Re: Higer-order plot abstractions

["B. Fallenstein" <b.fallenstein@gmx.de>]

Peter Weyhrauch wrote:
> 
> I like your idea of a "dramatic simulation".  It seems like a way to
> think about potentially useful story material, by filtering in some of
> the interesting world state and adding to it some abstract state.

Hi Peter,

I'm not quite getting you here. I have a hunch of what you could mean,
but... what does "filtering in some of the interesting world state and
adding to it some abstract state" *mean*?

> Higher-order plot elements seem like a plausible way to extend the
> number of concepts in this simulation.  They seem like a nice way to
> get a larger number of user options, especially since the author
> doesn't need to create by hand all these options.
> 
> The dramatic simulation state is defined by a set of "connection
> types," relationships being one example, with possible weights
> associated with those connections.  These connections, like
> higher-order plot elements, are higher-ordered.  That is, they can
> refer to other connections.

Right.

> > To be sure, the author of course has to specify every single action
> > (as a template) that a connection allows a character to perform:
> > this is part of specifying a type of connection.
> 
> I am confused by this for two reasons.
> 
> One, this seems to say that the author must now understand and take
> action on every connection.  Before, the combinatoric explosion was
> "free" in that the author specified a small number of components and a
> small number of rules.  In this case, there seems to be a problem of
> too much authorial work.  Can you unconfuse me?

My point is that when you specify a connection template, one subtask is
to specify all the action templates this kind of connection allows a
character to perform. So, in this model (not sure whether this point
about is good or not), every action is intrinsically tied to a
relationship that puts you into the position to exercise this action;
for example, if you are Bert's boss, you can tell Bert what he ought to
work on. So when you specify the relationship "X is the boss of Y,"
along the way you specify the action "X tells Y to work on Z," which X
is enabled to perform by virtue of being Y's boss. This is all I had
intended to say with the statement above; you seem to have interpreted
something else into it.

> Two, are such actions only simple actions, or complex actions that can
> refer to other actions ("X asks Y to do Z for them.").  If you mean
> complex action, I cannot understand how you don't have a double
> combinatoric explosion that the author must deal with.  That is, there
> are an exponential number of connections and for each the author must
> specify the allowability of an exponential number of actions.

I'm not saying that for an action with three slots X, Y, and Z, we must
specify "foo can go into slot X," "bar can go into slot Y" and so on,
but just for the subject of each action there being a connection they
must have in order to becoming that subject. This presupposes, though,
that for each action there is only one single type of connection
allowing you to become the subject of that action-- which probably does
not lead far enough. So we need something slightly more complicated.

> This leads to another point of my confusion.  If the characters can
> take any of these complex actions, how can we devise a weighting
> equations that can make sense of the combinatoric explosion of
> options?

The approach I had in mind was that for each action template, we would
have a formula for computing a "desirability" of the action. The actual
desirability would of course depend on the values in the different slots
of the action template. Higher-order actions could make use of the
desirability formula for the action they're acting upon. (e.g. in "Ana
agrees not to fire Mr. Gadwick," the desirability formula could make use
of the desirability formula for "firing Mr. Gadwick").

> As for your catch about character growth, it seems you may have your
> solution in the set of connections.  Why can't you create a set of
> connections that measure attitudes or coping ability, for example?  As
> you say, these are as important as character to character connections.
> So, as a methodology, build as many connections on those topics as you
> do on character to character topics.

Hm, yes, maybe this will suffice. That sounds nice. ;-) The challenge is
to archieve the combinatorial explostion for the player...

Um, I'm not sure if I really clarified my ideas. It's too late and I'm
too tired-- sorry, I'll try to reply at a more reasonable time of the
day next time. :/

- Benja