My weekend was devoted to an artist table at Anime Revolution in Vancouver and I haven’t had time to draft up a new entry. I want to talk about pacing next time. Hours are a little packed this week so I’ll get to something on Friday night.
Hello again all you excellent fans of D-pad studio. Here’s another part to my process of level design. My buddy Jonathan wrote an excellent short article about his design process about 2 years ago. I found myself picking up lots of great tips from it and began applying it to what I was doing with Owlboy at the time. I won’t summarize him here as it only takes a few minutes to read, but his point was clear about shapes. Using Tetris pieces as a backbone to room design is pretty clever, they’re versatile shapes. He talks about using the environment to suggest things to the player. Also, the flow of your levels to make a strong narrative of levels and always keep themes running. I noticed something really cool about his game (Ninja Senki) being played by speed runners. The repeating shapes were akin to a racetrack for them to memorize so that the unique enemies could be strategized. A focused and versatile set of room templates will engage the player and allow them enough familiarity, yet also allows you to defy their expectations.
Repeating room shapes are common in games, we need them to be or else every situation would be exhausting for both the player and the designer. Commonly these rooms are designed to suit a safe programming and design situation.
Alice Madness Returns almost always summons enemies in large open platforms and walled rooms. Square platforms are easy to notice in jumping sections because of the regular shape and size. It becomes an expectation to find them, fight a few waves, and look for the next way up. Killing the enemies opens up the next path. Doing this allows the player to gain familiarity with the environment and instead focus on strategies for defeating the enemies. Abnormally shaped and dynamic rooms are used to test what you’ve learned. These combat boxes also serve as ways to control the player’s pace. If you were to block it out, the maps almost become squiggly lines between puzzle and or combat boxes. That would be oversimplifying it though and there is a lot to say about 3D space and room shapes that will have to wait. If it helps, you’re just working with layered rooms. Carl Sagan has all kinds of apple pie related wisdom on this subject and is damn good jumping point.
In classic games you’ve probably seen plenty of common rooms for a purpose. The area before a boss fight is often different than the rest of the level and Megaman is well known for doing this, maybe starting it. The end of the stage meets a wall, an empty sealed corridor, and finally a closed and intimate space to focus on the boss. Tons of old games used boss rooms like this, and in a way it relates to my example of Alice. Boss fights are a confusion of patterns and phases at first sight. You needed to watch where every projectile fell, where the boss stood or what paths it would take. Bosses were complex compared to regular enemies and the room was supposed to be familiar.
What I want to get into more is how the room shapes can quietly evoke a response from the player. Enemies tend to have a simple set of behaviours, they’re programmed to move around the environment and perform their laundry list of objectives when you’re nearby. The easiest way to set them up for a player to learn is usually in a straight line. With nothing in the terrain to bias the player’s experimentation they can just figure out what works. This is important because you will need to express your themes to the player with player advantages and disadvantages through enemies and obstacles. This is all highly dependant on your player mechanics so lets look at a few games including Owlboy.
In Owlboy you can shoot in all directions. Because of this, the difficulty is getting a clean angle on your target. Uneven terrain like the Z block shape can hide enemies in a compact space. Twists and turns require the player to move to specific locations and those advantageous locations are sometimes protected. In contrast, a square room or straight hallway of any angle allows the player a clean shot. However, many enemies incorporate some form of protection and that provides conflict between the shape of the room and the angle of attack. Molts can’t be assaulted from the front, so they’re protected in a straight hall where your movement on a right angle is restricted. When you play the game of triangles, you win or you die.
In 2D Mario games the player needs to gain a height advantage. Enemies are designed to be jumped on, so anything like an uphill slope or a climb adds difficulty. With each attack you can only defeat one enemy at a time and learning to bounce and keep your height allows you to dispose of them a little faster, but there’s always risk. In contrast, being at the top of a slope while enemies trudge upwards presents itself with a huge advantage. The feeling of sliding down the hill to destroy a group in a single stroke is fun and relieving. Vertical climbs and upward slopes provide the most difficulty to the jumping mechanic.
What comes of this is tension and relief for the player as they experience the same behaviour in a different space. In part what happens is that the terrain signifies to the player what they should expect to tackle. An enemy might be presented in its most difficult to defeat position at the start and force the player to retreat, but later is found roaming around without any protection. It takes the tension away and this can be a good thing if you want to use it. There can be triumph at finally having the upper hand if the player has been regularly frustrated. If it’s done right the player should feel as if they fought their way into an advantage point, not that it was arbitrary like an invisible wall. The triumph could be soured on purpose though if that’s the feeling you want to get across.
I love those moments in games when I can finally plow through enemies. Even in small aspects like sliding down in Mario. The satisfaction of mowing down 3 enemies at once doesn’t change anything about the game, doesn’t gain me anything, but I enjoy it so much. In classic Sonic games the same happens when I can ball up and freely roll downhill. I think this is because in both sliding and rolling there is a temporary invincibility granted as a reward for learning to play and surviving. It’s a break in the action to hurry you along the goal post express. What’s also great is when the game pulls a fast one on you (but not an ugly cheap death). There might be a spiked enemy or a pit or something. It keeps you on your toes, but too much of it expresses something very different than before. Early in the game it expresses carelessness. Forbidding the player from enjoy it later expresses seriousness. This is usually a good way to make the game feel more difficult at the end, but it can also kill fun and the player’s engagement. Ultimately I think that these games are designed to apply pressure and disadvantage through most of the basic room shapes so that the few remaining are elating.
As a quick aside, look at Dark Souls (and Demon’s Souls). In a narrow hallway it would be stupid to do wide arcing swings, tactics require a thrust. Fighting up and down the stairs provide their own complexities with the wide variety of weapons. The shape of the room you’re in is of huge importance to the central mechanics of fighting. What I love about this game though is you can’t always be assured of an advantage or disadvantage just because of the familiar terrain. It does however play on what you might think you know about terrain and offers up many surprises.
I think what also makes the use of repeating structures fun to play with is the surprise for your player. Players will clue in to established patterns as some enemies will always appear in a certain room shape. The shape of the room then sets an expectation for the player like the boss room. Sometimes a fresh experience will never appear, likely because it wasn’t fun to play in testing, or glitch on certain terrain, or there just wasn’t enough time. Regardless of fun sometimes it offers a refreshing surprise to be confronted with an underplayed situation. Variation in itself can be fun.
I love secrets and a commonly played room is the vacant lot. There’s nothing in here to get you, nothing to distract you, so is there a secret door? Maybe. Or maybe it’s just a transition device, or maybe the surprise will drop down on you as you pass a certain point. The reasons for the empty room are numerous, it’s all about pacing. Here all of the room shapes built up by the player’s experiences are questioned.
Let’s look at Zelda games since they’ve inspired so much of Owlboy’s design. Looking at the very first one you can see how every room is the same size, this is both for the world map and in dungeons. On the world map these repeating shapes establish a way for the player to map out where they’re at. The player will become aware of the grid pattern and the system of secret dungeon doors becomes obvious yet the exact details remain elusive. The original game can get away with a world devoid of hairline cracks and ruptures because it relies on the player to figure out the way with hints and reasoning with the maps. There’s clear markings in dungeons for doors, but the world map is far more cryptic. Later games are far from uniform and it would be a cruel game without help or a total focus on cyphers (like Fez). The other 2D Zelda games do use a mix of Tetris shapes but I’m trying to be brief. The old formula for secrets transplanted into a new game would require copious amounts of text or added mechanics so the easiest solution would be a crack in the wall. Zelda games use repeating microstructures in the terrain with different enemies or secrets. In later areas the circle of stones and the 4 equally spaced blocks signify the presence of a secret passage. Central barriers provide defensive strategies for enemies. Water rooms provide them with offensive strategies, forcing you to directly confront them to progress. This can be a good or bad thing depending on what the enemy does as an attack. Outside the dungeons are gridded statues and bushes and gravestones. The repetition invites inspection.
Much the same as downward slopes in Mario and Sonic, I begin to hunt for rewards in the form of secrets. Relatively speaking, secrets and treasure are the rewards of enduring an adventure game. I get excited to try out what I know when I come across a room that resembles a secret.
Back to Mario, some of the most famous secrets in Mario are about defying the room. Getting above the ceiling can lead you to warp zones and items. Again, the height advantage rewards the player.
There’s a lot more to be said on the subject of how the shape of a room (and the terrain as I’ve worked it synonymously) contributes to the tension and relief of the player. I only use the terms together because once you divide up the traversable areas of a room, most fit these basic shapes. It’s not a rule I follow strictly, it’s just a way of understanding flow.
I find that room shapes are more impacted by gravity than anything else. I think the reason for this is because gravity forces you to follow the terrain closely. It could be to do with the fact that everyone has to deal with gravity in life and it’s immediately relatable. I’m just speculating on this one though, unless you’re a wizard or magical beastling or hoverboard cyber-knight, gravity can be a problem in life. Really it depends on the player mechanics being used. Friction is probably the best way to describe what I mean, look at how games like Dark Souls work conflict between walls and player mechanics. It’s about figuring out what about your mechanics causes players to struggle and what makes them happy. You might find that by locating grief it can help you retool a space to be as hostile or as friendly as the themes of your game demand.
