[MUD-Dev] Everquest player survey with interesting points

[Sasha Hart]

[Olof]

> My main reaction reading it is that it confirms many popular
> views, and it strikes me that it might be hard to tell to what
> extent the >replies simply mirror popular opinion rather than
> actual facts -

I agree about the free replies, he seems to put a lot of stock in
them.  That is okay, since he also asks questions which turn up
surprising results, I think when he asks them about their own
behavior without soliciting a long response or opinion.

---
My opinion about the survey.
---

I think it is a nice study. The statistics are quite decent. The
measures are not so bad at all. It is very good for this women
vs. men issue, at least within EQ. The sample is self-selected,
which means that it might be over-reflective of people who are
particularly opinionated and also patrol the EQ boards. So, for
example, my confidence in what they say about it being an addiction
is somewhat limited. The main thing that bugs me is the recruitment
using forums and such, but you do what you can.

I wonder how the results would differ between games. Particularly as
they vary in maturity, size, how they are marketed (the lurid
half-naked elf or whatever on the box comes to mind), and what
typical activities are like. This is not a criticism of the study.

Other than that, I had a problem with his skinner box section- there
are some errors here and there :)

> The most basic is a fixed interval schedule, and the rat in the
> Skinner Box is rewarded every 5 minutes regardless of whether it
> presses the lever. Unsurprisingly, this method is not particularly
> effective.

This is incorrect. The FI schedule rewards the first response after
the interval is up. It works just great. Basically what it does is
get the organism to time, vs. the FR schedule which in some sense
gets the organism to count.

Understandably, when you make the timing or count change, it is
harder to pick up when the timing or count has changed.  Variable
schedules can maintain more consistent responding probably at least
partially because of this difficulty detecting the change. The
author interprets this in terms of the feedback bars; I don't know
if he is right, but his comment that full experience feedback would
make it an FR and thereby less motivating is pretty open to debate.

> One important tenet of Operant Conditioning is that behaviors are
> not inherently rewarding - they are made rewarding through
> reinforcement.

Although there is no problem considering some behaviors as
"inherently rewarding"; such an approach has helped with certain
behaviors of autistic kids, etc.

> But it's hard for those of us inside the construct to realize this
> because the game has conditioned us to pursue these rewards.

Although this isn't an error, I like to point out that conditioning
is not magical mind voodoo. If EQ players are given good indication
that a certain behavior won't get them anything anymore, just as
anyone knows, they'll probably stop.

It isn't as if they're bound to do it forever just because they were
rewarded. Reward does not just create perseverance. Rather, as most
of us would expect, an animal will collect information that helps it
get the reward again in the future, but this doesn't mean that it
stupidly does the same thing over and over again no matter what.

Twd. the end there is some more imaginative stuff (hehe).
Basically, because EQ offers many concurrent schedules/ subgames,
people can find the one which appeals to them. He says that this
separates players, because he's using the metaphor of 1 skinner box
= 1 contingency for one organism.  of course, they still can
interact fully even if their schedules are different, in fact and
interact in a complex way (e.g. I choose to mine ore in the
mountains, you choose to hunt deer, we trade - both of our games are
affected even if we mostly do different things.) I seem to remember
some indication that you can treat whole groups as a single operant
"organism" (this has been done with groups of rats) and get similar
schedule effects (e.g. timing and counting.)
_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev