Re: Are there theorems for Agent Based Models

[Rick Riolo]

As per the citations in my other message,
this involved a little go-around with Nowak and May
making the origanal claims about the evolutin of cooperation
and chaotic patterns in their simple synchronous update model.
Its true that under some conditions cooperation
does not evolve in the non-synchronous verions huberman
and glance introduced (in fact what they did was make
an DES-type version, ala mousetraps, with encounters
being scheduled at various times in the future).
*But* nowak etal showed in the subsequent paper that
for a broad range of conditions cooperation still emerged
even in the asynchronous update version.

Still, the point stands that it does make a difference,
and one should think carefully (or at least a bit) about
what is more appropriate for the system one is modeling.

- r

Rick Riolo                           address@hidden
Center for Study of Complex Systems (CSCS)
4477 Randall Lab                
University of Michigan         Ann Arbor MI 48109-1120
Phone: 734 763 3323                  Fax: 734 763 9267
http://www.pscs.umich.edu/PEOPLE/rlr-home.html

On Mon, 12 Jul 1999, Kevin Crowston wrote:

> Date: Mon, 12 Jul 1999 22:50:41 +0200
> From: Kevin Crowston <address@hidden>
> Reply-To: address@hidden
> To: address@hidden
> Subject: Re: Are there theorems for Agent Based Models
> 
> At 3:37 PM -0500 7/12/99, Sven N. Thommesen wrote:
> >I believe Huberman and Glance had a paper where they investigated just this
> >difference, and they discovered that doing it in way #2 overturned previous
> >results achieved with way #1. Their conclusion was that it makes a Big
> >Difference how you arrange the order of actions.
> >
> >I'll dig out the reference if you need it.
> 
> 
> I think it's
> 
> Huberman, B. A. and Glance, N. S.  (1993).  Evolutionary games and 
> computer simulations.  Proceedings of the National Academy of 
> Science, USA, 90(August), 7716-7718.
> 
> 
> If I'm remember correctly, they reran a simulation that produced nice 
> ever-evolving patterns when run in way #1, I think having to do with 
> agents who were cooperative managing to survive even when surrounded 
> by non-cooperative agents (a nice story).  When run in way #2, the 
> non-cooperators took over completely (a sad story--at least, if my 
> memory is accurate--but one that seems a bit more likely).  And as 
> Sven says, a Big Difference in results.
> 
> Kevin Crowston
> Syracuse University
> School of Information Studies
> 4-206 Centre for Science and Technology
> Syracuse, NY   13244-4100               



                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================