It seems only yesterday that the teacher of a course on discrete-event simulation was limited in the choice of a text. The last couple of years have witnessed a deluge of introductory texts covering this topic. This one caught my eye even before I was asked to review it because it is written by a practitioner. Thus the focus is different from that of many texts written by academics.
The practical perspective of this text results in a strong emphasis on model building. A chapter called “Building a Simulation the Right Way” is the heart of the book. Here the simulation process is broken into 15 different phases, including the common steps of system definition, input data modeling, conceptual modeling, selection of a simulation tool, model construction, validation, and experimentation. The last two chapters show each of these phases at work in the context of a fairly detailed case study.
Much of the book is based around GPSS/H, and a 40-page chapter gives a detailed introduction to it. While the choice of language can be argued with, most introductory simulation courses still use some implementation of GPSS, and the presentation of the language is authoritative. Other practical aspects of the text include a chapter on being a professional simulation analyst, a nice chapter on the analysis of what the author calls a “sick simulation,” and a down-to-earth discussion on the objectives of simulation and the problems of determining the length and extent of a simulation project.
Given the objective of this text--a compact practical introduction to simulation model building--I see only two problems with it. The first is the lack of details on output analysis. Many simulation texts include pages and pages of excessive details on output analysis methods that are never used in practice, and the avoidance of this is to be applauded. This text goes too far the other way, however; I could find only a single paragraph on replications and estimation from multiple runs. The second problem is that the book is limited to simulation as preached and practiced in the United States. Thus a short history of simulation wrongly credits Gordon and GPSS as the dawn of simulation tools, rather than Tocher and GSP in England, and the only non-American tool that gets any mention is WITNESS. This shortcoming is a shame, since it limits the attractiveness of the text to those outside the United States.
Overall, then, McHaney has produced a good text with a different slant. I am not sure that I would use it as the sole text in a course, but it would make for good additional reading from a different viewpoint and a welcome antidote to texts that bore the reader with excessive statistics. It would also be a nice text to give away after a short one- or two-week training course, particularly one that used GPSS. Unfortunately, Academic Press has priced the text too high to make it a second required, or even recommended, text. At $25, this book would have a decent market; at $49.95, I doubt that it will sell well.