Designing a Mystery - one thought, and one mistake

My first adventure for the new campaign is a mystery - a young woman has been attacked and maimed, and people want revenge. So far, so obvious.

My approach for designing a mystery is basically to draw up a bunch of nodes and lines - the nodes show all the likely vectors for investigation and the lines show where they point. I try to make sure that there are at least three lines pointing to any node I want the PCs to reach, in keeping with the "Three Clue Rule".

The consequence of this is that the web that describes the mystery looks awfully like the web of room connections I would make prior to drawing a proper map for a dungeon.

Which is the observation I would like to share: designing a mystery is basically the same as designing a dungeon. It's just that instead of exploring the rooms of a dungeon, the party are exploring the web of interactions between NPCs.

(I don't hold with the notion that the DM should only design half the mystery, on the grounds that the players will come up with a more compelling resolution anyway. Because they might well not do that, and end up meandering aimlessly through the setting instead. It's much easier to provide them with an enjoyable journey if you know the landmarks along the way.)

Anyway, that brings me to the flip side of this observation, and the mistake I've made in my first session. In a dungeon, you have a bunch of rooms, and in most (though not all) of those rooms you want something to happen - some sort of an encounter, a trap, and what-have-you. And perhaps some wandering encounters in the corridors between rooms, or something like that, but I digress.

What I'm saying is that in dungeon design it's a mistake to have too many empty rooms - rooms where the PCs search and just move on. When designing a mystery then, and more importantly when designing a mystery adventure it's a mistake to have too many nodes where nothing happens. If the PCs go back to the scene of the crime to see what they can find, it's probably a good idea for them to meet some resistance at that point - maybe a glimpse of the killer (and an exciting chase), maybe some hired thugs sent to dissuade them from their investigations... something.

And so, the first "real" session of the new campaign got off to an okay start, but not a great one. It was just filled with too many 'empty' nodes, and therefore not enough action. Fortunately, that's easy to remedy for next time.

I hope.