Splitting Fiction: The Hardest Level Hazelight Studios Ever Built

Hello everyone and welcome to designing the final level of split fiction. My name is Hana and I am a designer at Hastelite which means that I do a combination of game design and level design and I worked on the final level of split fiction which I will be talking about today. kind of goes without saying, but spoiler warnings. If you fully want to avoid spoilers for the ending of the game, you have about one minute to leave the room. And I don't see anyone leaving, and that makes me very happy. Okay, why have I decided to talk about the final level of Split Fiction? Well, to start with, it was the highest rated level in the game in every single external play test that we did. and it has a unique gimmick that was initially meant to be used in the entire game.

This gimmick came with a lot of unique design challenges and forced us to think about co-op design in a different way. Okay, who in here has played split fiction?
That's quite a lot of you. It makes me very happy. U but for those of you who haven't, this is what split fiction is. So, Split Fiction is Hazellight's third game. All of Hastight's games, A Way Out, It Takes Two, and Split Fiction are co-op exclusive storydriven games, which means that they are specifically designed for two people and can only be played as such, and they follow a story from A to B. Now, to get you an idea of what Split Fiction is, I will now show you a small gameplay teaser for the game.
[cheering]

3 4 Engaging fire.
Now that a big boy shot me. A shot.
Attack formation.

[laughter]
It's showtime. In split fiction, you play as two writers. Mio writes sci-fi stories and Zoe writes fantasy stories. In hopes of getting a publishing deal, they partake in an experiment that throws them into a machine simulating their stories to life. But something goes wrong and they get stuck within their own worlds. Throughout the game, you play through their worlds, jumping between sci-fi and fantasy levels to try to find a way out of the machine. Okay, what is the final level of split fiction? As I said, throughout the game, you alternate between sci-fi and fantasy levels, but in the final level, the

machine simulating their stories breaks down. And now one is in a sci-fi version of the world and the other in a fantasy version. The final level is split into two parts. The first part is this split world concept and the second part is the final boss fight of the game with a lot of different crazy gameplay segments. I worked on the first part with the split world concept and the design process and challenges with designing this part of the level is what I will be talking about today. We will look at the origins of the concept and how it was initially meant to be used in the entire game. We will look at how we found the gameplay building blocks to design the level with. We will look at how we iterated on

puzzles to find the ingredients that would make them engaging and satisfying for both players. We will look at how we tried to encourage the unintuitive behavior of screen peaking. We will look at how we tried to create visual contrast between the worlds and how we had to design the level within the technical constraints set by how the level actually works. And finally, we will look at how we introduce the split world concept in a way that gradually reveals the different behaviors of the level. After the presentation, you will have learned how we managed to create a shared experience even when the players are in two different worlds. How we try to highlight what we think is cool about the concept. And you will have learned a lot of the design techniques we use when designing our co-op games. Okay, to start with, let's look at the

origins of the concept. And for that, we have to go all the way back to 2020. So that's my boss. The idea of playing together in two different worlds was actually the original pitch for the entire game that my boss Joseph Harris presented to the team in December of 2020, a few months before the release of It Takes 2. Here you can see some concept art from that initial presentation. And as you can see, it is quite similar to what the final level became. The players are in two different worlds, but compared to in the final level here, the idea was that they would have a portal that they could use to throw objects through to help each other out in the different worlds. With this concept art as inspiration, the team started prototyping different ideas, trying to utilize the concept of being in two worlds. And I will now show

you some videos of prototypes made during the very first months of pre-production. So here we can see a prototype using the portal idea. This was very early and we were using a copy of the it takes 2 project to be able to start prototyping as quickly as possible. Here is another prototype. And here the players are sort of in the same space but it looks different in the two versions. And here they also have something like the portal kind of to swap parts of the environment between the two worlds. Here is yet another take on the portal. And here we can see that the it takes

