I've already discusses the basics of a PCon3 here: http://www.indie-rpgs.com/forum/index.php?topic=17387.0
. Likewise I demonstrated an approach to reducing play goals to atomic pairs of topic and form. (Such as ethical business in a cultural form or military life in a declarative form.) That discussion can be found here: http://www.indie-rpgs.com/forum/index.php?topic=17514.0

What I would like to introduce now is the reason why reducing play goals to these atoms can be so useful. The first of these is a simple realization that most players are capable of maintaining multiple distinct atomic views, in essence viewing each in parallel. This generates the idea of view priority. Some views may be necessary, and player controls must work to ensure that learning occurs within those views. Others are desired, but not required, allowing greater flexibility of view satisfaction. A player may indeed only have desired atomic views, requiring only that some of those views be satisfied, rather than any particular atomic view. Lastly, there are accessory atomic views, which are not treated entirely as views. In this case satisfaction of the view is irrelevant, instead, these views serve as methods to enhance satisfaction of prioritized views.

For example, a stereotypical character channeling player would necessarily have an atomic view of "my specific character's responses and their consequences" in a declarative form. (Essentially necessitating a somewhat random walk through possible decisions constrained by the concept of that character.)  Such a player may also have a desired view of "experience moral conflicts during play" in a cultural form (meaning an interest in how moral affect all the players, including the one being discussed). In which case the player will attempt to satisfy both views at the same time. Lastly, a player may have an accessory atomic view of interpersonal conflict in a procedural form, in which case the player does not require a lack of interpersonal conflict, but is attending such a view to better make decisions in satisfying first his or her necessary view, and ideally the desired view as well. Learning may well take place in all three views, but the play controls of this player are focused on maintaining satisfaction of the first, and if possible the second view.

Now as more atomics views are added things can become more complex in general. Multiple necessary views can require priorities in terms of which is better satisfied, or whether equal treatment is preferred. 

Looking from a different perspective, views can be placed not only side-by-side, but also on top of each other. For example, it is possible to take an atomic view which based directly on another view. For example, if the player above also had a desired atomic view of "character achievement of specific in game events solely from my initial design" in a procedural form, then this would be a view realized only within the context of the initial necessary goal. Thus the patterns observed through the necessary declarative view become new content for this new procedural goal.

This process can be continued indefinitely, with a procedural focus on core mechanics giving patterns to a cultural focus on inter-player cooperation giving patterns to a declarative focus on world exploration. It is even possible for views to be self-referential in this respect, such as any view whose topic is my goals of play, not only does this view rest on all other views, it necessarily must derive from itself, since it too guides play. This stacking can then continue without bound, being realized as far as the player is willing to take it.

The result of combining atomic views in this manner is that we now have a rich language to describe the goals of a player. Unlike general classifications, this method does not gloss over inter-player differences, nor does it place strong constraints on the description. In particular, hybrid play is not simply describable in this language, it is an intrinsic part.

(As an aside to the more mathematically inclined: This language of views is actually analogous to regular expressions. In this case placing in parallel is union, stacking hierarchically is concatenation, and self-referential views use the Kleene star. This should give you an idea of exactly how descriptive the language of views can be.)

  - Mendel Schmiedekamp

Okay.  That's cool stuff.  I'd love to contribute something here, but you don't seem to be asking a question or soliciting opinions.  Is this meant to be a stand-alone statement (in which case, again, cool!  thought-provoking), or is there something that we can help you puzzle through?

Thankyou. I suppose I have been slipping into a lecture mode with these posts.

First of all, I am open to suggestions or corrections. I am particularly interested in people's opinions about the practicality of these methods. I've personally found them useful, but I can't be sure that this utility is generally the case. Most directly, do you find that you could use the views language of classification and do you think that such a language could be useful for therapy, simulation, or game design? Do you find views to be intuitive? Would you say that recall bias (tendency to recall specific portions or details of play) is at least as good an indication of what player's look for in a game as self-diagnosis?

I have several more pieces of this theory to relate, including a discussion of play control algorithms, the extraction of play constraints, and hopefully culminating in relating a more direct relationship between the process of design and the process of play. Would any of these topics be useful expanded sooner, rather than later?

    - Mendel Schmiedekamp

Mendel--I recall making the assertion some time back that what you are calling "recall bias" was an important element in diagnosing agendum. One of the key reasons why reports of actual play are so useful is that the person telling what happened will usually focus on the things they enjoyed, and thus when we see what it was they enjoyed we begin to grasp the kinds of things they seek in a game, and thus what is "fun" for them, their agendum.

I had in fact proposed an agendum test that would ignore techniques and instead present vignettes of game stories, asking the test take how likely such stories would be to emerge from their games and whether that is the kind of story they would like to have happen and to repeat to others. I haven't had the time to invest in devising such a test, but the idea is still granted some value as a possibility.

--M. J. Young

M.J.,

I'm fairly sure that I saw your post on that. I think I may have mentioned that it would be especially interesting to look at a retention questionnaire after a subject has read the vignette's. I suspect that the questionnaire would help focus our determination of the subject's interests. This is especially relevant since in PCon3 there is a wealth of categorization for play intent, which requires more discrimination than the writing of a vignette will likely be able to introduce alone.

On a side note, I'm less interested in classifying people as I am classifying behaviors. In that context the vignettes may be superfluous, other than to influence deductions about a subject's behavior in observed play. In particular, one of the most interesting things about RPG play is that players can modify their views over time, usually in response to other players. This self-reflective property makes classification more useful in the long run, but also makes it far less static.

That said, retention seems to be one of the best ways to bring home PCon3, by demonstrating directly to a player what portions of play are retained, that player can handle something more concrete than a therapeutic analysis often is. That may help people find new and better ways to approach their play, although it is more a knowing a problem than knowing the solution. (I.e., unlike the Big Model, discerning your view doesn't necessarily corelate to finding a better way to satisfy that view.)

  - Mendel Schmiedekamp