Introduction:
The dungeon design process can be a grueling one, but it is a necessary and rewarding practice that constitutes a sizable portion of game-play, especially in RPGs. There is a topic that poses the question "how well planned out should your game be?" Well, dungeons are complex creatures, and deserve considerable thought and planning. What follows is a list of steps that outlines my own personal creative process:

1.)Function in Story Arc
What is the dungeon's utility within the context of story at large? Why is it important to traverse its contents? In the example that I'll be using to demonstrate, The Dream Palace, it is the 3rd act in a training program that the main character, Toshi must overcome for personal mental and physical growth.

2.)Location/Concept
Establish the setting of the dungeon. This is more important than it may seem like at first. The environment that you create for the context of the dungeon directly affects its design in terms of the types of puzzles and enemies you'll have to face there, as well as contributing to the overarching atmosphere, which is also related to point 1. Emphasize its unique characteristics.

EX.1: The Chateau level in Uncharted 3. It's an old abandoned mansion in the woods, that is at least 100 years old. That's rich enough to stand as a satisfying setting, but midway through they set the place on fire, and it becomes even more exciting as you try to escape, it changes the spatial dynamic of the layout and alters the paths you can and cannot enter.



EX.2: For The Dream Palace, I'm using cliffs that are embedded with palatial features and caves, that have puzzles spawning between the interior and exteriors. There's a climbing aspect that highlights the utility of integrating vertical elements in 2D games. I explore that idea a bit in this post.

3.)Sketch Layout
Bust out the pen and pad. Call me old fashioned but I like to plan out my dungeons on paper. This is because it's nice break from staring into the depths of the monitor but also because it's traditional and immediate. If I have an idea, I can scrawl it out, however crudely, and see if it might work. The other benefit of drawing them out on paper is that you can see the entire layout at work at once and make decisions based on that omniscient view. It might be easier if you have a chipset in your head for the dungeon, so you can draw the maps with the tiles in mind to make it easier once you actually get to mapping.

EX.1: Here is the sketch for layout for the first section of the dungeon:

As you can see, artistic talent is not entirely necessary. I wanted to start off simple, to ease the player into the dungeon, so it can become more complex as you traverse it.

While I sketch the maps, I label them with numbers, and keep a second sheet dedicated to explanations, listing the rooms in order, and what their function is, the puzzles involved, contents of chests, enemies to battle, etc.


4.)Implementation
This part is made much easier since I already have the plans mapped/worked out. I just simply follow the plan, which means making a lot of maps based on small sketches. And if I forgot something, I have the explanation sheet to guide me back into my original thoughts. Because of this, I like to get the maps out of the way for a major section first, so my dungeon in game is also a rough shell of the final product.

Ex. Here is what those sketches roughly translate to in game:

(Exterior)


(Interior)

5.)Eventing
This is the set where you go back and make everything work as you originally planned for it to. This means, setting up puzzles/mini games, making fight-able enemies, filling in your chests, and setting up your traps for the player. The maps get filled in with events that allow each room to serve its function/role in creating the obstacles the player must face.

6.)Improvisation
Because of the benefit of hindsight and preparatory sketches, you may get better ideas once you start to make/implement the dungeon. Don't think that you have to follow your outline to the T, it's just an outline, a rough cut that is flexible enough for improvement and interpretation. Most people aren't going to get everything perfectly to how they want to on the first attempt, the importance of redux is paramount. Sometimes good ideas come to you in the heat of the moment, which is the beauty of improv.

Ex.In my notes (if you can read them) for room #7, it says that lighting the torches a certain way will make a new switch, and that switch will open a new path. When I got to making it, I thought that that extra switch was redundant, so I nixed it and made the torches open a new path directly.

6.)Playtesting
It is the responsibility of the game designer to make sure that the dungeon is playable, and works as you want it to. I play through my dungeons countless times diligently trying to break the game as much as possible. Redux and repeat so that it's airtight for when others play it.

7.)Beta-Testing
Even after you have gone through the dungeon a million times and fixed every conceivable glitch and it all works flawlessly, you need to make sure that everything is balanced, that other players will know what to do, that you're puzzles aren't overly complicated or difficult to solve and that all your battles are fairly balanced. This is where beta-testing comes in, ship it out to a few people, and let them see if they can figure it out for themselves. You need to have reliable people for this, with respect and knowledge of the game creation process, to offer rewarding criticisms and feedback.

8.)Redux
Continue tweaking based on feedback from above steps until everything is perfectly as you want it, and perfectly reasonable to expect a player to traverse the dungeon, that its overall difficulty isn't too demanding and that it fits its place within the greater context of the game.

Dungeon Content:
But what are the components that comprise a dungeon? How will you know which elements to plan for in your initial sketches?
Here is a formula based on the general over aching obstacles typically found in dungeons:

Dungeon= battles + puzzles + minigames + traps + (mini) boss + reward(s)

The order that they go in is variable, and you can play with the function/structure of each one, but remember to put them in an order that pleases game design, that is, it conditions the player into behaving a certain way. (That you as the game designer WANT them to behave in.)

Ex. battle>puzzle>reward, is a good set of 3 different content types. It challenges the hero/player physically (battles) and mentally (puzzles) and rewards them afterwards for their efforts, which positively conditions them into doing the same thing again. Using the prize of chests as an incentive for players to complete challenging tasks.

Likewise, negative conditioning comes into play as well. For example, if all of your chests had enemies in them instead of treasure, the player would actively avoid chests. You can use punishment to make sure the player doesn't keep making the same mistakes.

Ex. There's a spike trap that decreases HP by 10, so when the player walks into it, they lose HP, so they are conditioned to not walk into it again. It seems obvious, but only because it is a deeply rooted trope and because spikes show an immediate physical danger, but the principles behind it are sound.

So work the structure of your dungeon in a way that is gratifying and switches up game-play types to keep it interesting.

Well that's all for this post, which seems more like an article. The next one will be about different puzzle types, so watch out!