two characters have been swapped out with characters from a way out his life's first game. to more closely resemble the human main characters of split fiction. And finally, this is the prototype from pre-production that most closely resembles how the concept ended up being used in the final game. The players are in the same space, but one in a sci-fi version and the other in a fantasy version. And now they can also see each other, but compared to in the final level here, they see each other as a sort of ghost representation. And we thought this was really cool, but why was it in the end only used in the final level? Well, something that the team realized early on was that this was very

difficult to do. It was especially challenging for the artists since they basically had to art two games instead of one. And it was actually even trickier than that since the art in the two worlds had to align with each other with a high degree of precision. But it was also challenging design-wise, especially to create a shared experience when the players never saw the same things for an entire game. So, in the end, it was saved for the final level. And saving the concept for the final level made it less expensive and gave us a strong ending to the game. limiting the concept to one level turned it into an expensive level instead of an expensive game. And in the same way, limiting it to one level also really reduced the issue of the players feeling disconnected from each other. And I also think that the impact of

being in two worlds at the same time, sci-fi and fantasy, was heightened by the fact that the players at this point had become accustomed to only being in either sci-fi or fantasy for over 10 hours. Okay, next up, we will look at when I started working on the level and how we found the gameplay building blocks to design the level with. So I joined Hastite in 2022 and that was at about the same time as the team started working on the final level. The designer mainly responsible for the level at this point was my dear design colleague Gustav Linder and he prototyped a lot of different crazy gameplay segments with the help of our technical director. When we start working on a level at hastlight, we don't get this multi-page outline from Joseph of exactly what he wants the level to be. He rather tries

to act as inspiration, but gives us great autonomy and ownership over the segments. He lets us make something that we feel inspired by. And we then present these different ideas to Joseph and he gives feedback to make sure that what we are creating aligns with the larger vision for the game. So I started working on the final level about a year later. At the time there was only Gustav working on the level which had become quite ambitious. So it was clear that we needed more designer hands on deck. When I joined the final level, there was already an outline for the structure of the level. This is um uh actually an image from our mirror board of the actual level flow that we had at this point in time. The different gameplay segments that Gustav had made, some of them completely new and some of them inspired by the work done during pre-production had been distributed across the level, most of them as part of the final boss fight.

So, I sat down with Gustav and we discussed what was already there, what needed most work, and especially what I felt most inspired by. to Hazelite. This is really important because we demand a high level of creativity from our designers. So, feeling inspired about the work we're doing is crucial. Well, the part that got me most excited was the start of the level, the yellow part. It was meant to be using the split world concept, and barely any work had been put into it at this point. There was also a twist to the split world concept at the end of this segment which I also worked on and I will briefly talk about it at the end of the presentation. So when I started working on the level, this is how it looked and it was actually not much more than the room you see on screen now.

And one of the first things I often ask myself when I start working on a level is what are the unique mechanics of the level? I try to find the gameplay building blocks to design the level with. And to me, one of the most obvious and fundamental features of this level is what the players see. And this felt like it could be used as a mechanic since splitting the information between the players would force them to communicate. As an example, we have these two pressure plates that are very obvious in the sci-fi world, but they are not as obvious in the fantasy world. So, the sci-fi player has to tell the fantasy player where to stand. The second feature is objects being linked. I see these two worlds as sort of distorted mirrors of each other. So doing something in one world would naturally make something happen in the

other. And if we like in this case activate a mechanism in the sci-fi world, the fantasy world will try to interpret what is happening but not necessarily in exactly the same way. For the third and final feature that became the foundation for most of the game play, we will actually look at something that a lot of hazelite levels have in common, and that is letting the players interact with the world in different ways. Something we often do is give each player their own unique ability that allows them to interact with the world in a way that their friend can't. And this has a lot of nice design implications. It makes it very easy to force collaboration. It gives each player a unique role and identity and it really highlights that

this is not a single player game. Well, if we look at the final level of split fiction, allowing the players to interact with the world in different ways seemed natural since the players are not even in the same world. As an example, we can create machines that only the sci-fi player can operate because it doesn't make sense to interact with them in the fantasy world. Like in this case where we have a lever opening a door, which in the fantasy world is just a branch being broken. And the same goes for the fantasy player. Here we have water that only exists in the fantasy world. So the fantasy player can swim in the water to grab this floating/hovering platform and help their friend get across. So to summarize, the building blocks I identified for this level were what the players see, objects being linked, and

