[MUD-Dev] Game Economies

[Ling]

On Wed, 9 Jun 1999, Matthew Mihaly wrote:
> 
> I'm not an economist by a long stretch, but I've been trying to puzzle
> through some work that came out of the Santa Fe Institute in the late 80s
> and 90s on complexity, specifically as applied to economics, done by guys
> like Ken Arrow and Phil Anderson (both Nobel winners I believe). A lot of
> it, especially, the mathematics, goes over my head, but what I gather is
> that a real economic system cannot be properly modeled from the top down.
> Further, the idea of there being one (or indeed, any) equilibrium point in
> an economic system is bunk. It seems that the only way to model a real
> economic system (which no one has really done convincingly yet on any
> level as far as I can tell from the literature available), is to use
> fairly sophisticated techniques that produce emergent behavior among
> economic agents. It was sort of disheartening to read that, at least from
> the context of a mud developer, as modeling a real economic system,
> complete with positive feedback (which they have convincingly demonstrated
> is a real part of the economy, despite the party line spouted off by
> neo-classical economists), emergent structures, etc, is well beyond the
> capability of any mud that I've ever heard of. We all try to model things
> top-down, by providing pre-defined goals and methods for agents. Any
> comments on this?

The following link may prove enlightening (or not):

<URL:http://www.newscientist.com/nsplus/insight/ai/forecast.html>

Skip the first ten odd paragraphs to get straight to the stock market
simulator using moderately complex agents.  It's even in simple English.

Considering the simulation was conducted in 1987, I doubt the method
described above is beyond most muds these days.

PS: Do I append a copy of the text below?  I've never worked out if I
should.  Besides, I think I'd infringe copyrights for this article.

  |    Ling Lo
_O_O_  kllo at iee.org




_______________________________________________
MUD-Dev maillist  -  MUD-Dev at kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev