The Puzzle of Designing Puzzles
During the development of Puzzle Depot, I've received many questions about the design process. In this update, I would like to dig into that meaty subject and attempt to explain how I design Puzzle Depot's sinister rooms.
First, I'd like to show a comparison between the in-game visuals and the schematic mode which will be available in the level editor. Using simplified graphics, this mode allows the creator to focus purely on the mechanical elements of the puzzle:
Creating these simplified, 8-bit style variations of everything in the game has been a fun challenge! I'll be using them a lot for visual aids during this update.
Let's get into some of the questions I've received so far:
"How long does it take for you to build a puzzle/room?" -Ygor
This is something that varies wildly from room to room. Sometimes, it takes a couple hours or more, especially when I end up scrapping a design and starting over. In fact, the entire tutorial level was completely redone, the current version is actually the second draft and bears little resemblance to the original.
Other times, a room comes together almost on its own. In the case of this familiar room, it took less than an hour:
Very little changed from the first draft to the finished version, other than slight adjustments to the positions of the plates and crates. This was one of the few puzzles salvaged from the first draft of the tutorial.
So, in short, a single room can take anywhere from a few minutes to an entire afternoon!
"Do you know where you're going from the start with each, or how often is it freeform?" -Moth
A bit of both, really!
When designing the tutorial level for the second time, I had a rough map drawn out and a list of the different gameplay elements I needed to introduce throughout. It's really helpful to have a general idea of the "flow" of the level... in this case, the concept was that you would play through 3 different iterations of the level at different points in time, with more complex mechanics being introduced with each iteration. This would also allow me to show the world changing over time, and tell some of the story simply through the level design.
It's difficult to find a balance between the overall length of the level, and the introduction of new gameplay elements. I don't want to overload the player with too many new concepts at once, but, I also don't want to drag out the level forever. Overall, I'm happy with the length of the tutorial level, but some of it feels rushed in places. Especially in the 3rd iteration, you don't get much time with either the hazmat suit or the rusty can lid. Still, I feel it's important for the demo to show each type of item and how they work... these are the sorts of things I grapple with when designing a level.
For individual rooms, it's helpful knowing ahead of time if the room will be introducing a new object or mechanic, or if it will be building on something introduced in a previous room. I don't often have an idea of how the specifics of the puzzle will work out, so I generally will just start placing the objects that I know I want, and seeing what emerges... it gets much more "freeform" at this point. Once I have something more concrete in place, I'll tidy up the room, remove extraneous objects, and then add the aesthetic touches.
A common problem for me when designing a room is that I have a tendency to make it too difficult, and have to tone it down a bit. I'll often save a copy of the more difficult version for later, which will be useful when it comes time to design the Expert mode. Gauging the difficulty is quite... well, difficult! I'm sure anyone who designs puzzles can relate to this, and I'm not sure I have any real insights into it other than letting people play the level and observing how they perform. The most valuable feedback we've gotten has been from watching people play the game on Twitch.
Let's dissect my favorite room from the tutorial, "Trial", which is the culmination of many elements introduced throughout the level. Starting off, there are a few things we already know about this room:
- The layout of the walls and exits has already been established from the previous iteration of the level.
- The player will enter with a rusty crowbar in their possession. They may also have a spare hazcap, and possibly füdbars
- I want to introduce a new mechanic: Rust mites can eat items in your possession!
- Another new mechanic: Rolling barrels can be used as weapons.
- I want to allow for multiple solutions. This is part of the design of the entire game, really, but I want to reinforce it here.
- A one-way door is needed, to prevent the player from accidentally putting the entire level into an unwinnable state by backtracking at the wrong time.
The idea is that the rust mites by the door will rob the player of their precious crowbar when they leave, but there should be a way of getting past them, either by feeding them a spare barrel, squishing them with a barrel, or using a spare hazcap. I'd also like to further explore roach behavior in this room, so a cluster of roaches is a perfect candidate to demonstrate barrels as weapons. The placement of the roaches is pretty arbitrary, I likely moved them around a lot over the next few steps before settling on this position.
Currently, there's no way to solve this room, so we'll remedy that with some pressure plates and roaches with which to activate them, a plate controller for the exit door:
I also added a pit and some more mites around that roach cluster... barrels only do 1 point of damage, and the roaches have 2 hearts, so the presence of the mites is actually useful here in that they will remove the first barrel, allowing that middle roach to be hit with a second barrel. The idea is that the middle roach should be taken out, and the now-separated roaches should be pushed into the corners. At this point, the room is solvable, but it could be more interesting. The group of roaches on the left can actually be ignored entirely.
I'd like for this room to incorporate more lessons learned from earlier in the level, and I need to reduce the amount of "dead space":
I decide I don't mind the idea of circumventing the roach cluster as I want to encourage other solutions, but I've made it tougher to utilize the rusty barrels with the addition of toxic waste and more mites. The toxic barrels aren't strictly necessary, but offers another option for activating plates for the player with health to spare.
On the righthand side, a 2x2 crate was added. This is more interesting in that the barrel and roach are not immediately accessible, and the crate can also be broken down in such a way that it can be used to activate a plate. Since it may be difficult to get through this room without taking damage, there is a füdbar inside the crate as well.
At this point, the room is coming together, so I start adding details like information signs. There are currently many ways of approaching this room and I could definitely narrow the focus a bit, but for this early level I really want to reduce the risk of the player getting stuck, so I'm okay with leaving it more open-ended.
"How do you know when to say something is done?" - Angela
This can be really, really tricky for some rooms! Especially rooms like "Trial", where there are many options present. There comes a point when you just have to stop tinkering and make some decisions. Once a room has reached a point where it satisfies the goals it needs to, and has a minimum of extraneous elements, it's more about finishing touches and making it look pretty.
Still, I find that I often revisit rooms and make changes, especially as the level as a whole begins to take form and room-to-room interactions become more relevant. With each room, I have to consider how it factors into the rest of the level... how many hearts does the player have at this point? What items are they carrying? What items can be purchased at this point in the level, and will those items be too game-breaking?
Designing a room is a puzzle in itself, and designing an entire level is a puzzle within another puzzle. I don't know that I have a satisfactory answer for knowing when that puzzle has been solved... more often, I feel that it simply reaches a point of "good enough"!
"How many kinds of fish can you design?" - Soreth
I hope this has given some insight into the design process. I'm very much looking forward to designing (and subjecting people to) more puzzles... especially the Expert levels!
And don't worry, once you get your hands on the level editor, you'll get a chance to return the favor!