the most important one, letting the players interact with the world in unique ways. We identified the unique features of the level and found elements that had worked well in the past and these mechanics provided us with a strong gameplay foundation to design the level with. Next up, we will look at when we started to create gameplay with these mechanics and how we tried to find a satisfying co-op experience. So, one of the first puzzles I created looked like this. There is a robot arm/carnorious plant, and the plant is trying to bite the fantasy player. In the sci-fi world, there is an energy field blocking further progression, and the energy field is powered by a cable. The cable has a part that is exposed. And the goal with this puzzle is for the sci-fi player to tell the fantasy player where to stand to make the plant cut the

cable, disabling the energy field and allowing them both to progress. And I really like the idea of this puzzle. But in play tests, it had a few issues. So we will now look at an external play test. And I recommend you to focus on what the fantasy player is doing. And I guess you're not laughing because this is a great level design. This was absolutely terrible. A lot of players were accidentally solving the puzzle by the fantasy player just happening to be in the right spot, which left them really confused about

they were what they were just doing. And the puzzle was also a bit too simple. It only had one step to solve it. So even the people who solved it the correct way were not very impressed. So we had to go back to the drawing board. We wanted the players to not accidentally solve the puzzle, of course, and we also wanted to add some complexity to make it a bit more interesting. And I thought I had the brilliant idea. of adding a second cable. Hear me out. This is actually pretty good. So, it it would be very unlikely that the players would accidentally cut both cables without knowing what they were doing. And I also got the opportunity to add a twist to the second cable. The easiest way for the sci-fi player to tell the fantasy player where to stand was to stand there themselves. So, for the second cable, I added a danger zone around it to force

the sci-fi player to stand on the side and communicate to the fantasy player where to go. And when I tested this, the players were not accidentally solving the puzzle, but they were still a bit underwhelmed, and it didn't feel like both players were equally engaged. And I realized that there was actually a thing that both versions of the puzzle had in common. One player had the information of how to solve the puzzle and the other player was executing it. And in practice, this meant that the sci-fi player was just telling the fantasy player what to do and there was no back and forth communication. The fantasy player was not part of solving the puzzle and the sci-fi player had nothing to interact with which left them with an empty feeling. So, what I wanted to do was to give both players information and execution.

And I tried some different things and the puzzle went through a few more iterations. But I will now show you how it looks in the finished game. So, the first thing you will notice is that the plant is no longer automatically attacking the fantasy player. Instead, to give the sci-fi player something to interact with, I created this panel where they can press a button to decide when the plant will be attacking. The plant is still tracking the fantasy player at all times, but now to cut a cable, both players have to be involved. The fantasy player has to be in the right position, and the sci-fi player has to activate the bite. And this also really made sure that the players were not accidentally solving it. Now, to add some information to the fantasy player, I changed the twist of the second cable. I added some thorns around it that only the fantasy player could see. So, they would have to realize that they had to jump over the thorns and make the sci-fi

player cut the cable midair. And this version of the puzzle was way more appreciated by the players, both because they were communicating more and because they were both now mechanically involved. But there was also a third thing making the ending of the puzzle way more satisfying for the players. When cutting the second cable while jumping over the thorns, there was an element of timing and timing is a strong tool that we use throughout the game to create an intense feeling of collaboration. It requires the players to be actively in sync and there is also a tension building in anticipation for the execution. So these three ingredients, splitting the information, splitting the execution, and finishing with timing,

