So far we’ve been looking at level design purely from a gameplay perspective. I want to switch tracks for the next few parts and discuss the aesthetic and narrative aspects of level design. That is not to say that these elements don’t effect and inform how the player interacts with the actual mechanics of the game, which is a subject I’ll also be discussing in these parts.
Now, I want to make a disclaimer and say that I am first and foremost a game designer, and not a visual artist. This is not going to be a comprehensive lesson on the rules of composition as they apply to visual art. Instead, it will be a basic run-down of some basic elements of composition and how they can be applied to level design.
I want to step back a bit and discuss the term “Level Geometry.” Level Geometry refers to the physical space and construction of the level, and the shape and positioning of objects in the level. You’ll see the term most often used in relation to 3D games, although the term could be relevant to just about any level design. This article is going to be a rundown of some common ideas and terms that are useful when thinking about level geometry. It’s by no means comprehensive, but it should give you some stuff to think about when working with space.
This time around I want to talk about some ways to make navigation easier in non-linear space, and I want to specifically deal with the issue of backtracking. Backtracking is when the player has to go back over the same are they just went through to get to where they started. Here’s an example to demonstrate what I mean (note, in this example, you don’t start with the gun. You’ve got to find it):
There’s a level design idea which I’ve been trying to resolve into a concrete topic, but I was having trouble getting to the core of it. Then I played the new Strider and it resolved itself. So far we’ve been talking about non-linear level design in terms of each area being a challenge the player should want to go to and complete. In other words, in the best possible playthrough, the player would end up trying to go to every area.
However, there’s another option. You can have areas that exist for the player to end up when they fail at completing a challenge perfectly. You can design levels such that if the player completes it flawlessley, they might only see a small percentage of the level.
How Strider accomplishes this is by making it so that there are no bottomless pits for you to fall down. Instead you just fall to a lower part of the level and have to fight and climb your way back up. Large areas would often have multiple tiers, with the upper tier offering the fastest path, but requiring perfect timing and platforming, and lower tiers possibly taking longer to get through, but offering more hiding places and movement options.
This sort of design can be seen throughout the original Sonic the Hedgehog games. The game was about maintaining your momentum, and doing so would allow you access to a certain path through the level. If you timed everything perfectly and maintained your momentum, you’d be able to stay on the upper path. However, if you either fall down, or don’t have enough momentum for certain jumps, you’ll end up on one of the lower paths.
Here’s an example where there is an upper path of platforms with shooting enemies on them. If you miss a jump, you’ll fall to a lower path with hopping enemies and a climb back up to the goal.
The game Antichamber by Alexander Bruce is a fantastic example of this sort of level design. The game is full of strange areas that you reach if you make a wrong turn or get lost. You’re likely to see most of the game world in your first playthrough, which could take hours, since you’ll be making plenty of wrong turns. The game plays around with illogical space and non-euclidean geometry, so getting lost is a large part of the experience. However, if you know what you’re doing, it’s possible to speed-run the game in less than 10 minutes, and see only a small fraction of the game world.
This sort of design works in more subtle ways in games where enemies can move around the level quite a bit. You can make areas whose purpose is to give the player a place to run or hide if they get spotted by enemies, or places for them to be chased by large groups of enemies. There might be an optimal path through the level which doesn’t involve these side-areas, but the player either doesn’t know that route, or has botched the timing of it, and now has to improvise. Think to games like Metal Gear Solid, which would contain side-rooms which served as hiding places in case you got spotted by an enemy, or games like Bioshock, where each area was a fairly open space with a network of hallways and rooms which could provide the player with strategic options if the situations turned against them.
In addition, you can use areas that the player reaches through failure in interesting ways. Ninja Gaiden 2 has an area which requires successful timing on wall-running and jumping. If you fail, you fall down a pit. The pit contains a broken staircase which makes you practice your wall-running and wall-jumping before returning you to just before the challenge you failed. Instead of automatically assuming that everyone needs the same amount of coaching with the skill, they put a test in place, and give an extra learning challenge to the people who fail that test. See the following example:
The bottom area is an easier, lower risk version of the challenge on top. (Also, take note of the way I block off the bottom path. The descructable blocks are placed in such a way that you can’t get through them from above. This serves as a sort of one-way gating mechanism to keep the player from intentionally exploring the bottom area, expecting some sort of reward. If they fall down there, the reward is getting back up, and thus the area makes sense. However, if they notice the path down at the beginning, they might venture down there and feel cheated that they didn’t find anything)
You can also build in negative feedback into your level design by utilizing areas that the player reaches through failure. (For those who don’t know, negative feedback is when a system tries to push itself away from extremes. For example, a mechanic where you deal more damage the less health you have would be a negative feedback loop, as the worse you’re doing, the more chance you have to recover and do better. The blue shell in Mario Kart is another example, since it specifically targets whoever is in first. This is as opposed to positive feedback, where the system pushes itself towards extremes, for example, a mechanic where the less health you have, the less damage you deal.) For example, you could put health items and ammo in a room the player might be chased into by enemies, or extra health items in an area they might reach by failing some challenge. Although be careful, or you might end up making what you intended as a failure area into the optimal path. And, of course, something could be a little of both. You could have an area that’s there to either reward exploration, or provide a little boost to someone who has been pushed there.
In other words, when designing an area, it’s important to think “what would bring the player here.” However, the answer to that question isn’t necessary “a shiny new item” or “progressing the story”. Sometimes it could be “because they missed the jump and fell down here” or “because they bit off more than they could chew, and were chased here by a gang of enemies.”
I want to take what I’ve talked about so far in terms of non-linear level design and show a practical application. Thus, I’m going to do another “Planning out your Levels” article, this time showing my process behind the level design from Treasure of the Abandoned City, specifically the space between starting the game and actually entering the city.
We’ve talked a bit about arranging challenges in a non-linear level. But what about the person actually playing your game? How do they find their way around? In linear games, it’s easy. “Which way did I come from? I’ll go the other way”. In non-linear games, it’s a bit more complicated (hell, sometimes the answer IS “back the way I came.”).
Now, the easiest way to do this, which a lot of modern games do, is to give the player a map, and put a marker on the map telling them where to go and an arrow pointing them how to get there. Now, this works to get the player from point A to point B, but it takes away a lot of the fun of actually exploring the space and finding your own way to the goal. It can also feel unnatural and immersion-breaking. On the other hand, if you give no guidance, there’s the possibility of the player getting lost and frustrated and not even know where they’re supposed to be going.
As I said in the last part, there are a lot of benefits to linear design. You can control the pace at which the player is introduced to new content and harder challenges. What if we want the feeling of exploring a non-linear space, while also creating a sort of linear-experience of introducing challenges to the player?
We’ve talked so far about stringing together challenges in a linear order. Challenge 1 is followed by challenge 2 is followed by challenge 3 and so on. There are many advantages to building a game around this structure: you can tightly control the difficulty curve, easily make sure everything is introduced in the right order and you can know where the character will begin a challenge and design all the obstacles to block that specific path. But, of course, many games benefit from giving the player more freedom in how they explore the world of the game. And you can still apply much of what you’ve learned about linear level design to non-linear level design, it just takes a bit more consideration.
I think one of the biggest stumbling blocks games have as an art form is the idea that the author of the game does not have total control over the game due to the element of player input, and the fact that unexpected events can emerge out of the interaction between the player and the systems within a game. A lot of designers see a sort of wall between the parts of the game that they can control, and the parts of the game that, due to their interactive and emergent nature, are out of their control. There’s an approach to emergent gameplay that basically says “put a bunch of interaction-rich mechanics together and see all the weird things that happen when someone plays it”. There’s a kind of disavowing of responsibility for the outcomes of an emergent system.