Justin,
I reviewed and enjoyed your PPT presentation and I must confess got a
little lost in the economics jargon, but felt impelled to share an
observation and pose some questions...
At one point you complain of not being able to get the message out to
the business world and provided a number of good reasons. However,
competent businesses are forced to justify their decisions (at least
internally) while balancing risk. In any application of computer
modeling, regardless of the underlying technology (from well established
engineering principles to system dynamics), for it to be useful it has
to have predictive capabilities or characteristics. No one, or at least
very few in your field, gets specific about predictability. (In some
fields there may be a good reason: If I knew of an accurate way to
predict stock market behavior and told the world it would probably be
self defeating. The presence of the model defeats the model.)
Competent business people will seek assurances of predictability.
You point to the need to pursue the "Brand Experience Cycle" when
promoting the discipline of SD/ABM/etc. Have you applied your own
technological approach to modeling your own success? Did you estimate
the numbers game sales professional insist on knowing: from x contacts
y% move to the next stage, z% to the next and so on? Did you back
calculate the effort needed to meet your two year goal and what did it
look like?
In the spectrum of effective computer models, that can vary from 100%
totally inaccurate to 100% totally accurate, I suspect No Model At All
is slap bang in the middle. Convincing managers that your model lies
firmly on the (predictably) accurate side seems all that is needed to
close a deal. I've been there and done that (but not always.)
As a model (sic) for your discipline (ABM/SD/etc) have you looked at
the field of Process Control that uses Model Predictive Control to
manage difficult manufacturing control problems
(http://en.wikipedia.org/wiki/Model_predictive_control). It is now, I
believe, a well established engineering sub-discipline.
Robert C
Justin Lyon wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
As promised, here is the URL for the Brazil talk I gave.
http://s158641480.onlinehome.us/public/Justin_Lyon__Future_of_Simulation_Brazil_Conference_2006v2.ppt
Download the PPT so you can read the notes on some of the slides.
Any mistakes are mine.
Any surprising insights are the work of other people smarter than me.
:-P
Please point out mistakes to me so I can fix them.
- --
Best regards,
Justin Lyon
M: +423 663 168892 (Worldwide)
M: +44 781 480 2797 (London, UK)
O: +1 210 787-3498 (San Antonio, USA)
O: +44 20 8144 4072 (London, UK)
E: [EMAIL PROTECTED]
W: http://www.simudyne.com
Justin Lyon wrote:
Owen and Phil,
Traditional economics borrowed ideas from equilibrium physics and tried
to force-fit reality to make the math tractable prior to the emergence
of computers sufficient to solve the math. No need to do that now.
Also, Pareto optimitality is a sacred cow that must be slaughtered.
And, the myth of rational humans....hah...that is a good one.
I'll find the PPT for the Brazil conference, type up my notes and post
it to the web for the group's reading pleasure. The conference
organizers recorded the speech and I'll try to hunt down the URL.
I will be giving a workshop on simulation in London for a Gartner Group
conference where I will elaborate in more detail.
http://www.europe.gartner.com/bpm
============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFFt20LhfoqghrmIrARCoO9AKCTBaGMHPqaDgXo6I023bdJQYDPRACgjOj4
fZfjb9UkmOGhZH6kmijtpZU=
=vfNd
-----END PGP SIGNATURE-----
============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org
============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org
