• Complex Adaptive Systems Demos (Macintosh)
  Robert Goldstone (Department of Psychology, Indiana University) maintains a site that provides Complex Adaptive Systems Demonstration Software for an undergraduate course he teaches on complex adaptive systems. The software programs available at this site, designed to be run on Macintosh computers, permit students to explore complex systems interactively by setting experimental parameters, conducting experiments, and instantly seeing the results of the experiments.

  Emergence: Interactive Tutorial

  Mitchel Resnick and Brian Silverman (Epistemology and Learning Group, MIT Media Laboratory) have developed an interactive tutorial on Emergence. The authors examine how objects and patterns can arise from simple interactions in ways that are surprising and counterintuitive.

  Note: A Java-enabled browser is required to run the applets accompanying this tutorial.

  Game of Life

  Paul Callahan maintains a site called What is the Game of Life?. This site explains and presents, in interactive form, a delightful game invented by John Conway and his students in 1970 demonstrating that even very simple rules can lead to extremely complicated outcomes.

  Johan Bontes (University of Wisconsin) maintains a site titled Life32 where people can go to download and install freeware for the Game of Life suitable for Windows Win9x/NT/2000/XP. Bontes describes Life32 as the "best, fastest, most user-friendly life player around."

  Mandelbrot and Julia Sets

  David Joyce (Department of Mathematics and Computer Science, Clark University, Worcester, MA) maintains an interactive website where interested parties can create and view images of Mandelbrot and Julia Sets. The website also includes an archive of recently created images as well as pointers to related websites.

  Newton Basins

  Nonlinear equations F(x)=0 typically have multiple solutions x. Newton's method for finding solutions to nonlinear equations is a successive approximation technique with the property that different starting points can converge to different solutions. A Newton basin is a set of starting points for Newton's method that all lead to one particular solution. It turns out that Newton basins are "fractals" in the complex plane with beautifully intricate boundaries that can be visualized by applying distinct colorations to starting points converging to distinct solutions (or to no solution). David Joyce (Department of Mathematics and Computer Science, Clark University, Worcester, MA) maintains an interactive website where interested parties can create and view images of Newton Basins.

  Schelling Segregation Model (Windows, C#/.Net)

  Chris Cook (Computer Science Department, Iowa State University, Ames) has developed an interactive demo of the Schelling Segregation Model (SSM), due to Thomas Schelling (Micromotives and Macrobehavior, Norton, 1978). Targeted for pc's operating under the Microsoft Windows operating system, the SSM demo has an easily manipulated graphical user interface. Agents are located on a chess board with 64 locations. The user determines the population mix from among three agent types (red, green, and blue), or chooses from among various default settings. The user can also specify a "happiness rule" for each agent type or select a default setting. The happiness rule determines when an agent is happy with his current board location, taking into account both the number and the types of his neighbors. If unhappy, the agent either attempts to move to a more desirable board location or exits the board altogether.

  Chris Cook has released his SSM demo as freeware under the GNU Public License. Automatic installation software for his SSM demo can be obtained at The Schelling Segregation Model: Demonstration Software Also available at this site is a more detailed description of the SSM, a description of the SSM demo's capabilities, instructions for using the automatic installation software, a link for accessing the SSM demo source code (C#/.Net), and copyright information.

  Schelling Segregation Simulator (Windows, Java Applet)

  Mark Fossett (Department of Sociology, Texas A&M University) has developed SimSeg, an interactive software program simulating the dynamics of residential segregation by ethnic and socioeconomic status. Program files and documentation files for downloading, installing, and running SimSeg are available at the SimSeg site. In addition, Fossett has developed a more limited Java applet version of the SimSeg program called SimSeg Lite, that can be run on-line via a java-enabled browser.

  Schelling Segregation Model: Computer Exercises (Visual Basic/Pascal)

  Robert Axelrod (Political Science and Public Policy, University of Michigan) has developed a series of computer exercises related to the Schelling Tipping (Segregation) Model, implemented in Visual Basic and Pascal. Materials available at this site include documentation and software.

EMERGENCE OF COLLECTIVE BEHAVIOR

• Boids: Simulation of Flocking Creatures
  Craig Reynolds (Sony, Research and Development Group) maintains a web site titled Boids featuring his simulations of flocking creatures called "boids." His basic flocking model consists of three simple steering behaviors possessed by each individual boid that govern how each boid maneuvers itself based on the positions and velocities of its nearby flockmates. As illustrated by the Java applets at this site, the model results in amazingly life-like collective flocking dynamics. Also available at this site is a link to work by Reynolds on an interactive system permitting user interaction with large groups of autonomous characters. The characters respond in real time to the user's interaction as well as to each other and their environment.

  Local-Interaction Games (Windows)

  Giorgio Fagiolo (St. Anna School of Advanced Studies, Pisa, Italy) has developed user-friendly software for the Windows desktop (9x/ME/00/XP) implementing several forms of Bilateral Games with Local Interaction Structure. The interation structure is exogenously given and can be either a lattice or a stochastic graph. The user can set all parameters (including the form of decision rule employed by the agents) and plot and save the simulation results. The software can be downloaded in a zip file. Once the set-up file is executed, an installation wizard provides guidance for all subsequent installation steps.

  Sand Pile Model

  Jos Thijssen (Applied Physics, Delft University of Technology, the Netherlands) maintains a site on Per Bak's sand pile model titled The Sand Pile Model and Self-Organized Criticality. Linked to this site is Thijssen's interactive sand pile demo in which the sand pile is modeled as a regular array of columns consisting of cubic sand grains.

EVOLUTION

• Biomorphs: Simulation of Biological Evolution
  In The Blind WatchMaker (W.W. Norton and Co., New York, 1987), Richard Dawkins introduces the concept of "biomorphs," i.e., encapsulated software programs capable of breeding. Those bred within a particular "pedigree" have a precisely known relationship to each other and a precisely known genetic make-up. Dawkins (Appendix, p. 15) claims that these features make biomorphs "ideally suited to verifying methods of taxonomy, and also to teaching the principles of taxonomy."

  A large variety of interactive biomorph demos are available on-line. For a simple example, see the demo at Richard Dawkins' Biomorphs. An interactive biomorph demo with interesting additional features can be found at Introduction to Biomorphs. An annotated list of biomorph demos and related support materials is available at Richard Dawkins: Software and Biomorphs.

  Exhibits on Evolution

  The Museum of Paleontology at the University of California at Berkeley maintains an on-line Museum of Paleontology Evolution Wing that permits visitors to explore exhibits and other resources related to the theory of evolution and the history of evolutionary thought.

LEARNING

• TNG Lab: A Computational Laboratory for Evolutionary Trade Networks (Windows, C++/Visual Basic)
  Leigh Tesfatsion (Iowa State University) and co-authors have developed the Trade Network Game (TNG) Laboratory for exploring the evolution of trade networks among strategically interacting traders. The TNG traders repeatedly choose and refuse their trade partners on the basis of expected payoffs, and trade interactions are modeled as two-person games. The TNG traders use genetic algorithm learning to evolve their trading strategies over time. Users can vary key model parameters, including game payoffs and learning parameters, on an easily operated graphical user interface and see run-time displays of the resulting experimental outcomes. The TNG Lab is targeted for the Microsoft Windows desktop (9x/00/ME/NT/XP). Automatic installation software for the TNG Lab, together with tutorials and research articles, can be obtained at the TNG Home Page.

NORMS

• Axelrod Tournament (Windows, C#/.Net)
  In 1979, Robert Axelrod (University of Michigan, Ann Arbor) posed an intriguing question: What type of strategy (if any) ensures good individual performance over the long haul when one is engaging in repeated Prisoner's Dilemma game play in round-robin fashion with multiple strangers whose strategies are not known in advance? Axelrod explored this question by conducting a computer tournament with strategies solicited from game experts from all over the world.

  Chris Cook (Computer Science Department, Iowa State University, Ames) has developed the Axelrod Tournament, a computational laboratory that captures and extends the salient aspects of Axelrod's 1979 tournament. The Axelrod Tournament permits a user to specify as treatment factors the number of tournament iterations, the total number of agents (strategies) comprising the initial tournament population, the types of agents in this population, and the form of payoff matrix for each agent type. The latter feature extends Axelrod's original tournament by permitting matched agents to engage in a wide variety of two-player games (Chicken, Stag Hunt, etc.) depending on their specified types. The user can then observe the comparative performance over time of each agent, and each agent type, by means of run-time graphical and text displays. The Axelrod Tournament is available as freeware for Windows desktops under the GNU Public License. Automatic installation software, together with more detailed information about its capabilities and implementation, can be obtained at The Axelrod Tournament: Demonstration Software (html).

  Evolution of Cooperation and the Prisoner's Dilemma

  The Complexity of Cooperation Website provides software and documentation (as well as other resources) related to two books by Robert Axelrod: (a) The Complexity of Cooperation: Agent-Based Models of Competition and Collaboration, Princeton University Press, 1997; and (b) The Evolution of Cooperation, Basic Books, 1994.

  Evolution of Cooperation: Interactive Tutorial

  The University of Toronto maintains an interactive tutorial on the Evolution of Cooperation, using the Prisoner's Dilemma Game for concrete illustration.

  Evolution of Trade Networks Supporting Cooperative Behavior

  Leigh Tesfatsion (Iowa State University) and coauthors have developed the Trade Network Game (TNG) Laboratory for exploring the evolution of trade networks among strategically interacting traders. The TNG traders repeatedly choose and refuse their trade partners on the basis of expected payoffs, and trades are modeled as two-person games in which the traders can either "cooperate" or "defect". The TNG user can set trade payoffs to mimic a wide variety of game configurations (e.g., prisoner's dilemma, chicken, stag hunt,...). The TNG lab permits systematic experimentation to determine conditions under which the trade networks that evolve over time support cooperative behaviors among buyers and sellers (e.g., workers and employers).

MARKETS

• Endogenous Growth Model (Windows)
  Giorgio Fagiolo (St. Anna School of Advanced Studies, Pisa, Italy) has developed user-friendly software for the Windows desktop (9x/ME/00/XP) implementing Islands: An Endogenous Growth Model earlier developed by Dosi and Fagiolo (2002). The software simulates an open-ended economy where boundedly rational firms can exploit existing resources, imitate other firms, explore the technological space, and/or endogenously introduce innovations. System parameters controlling technological and behavioral regimes can be set by the user or loaded from a text file. The software can be downloaded in a zip file. Once the set-up file is executed, an installation wizard provides guidance for all subsequent installation steps.

  Multi-Agent Economy Simulation (Java)

  Erich Kutschinski and Thoman Uthmann (University of Mainz, Germany) together with Daniel Polani (University of Hertfordshire, UK) have packaged and made publicly available the DMarks II software to accompany their article "Learning Competitive Pricing Strategies by Multi-Agent Reinforcement Learning," Journal of Economic Dynamics and Control, Volume 27 (2003), 2207-2218. Materials provided at this site include the source and compiled code, javadoc, some scripts showing how to run simulations, and a powerpoint presentation. The article is available online at Science Direct.

  Santa Fe Artificial Stock Market (Swarm and Objective C)

  The Santa Fe Institute (SFI) has open-sourced the Santa Fe Artificial Stock Market (ASM) simulation model, originally developed by a number of SFI researchers in Objective C using the Swarm toolkit. The latest ASM-Swarm file releases can be downloaded from an ASM sourceforge page and repository maintained by Paul E. Johnson (Political Science, University of Kansas, Lawrence).

  Zero-Intelligence Trading Demo (Java Applet/NegLogo Model)

  Mark McBride (Economics Department, Miami University, Oxford, Ohio) has developed Zero-Intelligence Trading Demonstration Software that is based upon the double-auction trading experiments conducted with "Zero-Intelligence (ZI)" traders by Gode and Sunder (Journal of Political Economy, 1993). A Java Applet is available at this site that permits you to run ZI trading experiments in a Java-enabled browser (Java 1.4.X or later). Alternatively, a NetLogo model can be downloaded for running on a personal computer with a NetLogo 3.1.1 (or later) installation.

NETWORKS

• Small-World Trade Network Demo (NetLogo)
  The Small-World Trade Network Demo developed by Mark McBride (Professor of Economics, Miami University, Oxford, Ohio) implements a trade model designed by (Alan Wilhite (Computational Economics, 2001). The purpose of the Wilhite trade model is to examine the effects of a small-world network structure on the market efficiency of bilateral trade. The Small-World Trade Network Demo can be either be run in a web browser or downloaded as a NetLogo file to run locally using NetLogo 4.0.3 (or higher). Possible network structures available for testing include globally (completely) connected, locally connected, locally disconnected, and small-world (local connectivity with global reach). See, also, Notes on the Wilhite Trade Model.

  TNG Lab: A Computational Laboratory for Evolutionary Trade Networks (Windows, C++/Visual Basic)

  Leigh Tesfatsion (Iowa State University) and co-authors have developed the Trade Network Game (TNG) Laboratory for exploring the evolution of trade networks among strategically interacting traders. The TNG Lab is targeted for the Microsoft Windows desktop (9x/00/ME/NT/XP) and is both modular and extensible (C++/Visual Basic). It permits visualization of the formation and evolution of trade networks by means of run-time animations, run-time charts, and run-time data displays. A clear, easily operated graphical user interface permits testing of key parameters pertaining to learning, payoffs, types and numbers of agents, and the physics controlling network displays. Automatic installation software for the TNG Lab, together with tutorials and research articles, can be obtained at the TNG Home Page.

MODELING TECHNIQUES

• Steven F. Railsback, Steven L. Lytinen, and Stephen K. Jackson have developed StupidModel: A Template Model for ABM Platforms. The template model is implemented in five different platforms: NetLogo; RepastJ; MASON; Java Swarm; and Objective C Swarm. Although relatively simple, StupidModel includes many commonly used features of agent-based modeling (ABM) platforms. Sixteen versions of StupidModel are implemented for each platform, beginning with a bare bones version and ending with a relatively sophisticated version that involves two agent types, a full agent life cycle (birth, reproduction, predation, and death), and a habitat with data read from an input file. Each implementation is made available as freeware with accompanying implementation notes. In addition, the authors include at this site a pointer to a paper titled "Agent-Based Simulation Platforms: Review and Development Recommendations" that reviews and compares the five ABM platforms and seeks to identify key development priorities both for these specific ABM platforms and for ABM platforms in general.

GENERAL SOFTWARE SUPPORT

• Software and Toolkits for Agent-Based Modeling

• Agent-Based Computational Laboratories and Demonstration Software

• Matt Weisfeld, The Object-Oriented Thought Process, SAMS Publishing (Division of Macmillan), Indianapolis, Indiana, 2000, ISBN: 0-672-31853-9 (paperback).
  Agent-based computational modeling is often implemented using languages with object-oriented programming (OOP) capabilities, such as Java, C++, and C#. This book is designed to help newcomers learn OOP guidelines for solid class design and master the major OOP concepts of inheritance, composition, interfaces, and abstract classes. The author motivates and illustrates his points by taking readers step by step through simple concrete examples.
  Matt Weisfeld has more than 18 years of experience as a professional software developer and project manager using COBOL, Pascal, C, C++, smalltalk, and Java. He currently teaches Java programming and project management at Cleveland State University.

Copyright © 2009 Robert Axelrod and Leigh Tesfatsion. All Rights Reserved.