seem to create a very satisfying co-op puzzle. And I will now show you how a second puzzle that was also very appreciated by the players uses these same ingredients. So the goal with this puzzle is to get through this entrance blocked by a gate on the sci-fi side and a fire on the fantasy side. The sci-fi player will find this turret that they can operate, letting the fantasy player know that they can shoot rockets. Information on the sci-fi side. The fantasy player will see these pots being transported along a rope over the fire. And these pots are only visible to the fantasy player. There is no visual representation of them in the sci-fi world. The fantasy player might get the idea to ask their friend to align their crosshair above the gate to try to shoot one of the pots. And if they do so,

nothing will happen. But if they follow the rope backwards, they will find this lever that they can pull to pour water. and they now each have part of the information required to solve the puzzle. And if they communicate, they might understand that they can fill one of the pots with water and then shoot that pot as it crosses the fire to put the fire out. Now for the execution, the fantasy player fills one of the pots with water and then quickly repositions themselves to be able to track the pot as it's heading towards the fire. The sci-fi player aligns their crosshair above the gate where the pot will pass and get ready to fire a rocket as soon as the fantasy player tells them to. And this is of course the element of timing. Oh shoot. Completing the puzzle if both players are successful in their communication and execution.

So giving both players information, execution, and finishing with timing created satisfying co-op gameplay. This specific formula is of course only applicable if, let's say, information is part of your toolkit. But I think the most important point is that we want to give the players a reason to communicate. We want them both to interact with the game. And we want to build a shared tension that is then released to give the players that satisfying high- five moment. Next up, we will look at how we tried to encourage screen peeking. So to make the players experience both worlds, we wanted them to look at each other's screens as much as possible. And this was something that we had a unique opportunity to do since the entire game can be played in split screen no matter if you're playing couch co-op or online.

And this is not because of the split world. This has been the case in every single Hazelite game. Being able to always see the same things as your friend creates a stronger feeling of a shared experience closer to that of couch co-op, even when playing online. And I thought that in this level, the players would surely want to like utilize this opportunity to look at each other's screens as much as possible to compare the environments. But making them look away from their own screen was trickier than we anticipated. So something that I was trying to do with the first versions of the level was that I had created a lot of traversal segments between the puzzles. And the idea was that the players would have time between the puzzles to look at each other's screens and compare the worlds. I had created a lot of traversal elements that worked the same in both worlds, like the fan launching the players in sci-fi and the geyser launching the players in fantasy. And

the idea was that they would like talk about these things. But when I watched play tests, I realized that as soon as the players had basically any gameplay whatsoever, they were looking at their own screen. which made sense. If they were doing something, they had to see what they were doing. And if they would be looking at their friend screen, they would most likely not be seeing similar views either. They would more likely be seeing something like this, being in two completely different spaces and maybe even looking down into the ground. Once again, it had to be scrapped. So, we wanted the players to do the unintuitive thing of looking away from their own player character, and we wanted them to see a similar view when looking at their friend's screen. This took a lot of iterations, but I think we managed to do it quite well in the end, and these were some of the things we did. So, first off,

we couldn't have tricky traversal. It required way too much attention from the players and didn't allow them to look away from their player character. Instead, the traversal between the puzzles had to be simpler and less mechanically engaging. And here are a few examples. We have this bridge that the players have to cross to get to a puzzle. And here there is basically no gameplay, so they can look away from their own screen without the risk of falling off a ledge and dying. The bridge is also quite long. It takes a few seconds to cross the bridge. So, it will be more likely that both players will be on the bridge at the same time, and it will be easier to compare their views. We also have this flower/propeller that the players ride on to get to the next area. And here there is no gameplay. They are not controlling the flower. And since both players are in

exactly the same spot, it also makes it a lot easier to compare the two worlds. So the second thing we did was to add what I call Zelda cameras. Sometimes when you are in a Zelda dungeon solving a puzzle, the camera will leave the player character to frame a door opening or a chest being revealed. And of course, these camera moves are not unique to Zelda, but that's where I got the inspiration, and this is very nice. It both allows us to remove movement and camera controls from the players, and it [laughter] okay, it also allows us to frame the environment from exactly the same angle, which is the ideal setup when comparing the two worlds.

