Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Testing theories of player types

10 views
Skip to first unread message

Joshua Von Korff

unread,
Jul 14, 2009, 8:10:40 PM7/14/09
to
I've noticed a number of theories that divide players into types: GNS,
Robin's Laws, perhaps others.
In 2002, this person: (http://www.nickyee.com/facets/home.html) seems
to have tested such a theory for MMORPGs. The original theory divided
players into 4 categories: achievers, explorers, socializers, and
killers. Using a mathematical technique called factor analysis, he
modifies and sharpens his understanding of the types. He also splits
one of the types ("socializing") into two ("leadership" and
"relationships"). This results in a fairly objective picture of
player motivations.

Has anyone considered doing this sort of analysis for tabletop RPGs,
for instance with GNS or Robin's Laws? According to my limited
explorations, many people find the GNS categories problematic, and
there is a simple mathematical way to address the question. One
writes up a survey, listing questions related to player motivation;
then one uses factor analysis to test whether gamist question #1 ("I
enjoy selecting the most powerful abilities for my character") is
correlated with gamist question #2 ("I enjoy defeating monsters in
combat"). If not, then perhaps they belong to different factors.

Again I don't really know much about GNS, but I find the idea of
mathematically testing such a theory to be very interesting. All one
would need is hundreds of players to fill out the survey. Not sure
how that would be accomplished. In the MMORPG, I imagine he somehow
got it to them electronically.

Let me know if this is the wrong forum for this, and suggest another
if possible.

David Lamb

unread,
Jul 14, 2009, 11:46:21 PM7/14/09
to
Joshua Von Korff wrote:
> Again I don't really know much about GNS, but I find the idea of
> mathematically testing such a theory to be very interesting. All one
> would need is hundreds of players to fill out the survey. Not sure
> how that would be accomplished.

I suppose one might try to get people to fill out surveys at a big
gaming convention.

> Let me know if this is the wrong forum for this, and suggest another
> if possible.

It fits. This has been more of a "theory about gaming" newsgroup than
actual "advocacy."

gleichman

unread,
Jul 15, 2009, 7:14:17 AM7/15/09
to
On Jul 14, 10:46 pm, David Lamb <dal...@cs.queensu.ca> wrote:
> I suppose one might try to get people to fill out surveys at a big
> gaming convention.

That might give you an idea about people who attend gaming
conventions, but little about role-players in general. IME, across
decades and over a hundred of player I've gamed with- not one ever
went to a rpg convention.

Additionlly Prof. Bartles (the likely source of the MMORPG study
although he wasn't named) had a serious advantage over the table top
hobby. He ran the world's first MUD, and thus could select from his
entire player base. An option difficult to say the least for table
top.

He also as I recall didn't go into it with the theory already
complete- but rather collected data and then came up with his player
types (I'm not certain on that, I could check if there's interest).
This was also the claimed method in the WotC study that split gamers
into 4 groups.

Frankly you likely should give the WotC study more attention than GNS,
an overview is found here: http://www.seankreynolds.com/rpgfiles/gaming/BreakdownOfRPGPlayers.html

Joshua Von Korff

unread,
Jul 30, 2009, 1:08:02 PM7/30/09
to
> Frankly you likely should give the WotC study more attention than
GNS,
> an overview is found here:http://www.seankreynolds.com/rpgfiles/gaming/BreakdownOfRPGPlayers.html

Thanks! This is really interesting, and I agree it seems better than
GNS. It doesn't address S = "simulationist" role playing, but I have
never been part of a group that was really into simulating a plausible
world using the game, so I don't mind that it's not represented.
(Assuming that's what "simulationism" means, which I'm not sure
about.) In the WotC study, I think I have a strong dose of "thinker",
but wish I could be a better "method actor" or "storyteller."

The one thing I have a hard time with is the method actor /
storyteller distinction. What's the difference?

I came up with one example: in one LARP, my character had the
opportunity to poison another character, killing that person off. As
a thinker, I should decide whether that character's death is in my
character's best interests. As a method actor, I should decide
whether my character hates the other guy enough to kill him off,
regardless of whether it's in my best interest. As a storyteller, I
should kill him off if I think it would be interesting to have a big
investigation into the murder ... which it almost certainly would
be.

Problem is, I was imagining the other players asking "why did you kill
off so-and-so" at the end of the game, and I can't imagine myself
saying "because it would make for a better plot." I guess that's just
what I'm used to -- characters need in-character reasons for
everything they do. Also, this seems to be an exceptional case ... it
was clear that the GM had planned for the eventuality that my
character would poison the other guy, so I knew that it would lead to
an interesting story. In a tabletop RPG, you don't really know what
will lead to an interesting story. Usually the GM has planned for a
particular chain of events, which assumes that the PCs are led by a
"thinker". The most interesting story will result from following that
chain. If you deviate, you will just mess the GM up and lead to
uncharted territory where the GM is less prepared. Basically I am
saying that the GM, not the PCs, is the storyteller. What do you
think?

Rick Pikul

unread,
Jul 31, 2009, 3:57:11 AM7/31/09
to
On Thu, 30 Jul 2009 10:08:02 -0700, Joshua Von Korff wrote:

> Thanks! This is really interesting, and I agree it seems better than
> GNS. It doesn't address S = "simulationist" role playing, but I have
> never been part of a group that was really into simulating a plausible
> world using the game, so I don't mind that it's not represented.
> (Assuming that's what "simulationism" means, which I'm not sure
> about.)

Not quite, it's more like letting the world go the way the world would
without consideration for the story you get out of it or maintaining the
right level of player challenge. Also, the world only needs to be
internally consistent, not plausible.

--
Chakat Firepaw - Inventor & Scientist (Mad)

gleichman

unread,
Jul 31, 2009, 12:34:58 PM7/31/09
to
On Jul 30, 12:08 pm, Joshua Von Korff <joshlegan...@hotmail.com>
wrote:

> Thanks!  This is really interesting, and I agree it seems better than
> GNS.  It doesn't address S = "simulationist" role playing, but I have
> never been part of a group that was really into simulating a plausible
> world using the game, so I don't mind that it's not represented.

Given that the WotC project didn't come up with their axis and naming
conventions until after the data was collected and analyzed, it may be
that Simulation as talked about in GNS and on this newsgroup isn’t
that important of concern for most gamers- at least compared to the
Combat/Story and Strategic\Tactical split.

Or (more likely in my opinion) it could be hidden in the terms used.
Storyteller seems to contain much of the same concepts (to quote the
article referenced, the “logical progression of the narrative of the
scenario”).

The bloodshed and debates here and elsewhere has poisoned terms such
as Story and Narrative, and it may difficult to understand how WotC is
using the terms (although I believe them to be using them more
correctly).

For example, the Threefold assumed by its own definition that Story
and Narrative were not ‘the telling of the outcome of whatever
happened’, but meta-game creations of the players forced upon the game
reality. If one takes the former and more traditional viewpoint,
classic simulation (i.e. what would happen if the gaming world was
independent and consistent) is contained within WotC’s Storyteller
group, if only as a subset.

One has to remember that GNS and the Threefold were both creations of
ego, in that the creators of each had a goal in mind before making the
model. In the case of the Threefold, it was to defend their version of
Simulation. In the case of GNS, it was to justify the types of games
interesting to Ron Edwards. Bad foundations make for sloppy theory and
the twisting of words.

> The one thing I have a hard time with is the method actor /
> storyteller distinction.  What's the difference?

Try viewing it as 'where the player's enjoyment comes from'. The
method actor at the extreme would be completely focused on his own
character, and would only be interested in those things he could
immediately react to. Thus consistency over time (of those things
outside his character) isn't as important, nor are the world events as
long as he can 'act' in response to them.

The Storyteller on the other hand is more interested in existing in a
cool world (however he defines cool). And he looks at all that stuff
outside the character as more interesting. He may or may not be making
decisions to alter the Story, as the model doesn't tell us if making
Story or exploring Story is his actual goal. From what I've read
there- it could be either.

So, to use your example of poisoning another character-

One Storyteller may be interested in how the world is affected either
way. He has no stake, and only wants consistency of the game reality.
He views the 'Real Story' as what's important.

Another may be interested in creating the best Story (in his opinion)
and may make his decision based upon that.

The two types likely wouldn't get along over the long time.


> Basically I am
> saying that the GM, not the PCs, is the storyteller.  What do you
> think?

I think that would depend upon the tastes of the players.

Andrzej Jarzabek

unread,
Aug 5, 2009, 1:39:16 PM8/5/09
to
Also sprach gleichman:
> On Jul 30, 12:08�pm, Joshua Von Korff <joshlegan...@hotmail.com>
> wrote:
>> Thanks! �This is really interesting, and I agree it seems better than
>> GNS. �It doesn't address S = "simulationist" role playing, but I have

>> never been part of a group that was really into simulating a plausible
>> world using the game, so I don't mind that it's not represented.
>
> Given that the WotC project didn't come up with their axis and naming
> conventions until after the data was collected and analyzed, it may be

Actually without knowing their dataset and methodology it's difficult to
draw viable conclusions of this type just from the fact that they
gathered some data and drew conclusions from it.

I would imagine the data was gathered by having players fill out
questionnaires. In that case very obviously the kind of results that can
be obtained is very much shaped by what questions they choose to ask.

> that Simulation as talked about in GNS and on this newsgroup isn?t


> that important of concern for most gamers- at least compared to the
> Combat/Story and Strategic\Tactical split.

Also, GNS and GDS aren't really classifications of player types. They're
classifications of game goals or "creative agendas". You wouldn't
neccesarily expect that even an objective classification of player types
would give you groups of players by preference of one threefold goal
over another, i.e. "simulationists", "dramatists" and "gamists" (however
defined).

> One has to remember that GNS and the Threefold were both creations of
> ego, in that the creators of each had a goal in mind before making the
> model. In the case of the Threefold, it was to defend their version of
> Simulation. In the case of GNS, it was to justify the types of games
> interesting to Ron Edwards. Bad foundations make for sloppy theory and
> the twisting of words.

Many theories introduce their own terminologies, not just sloppy ones.

And whatever their foundations were, the reason these theories are the
popular ones (as far as RPG theories go) is that they reflect the
experience of many players, whether or not they like games that are
interesting to Ron Edwards.

Now, I'm not saying they aren't sloppy, but take them for what they are.
They're not classifications of player types. They're also not a
classification of all possible player preferences: players may prefer to
act out the voices and gestures of their characters, or they might not,
they may prefer to dress up or not, to have tangential conversations or
not, they might prefer lots of violence or little violence, adult themes
or juvenile escapism, fantasy or science fiction, cheese and onion or
salt and vinegar, Pepsi or Coke. They are all player preferences, but
they're outside of scope of threefold or GNS.


--
"I'm an idea man. Hard work is not my forte."

gleichman

unread,
Aug 5, 2009, 2:02:46 PM8/5/09
to
On Aug 5, 12:39 pm, Andrzej Jarzabek <andr...@hoodoo.hell.pl> wrote:
> Many theories introduce their own terminologies, not just sloppy ones.

Good theory does it after data is collected, and does not make up new
meanings for old words while doing so.

> And whatever their foundations were, the reason these theories are the
> popular ones (as far as RPG theories go)

I'm not very interested in populart.

For one thing, there is no data that indicates any of these are
popular. Indeed, I would say they are anything but as the word
'popular' is typically used.

But further, popular only indicates that it is something desired and
speaks in no way to quality or correctness. People for many reasons
desire that which is bad for them. And I'd say that GDS and GNS both
match that approach nicely, they feed ego and self-justification, cut
off inconvenient communication and generally allow one to hide from
reality.

I can understand why that would be popular, but I don't consider it a
virtue.

Andrzej Jarzabek

unread,
Aug 6, 2009, 7:25:08 AM8/6/09
to
On Aug 5, 7:02 pm, gleichman <fox1_...@hotmail.com> wrote:
> On Aug 5, 12:39 pm, Andrzej Jarzabek <andr...@hoodoo.hell.pl> wrote:
>
> > Many theories introduce their own terminologies, not just sloppy ones.
>
> Good theory does it after data is collected, and does not make up new
> meanings for old words while doing so.

It's not something I want to argue about, but it certainly does not
seem true of many well-established scientic theories.

Also, note that the authors of these theories have had years of
personal experience playing role-playing games themselves, as well as
discussing play with other players. I'm nor expert at this, but from
what I know participation and interview are both valid methods of
gathering data in social studies.

> > And whatever their foundations were, the reason these theories are the
> > popular ones (as far as RPG theories go)
>

> For one thing, there is no data that indicates any of these are

Data in what sense? Is there data to indicate anything you wrote about
these theories in your postings?

I base this on the fact that these are the teories I repeatedly see
mentioned, used or discussed on various forums and websites dedicated
to role-playing games.

> popular. Indeed, I would say they are anything but as the word
> 'popular' is typically used.

What I said is they're popular compared to other theories of role-
playing games. I wouuld say no theories are popular, but some are just
a little bit more popular than others.

> But further, popular only indicates that it is something desired and
> speaks in no way to quality or correctness.

One possible reason something might be desired is its quality or
correctness.

> People for many reasons
> desire that which is bad for them. And I'd say that GDS and GNS both
> match that approach nicely, they feed ego and self-justification, cut
> off inconvenient communication and generally allow one to hide from
> reality.

Can you explain how you came to that conclusion? Because my
impressions are exactly the opposite.
How does adopting GDS or GNS feed someone's ego or slef-justification?

It may be that the wording of these theories lacks precision, that the
problem is ill-defined etc., but the basic observation does adress a
real issue, that if left unadressed often leads to conflict, no fun
and sometimes no play. The other way of adressing this issue, that
I've seen is to come up with a "correct" way to play, which not only
feeds ego, cuts off inconvenient communication and allows to hide from
reality, but also has a more practical implication when playing with
people outside of one's usual group.

gleichman

unread,
Aug 6, 2009, 7:50:38 AM8/6/09
to
On Aug 6, 6:25 am, Andrzej Jarzabek <andrzej.jarza...@gmail.com>
wrote:

> It's not something I want to argue about, but it certainly does not
> seem true of many well-established scientic theories.

By the strict definition of science, an idea of something is presented
first and then data is collected in an effort to disprove that idea. A
method certainly not used in any RPG theory, and sadly not always used
in science itself.

> Also, note that the authors of these theories have had years of
> personal experience playing role-playing games themselves, as well as
> discussing play with other players.

Anecdotal Evidence isn't acceptable and is never a valid method. And
further, I see as much if not more of that type of evidence online
from people who also have years of "years of personal experience


playing role-playing games themselves, as well as discussing play with

other players." who think those theories are complete junk.

As does actual studies such as those carried out by WotC.


> Data in what sense?

What percentage of the greater rpg population believe and like these
theories? What percentage of even those who post online (a small
percentage itself of the greater total) like and believe these
theories?

You and I both exist online, we both see conversations like this. You
think it makes the theories popular, I say they are nothing more than
a very vocal subgroup who speak louder than their importance.

Since you're the one claiming validity for an idea, the burden of
proof is on you. Not me.

Exactly the same as if you wished to prove magical unicorns exist.

> One possible reason something might be desired is its quality or
> correctness.

You seem to having a very idealized view of mankind. But I suppose a
serial killer may will desire 'quality' in the deaths of others... no
wait, that would be a foolish assumption- as would yours.

> Can you explain how you came to that conclusion?

I was there when these theories where developed, I spoke extensively
to the primary movers behind them. Over months their motives became
clear, as did the effect of their theories. That why I believe what I
do about the reasons for their development.

As for those believers who came later...

...take yourself for example, you don't seem to understand how science
works, or marketing, or data collection. But yet you've launched
yourself into a theory debate when it's clear you don't have the
knowledge to do so.

Why?

Because it fills some need in you, not because you understand it. What
need? I don't know yet, but I bet if this goes on- I'll find out. It
could be an honest one, i.e you're an example of the small slice of
gamers who the threefold theories actually fit. Or it could be other
things.


> Because my
> impressions are exactly the opposite.
> How does adopting GDS or GNS feed someone's ego or slef-justification?

You think these theories don't push a correct way, you're wrong. They
do.

And as such, it lets people claim that they have higher insight and/or
more knowledge. It justifies their own play style at the expense of
others.

Andrzej Jarzabek

unread,
Aug 6, 2009, 10:33:24 AM8/6/09
to
On Aug 6, 12:50 pm, gleichman <fox1_...@hotmail.com> wrote:
>
> > It's not something I want to argue about, but it certainly does not
> > seem true of many well-established scientic theories.
>
> By the strict definition of science, an idea of something is presented
> first and then data is collected in an effort to disprove that idea. A
> method certainly not used in any RPG theory, and sadly not always used
> in science itself.

As I said, I don't want to discuss science or scientific theories. I
agree neither GNS or GDS are science.

> > Also, note that the authors of these theories have had years of
> > personal experience playing role-playing games themselves, as well as
> > discussing play with other players.
>
> Anecdotal Evidence isn't acceptable and is never a valid method. And

I could ask how data gathered by participant observation and
interviewing is different anecdotal evidence.

I may be wrong, so correct me, but from what little I know, methods
that are valid for anthropology or sociology, would not neccesarily be
acceptable for, say, medicine, and data gathered using these methods
would be considered anecdotal evidence.

But if your point is that this kind of data gathering does not meet
the rigours of scientific research, you are right. It doesn't make it
invalid for general discussion though.

> further, I see as much if not more of that type of evidence online
> from people who also have years of "years of personal experience
> playing role-playing games themselves, as well as discussing play with
> other players." who think those theories are complete junk.

And this evidence you see invalidates GDS/GNS? How exactly?

> As does actual studies such as those carried out by WotC.

What does what?

If you mean to say that the WotC study of player types falsifies GNS
or GDS, then it doesn't.

And as for it being an "actual study", we have no more reason to think
it qualifies as scientific any more than GDS or Ron's theories.

> > Data in what sense?
>
> What percentage of the greater rpg population believe and like these
> theories? What percentage of even those who post online (a small
> percentage itself of the greater total) like and believe these
> theories?

Yes, there is not data in that sense. But quantitative data isn't the
only kind of data there is.

Also, my claim of popularity wasn't about either how many people
"believe" the theory, or its absolute popularity.

> You and I both exist online, we both see conversations like this. You
> think it makes the theories popular, I say they are nothing more than
> a very vocal subgroup who speak louder than their importance.

So if there are theories more popular than these, why aren't they
voiced online? And how do people communicate these theories? Surely, a
theory passed by word of mouth must be known to, and consequently
liked by a much smaller number of people than theories that are often
discussed online?

> Since you're the one claiming validity for an idea, the burden of
> proof is on you. Not me.

Are you then only willing to discuss things that are scientifically
proven? In that case I would remind you, that no proof has been
presented of the validity of the WotC study either.

I can certainly tell you why I think GNS and GDS are popular among RPG
theories:
* Other theories are hardly ever voiced online.
* While it may be true, that there is a correlation between being
receptive to GNS or GDS, and being willing to talk about or explain or
defend these theories, the very fact that a theory is not talked about
means others get less exposure to it.
* As far as I know there are no other mass communication channels over
which RPG theory is regularly communicated.

I'd like to know why you think this is not true? Is this because you
know of some theories that you think are more popular than GDS or GNS,
or that you think there may be theories that are more popular, even
though you or I have never heard about them?

> Exactly the same as if you wished to prove magical unicorns exist.

I don't wish to prove anything.

> > One possible reason something might be desired is its quality or
> > correctness.
>
> You seem to having a very idealized view of mankind.

I would say you have an overly cynical view, if you discount these as
even possible reasons. But in that case, I have no intention of
proving you wrong.

> But I suppose a serial killer may will desire 'quality' in the deaths of others... no
> wait, that would be a foolish assumption- as would yours.

Was that supposed to be a 'proof' that my assumption is foolish?

> > Can you explain how you came to that conclusion?
>
> I was there when these theories where developed, I spoke extensively
> to the primary movers behind them. Over months their motives became
> clear, as did the effect of their theories. That why I believe what I
> do about the reasons for their development.

I would say this is probably right, but this is irrelevant.

> As for those believers who came later...
>
> ...take yourself for example, you don't seem to understand how science
> works, or marketing, or data collection.

I know a little, with the possible exception of marketing, of which I
know very little.

Enlighten me, please, if you think anything I said is factually wrong.
Do we, for

> But yet you've launched
> yourself into a theory debate when it's clear you don't have the
> knowledge to do so.
>
> Why?

Because honestly I do not believe knowledge of science methodology,
marketing or data collection is relevant to RPG theory (and we are not
talking about scientific theories).

> Because it fills some need in you, not because you understand it.

Obviously, the same could be said for you.

> What need? I don't know yet, but I bet if this goes on- I'll find out. It

I could tell you that right away. I was curious about why you dislike
these theories so much.

I also have a deeper motive. I feel that having been exposed to these
theories and thinking them over has greatly enchange my role-playing
experience. I thought you might present an argument against them that
might do the same for me, or perhaps even point me to a theory (or
something else) I could apply with better results in its place?

> could be an honest one, i.e you're an example of the small slice of
> gamers who the threefold theories actually fit.

Can you describe, what in your opinion are the characteristics of a
gamer that threefold theories fit? It deos seem strange to me, that
not only I am an example of this small slice, but almost everybody I
ever gamed with?

> Or it could be other things.

Let me guess - to inflate my ego etc.?

> > Because my
> > impressions are exactly the opposite.
> > How does adopting GDS or GNS feed someone's ego or slef-justification?
>
> You think these theories don't push a correct way, you're wrong. They
> do.

And what way would that be?

> And as such, it lets people claim that they have higher insight and/or
> more knowledge.

Any theory lets people do that, if they are so inclined.

> It justifies their own play style at the expense of others.

How so?

gleichman

unread,
Aug 10, 2009, 2:13:59 PM8/10/09
to
On Aug 6, 9:33 am, Andrzej Jarzabek <andrzej.jarza...@gmail.com>
wrote:

> But if your point is that this kind of data gathering does not meet


> the rigours of scientific research, you are right. It doesn't make it
> invalid for general discussion though.

I'm not interested in pointless discussion, if you don't have anything
but your personal experience to dwell upon- you're not worth talking
about.


> If you mean to say that the WotC study of player types falsifies GNS
> or GDS, then it doesn't.

I feel that it does indeed do so, using more data and information than
anything that supports it. While it was marketing research, I consider
that perfectly acceptable in a soft area like rpgs.


> So if there are theories more popular than these, why aren't they
> voiced online?

Good theory doesn't draw the flamewars, so it's boring. Commented upon
at the time and then dropped. This was the fate of the WotC Study.

Bad theory, driven by ego is the very stuff of flamewars. And will be
defended at all costs. Hence it will always be talked about online.


Really, it's a basic dynamic for the Internet.


> Are you then only willing to discuss things that are scientifically
> proven?

I'm willing to talk about valid marketing studies, or studies backed
by accepted Academic research. I'm not interested in your personal
opinion.


> Because honestly I do not believe knowledge of science methodology,
> marketing or data collection is relevant to RPG theory

Then I have nothing further to say to you. I hope you enjoy your
fantasy world you've created for yourself.

Andrzej Jarzabek

unread,
Aug 11, 2009, 7:33:21 PM8/11/09
to
On Aug 10, 7:13 pm, gleichman <fox1_...@hotmail.com> wrote:
>
> > If you mean to say that the WotC study of player types falsifies GNS
> > or GDS, then it doesn't.
>
> I feel that it does indeed do so, using more data and information than
> anything that supports it. While it was marketing research, I consider
> that perfectly acceptable in a soft area like rpgs.

Honestly, if I can't see the data and can't see the method, the
conclusion tells me very little.

Here's for example an alternative interpretation:

The results are strangely similar to the so-called Fourfold Way, first
proposed by a chap called Glen Blacow in the article "Aspects of
adventure gaming", in Chaosium's magazine "In Different Worlds", issue
10, in the year 1980. This article had no pretense of being scholarly
or quantitative market research. Maybe it's because Blacow was so
insightful (and by the way this original theory did generate a lot of
what in these times counted for as flamefest).

But another possibility is that the author of the study knew about the
Fourfold Way and approved of it, and either consciously or
subconsciously chose their method to reflect this pre-conception about
player types. For example by writing a questionnaire that had many
question about the aspects that the fourfold way considers important,
and specifically questions delineating along fourfold way categories.
Which we don't actually know, as we haven't seen the questionnaires,
or even know if questionnaires were used to gather this data.

But even if I do believe these results, they still do not falsify GDS
in any way. While some of the topics of this study cover GDS scope,
others do not. A power-gamer might certainly have preference for
"gamist" priorities, or for "simulationist". Someone, who likes
"combat" in games might like a tactical game, or think that violence
and the possibility of death create better suspense, or both.
Certainly GDS does not make any claims about quantitative distribution
of preferences among players.

> > So if there are theories more popular than these, why aren't they
> > voiced online?
>
> Good theory doesn't draw the flamewars, so it's boring. Commented upon
> at the time and then dropped. This was the fate of the WotC Study.

If the data is not available and the methods and reasoning used to
reach this conclusion is not available, what is there to discuss?

And also, as I mentioned, the taxonomy in that model has been
flamewared to death in magazine articles in the 1980s.

And I'll throw in a little flame myself: the existence of the "fifth
group" of "general roleplayers" might indicate that this was a test
designed with the preconception of the four groups of the fourfold
way, but in fact failed to prove the existence of these groups.


> > Are you then only willing to discuss things that are scientifically
> > proven?
>
> I'm willing to talk about valid marketing studies, or studies backed
> by accepted Academic research. I'm not interested in your personal
> opinion.

I was only telling this because I was curious on what you base your
claims about GNS and GDS. I suspect these claims are not backed by
accpeted academic research or valid marketing studies.

In fact, I think my curiosity may have been satisfied, as I came
across this on the Web:
http://home.roadrunner.com/~b.gleichman/Theory/Threefold/rpg_theory_bad_rep.htm

It looks like it might be yours, am I right to assume that this is
your opinion at this time as to why these theories are bad? I assume
you have no interest in discussing the contents of the webpage, or
your opinions of GDS, GNS, since your opinions neither academic
research not marketing studies.

So I will just state that I reject the argument on that webpage
because of the serious fallacies it's based on.

As for academic research, I'll drop this link: http://www.journalofroleplaying.org/

Several of the papers touch upon subjects of GNS/GDS and the WotC
study, in particular:
http://marinkacopier.nl/ijrp/wp-content/uploads/2009/01/montola_the_invisible_rules_of_role_playing.pdf
http://marinkacopier.nl/ijrp/wp-content/uploads/2009/01/harviainen_hermeneutical_approach_to_rp_analysis.pdf

> > Because honestly I do not believe knowledge of science methodology,
> > marketing or data collection is relevant to RPG theory
>
> Then I have nothing further to say to you. I hope you enjoy your
> fantasy world you've created for yourself.

I hope the same for you too. Sadly, in my fantasy world tabletop
roleplaying is not a subject of much scholary interest.

And as for studies on large number of subjects, that do not disclose
methods or data, in my fantasy world studies like that have been used
to prove that sugar pills cure cancer.

Peter Knutsen

unread,
Aug 12, 2009, 12:22:43 PM8/12/09
to
Andrzej Jarzabek wrote:
[...]

> But another possibility is that the author of the study knew about the
> Fourfold Way and approved of it, and either consciously or
> subconsciously chose their method to reflect this pre-conception about
> player types. For example by writing a questionnaire that had many
> question about the aspects that the fourfold way considers important,
> and specifically questions delineating along fourfold way categories.
> Which we don't actually know, as we haven't seen the questionnaires,
> or even know if questionnaires were used to gather this data.

Yes.

I also find it intensely suspicious that the division of results is so
perfectly neat. IIRC 24% in every single quadrant, and 8% in the middle.
Coincidence or cheating?

> But even if I do believe these results, they still do not falsify GDS
> in any way. While some of the topics of this study cover GDS scope,
> others do not. A power-gamer might certainly have preference for
> "gamist" priorities, or for "simulationist". Someone, who likes
> "combat" in games might like a tactical game, or think that violence
> and the possibility of death create better suspense, or both.

Note that there can also very well be a *radically* different reason,
for why many players gravitate prefer combat over other types of conflicts.

> Certainly GDS does not make any claims about quantitative distribution
> of preferences among players.

--
Peter Knutsen
sagatafl.org

gleichman

unread,
Aug 12, 2009, 1:21:45 PM8/12/09
to
On Aug 12, 11:22 am, Peter Knutsen <pe...@sagatafl.invalid> wrote:

> I also find it intensely suspicious that the division of results is so
> perfectly neat. IIRC 24% in every single quadrant, and 8% in the middle.
> Coincidence or cheating?

Those are rounded values by the way.

Or it may simply be the natural backdown of how people approach games.
One may also claim that it's suspicious the US is so close to 50%
'blue' and 50% 'red', sometimes that is just the way things are.

In effect to reject the study without cause (by assuming ill intent on
the part of the marketing group who made the study- not through any
data of your own) is not unlike someone saying that the Bell Curve is
suspicious in how perfectly formed it is. It's a statement of
ignorance.

Now there is a problem here in that we can't look at the data, and
that no other study has been made to confirm it, or disprove it. It
did however represent WotC view of the hobby, and informed it's
actions during the start and early part of the 3.x marketing which was
in rpg terms highly successful. And that beats anything from the
Threefold or GNS by miles.


Warren J. Dew

unread,
Aug 14, 2009, 1:41:50 AM8/14/09
to
Nice to see this group still has interesting discussions when someone
posts.

Joshua, before going into table top games, I'd like to point out a
couple of things about the MMORPG research.

Richard Bartle was the one who came up with the achiever/explorer/
socializer/killer groups. This was based on running the world's first
online roleplaying game, and analyzing his player base, as Brian
pointed out. This was only on the order of a hundred players at a
time, but it was all the players in the world at the time.

Nick Yee's analysis, at the website you link to, does try to do a
similar analysis. He uses a different method, however; he relies on
self reported answers to surveys he's prepared. While the mathematics
of his analysis is more rigorous, his data collections methods are
weaker; he didn't actually follow his subjects directly over months,
and he obviously can't collect any data he doesn't think to ask the
questions for. In science, self reported data are always suspect.

Bartle has looked at some of Nick's results, and has commented on
them; the bottom line is that Bartle doesn't think Nick has managed to
capture the important information. I agree. I've suggested a
question for Nick to use to help remedy this, but when Nick ended up
using a similar question, it was subtly different in a way that caused
it to completely miss the point I was hoping it would investigate.
While I think Nick's efforts are clearly valuable, especially in the
area of more easily quantified data - how large "guilds" tend to be,
age and gender variations in the play of particular character classes
- I really think Nick is one of those people who only understands his
own style of play and just doesn't "get" the full variety of different
player characteristics that are out there.

There is one other difference. Bartle came up with categories of
players, with the implication that they were largely mutually
exclusive. Nick came up with motivations of players, with the
assumption that they were not mutually exclusive. They might be
looking at things that are much more different than even they
realized.

There's one other thing to note, too. Bartle believes that the
distribution of players in his categories is extremely asymmetrical;
he thinks that at most about 2% of the players are "explorers", for
example. I have reason to believe, based on investigations of player
killing, that killers constitute a comparably small percentage of
players. Bartle's model includes these types because they have a
disproportionately large effect on the culture of an online game, but
such small numbers could easily be missed in surveys such as Yee's.

Sorry about the digression, but I thought a bit of background on the
rigor or lack of rigor in the "science" done on online games would be
good to keep in mind when discussing similar models that have been
developed for table top games.

Warren J. Dew

unread,
Aug 14, 2009, 2:00:00 AM8/14/09
to
Andrzej Jarzabek posts, in part, regarding the gamer types derived by
Wizards of the Coast:

The results are strangely similar to the so-called Fourfold Way,
first
proposed by a chap called Glen Blacow in the article "Aspects of
adventure gaming", in Chaosium's magazine "In Different Worlds",
issue
10, in the year 1980.

Well, it's superficially similar in the way it's presented, at least,
so yes, WotC might have been doing some borrowing there. However,
Blacow's model did not have axes and from a mathematical standpoint
was better represented in three dimensions rather than two; in
addition, Blacow meant something rather different by "power gaming".
It actually looks kind of like the WotC model borrowed the axes idea
from Bartle's model.

Peter Knutsen

unread,
Aug 14, 2009, 1:45:08 PM8/14/09
to
Warren J. Dew wrote:
> Well, it's superficially similar in the way it's presented, at least,
> so yes, WotC might have been doing some borrowing there. However,
> Blacow's model did not have axes and from a mathematical standpoint
> was better represented in three dimensions rather than two; in
> addition, Blacow meant something rather different by "power gaming".
> It actually looks kind of like the WotC model borrowed the axes idea
> from Bartle's model.

It would be interesting to see a variety of definitions of "power
gaming". I think it is an overlooked preference, and quite possibly a
mis-defined one.

--
Peter Knutsen
sagatafl.org

Andrzej Jarzabek

unread,
Aug 14, 2009, 8:33:31 PM8/14/09
to
On Aug 14, 7:00 am, "Warren J. Dew" <psychoh...@aol.com> wrote:
>
> Well, it's superficially similar in the way it's presented, at least,
> so yes, WotC might have been doing some borrowing there.  However,
> Blacow's model did not have axes and from a mathematical standpoint
> was better represented in three dimensions rather than two; in
> addition, Blacow meant something rather different by "power gaming".
> It actually looks kind of like the WotC model borrowed the axes idea
> from Bartle's model.

In any case, the striking similarity to both these models is an
indication that they were known to the people conducting the study,
and they were consciously or unconsciously used as a pre-conception to
whet the results should be. This might affect how data collection was
done, for example if there were questionnaires, then what questions
were asked and they were worded, and also very possibly analysis of
the data.

The reason I'm thinking this might be so is that there are multitude
of methods for gathering data. Questionnaires are a popular way of
conducting marketing studies, but from a cognitive point of view they
have a lot of problems. First, as you noticed, self-reported data is
always suspect. People are sometimes untruthful and innacurate, and
what's worse, there's a statisticly significant bias as to what they
lie about, forget or misrepresent. Second, choosing a representative
sample is important, yet difficult. Getting even to _ask_ the
representative sample of the gaming population would be a difficult
and costly affair, and then people who actually _agree_ to answer
aren't necessarily representative of the group that are asked.

But of course the biggest problem are the questions themselves. There
is a virtually infinite number of questions that are asked, but the
number of questions you do actually get to ask is very limited.
Especially that the more question you ask, the less representative
your data becomes. The longer the questionnaire, the less people agree
to fill it out and the more people who do agree become bored and do
not complete theirs. But the shorter the questionnaire, the less
qualitative data you get.

So given there will only be a very limited number of questions, how do
you choose your questions in particular. In practice this kind of
study, even if done with scientific rigour (which this one most likely
was not), starts with an qualitative assumption and is aimed at
measuring some quantitative parameters within that assumption.

Lastly, analyzing the data is something that can be done in again a
huge number of ways. Questionnaire results are in fact a set of
possibly thousands of elements in a space of possibly several dozen
dimensions. There is an infinite number of properties that can be
aggregated from such data and you have to know what you're looking
for. "What types of players are there" isn't likely to be good enough,
you need to be more specific. This is even assuming that the analysis
is intellectually honest, and there are no considerations of the type
that clusters of dots in 3D or (worse) 4D space would be difficult to
present convincingly in PowerPoint when you resent the results to
managers and marketers.

For example, I wonder if the two axes do actually mean, that the
number of responders who are somewhere between Power Gamer and
Character Actor is much greater (since they may be represented on the
graph) than the number of Thinker/Actors (which have no place on the
graph). Or would it be just, that someone who has some preferences of
Thinker and some of Actor, but very few of Power Gamer or
Storryteller, would come up as close to 0,0 on Combat/Story and
Tactical/Strategic scales and be lumped together with "mixture of all
four".

For a marketing study, this is perfectly acceptable. It is more
important for a company getting products to market, what part of their
customer base would approve or disapprove of features their products
might have, than whether their taxonomy or the way they represent it
on a graph.

I will not even be cynical at this point and consider the fact that
releasing these results was a PR ploy and the type and form of
presented conclusions was such that was deemed to best suit WotC's
corporate interests.

Andrzej Jarzabek

unread,
Aug 14, 2009, 9:23:30 PM8/14/09
to
On Aug 12, 6:21 pm, gleichman <fox1_...@hotmail.com> wrote:
> On Aug 12, 11:22 am, Peter Knutsen <pe...@sagatafl.invalid> wrote:
>
> In effect to reject the study without cause (by assuming ill intent on
> the part of the marketing group who made the study- not through any
> data of your own)

I for one was not assuming ill intent (merely taking the possibility
into consideration). I also do not reject this study, merely doubt it.

But I actually do have data of my own to reject it:
Datum 1: The data for this study has not been made available.
Datum2: Methods used in the study were not published.

> is not unlike someone saying that the Bell Curve is
> suspicious in how perfectly formed it is.

In fact there are situations in statistical analysis where data is
suspect for fitting a theoretical distribution curve too well and a
possible explanation of these is that data has been massaged.

> It's a statement of ignorance.

We are all ignorant about this study really, and that's the problem.

> Now there is a problem here in that we can't look at the data, and
> that no other study has been made to confirm it, or disprove it. It
> did however represent WotC view of the hobby, and informed it's
> actions during the start and early part of the 3.x marketing which was
> in rpg terms highly successful.

Nobody was contesting the usefulness of the study as a marketing tool.
And if it did contribute to the success of the marketing campaign, it
still telss us very little about the validity of the taxonomy
presented. Even if it were completely false, the survey would still
contain very valuable data about quantitative player preferences
concerning specific decisions that WotC was faced with when developing
3.x and its product line.

> And that beats anything from the
> Threefold or GNS by miles.

That's an opinion.

Someone else might say, that GNS and GDS have had more impact on
scholarly research of role-playing games, in particular qualitative
studies, which beats conclusions from an unpublished survey by miles.

Peter Knutsen

unread,
Aug 15, 2009, 8:17:08 AM8/15/09
to
Andrzej Jarzabek wrote:
> On Aug 12, 6:21 pm, gleichman <fox1_...@hotmail.com> wrote:
>
>>On Aug 12, 11:22 am, Peter Knutsen <pe...@sagatafl.invalid> wrote:
>>
>>In effect to reject the study without cause (by assuming ill intent on
>>the part of the marketing group who made the study- not through any
>>data of your own)
>
>
> I for one was not assuming ill intent (merely taking the possibility
> into consideration). I also do not reject this study, merely doubt it.
[...]

Andrzej, you need to learn to quote properly. There's not a single word
from my post in the material you've quoted, yet you failed to delete the
attribution to me above.

--
Peter Knutsen
sagatafl.org

gleichman

unread,
Aug 16, 2009, 1:13:46 PM8/16/09
to
On Aug 14, 12:41 am, "Warren J. Dew" <psychoh...@aol.com> wrote:
> Nice to see this group still has interesting discussions when someone
> posts.

Thanks for the post Warren, it's nice to see that you're still around.


> Bartle's model includes these types because they have a
> disproportionately large effect on the culture of an online game, but
> such small numbers could easily be missed in surveys such as Yee's.

As a side note, I tend to think of them as representations of
extremes, and it's typically the extremes that drive MMO play. I don't
think that's the case in table top, or at least isn't a much the case.


0 new messages