[MUD-Dev] On balance and reality

[William Leader]

"Ola Fosheim Grøstad" <olag at ifi.uio.no> wrote:

> In some of the previous discussions on the list it has been argued
> that MUDs are different from the physical world and that you
> therefore have to stick to a faucet-sink design (spawn-points and
> so on). The assumption seems to be that you cannot design a
> virtual world by modelling the physical world. I don't think this
> is correct and cannot really find any arguments for the
> hypothesis.

....

> There is no reason to believe that one cannot have simulations
> that are fun and are at least as well balance as the systems we
> see in the physical world. As designers we have more freedom than
> politicians. We can come up with models which are more interesting
> than faucet-sink and which balance wise stays within
> "experientially acceptable" boundaries.

>   1. We can obviously draw on techniques from applied mathematics:
>   mechanics, control and signal processing, "computational
>   biology" etc.

>   2. We can relate the mathematical models to the discrete event
>   like nature of MUDs by drawing on simple statistics. You do that
>   in ray-tracing (rays are discrete).

>   3. We can draw on research on multi agent systems (MAS)

About a year ago I got interested in MUD economics after having
completed macro economics at school. I started wondering about
things like inflation and all those other economic problems the MUDs
face. This began a six month personal journey. When I set out I
wanted to figure out how to fix all those problems with my new
arsenal of knowlege from the macro economics course.  About 3 months
into it The result I was finding was that no matter what clever
mechanism I tried to come up with to affect these problems, all I
could ever do was slow down or speed up the economic activity in the
simulations. It seemed things were doomed to be out of balance.

Now, how does this relate to your question? All my simulations where
based on a faucet sink design. Players earned treasure from and
endless supply of mobs, and spent it on a limited merchant
inventory. Even if the merchant changed prices, the net result was
that simulated players bought everything and then just got rich. The
consumed all the content, and accumulated fabulous hoards of
money. It points to a failing system because I couldn't suck as much
treasure out of the players as I put in.

I began to feel as you that the faucet-sink paradigm must be wrong
and that there had to be a better way. I set forth looking into the
ways that the real world fights inflation and such. There a couple
of ways this is done, but the federal reserve is one way and its
easy to model. So I set forth on making the simulation more
realistic by seeing what happens when I add a reserve system and all
the other things it implies like depository institutions. The
simulation was a huge failure. It was too complicted to be easily
modeled in a game. It broke. It broke alot. I kind of gave up the
problem for a few months.

When I came back to the problem, I came back with I different
idea. I wanted to control inflation by controlling the money
supply. I was trying to control the money supply with a reserve
system which in hindsight I think was kind of dumb. I decided that
since what I wanted to do was control the money supply I should just
control the money supply. When prices when up, I just handed out
less treasure, when they went down, the simulation gave out more
treasure. The prices stablized. They wobbled, but were stable. It
worked. It was at this point that I learned something about reality
in MUDs.

The lesson I learned was that real conecpts like the relationship
between inflation and money supply do exist in MUDs. I feel that had
I been more persistant at that reserver bank simulation I might have
worked out all the bugs and it may have worked. But that would
require an awful big effort on my part, and there would be no
certainty that it wouldn't have bugs. The simple solution was that I
didn't need to model that specific detail, just what the result
would have been if the detail were there. It was simpler more
reliable code, and it gave the same results.

My take on the whole thing is that we can have better balanced and
realistic feeling worlds. We just don't need to track every fish to
have varying cod populations, nor do we need build a reserve system
to control inflation. The only reason we can do this is in virtual
worlds we are allowed to have black holes and miracles. We don't
need to explaine where things come from or where they go to. If in
the real world the answer to inflation was to just pay people less
that would blow up faster than you can say revolution, but in the
game I don't need to explaine why monsters suddenly start carrying
less cash. I've been running with that idea so far and it hasn't
failed me yet. I don't need to explain why things happen so long as
when they do happen they appear plausible.

That was longer than I thought it would be.

-Will Leader
k2wrpg.org
_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev