On Wed, 24 Apr 1996 George Richardson wrote:
> Eds point is that simulation allows more people access to such models.
This is just the aspect of SD that first attracted me. Coming from a
physical science background into economic forecasting in 1981, I was
disappointed that econometric models omitted feedback and were
largely linearized. This seemed too restrictive for the long term
studies required for policy analysis in capital intensive markets.
Getting turned on to the tradition of World Dynamics as a role model
and Dynamo as a toolkit (instead of single equation elasticity models
and SPSS) was just what I was looking for.
Many in the sciences truly believe that diff eqs are the stuff that
the universe runs on. SD methodology respects this "religion" and
gives believers a version of the creed and ritual in the vulgate so
they can understand and use it. (Sorry about the metaphor, George
Backus, but I couldnt resist.)
William Steinhurst
wsteinhu@psd.state.vt.us
System dynamics is not ...
-
William Steinhurst
- Junior Member
- Posts: 3
- Joined: Fri Mar 29, 2002 3:39 am
-
gallaher@teleport.com (Ed Gallah
- Member
- Posts: 39
- Joined: Fri Mar 29, 2002 3:39 am
System dynamics is not ...
>On Wed, 24 Apr 1996 George Richardson wrote:
>> Eds point is that simulation allows more people access to such models.
>
>This is just the aspect of SD that first attracted me.
Me too!
Ill have more to say on this in the future, but right now Im going to
mostly mull all this over.
With one exception:
I believe that the power of stock-and-flow diagrams per se are largely
unrecognized. They are considered to be a convenience, but otherwise, no
big deal. I think they are a very big deal.
Leibniz is generally credited with presenting the modern notation for
calculus, in such a way that the underlying concepts were much more clear
and consistent. This *notation* is not an unimportant aspect of calculus
and diffeq as we know them today.
Feynman received the Nobel prize for quantum electrodynamics (QED). What
he did (among other things) was develop Feynman diagrams, which illustrated
intuitively the underlying physical concepts, which were then confirmed
using traditional math techniques. Schwinger had been doing the same work
in the traditional way, but no one realized the overlap until Feynman
described what he was doing to Freeman Dyson, and Dyson spent the rest of
the summer comparing Feynams work with Schwingers math.
Again, the *notation* was a huge step forward.
Stock-and-flow diagrams allow us to accomplish a great deal, with about two
years less math background than would otherwise be required.
For example, the winning SyM Bowl model, on the loss of body temperature
after death is found as a standard problem in many diffeq texts. This same
work was accomplished by two high school seniors who are totally ignorant
of this fact.
They may be surprised as sophomore college students, in their 4th quarter
college math class, to discover this. They might also say, "I took 20 math
credits to do this.....? I was doing this two years ago in high school!"
Stock-and-flow diagrams easily handle what would otherwise be partial
differential equations, and nonlinear diffeqs, which would not be covered
adequately until at least several classes later.
The notation makes the difference!
We can debate the relative merits of Euler vs RK2 and RK4, and the
challenges of stiff models and variable dt, and so on, but the fact remains
that a great deal can be accomplished with simple S&F diagrams and simple
math techniques.
In fact, if we did not have a computer, once we developed a S&F diagram we
would be forced to use pencil and paper numerical analysis methods. Just
as Feynman did on the Manhattan project at Los Alamos during WWII.
I would assert that we would put pencil to paper much sooner if we used S&F
diagrams.
>Many in the sciences truly believe that diff eqs are the stuff that
>the universe runs on. SD methodology respects this "religion" and
>gives believers a version of the creed and ritual in the vulgate so
>they can understand and use it.
>William Steinhurst
Re the final paragraph, I couldnt have said it better myself.
ed
gallaher@teleport.com
>> Eds point is that simulation allows more people access to such models.
>
>This is just the aspect of SD that first attracted me.
Me too!
Ill have more to say on this in the future, but right now Im going to
mostly mull all this over.
With one exception:
I believe that the power of stock-and-flow diagrams per se are largely
unrecognized. They are considered to be a convenience, but otherwise, no
big deal. I think they are a very big deal.
Leibniz is generally credited with presenting the modern notation for
calculus, in such a way that the underlying concepts were much more clear
and consistent. This *notation* is not an unimportant aspect of calculus
and diffeq as we know them today.
Feynman received the Nobel prize for quantum electrodynamics (QED). What
he did (among other things) was develop Feynman diagrams, which illustrated
intuitively the underlying physical concepts, which were then confirmed
using traditional math techniques. Schwinger had been doing the same work
in the traditional way, but no one realized the overlap until Feynman
described what he was doing to Freeman Dyson, and Dyson spent the rest of
the summer comparing Feynams work with Schwingers math.
Again, the *notation* was a huge step forward.
Stock-and-flow diagrams allow us to accomplish a great deal, with about two
years less math background than would otherwise be required.
For example, the winning SyM Bowl model, on the loss of body temperature
after death is found as a standard problem in many diffeq texts. This same
work was accomplished by two high school seniors who are totally ignorant
of this fact.
They may be surprised as sophomore college students, in their 4th quarter
college math class, to discover this. They might also say, "I took 20 math
credits to do this.....? I was doing this two years ago in high school!"
Stock-and-flow diagrams easily handle what would otherwise be partial
differential equations, and nonlinear diffeqs, which would not be covered
adequately until at least several classes later.
The notation makes the difference!
We can debate the relative merits of Euler vs RK2 and RK4, and the
challenges of stiff models and variable dt, and so on, but the fact remains
that a great deal can be accomplished with simple S&F diagrams and simple
math techniques.
In fact, if we did not have a computer, once we developed a S&F diagram we
would be forced to use pencil and paper numerical analysis methods. Just
as Feynman did on the Manhattan project at Los Alamos during WWII.
I would assert that we would put pencil to paper much sooner if we used S&F
diagrams.
>Many in the sciences truly believe that diff eqs are the stuff that
>the universe runs on. SD methodology respects this "religion" and
>gives believers a version of the creed and ritual in the vulgate so
>they can understand and use it.
>William Steinhurst
Re the final paragraph, I couldnt have said it better myself.
ed
gallaher@teleport.com