEWER: John Greenwood has around fifteen years of project management experience gained in the engineering and IT industries, and has been an active member of the PMI UK Chapter. He holds a degree in Physics from the University of Birmingham, and has worked for a number of years as a Systems Engineer, where he has used Monte Carlo techniques to analyse and predict the performance of radar systems.