If you want to eat a book over the course of a campaign, the AD&D player's handbook is a good book to eat. The classes will be referenced non-stop, and the races have abilities and contexts totally relevant to a conventional AD&D game. "Detect grade or slope in passage" is a useless ability in many games, but not in a modal game of AD&D
What parts of the PHB go uneaten? All the rules you never touch. All the fiddly bits you ignore and the high-level abilities neither the players nor the NPCs fall short of. All the weird weapons the DM eschews for swords and long swords and bastard swords. Bards. This is still a pretty good rate.
|The bard, an undigested lump|
Things like short stories, example adventures, maps, and flavor gets eaten when it helps people to run their games. They go uneaten when the don't grab attention, or when they are irrelevant to the kind of game the system lends itself towards. I have read rulebooks which seem promising, but which I had no clue how to run. An example adventure in the back, written as a careful model which cashed all the checks written in the rest of the book, can be totally nutritious. Of course, the adventure needs to be relevant. "Tomb of the Serpent Kings" helps people understand how to run, for instance, Many Rats on Sticks. "Tomb of Horrors" would be a miserable example scenario for D&D 3.5.
My GLOG hack Vain the Sword tries to be as edible as possible, and the rules as I've written them are my earnest attempt to suggest what kind of stories the ruleset is good for. Just as AD&D player characters should expect to encounter locked doors, spell scrolls, and orcs bearing becs de corbin, VtS player characters should expect narcoleptic monsters, gigre recipes, and morally corrupt bats. The example dungeon at the back of the book features battle cries, eunuchs, visions, and drugs because that is what the rest of the book promised. I want players to be able to say "Wait, since I'm a dreamer, can I--" and "If we survive, I want to learn how to--" and "So these guys have the same ability as me?"
Not everything will be relevant to every campaign. Someone might run an AD&D game without monks and bards and polearms, or run a VtS game without dreamers and eunuchs and battle cries, and enjoy themselves all the more. The goal of the person writing the book, however, should be to make each part appealing and relevant. If there's a feature that players routinely ignore without affecting how they play the game, why not spare them the trouble and debone the book before you serve it to them?
This is a different concept from conceptual density. The principle of conceptual density is about making the ideas found in your book better. The principle of eating a book is about making the ideas found in your book relevant. A book can have a great concept that it makes difficult to eat, and it can also have a generic idea that is totally apropos. The warrior class in VtS is an example. It is a totally boring GLOG fighter, similar to other generic GLOG fighters. But hopefully it is still relevant. My players make warrior PCs because they are fun to play even if they are generic, and because it enables the concept they pursue. Players encounter warrior NPCs, and the generic abilities they possess are enough to keep the threat of those NPCs interesting. It does this while taking up very little space.
- books should be relevant
- they should not be irrelevant
- relevance can be broadly applied
- Not every campaign will use content designed to be use