Oh man, I made it through this whole article without a single Tommy Wiseau joke, Lisa would be proud of me if she didn’t betray me like everyone else. Haeh Haeh Haeh!
I get a number of emails from our fans asking questions about Owlboy’s development, and while I love answering each person when I can, some of these questions are a little too much for an email. Today I will answer a common question about level design and world building. Later I’ll expand this with entries about transitions and theming.
Design is something you can approach from a variety of fields, each with their own strengths and weaknesses. My background is in English literature and it means that I look at everything through a story and thesis lens. Much of what I do involves critical thinking and close readings of material to tease out an interpretation. I don’t however stay within my narrow field. For example, when I took courses in geology I saw the process of building mountains and rivers as a creative story to fill in the data gaps. I find these stories about erosion and weathering to be highly fascinating. No one was there to see any of this and we only have educated guesses. A data sheet charting out these changes in size and water content won’t work for anyone but a geologist. A nature documentary about animals wouldn’t be nearly as interesting without Morgan Freeman’s velvety Morgan Freeman voice. Morgan Freeman~.
Footage without Morgan Freeman would be snippets of data for biologists who understood the details not shown, or we as an audience would interpret that sequence in our own ways.
Stories are important because they package information neatly for others. It’s an efficient way to send information, with context included, so the receiver can understand and interpret it in a meaningful way. However, I’ll never efficiently convey what I want until I hone my ability to tell a powerful short story. I can’t expect every player to sit and think about every detail I’ve provided; that’s if I can express them through the tile art I’m using. You only have so much time and money to produce all the art you’ll want to work with and you need to be as creative as possible with limitations. Make your unique tiles after you have found a rough unrealized potential idea in the scenery. Another problem is that stories can be based on obsolete ideas or ill assumptions. I learned to tell stories about the environment through the sciences, and since these stories can become obsoleted at any time my understanding becomes encoded in that time and place. For example, steampunk worlds are built on Victorian era science and philosophy. The same applies to any culture that happens to see the world differently than I do. It just happens, there’s little you can do to avoid this, but why would you? If the world you create sticks to a set of rules and is consistent enough to suspend disbelief then you’ve succeeded. Pick your rules and stick to them else your environments won’t feel as if they belong together in the same world.
If you’ll indulge me for a little while I have a short story to tell of the Owl Temple (not shown in the old demos). I’ll have more of my design process at the end. I rarely write things out in this much detail. The following is what I try to tease out of the maps as I build and test and redesign with each play through. Most of the time I work from a loose outline of what I want to accomplish and why.
Otus descends the twisting maze of rocky tunnels clutching Geddy firmly with each change in direction. Their path leads to a small flooded chamber, cylindrical in shape. They pass through an opening above the water line. The door was once perfectly crafted, but now it seemed as if time had become too impatient and too large for the business of doors. The great basement hall was open to the eye of the storm on two sides, the ceiling held up with large evenly positioned stone arches. All of them were engraved with faint runes, or was it just the imagination making sense of weathered scars in the awe of this place. Great owl statues stood defiant of erosion but the slow relentless dripping has claimed their numbers. The collapsed became a thousand scattered islands in the crystal clear lake under the temple. Moss clung to whatever had not been claimed yet.
Molts howl and ready the fires deep in their bellies for the flying intruders. Otus dives and climbs quickly. Flashing amber coloured beasts pull themselves from their watertight burrows. Clouds of smoke and fire erupt from their mouths obfuscating Otus’s eyes and nose. They push through into to the stairwell at the end of the ruins.
The walls in this place lay bare, the stairs resting along the floor looked as if they were never raised, but the door at the top remains waiting patiently as ever. There isn’t much light in the hallway above. This new place is heavy with shadows and vines. The odd pink vines have taken full advantage of this to strangle the temple rocks further. Cascading water droplets are flung from Otus’s wings as he flies on. Geddy shakes the moisture from his hands to reaffirm his weapon’s grip. The vines are getting thicker and harder to see through. Bulbs growing from some of the vines glow with a pale luminescence, a sickly cold imitation of the sun. At the end of the hall is the familiar rubble that had blocked the path earlier. They head through another hole in the ceiling.
The vines here carpet every surface, working up around any protrusion from the floor. The two come across a large purple pod at the heart of this confusion of vines. This might be the heart of the strangling plant, or one of many. Withered vines create a stubborn mesh around a spot in the ceiling. Geddy suggests to Otus that he might loosen the mesh with water. Otus rummages through the pile of nearby vases. Most all of them are cracked and hold no water.
Otus returns from the small pool and releases the vase’s contents. It worked, the dry vines swell and surge to life for a moment. The dislodged rocks bounce freely only to be caught once more between the vines on the floor.
The dark passage is narrow, wet and rocky. Darkness expands the otherwise small space. The ground crunches and snaps under foot. Scattered everywhere is an unrecognizable mess. Bones. They’ve been stripped clean and don’t belong to any one person or creature. Otus and Geddy disengage their interest for a moment. An ominous carved stone gate hangs above ]them looming like a cold finger wisping the backs of their necks. The pitch hole inhales slowly without pause. Reluctantly they continue on without any other options.
The air is quiet and afraid to move. Water streams quietly as it can down the back wall escaping through the cracks. A sickly green flame grows in the distance. It dances atop a short pike held by a tiny round creature. Gurgles of hunger and splashes of saliva erupt from a mouth already too full of jagged needle teeth. More of them awaken higher up. The gate seemingly takes advantage of the moment and closes tightly. High above a spec of light beacons the only way out alive.
By now Geddy has already hidden himself. Otus begins his climb, too afraid to fly for his wings will betray him with both sound and smell. There are dozens of these snarling creatures and dozens more unseen. After much climbing the last platform is under his feet, but the exit is still a great distance higher. Otus will have to fly hard and fast or they will catch him with their teeth and each will have it’s own direction to go with that mouthful in the dark. At the very first flap of his cloak the pack is sent into bloodlust. The scrambling terrors gain on him. The snap Otus hears is the stone gate behind him. Otus fumbles for the Owl relic and brings Geddy to his side.
Two great waterfalls crash down down behind them, the source of all the flooding below.
Usually when I redesign an area I take into account where it is in immediate relation to others and the whole map. As I come up with something I look at what I need to change to transition the rooms smoothly. When I began redesigning this part of the map I needed to do something about the pieces already set in place. I had a gnome room and the waterfalls above. I needed a way to connect these pieces in a coherent way. There was nothing to capitalize on these pieces. I decided to tell a short story of a great hall that had been destroyed by water and time. It may not be much of a story really, but a flooded basement isn’t in itself visually striking or interesting. It makes sense to put a great lake below all this water, there is no way it would remain watertight after the uncounted years of existing.
This all sounds really simple and basic when I explain it like this, and it is, I don’t mean to beat anyone over the head with it. The rooms in your game are a part of a larger structure be it a temple or lava mountain fortress or a spaceship or a cake or a wizard’s beard. Keep in mind the macro structures and decide on what kind of rooms should exist in such a place. Once you’ve figured out a loose of idea of that, begin linking these places. Write out the room ideas you have. Overall shape can help draw out themes. Also by joining a few areas at a time it can help create direction and small goals for the player. Hopefully if done right by giving an area purpose it becomes a landmark to remember and all of this working together will orientate the player in a confusing dungeon.
Also, some rooms have elements that follow a natural direction. Water falls down, easy. But what about a structure with materials being piped in and out. Symbolic structuring requires you to think before you lay these rooms down, or choose from the most appropriate ideas in you sketchbooks. These things get complicated quickly so it’s best to leave the complexity to the sum of the whole structure. Be elegant and use the smaller groups of rooms to build your grand plan.
A huge issue I call “tubes and boxes” is when you have all these ideas, but the means to joining them are passages of nothing in them and no purpose to the downtime. A breather is fine but the game’s pace becomes erratic, spikes of life and flat lines, evenly pulsing, on and off. So. Boring. Tubes are a sign of a poorly organized map. In stories you have something to develop in the downtime, or you leave it out with a good transition. RPGs are able to hide behind a wall of walking between key locations with a smattering of battles. There’s rarely any idle dialogue between characters for these long walks, but I’ll talk more about transitions in another article.
There’s no formal rules for how to do all of this, you just have to be a practised designer and work this stuff out in your head or paper or excel sheets or with other people or whatever works best. I work these problems out through stories and graph paper, but many of the details will be uncommunicated so I need to stick to what actually comes across. Help your player out and give them a reason to remember where they are and what they should be doing. Don’t just resort to a map conveniently located in the dungeon.
What I want to achieve in my level designs are not necessarily obvious to the player. I want to make subtle collections of logical structures that help immerse you. I want this place to tell you about the world you’re in and that it existed long before you came tromping in and will continue to later. Environments can be characters too and by developing your environment you create situations that can only happen in the world you’ve created. Puzzles and action sequences become memorable and unique because they’re bound to your world’s rules.
Thank you for taking the time to read this.
.png){kind=link}