[MUD-Dev] Fw: [MUD-Dev] Self-organizing worlds (was: Elder Games)

Matthew Mihaly diablo at best.com
Sat Apr 10 23:40:26 CEST 1999

On Sat, 10 Apr 1999, Kylotan wrote:

> I would argue this totally invalidates the experiment. When you tell
> people that they can earn money based on their performance, they will then
> try and guess the way in which you want or expect them to perform... which
> will generally mimic real-life, maybe not for the reason that it's natural
> behaviour, but because it is the behaviour that the participant thinks the
> experimenter expects, and will thus stand to earn the participant more
> money.
> As soon as the participant in an experiment knows what the experimenter
> 'wants' from the experiment, the study loses validity. -Especially- if the
> participant stands to gain from it. This is known as 'demand
> characteristics' and references/data can be found in many experimental
> psychology textbooks.

This isn't true. If you are trying to model the way a group of people will
compete or organize to gain resources, the only way TO validly do it is to
make the reward (the resources they are competing or organizing to gain)
something that the people want, and money is the most generalized
expression of that.

Example: Game theory. Stick a pile of money in the middle of a table, with
a group of human lab rats sitting around the table. Tell them that every
minute, you will double the amount of money that is on the table. Also
tell them that any of them can take the money at any time. See how long it
takes them to learn that by cooperating they can, eventually, all end up
with more than if one person happened to get all the money at the centre
of the table. Valid experiment and the sort of thing that is very
important to think about when setting up an economic system (that's a
simplistic example, but it sort of underlies most of game theory).


MUD-Dev maillist  -  MUD-Dev at kanga.nu

More information about the mud-dev-archive mailing list