John H. Miller (1) (1959–)

This book provides a good introduction and explores several concrete models, using them to develop theory and point out strengths and weakness of various approaches. I would divide it into the following three sections:

Chapters 1-7 provide a slow and easy introduction to complexity and agent based models. Many interesting ideas are introduced but I had seen most of them in earlier books and found myself getting restless for something new. Since the book’s title had “introduction” in it I couldn’t fault the authors for taking the time to build up the concepts for people who had never been introduced to them before. I did enjoy the Buddhist “Eightfold Path” as applied to agent based models, where the authors describe the “Right View”, “Right Intention”, “Right Speech” etc . in terms of a good model. For instance “Right View” encompasses the information the agent receives for the world, “Right Speech” accounts for the information agents send to others and so on. Some of the analogies are a stretch as the authors point out, but it was a very nice way to think about modeling.

Chapters 8-12 cover more meaty examples and develop several “claims” (theorems) about the models. This was not quite what I was anticipating. These chapters are very practical for those working in the field and as the authors say later in the book “Great artists study the masters; so too must great modelers”. But I found the details of the models tedious and the mathematical formalism was in stark contrast to the first part of the book. Going from introductory material very accessible to a layperson to mathematical formalism, proof, Game Theory, Cellular Automata, and State Machines (in rule table format) is quite a jump. If you have been introduced to these subjects elsewhere then the material is not difficult, but for the authors to assume their readers will have this familiarity is expecting too much for an introductory text.

The book closes with two appendices. The first “An Open Agenda for Complex Adaptive Social Systems” provides a framework for the field and was full of all the interesting questions I was hoping the book would explore further. In my opinion it would have served as a great first chapter. The last appendix “Practices for Computational Models”, along with chapter 7, provided very practical and useful pointers for constructing computational models.… (altro)