The third thing we did was to add dialogue between the player characters. And this was done throughout the level. And the dialogue would not only highlight when to look at your friend screen, but also what to look for. I will now show you some examples from the game. And you will notice that the dialogue is a bit on the nose. But in this level, it was really important to us to highlight what we wanted the players to notice. Oh, the trees are spinning. Trees. Wait, wait. How are you jumping like that?
It's a bouncy mushroom.
Weird. Weird.

I've got a carnivorous looking plant gazing at me.
What? Be careful.
Don't jump. No, Zoe. No.
What? How do you Never mind.
So, these three things, making traversal less engaging, framed camera moves, and dialogue between the player characters, were used throughout the level to highlight what we thought was cool, and encouraged the players to look at each other's screens. Moments of reduced player engagement allowed the players to look away from their player character. We really underestimated how difficult this would be and we had to use pretty blunt tools to achieve it. But as a result, the players looked at uh they looked at each other's screens which improved their experience. So I think it

was definitely worth it. And to make the effect of looking at each other's screens as strong as possible, we wanted the two worlds to be as different as possible. So next up, we will look at how we tried to create visual contrast between the worlds. So there was a lot of thought put into the choice of environments. The worlds both had to be similar and as different as possible. It had to be possible to translate shapes between the two worlds, but you should also be able to see the contrast even when squinting your eyes. Early on, we also had some issues of making these environments distinct from the rest of the game. A lot of players thought that they had returned to a previous level, which gave them the wrong impressions of what the final level would be. A brutalist sci-fi city with tall and tangled structures seemed to be matching quite well with a big fantasy forest

with large mushrooms and trees and ancient ruins. We wanted the contrast of the cold and man-made sci-fi against the lush organic fantasy. To easier orient the players when looking at each other's screens, we wanted the broad strokes to be similar. The big structures would use similar shapes while the smaller details and the feeling you got when looking at the two worlds would be as different as possible. Now to fully understand the challenges with creating visual contrast between these two worlds, we have to understand how the level actually works. So this is the fantasy world in the editor. And if we fly away from the playable space, we will see the sci-fi world exactly 5 km offset. With the press of a button, we can swap

between the two worlds in the editor. And we also have a really cool tool to show the other world as an overlay to make it easier to align the geometry. One more time. Okay, now now to the main tool used to sell the effect of being in the same space. Your friend is not your actual friend. They are just a mesh playing the same animations as your friend 5 km offset. So if the geometry doesn't align, the players will either float midair or clip through the geometry. So to not break the illusion, the collision in the two worlds had to be matching very well. And this was not easy, especially during the times when we had different artists working on the different worlds at different times.

As soon as an artist had added an element in their world that was difficult to justify in the other, it had to be removed. And it was also really tricky to create two very different environments using almost exactly the same geometry. But these were some of the things we did. So, first off, the use of colors. Lucky for us, colors don't impact the collision. And it is one of the most obvious things you see even when squinting your eyes. The sci-fi world would use a lot of blue and pink and colder colors, while the fantasy world would use a lot of green and warmer colors. And this was done through a combination of materials, lighting, and post-process. The second thing we did was to utilize the backdrops because this was an area where the players could not be at all and where the collision didn't matter. So here we could have very different geometry. And we also made sure to add distinct

focal points in the two worlds like the pink planet and the big frog that the players could talk about and compare. Another thing that didn't impact the collision is foliage. So, we added a lot of foliage in the fantasy world to give it a more soft, organic feeling. For the collision in the playable space, we made sure to give the objects slightly oversized, bulky collisions to give the artists the opportunity to give the objects slightly different shapes in the two worlds, but maintaining the same collision. And we made almost all of the objects in the playable space tall enough so that you couldn't stand on top of them. And this allowed us to give the objects different silhouettes without the risk of letting the player stand on top of the oversized collision. Another thing we realized was that it was easier to match the fantasy world to the sci-fi world than the other way

