[MUD-Dev] Something in the water

[J C Lawrence]

On Wed, 25 Jul 2001 15:49:29 -0400 
Travis Casey <efindel at earthlink.net> wrote:

> Wednesday, July 25, 2001, 1:57:08 AM, J C Lawrence wrote:
>> Caliban Tiresias Darklock <caliban at darklock.com> wrote:
>>> J C Lawrence <claw at 2wire.com> wrote:

>>>> How can you mechanically determine when RP is occuring?

>>> I don't think you can. I've thought about that a lot.

>> That was the conclusion of previous attempts on the topic on this
>> list.

> I'd say that the first thing you have to do is define what RP is.
> I've given my own definition of RP many times here before, but to
> save anyone from having to dig in the archives, here it is:

>   Roleplaying is making decisions about "what does my character do
>   next" in character.  That is, such decisions should be based on
>   the question, "What would this character do, given his/her
>   knowledge of the situation, current physical/mental state, etc?"

> There are, of course, other people who have different definitions.
> One common definition seems to be:

>   Roleplaying is interacting with others in character.

> I'll refer to these as definition #1 and #2, respectively, in this
> message.

The essential difference being that the former is an internal
perception (you of your character) and the latter an external
(others of your character).  In essence they are both attempting to
define the same thing -- its just the viewpoints are different.

>>> The problem is that the *propriety* of roleplay is also at
>>> issue:

>> Taking the definition of propriety in its "suitable for a given
>> context" form, yes, quite, very.

> Under either definition, determining whether or not something is
> "proper" roleplaying requires knowing a lot of information about
> the character and his/her situation (in the general case, that
> is).  Further, it's going to require a lot of world-knowledge and
> the ability to use it -- e.g., the ability to reason from premises
> like:

<nod>

Big pain in that direction.  Yup.

> I'd venture to say that it's an AI-complete problem.

I'd venture to concur.

>> Posit:

>>   Player value requires communication.

>> Counter:

>>   No communication equals no value.

>> The silent type who stands about and says nothing communicates
>> nothing and therefore (arguably) has little/no value to the game
>> or other players within the game context.

> Which matters, *if* the purpose of the system is to reward those
> who are of value to the game and/or other players.  With a stated
> goal of "rewarding roleplaying", though, we can't make that
> assumption.

So:

  If a bear roleplays in the woods, does anybody notice?

Or perhaps more significantly:

  What value are RP points to a character who doesn't observably RP
  in the presence of others?

> Using definition #1, such a system has nothing to do with
> roleplaying at all -- you can roleplay without ever interacting
> with another character, and you can contribute great value to the
> game without ever actually roleplaying.

Quite.  

Does the fact that you have a saintly soul or are in fact a chess
genius have any relevant value if that fact is never revealed or
detectable?

> IMHO, "rewarding roleplaying" is a very bad specification for what
> you want a system to do.  In order to implement a real system, the
> goals ought to be better defined.

Underneath is a semi-Pavlovian model which assumes that positive
rewards for Good Behaviour will encourage further Good Behaviour
(Pavlov concentrated more on negative reinforcement).  Central to
such an approach is the idea that the Good Behaviour can be both
defined and quantified, and then, that given such (which is tacitly
assumed) that a system can be arranged which created the positive
reinforcement loop.

  Not sure how many of you have kids: the cherry and stick approach
  does NOT always work, or even work reliably in the cases it does
  work.

You've defined two definitional models for RP: one player viewpoint
centric and one player base viewpoint centric.  I'd argue that under
definition #1 RP is neither mechanically definable or quantifiable,
and that under definition #2 the problem is both large and
unmalleable.  

However that doesn't seem the critical point.  The critical aspect
seems to be:

  Type #1 RP is explicitly undetectable and undefinable by
  mechanical systems.

  Type #2 RP *may* be detectable and quantifiable, but its known to
  be a hard problem.

So, in talking about mechanical RP reward systems, even if they are
based on player collusion, we're explicitly talking about definition
#2.  We can't be talking about definition #2 as that's an
undefinable question (eg what colours is that hard vacuum?).

Now, if we're only talking about externally viewed RP that
significantly simplifies the problemscape -- __ALL__ we are
interested in are the activities of players that are subject to
being viewed by other players and thereby judged.

That doesn't implicitly select against the taciturn (other than
socially), but it does select against the wall flower of the
Shakespearean wanna-be who recites to empty halls where nobody can
hear.  This latter case however doesn't fit definition #2, and so
doesn't apply.

I can see why this doesn't make many RP'ers happy as there seems to
be a common confusion between the two definitional types such taht a
firm believer in definition #1 will then think he is applying that
definition to others in his game when in fact the simple fact of the
application of the definition to a character not his own EXPLICITLY
makes it an applications of definition #2.

  Note: Questions of IC as versus consistent with externally
  observed character, background and mental state are arguably moot
  and irrelevant as the two are effectively synonymous.

> I've said here before: people do not actually want to reward
> roleplaying (at least, not by my definition).  

Which makes sense given the above.

> They want to reward a set of behaviors that they associate with
> roleplaying.  Roleplaying, however, does not require those
> behaviors, and those behaviors can be engaged in without
> roleplaying.

Yup.  

> Or, to put it another way: The appearance of roleplaying is more
> valued than the actuality of it.

Rephrasing:

  Things are only important if groups of people agree they are
  important.  If a group decides that a certain form of behaviour is
  RP and is desirable, then, well, it is.

  The solo guy need not apply tho this equation.

>> Taciturnity is not the problem.  Social connectivity and
>> representative perceived social value is (ie how well is he
>> known, what is his perceived value).

> This only matters to roleplaying, though, if you define
> roleplaying in a social context.

Of course, as above.

>> I've done some messing about with this list attempting to graph
>> and determine aspects of the peerage and meme transfers by
>> isolating and tracking adjectival phrase usage across posts and
>> posters.  I'm not convinced of the validity of the model (MUD-Dev
>> is a very small sample base) but I get some nice graphs out of it
>> which subjectively feel "right" for the list.  The suggestion is
>> that automatic analysis of character actions and resultant
>> memeticly similar actions by others can both be detected and
>> measured, and would be a reasonable indicator of effective social
>> activity.

> But not of roleplaying, by my definition.  Of course, that doesn't
> really matter too much -- I'm not a zealot saying mine is the only
> true way.  All I'm saying is that "roleplaying" is not a very good
> descriptive term.  If you want to reward people for effective
> social activity, then say that, instead of saying, "roleplaying"
> -- it'll be much more clear to everyone what's expected, then.

<nod>

There is a significant subjective divide between the two
definitions.  They really are different terms.

>>> Technical processes can rather easily determine that something
>>> MAY BE roleplay, but they can't easily determine that something
>>> IS roleplay.

> From my point of view, technical processes can't even do that much
> -- at least, not in a useful fashion.  Since roleplaying takes
> place at the level of decisions, anything a character does *or
> doesn't do* could be a roleplaying choice.  Technically, it's very
> easy to always pop a "yes" flag up, but it's not very useful.  :-)

But, but, but, but its right more often than my stopped watch is!

>>> An active staff involved in maintaining a desired environment is
>>> helpful for RP, while it tends to get in the way of
>>> game-oriented play.

>>   You can't kill the beggar!

>>   Why not?  He's worth 15 XP!

>>   But he's a beggar!  You're supposed to pity him and give him
>>   spare change and maybe some scraps of food.

>>   Screw that!  I only need 20 more XP to level and he'll do
>>   nicely for most of it.  How many XP is the little orphan worth?
>>   Maybe I'll just stab the leper.  That should be worth at least
>>   5.

> Logical consequences within the game world can help as
> well... and, for that matter, getting rid of the whole
> "shoot-and-loot" D&D-style experience structure can as well.

John Hopson's recent post would seem to have hit that one.

--
J C Lawrence                                    )\._.,--....,'``.	    
---------(*)                                   /,   _.. \   _\  ;`._ ,.
claw at kanga.nu                                 `._.-(,_..'--(,_..'`-.;.'
http://www.kanga.nu/~claw/                     Oh Freddled Gruntbuggly
_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev