I’d like to suggest an extension, or perhaps re-imagining of the “complete” model, which still makes goals a requisite, but defines three types of goals: explicit, implicit and induced. Explicit are what we see in missions and quests et al., implicit are what we see in scores or earning more moeny, and induced are what we see in aesthetic conditions or “success states”.

Puzzles then, in the Murray sense of an optomizitation of a problem includes explicit and implicit goals, whearas an induced goal correspondes to the Rhizome. I think there’s a lot of interesting potential in the combinations and nesting of these different goal-orientations, as well as in pure exploration of inducive goal-orientations (which to a layman, or any theorist who disagrees with my usage, is not a goal-orientation at all).

I’m currently producing a casual game that is a series of explicit goals with a ditchomous inducive goal (and a very interesting dichotomy, from a fiction perspective). My next goal is to design and/or produce a casual game with an implicit goal and a number of induced goals, but no explicit goal whatsoever.

A tiny bit of vocabulary that I’m fond of using is that of “success states” – those may not be goals per se, but simply things that the player recognizes as being expected of him, or simply things “well done.” They are typically (but not necessarily) associated with rewards: plain advancing or killing an enemy in Scramble, finding a hidden room in Doom, etc.

It is often that the player is forced to choose between different success states, all equally viable (or even irrelevant) to the overarching goal, and that’s where he can define his style of playing.

In contrast, I really couldn’t get into Second Life precisely because it was too open and undirected. Whereas WoW provides optional goals, Second Life provides none, and I failed to find that satisfying.