around. It was easier to have flat collisions in an organic environment than to excuse organic shapes in the man-made sci-fi world. So, this is an area in the level that is very organic in the fantasy world. And I will now show you how this area looks from exactly the same angle in the sci-fi world. And as you can see, it is not at all as organic. If we go back to the fantasy world, we can now look at how the collision looks in the fantasy world. And as you can see, it is much closer to the geometry in the sci-fi world. And this is how it looks in game. So you can also see on the sci-fi screen how it looks for them. And you will notice this small frog in the fantasy world that excuses the collision of the pillar that sticks out in the sci-fi

world. And I think this really works. This is also me torturing the level. Like most players are not hugging the walls to look for imperfections in the collision. But yeah, to make the worlds as different as possible without breaking the illusion of being in the same space. These were some of the things we did. We used colors. We used the backdrops where the players could not be, foliage, and other things that didn't impact the collision. And we gave both worlds slightly oversized collisions to give uh the artists the opportunity to give the two worlds slightly different shapes. We added as much contrast between the worlds as possible without impacting the collision in the playable space. So when you looked at your friend, they had to reasonably fit into your environment. And we realized that the collision matching the collision was the most important thing to achieve that goal. Okay. Finally, we will look at how we

introduce the split world concept. Since the concept of being in two worlds at the same time was very unique and the effect very trippy when you saw it for the first time, we wanted to gradually reveal the different features of the world to keep surprising the players and allow them to take in each element of the concept one at a time. One of the first ideas that my dear boss Joseph Harris had even before I started working on the level was that the players would start the level separated and not know that they are in the same world. They would then do a bit of traversal by themselves and eventually get to the same area where they would get this mind-blowing surprise of realizing that they're actually in the same world. And creating this moment seemed pretty simple. But it was actually pretty challenging to do. The first thing we had to do was to

funnel the players into a small area where they would not be able to progress further without their friend. And that part is actually very easy. We do this quite a lot throughout the game. We have a lot of side interaction areas or traversal segments where we let the players get through it at their own pace, but where we then want to reync them for the next gameplay segment. And something we do then is to add what we call a double interact. And a double interact is a progression blocker that requires both players to get through it. It can be anything from a double door with two interaction points to a puzzle that requires the players to use their level specific abilities. Most double interactions are very obvious as double interactions. We don't want the players to get stuck and confused. We want them to know that their friend is required to proceed. But in this level, we actually don't want the players to

know that their friend is required to proceed because that would kind of spoil the surprise of being in the same world. So, what we wanted to do was to create an invisible double interact. One of the first things we tried was to create an event that would trigger when both players had reached the area. When the first player got there, nothing would happen. But when the second player got there, we would trigger an event that would eventually lead to them both being able to progress. This is actually how the level looked when we had this event, but I haven't managed to find video footage of it. So, I have instead tried to recreate it. And this is kind of how it looked. And the players would then be able to use this antenna/tree as a bridge to progress forwards. And this seemed like an okay idea. It would be completely invisible as a

double interact. And there would also be an action-packed event happening which fitted quite well with the tone of the rest of the game. We would also highlight that objects were different in the two worlds like the dragon being a ship and the tree being an antenna. But this didn't work. there were too many things happening and the players were really not able to take it in, especially since they were in the middle of realizing that they're in the same world. And worse than that, this event was kind of taking attention away from that cool moment of realization. And we really didn't want that. So once again, back to the drawing board. We wanted to make the interaction something that would not be obvious as a double interact for the first player arriving. We wanted it to be simple, not even a puzzle, because we didn't want to take attention away from this code moment. And we wanted it to be something that

the players could digest at their own pace without being interrupted. And we thought long and hard about this and we realized that that there was a tool core to the concept that we could utilize and that is what the players see. So we created one of the most obvious double interactions in the world, two pressure plates. But in the fantasy world, we masked this as two tree stumps which would not have the same affordance of a double interact. We then made the traversal segment for the fantasy player a bit shorter to make it very likely that they arrived first on the scene. They would only see two tree stumps and not expect to soon meet their friend. And when they found each other, they would have this uninterrupted moment of realization that they could digest at their own pace. And when they were done with all of that, the sci-fi player would let the fantasy player know to stand on the buttons. And that leads into how we

