[MUD-Dev] Understanding Simulation (was: Point of View)

[Ted L. Chen]

Sasha Hart wrote:
> [Ted L. Chen]

>> Raph and some of the boys from UO in the early days can probably
>> attest to the stability of programmed ecology systems.

> They can certainly attest to the stability of _their_ programmed
> ecology systems. UO's model is not representative of all possible
> programmed ecology models.

> There is a vast space of possible ecology models which UO does not
> even begin to exhaust. There is no demonstration that it is
> impossible to devise a model that would have worked well in UO,
> even on a practical level. There is not even a demonstration that
> different parameters wouldn't have worked OK. My assumption has
> been that fixing it wasn't amenable in terms of cost:benefit at
> that time, and thus that it was ditched in favor of something
> which more neatly met the needs of the players.

Oh, I'm sure there are stable ecology models out there somewhere.
But chances of you hitting it if you create one blindly are about as
much as you creating a good airplane without doing any analysis.  I
think the analogy is apt in that if you try to put more "parts" into
it, the more important that analysis and design cycle becomes, and
the less likely the plane would fly well.

> The uses of a model, and therefore the dynamics which qualify as
> undesirable, vary. Even the UO model might have been perfectly
> usable with zero change, if it were not important that there be a
> stable supply of animals to kill. This stability thing isn't even
> a consistent requirement for ecology models.

Comon', that's like saying "I meant to do that."  Of course you can
flag anything as "successful" if you change the
specifications/requirements for it.  That's not very useful for
avoiding that same situation in the future.  As for stability being
important in an ecology model?  In academia, the answer is no.
Seeing a population go extinct is just as informative as seeing one
last forever.  In MMOGdom, it's mostly yes.  You'll be hard pressed
enough to come up with different species to populate your ecology.
Having to watch your work evaporate as they go extinct one at a time
is not a very fun thing to do.

> The UO example is instructive for people who contemplate models
> like UO's for similar purposes. For different models or ones with
> different purposes it really says very little. (Although trying to
> use it as a proof of the bankruptcy of ecology systems makes me
> roll *my* eyes).

Heh, I deserve that.  :) I would say that UO's failure was due more
to design methodology than it was the actual model.  Oh hell, the
model was bad for what they were trying to do, but liping 'their
model is different from ours' and leaving it at that won't keep you
from repeating the steps that lead to their ill-fated model.

I do think ecology systems are doable in code.  However, I also get
the impression that a majority of 'ecologists' on this list simplify
/or ignore the dynamics.  The rest, already have an inkling of what
components might be causing the dynamics (eg. predators) to avoid
them.  But that's wrong too as that throws the baby out with the
bathwater.

In either case though, talk still degenerates into "if we use this
lego-block with this lego-block, tune it, then it'll work" variety.
Or the ever more scary (to me), "to keep this A check, we'll add B." 
Which I think greatly oversimplifies things.  In some ways, that's
the design methodology the UO model followed.  There's a grand
difference between "including a factor" and "including a factor that
improves the system," and that's where I hoped the System Dynamics
references might be of some use to people.

> On the same grounds, don't discount the fact that there are still
> animals after millions of years of inter- and intra-species
> gobbling, either. Or after thousands of years of man, or a number
> of decades of egregious pollution. Thus, if you are arguing that
> the real world is a demonstration that extinction is inevitable
> (perhaps for sufficiently realistic models), you also have given a
> demonstration that a single extinction isn't a terminal
> catastrophe for everything in the whole world. Of course, in
> either case we have the advantage over nature that we can make the
> rules nearly however we want them.

Thanks for bringing up the point.  And the difference is all in
scale.  The original comparison I made between simulated ecologies
and nature are on the level of a few selected species.  The closest
analogy would be species from the semi-isolated Galapagos islands.
At that scale, species emerge and die of their own accord.  If that
was your whole ecology, it would be a very precarious state.  Heaven
forbid any large external disturbances.

The resilience of Earth's ecology, on the other hand, works on a
vastly larger scale.  One, which I admit is amazing.  But one which
I don't see much parallel in MMOG development.  The main reason
again is due to the fact of content development, something no one
can do at that scale.  And if you're thinking "hey, use GA's to
evolve the content!" then I'll roll my eyes in another direction :P

> The real world is resilient enough to be interesting despite
> single extinctions. They are very sad, but they do not constitute
> total failures. In much the same way, banking survives despite
> single bank robberies, and people still know the ABC despite the
> deaths of single people who knew the ABC. That kind of resiliency
> isn't mysterious. No one should be discouraged from using
> ecologies just because they have trouble keeping single species
> from eventually going extinct. (Although actually, I haven't seen
> that it's quite so hard to avoid extinctions, even just by
> adjusting parameters. Even

There's a fine line between discouraging people from using ecologies
and warning people about the pitfalls of ecologies.  So stop saying
I'm discouring people from ecologies.  Other people might get the
wrong impression ;)

As for the bank robbers example, the reason society survives is the
ratio between bankrobbery/population.  One can argue many cases in
the modern world where that ratio is high enough to actually degrade
society.  :P Noise... equilibria... oscillations...

> if you don't believe me, the real world doesn't have any trouble
> keeping a bunch of species going for thousands upon thousands of
> years, let alone the incredibly tiny time scales expressed in a
> typical MUD's career. If extinction is in your model, it should be
> there because you want it there - it is NOT inevitable unless you
> make a lot of rigid and unnecessary demands on how things have to
> be).

Large difference between extinction being 'inevitable' and it being
'most likely' and my apologies if I wasn't clear.  If I had thought
it inevitable however, I wouldn't bother with all this, nor give you
references to make the 'most likely' into simply 'likely' :P

And besides, I don't think human interaction nor interaction with
other species is that much of an unnecessary demand.

TLC


_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev