[MUD-DEV] Re: Spouse play

Sasha Hart Sasha.Hart at directory.reed.edu
Fri Feb 15 17:41:58 CET 2002


[Ola]

> The games are designed to attract men (boys), women watch their
> spouse play, their husband is eager to get them on (to not feel
> guilty about playing, to share or whatever) and buy an extra
> account.

I saw at least in the EQ data indications that this might not even
be the mode, just something which happens a substantial part of the
time.  Perhaps more with older women than younger (that's
speculation from the age distribution of spouse play, if I
remember).

Although it's a big number I would be loathe to brutalize the
pattern (even the relatively clear-cut pattern I might expect out of
EQ) by making it into a sweeping generalization about men and women.
Especially if I am changing my design on this basis, I want to know
how is in some detail rather than just be able to say that men and
women play differently (and then go zooming off to weird conclusions
about how I need to add more mind-numbing crafting or something.)

> Indeed I have yet to meet couples where it is obvious that the
> woman spends more of her free time playing than her husband.

I'll take your word that you have seen this. But I do wonder how
younger women who play a lot and aren't paired off fit in. Here we
run up against the problem that age is confounded with cohort:
perhaps as time goes on, "young women" and "older women" will differ
less because the difference here was actually one of age of exposure
to computing, early learning, whatever.

> (When guessing about player gender my basic heuristic is

...

> Not reliable at all, but there seems to be some differences out
> there.)

The data say to me that there are differences, I couldn't agree
more. But also that each group is heterogenous. Because we are
looking at means, we have a terrible idea of how big individual
differences are as compared to these sex differences.

This might be a funny question, but do you keep track of your
predictions, like in a journal? I can imagine this being a great way
of exploring MUD populations, without a lot of setup but with pretty
good validity/reliability - at least for things that I can trust
myself to be honest about ;)

> Current games are not designed for "spouse play". Group level
> limits means that the man will have to level his wife's
> character. No valid alternative advancement paths. No useful
> classes that are easier to play up (but less powerful in terms of
> brute force). Etc..

This is very frustrating in the more general case of having friends
and acquaintances in the game. I notice it a lot in AO, and as a
result see lots of character-switching to be the level of
friends. (On the other hand, this might simply be people who care
less about the game and would rather help their friends out.)

Freeform roleplay games don't seem to suffer it as much. I think the
big problem is that the player's social environment is not well
enough pried apart from his character's status. Cf. the problem in
PD games of, when you die and change your name, losing the identity
you have built up in the game as a player. Everyone I have actually
watched play MUDs has used an instant messenger window concurrently
(for what the anecdote is worth.)

The basic problem in maintaining social continuity might
substantially be driving our playing-with-boyfriend-from-outside-EQ
effect, too.

> I would a priori expect player frequenting websites often enough
> to become aware of a survey and also willing to take the pain of
> filling out surveys to be rather involved in the game.

Yeah, that is terrible. I still have faith in some of the answers,
like I think I can believe that players are mostly not lying about
how much they play a week. There is some noise but it is pretty
unparsimonious to assume that women are all inflating their
estimates or something.

I am pretty sure that the little essays the players filled out are
REALLY influenced by this kind of thing, given the kind of stuff
that goes on in forums.

> Of course, I am also rather sceptical about quantitative studies
> where people assess themselves and put themselves on a scale.

Yeah, it sucks but it's even worse to have totally unclassifiable,
or hopelessly laborious to code, data. I just put less confidence in
the things where people are more likely to manage their
presentation, e.g. to try to be representatives of their gender
(e.g. I am a girl, this survey is asking about girls, I should take
this opportunity to put in that I hate GoP. Or love it. or whatever)
or when the question is asking about an opinion versus something
more or less objective like how much they RP cross-gendered.

Even if different subjects interpret the scales differently, all
you're doing is analyzing them like any number, so the individual
differences will mostly just add noise and prevent you from getting
significance, if anything.

> Why not expect gender differences in how people interpret
> questions and approach surveys? Gee, all the data are ruined even
> before you got started.

I think I disagree just on heuristic grounds. It is a good critique
but often enough I read something, say basically what you just said,
and then read a followup where a nice control nixes it.  Hopefully I
am not repeating myself ad nauseam, but I don't put much stock in
opinion questions to begin with, and mostly just ignore everything
that isn't completely straightforward to answer. After all, it's not
my thesis, it's just a dataset to me :)

> What really bugs me about the site is that there is no real
> analysis, just some averages of individual variables. Does the
> fact that women are less willing to sell their characters tell me
> anything?  I do need to know something about the correlation
> between personal finances variables and this willingness to sell?

Hmm, there is a lot of correlational stuff in there.. I just thought
it was interesting because this is the kind of topic which spawns
endless opinion wars, even when some information is out there,
abundant information in some cases.

> Anyone that plays and socialize on EQ do have information on this.

Yes! Of course, your average EQ player probably has a set of friends
she deals with who are not randomly picked (I would expect at the
least some self-grouping by age.) Their sample is going to be just
as biased as this study's, if not more so.

And "anyone" is not necessarily going to be perfectly honest or
unbiased in how she derives conclusions from the information at
hand.

So the information I could get by asking any EQ player what they
think about finer points (I'm sure they have a good idea of how
often they are attacked, how much they play etc. but other things
can get difficult) is not so useful, if for no other reason than
because the informal way the data was collected & analyzed could
obscure any kind of manipulation or bias at all, prevents
replication, etc. and this short-circuits many arguments I'd make to
others even if I believe it.

> I personally trust that in-depth knowledge in a smaller set of
> cases more than surveys like these that tend to be shallow and
> hide a lot of bias and misconceptions.

Well, this survey did solicit rather a lot of free responses, and
the page goes on and on with them. But even with just one of these
reports, it is very hard to analyze them in a way that isn't
self-serving, and even if you can, it is difficult to convince
people that you weren't serving yourself. I don't really know what
it means if half of the sample says that they think the game is too
violent, especially when they are drawn from forums which probably
deal with the same issues over and over again.

The bias situation is a lot better when you at least have exposed
your methods and data to some extent, because then people can figure
out the biases by looking at your paper.
_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev



More information about the mud-dev-archive mailing list