introduced the second feature, objects being linked. So standing on the buttons would activate two crates that started rotating in the sci-fi world. And it didn't make any sense that the trees in the fantasy world would rotate because you stand on two tree stumps. But this was a feature, not a bug. I promise it. It forced the players to communicate or look at each other's screens to make sense of what was happening. And that's what we wanted. Now, I will show you a clip of someone playing this in the released version of the game, and you will see how all of the things I've talked about come together to create a strong start to the level.
Okay, so you're in fantasy and I'm in sci-fi.
We can meet up.
Oh, this is really weird now that we played together for so long.
The fiction is true. I
kind of want to eat this place except for this one.

So we don't depend on each other anymore.
I guess not in this section.
There's froggy. Oh, look. Dude, I al Wait, is that you?
Froggies.
There you are.
What the hell has happened to the frog?
Wait. Oh my god. This is
Oh, it's the same. No way. This is same.
It's the same but it's different. So cool. Okay, that's pretty mind-blowing to be honest.
That is pretty dank. How is this going to work?
So, we're going to help each other. Wait, helps.
Oh, the trees.
Trees.
So, the final thing I wanted to introduce at the start of the level was that the fact that the players could interact with the world in unique ways. And for this, I wanted the players to interact with the same object because I

wanted both players to be able to see both player characters on both screens. And I wanted the way you interacted with object to only make sense in your world. And I decided to go with something very gametropy. A bouncy mushroom and a pushable crate. And pushing the mushroom looked really weird. and bouncing on the crate looked weird and it created this comedic and trippy moment and made the players talk about what was happening. Now, something that all of these interactions at the start of the level, the uh pressure plates, the bouncy mushroom, and the pushable crate have in common is that they are very simple. Something I think you're always trying to do as a level designer is to fight for the players attention. And by using things that the players recognize, things that are by themselves very unimpressive, we let the focus be on the new ideas.

And speaking of new ideas, as I promised at the start of the presentation, I will end this presentation by showing you how we twist the concept at the end of the segment and especially how we introduce this part of the level. So the big thing throughout the level so far has been that the players are separated. They are in their own worlds. But there was a prototype made in early pre-production where we aligned the cameras essentially turning them into a full screen camera which allow the players to walk from one side of the screen to the other to move between the two worlds. And this was made by my dear colleague P who's here within the audience with us today. Applause for Pat. This seemed like a very cool way to end the segment. This is how the prototype looked when I started working on this segment. And as

you can see, it is not only the players crossing the split screen, but also other objects seamless seamlessly transitioning from one thing to another, which was a really cool visual effect. And instead of trying to explain to you how this works in the final level, I will just show you a playthrough of it. It's only about one and a half minutes. Did he just THROW A BLOODY PLANE at us? This is going to get ugly.
Here we go. Helicopter incoming. Woo! Dragon, hold on.

Of course, it's fire breathing. Not good. Not good. My kingdom, my rules.
[applause]
What a nice audience. [laughter] Thank you. And the players also thought that this was really cool. But what actually got the biggest reaction from the players was not any of the things you just saw. It was how it transitioned

into this segment of the level. It got the biggest reaction in the level and dare I say maybe throughout the entire game. Okay, to start with, the initial idea was that the segment was going to be entirely action-packed with a constant forwards momentum. But a lot of players didn't realize that they could move over to the other world and instead ended up dying on the first obstacles, which was a really awkward and bumpy start to the segment, which should have been a wow moment for the players. So to fix this, I added a calmer moment at the start of the segment. I put the players on a bridge and the bridge was broken in the fantasy world. So the only way to progress further was to eventually realize that they could move over to the sci-fi world and cross the bridge that way. And this completely changed the experience for the players, especially for those who were previously unsure of what to do. They were now surprised in a

positive way instead of thinking how the hell was the game expecting me to know that I all of a sudden can move between the two worlds. And this surprise reminded me a lot of the experience that the players had at the start of the level when they realized that they were in the same world. To me, that was the strongest part of the level so far. And I somehow wanted to recreate that experience here. Something I thought worked really well with the start of the level was how understated it was. Instead of showing the players in a cutscene exactly how the mechanics work, we let them just discover it by themselves. And by removing all of the unnecessary distractions and making the moment quite simple, I think we gave them more room to be blown away. So, what I wanted to do was to align the two cameras and keep the players in their separate worlds throughout it to not spoil the surprise of being able to cross the split screen.

