December 18, 2003

Seek Ye the Gnarl – [books]

I am working my way through Salen & Zimmerman's Rules of Play. (I discovered this book by way of andrew's reading list [grandtextauto].) I've just completed Unit 2, which puts me somewhere around the halfway mark. Some preliminary notes:

- plenty of typos (a useless thing to say, really)
- I'm not sure I like the writing style. Lots of remarks like "Do you remember when we talked about..." and "As a game designer, you need to..." I think this was a choice to give the book a focused target audience, but it could have been taken up a level. Add a little redundancy for things like key definitions, or at the very least add page number references instead of "back in Chapter 3". These are all things I'm sure will be taken care of in the second edition.
- It's a textbook.

Now, as far as content goes...

I am disappointed with the "commissioned game" sections that follow each unit. I think perhaps they are overedited, as they are terse and lack depth. The commissioned essay by Reiner Knizia about the Lord of the Rings Board Game was quite emphatic about the designer's loyalty to the license. Now I haven't played the game, but I suspect it will quickly find itself on the forgotten trash heap of American culture. Chris Crawford's Lesson 28: Build licensed products for money, not creative challenge.

I think there's an idea here that needs to be drawn out, the "missing schema" of Games as Commercial Products. It is striking to read an analysis of tic-tac-toe followed by remarks about Starcraft. The latter was quite specifically designed, while it is impossible to say who designed the former, if it was designed at all. In fact, in some ways I think that "game design" may be an impossible task, with respect to games as cultural tradition. As I haven't gotten to S & Z's chapter on Culture I'll leave this as a topic for later.

Chapter 14: "Games as Emergent Systems". I have long been critical of the way "emergence" as a property of designed systems is constructed. In particular, the idea of "designed emergence" is in itself contradictory – if you have arrived at some specific behavior as a result of careful design tuning, then you haven't found emergence, you have instead redirected the emergent properties of the system back into the form. It is the ability to use emergent behaviors to some other end that people desire, not the behaviors themselves. Emergence is a kind of annoying "Abort, Retry, Fail?" mechanism that appears when a system gets to an unexpected state.

I also don't believe that complexity is the result of adding elements to a system. (Complexity being the sweet spot between periodic systems and chaotic systems in Langton's categorization.) I think the natural order of systems is chaotic, or rather that there is a lot of chaos out there and we want to identify some of them as systems, so they become chaotic systems. The move from chaotic to complex, and eventually to simple or fixed systems, is a matter of adding contraints to the system, of removing freedom. A scheme with only three elements, where one element has a changing relationship with the other two, is simple because it has been constrained down to only three elements. The addition of more elements is not a change to the system, it is a change of constraints and the definition of a new system. (A change of contrants as opposed to a loosening of constraints, as in "now there are 5 elements" as opposed to "now there are an unspecified number of elements". The change from 3 to 5 is not a loosening of contraint, but a change to a different constraint from the same class.)

Rudy Rucker refers to the sweet spot between chaotic behavior and periodic behavior as gnarly. The goal of designing systems, for artificial life or for games or perhaps even cultural systems, is to find the gnarl. Does Conway's life fit in the gnarl simply because you can create glider guns? You have to look at the particular constraints of Conway's life, the 2+3 rule that makes it work, next to the less interesting contraints, 123+4, 24+5, +6, to see why the 2+3 rule produces interesting results.

Getting back to Rules of Play, the rest of Unit 2 just scratches the surface of a lot of areas. Ten or twenty years from now a lot of the game references will be obsolete -- not just the computer games, but board games (LotR) as well. At the same time, I think the Rules schemas would benefit from from encapsulated analysis of a larger number of games, a sort of reading companion. This needs to include some of the "terrible" games that the author's admonish us to play and study (which are strangely absent from their own discussions).

Part 1
Part 2
Part 3

Posted by B Rickman at December 18, 2003 12:30 PM | TrackBack