One guiding principle: I am not going to try to publish this a book, in print, to make money. Doing so is (1) to damn much work. I have a job thanks. (2) it brings in considerations of page count and binding that I would sooner not worry about, especially where a need for economy compromises design, especially in the use of white space. If produce a good set of rules, I expect people to want to download them and print them -- but the complaint I usually read is not about volume of ink rather than page count.
Writing style is also important. Early on, I was told that rules (and business documents) should be as short as possible since every word you write increases the risk that you will contradict yourself. I've considered that for a while, but I think that there is a more important principle: learning styles vary.
Some people prefer find it easier to learn precisely written procedures reminiscent of the best boardgames. Others prefer more loosely written rules that discuss the procedures rather than specify them.
The experiment I will try (which I vaguely recall was actually used to some extent by Young and Lawford's Charge!) is based on a principle I learned for presentations: Tell them what you are going to tell them; tell them; tell them what you have told them.
How could this apply to a set of rules? The structure I propose is:
- An overall orientation that should set out all the major points of the game clearly and economically, but not too precisely.
- The balance of the document set out into sections that clearly divide what you need to know to prepare for play from what you need when playing.
- Within those sections, each rule will be stated in very precise language, with paragraph numbering for cross reference. The "precise" statement will be followed by a discussion of the rule, using examples, showing the reasoning behind the rule, and how it interacts with the other rules. Instead of risking contradiction I see this as both a chance to address different learning styles, and to provide a proof reader or play-tester to identify points where the rule I have written does not match the explanation I have provided for it.
- As a distinct section or a side document, collect the precise (or "summary") rules into a single section that can be printed for use during play.
- The final "what you told them" are play aids, which should cross reference back to both other rules sections and, of course, provide everything a practiced player needs during play.
Stating the same text multiple times invites the thought of a tool-driven printing approach, perhaps involving XSL-FO transformation of document fragments.
Now I have written some huge documents with word -- hundreds of pages with dozens of illustrations and scores of cross-references. Those exercises have mostly left me convinced that we have not begun to consider the next software to ease the production of documents -- granted we have eliminated the typing pool (yes, I have had documents prepared by a typing pool. I am not young), the typesetter (yes, I remember linotype machines too) and cut and paste with scissors and tape. A modern word processor used to it's full potential by someone who understands its features is an amazing thing. But I am sure that there is a next step begging to be taken.
But I'll probably just start with Word and PDF.