Searching \ for '[OT] Hacker's Diet as Systems Engineering?' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=hackers+diet+systems
Search entire site for: 'Hacker's Diet as Systems Engineering?'.

Exact match. Not showing close matches.
PICList Thread
'[OT] Hacker's Diet as Systems Engineering?'
2005\12\20@121025 by Gerhard Fiedler

picon face
M. Adam Davis wrote:

>> Seriously, from a systems engineering standpoint that approach fails
>> quite a bit.

> I what ways does it fail?  Like any approximation it can't take
> everything into account, but from a general overview what goes in must
> come out as either energy or matter.

Right. But he equals energy (kcal) with matter (weight). That's probably
quite a bit short of the physiological reality, and exactly the difference
between energy and substance (that is, that the metabolism does see a
difference between e.g. sugar and olive oil) is the basis of many other
dieting approaches that he so quickly discards. Then he completely ignores
what goes out; there is the first picture with something going in,
something going out, and something being consumed, but after that, he never
talks about it again and just ignores what goes out, apparently assuming
that it is either constant or doesn't matter.

Two major issues, IMO. Every model is an approximation, I know that, but I
also know that from a systems engineering standpoint you can't expect too
much from a model that doesn't cover essential elements of the system.

Gerhard

2005\12\20@130601 by M. Adam Davis

face picon face
Interesting.  So you're saying that it's a poor approximation because
1) He just counts incoming calories, but there is a difference between
sugar calories and olive oil calories (for example) and
2) He ignores the output of the system

As far as I can tell, the basic process is this:
Count incoming calories daily
Take weight daily
Note extra exercise or activity

After a suitable history of calories, weight, and exercise is logged
the algorithm then indicates how many calories you should be eating
daily to maintain a stable weight.

This appears to be a fairly closed system.  Your points are valid - he
doesn't differentiate based on the type of food, nor does he
accurately track the output of the system.  Knowing the input and the
weight change, one can only approximate the output.

However, it appears to be self adjusting - if you spend three weeks
eating nothing but sweets, the daily calorie value will change
accordingly.

So I fail to see how it's not systems engineering.  If I'm working on
a system and I can only conveniently measure a few inputs and a few
outputs, then I still consider it systems engineering even if I don't
have a complete set of data.  It's a rather loose loop, but it is a
closed loop system.

I suppose the issue, however, is what is considered an essential
element of the system.  A more rigorous examination of the problem
would show just how much better the system could be controlled when
the output and input are more accurately measured.  I suspect that a
large increase in accurate data would only yield a small
performance/efficiency increase.

-Adam


On 12/20/05, Gerhard Fiedler <spam_OUTlistsTakeThisOuTspamconnectionbrazil.com> wrote:
{Quote hidden}

> -

2005\12\20@163240 by Gerhard Fiedler

picon face
I think you may have misunderstood me. Note that I'm not saying his method
is bad. Seems interesting, and probably works for many people. (Whether or
not it works seems to be much more a question of personality than anything
else.) But he's discrediting other methods based on his "system
engineering" -- which I think is not good enough to tell anything about
many of the other methods. His model is not good enough to explain why his
method works (he doesn't really use it for that either) and it definitely
does not explain why the others should not work. So I'm not talking about
the method (which is what you are describing), I'm talking about the model.

His "method" doesn't use his model at all. Basically, he measures weight,
and if it's too much, he reduces the intake. It's as simple as that. This
done rigorously can probably be quite effective. But it has not much to do
with his model. And I'm pretty sure that pretty much everybody could tell
you that this could work, without any system engineering background.

A number of other methods are based on assumptions about digestion and
metabolism (responsible for what happens with incoming matter) -- whether
that matter gets transformed into energy (which gets used), whether it gets
"stored" (that is, weight gain) or whether it gets exhausted (that would be
the output that he ignores). He says those other methods are all crap,
basically -- yet his model doesn't even start to describe the mechanisms
that are responsible for transforming incoming matter into stored matter or
into energy or leave it going out.

His method doesn't include modifications to the metabolism or to the
digestion at all. This possibly can be explained with a relative lack of
familiarity with the involved processes, as opposed to a higher familiarity
with simple systems engineering. Yet it is quite possible that a method
that results in a modified metabolism or digestion could have equally good
effects, or maybe better ones, and he simply ignores that.

Again, I'm looking at it from a system engineering point of view. The model
is never perfect, but I need to know what the model can do and what it
can't do -- and use it properly. Using it for what it can't do is bad
system engineering. I can still get to results that work; the fact that it
works doesn't make it necessarily good system engineering.

I think for further discussion we would have to take his text and get into
the details. Not sure this is the right place here (even though the
approach would be a system engineering approach that is useful for many
engineering tasks).

Gerhard


M. Adam Davis wrote:

> Interesting.  [...]

More... (looser matching)
- Last day of these posts
- In 2005 , 2006 only
- Today
- New search...