Last week I said I was going to work on the concept of difficulty. I wound up spending an awful lot of time reading and performing other research, and comparatively little time writing about it. Not sure if I’ll wind up writing a lot to make up for it, or just plowing forward with the next subject.

Continuing in the vein of “dungeons like video game levels,” I tried to come up with a purely mechanical justification for the existence of doors. Doors often represent one of the dumbest pass/fail elements in dungeon design.

“You come upon a door. Do you open it?”
“…”
“What do you mean ‘no?’ I guess that means no adventure for you!”

You should probably never use a door when a simple hallway will suffice. Seriously, the inclusion of a door does insert a number of possible choices for the players, not to mention approaches to the problem of the door.

Here’s what I’m going to propose however: the adventuring party should be able to complete an average of say, eighty percent of a typical dungeon exploration without opening a single door. Why? Doors should be special. A door ought to represent an actual choice, not a speed bump.

Let’s say for example that a typical dungeon of twelve encounter areas (one-third of which are “blank”), features six doors. One of those doors should lead to a “blank” encounter. There’s always a chance that a door will lead to nowhere, for the same reason that characters should be able to die as a potential setback.

The other five doors should always go somewhere. At least one of those doors should contain a secret (the other end of the extreme from the “blank” room), so the players will know that every dungeon has at least one secret, and a general idea of where to look for it. There’s only one guaranteed secret however.

So far, we have door results for 1 and 6 on a dee-six. Good start.

There should be two totally normal doors leading to two totally normal areas, which include totally normal encounters. Add those together with the Door To Nowhere and you have half your dungeon’s door-related encounters covered.

The other two doors should probably be locked or trapped, but that’s going to be up to you. By having one-third of all doors be locked or trapped, you guarantee the guy with lockpicks has something to do, and the party is denied no more than two doors’ worth of content (5% dungeon) if they didn’t bring a thief.

Don’t lock or trap the Door to Nowhere. Don’t be that guy.

Here’s your random door table:
6 – Empty (contains a secret)
5 – Encounter.
4 – Encounter.
3 – Locked or trapped.
2 – Locked or trapped.
1 – Empty.

Two doors lead to an area which may contain a secret — they certainly don’t have anything else — and two doors lead to an encounter. The last two doors are an encounter, and have a reward of some kind on the other side.

You’ll notice that these doors don’t indicate whether they link to other parts of the dungeon or not. They could hypothetically consist of two doors sealing off one area of the dungeon from the rest, perhaps creating a loop. These doors technically have less to do with layout, and more to do with content.

These are for a hypothetical “average” dungeon. Their importance lies in establishing a rapport with your players. If they don’t have some idea of what to expect from a dungeon, you can’t subvert or play with those expectations.

Don’t put lots of pointless doors in your dungeon. Don’t do it.