Two years ago I wrote “A Use for Bookshelves.” It’s one of my favorite game ideas that I’ve ever come up with.

But it’s a clunky system. I’ve been using it in pretty much every dungeon I’ve designed in the last two years, and it’s always a hassle. Preparing a table of interesting information is a lot of work. It’s tiring, and I end up resorting to options that feel cheap. Like hollow books with a few gold coins in them, or completely random information that has nothing to do with the ‘theme’ of the bookshelf.

When I wrote the bookshelf system, I believed in front-loading game content. I wanted to create a huge game environment with mountains of detail. A fleshed out sandbox where the players could pick any direction to move, and there would be something for them to find there. But attempting to put that philosophy in practice has meant a lot more work for very little apparent benefit. Now, I think player interest should drive the referee’s worldbuilding. The sandbox still needs to be structured in advance, but the details don’t need to be at the referee’s fingertips before play begins. Take, for example, bookshelves.

Each bookshelf has a subject, and a number. That’s the only thing  prepared in advance.

All the books on the bookshelf fall within its subject. So a bookshelf about fishing might have books about types of fish, methods of fishing, history of fishing, ways of cooking fish, etc. The actual books don’t need to be enumerated or titled. The shelf’s theme just gives you a range of possibilities.

The number has two functions. First, it is the number of encumbrance points (not encumbering items) required to haul the books off. So “Bookcase: Fishing 12” will require the party to use 12 encumbrance points to get the books out of the dungeon. If they like, they can break the library up, taking as few as 1 encumbrance point worth of books with them. But if you only take 1 encumbrance point worth of books, then you only have Fishing 1 at your disposal, rather than Fishing 12.

The number is also the amount of questions that can be answered by the books before they’ve absorbed all of the knowledge the shelf has to give them. The question must be reasonably within the bookshelves’ subject, but other than that any question is fair game. Answering a single question requires 12 hours of game time book studying. If the referee doesn’t already know the answer, they’ll figure it out by the start of the next game session.

So, for example, the players find bookshelf: Religion 8. Between everyone, the party only has 6 available encumbrance points, so they haul Religion 6 back to the surface.  The cleric wants to know if there are any holy relics of his deity within 100 miles that he could recover for the glory of his god. The referee happens to have included such an item in a nearby dungeon, so after the character takes 12 hours to study the books, the referee tells the players about the dungeon, and that there should be a relic there. The fighter, meanwhile, wants to know if there’s any god badass enough for her to want to worship. The referee doesn’t really have a good answer, so after the session ends he comes up with a heavenly dog who has swords for legs and eagles instead of eyes. He presents the fighter with that info at the start of the next session.

Not only does this dramatically reduce the amount of work that the referee has to do, but it also gives the players a unique tool that they’ll be interested to use. The old system was a means of dispensing clues and quest hooks that would spark the player’s interest. This system is a tool to satiate an interest the players already have.

I’m eager to test this out.