And I got the idea to do this with a classic simple double interact. A double door allowing me to align the cameras during the opening animation and keep the players in their separate worlds throughout it. This worked really well. And to prove it, I will now show you some people playing this in the released version of the game.
He's being awfully quiet. Still no glitches.
He's definitely a boss somewhere. Do you think we're going to kill him?
I hope so.
Get arrested for man.
Look at that.
Wow. That is
What the And can I Oh my god.
Oh.
Oh, weird.
That's one of the craziest things I've seen a game do. Oh,
what? That is so cool.
Otherworldly, bro. Wait. Oh my god.
Wait, it's both.
Oh, I can Wait, can I be half Look at my

outfit.
Oh my god,
it's This game's got
This game's got my uh vote for game of the year right now. What the freak? expedition. [laughter] Of course, not everybody plays the segment like we want to.
Does that mean I can go?
Damn sure not.
Wait, how did you come over here?
What? Wait, what? We can just run through each other's world.
Oh boy.
This is crazy.
Creepy, bro.
Like, how do you even create this? the game. Like, whoever designed this level is a genius. [applause] [cheering] [applause]
It's his words, not mine.

So, for the final takeaway, reducing complexity can give the players more room to be blown away. Something I think is always very tricky is to direct the players attention. And especially when creating co-op games, when we want the players to notice something, something that we think is really cool and that we might have put a lot of effort into, the players might be talking about their plans for the weekend. So removing all of the unnecessary distractions is often very important to us to give the players the intended experience. This was for sure the most challenging level I worked on in Split Fiction, but it was also the most rewarding when seeing the players reaction to it. I have gone through some of the challenges we faced and how we solve them to make this the most appreciated level in the game. Now, one of the worst things with working at Hastlight is that there are

fewer co-op exclusive games that I can enjoy as a gamer. So, I pledge to you, dear video game industry, please make more co-op exclusive games. It is such an unexplored genre with lots of potential. And this level that I have talked about today is just one example of what is possible to do and only possible to do in a co-op exclusive game. Thank you for listening. [applause]
[applause]
So, [applause] okay, here is a summary of the talk for anyone who wants to see that. And please remember to fill out the the survey afterwards. Also, if you're leaving the room, we probably I guess we have time for how many minutes do we have? Five minutes. We have some minutes for questions. If you're leaving the room, please be a bit quiet so that we can hear the questions from the audience as

well. Uh but yeah, we'll go to questions. I guess you can go to the mic stands as well if you want to ask questions.
Uh I guess we are ready to start with questions. So uh who should we start with? Let's start with the right side.
Hey, how's it going? Um I just had a quick question that I've always kind of wondered about at your studio and how other studios do it. But at Hazlet, when you guys come up with specific level mechanics that you can build the level around,
are you thinking more about the narrative first or you do like an action blocking phase in your pre-production to like kind of come up with ideas? Like how does that work?
Yeah, great question. Uh, so when we come up with mechanics, it's often like it's a mixture. Sometimes I know for like it 2 for example, Joseph had a lot of ideas of trying to utilize time in a way that would both tie into the narrative and the gameplay. But for most of the gameplay, we are just creating it like we during the first months of pre-production, everyone just gets to create whatever they want that fits

within the context of the world. And then like we we just do very quick iteration and find what things that are most excited trying to present these different ideas to Joseph and to the rest of the team. Uh and there we just kind of decide what is most fun. So I think for us generally like the fun aspect and the gameplay aspect is what is most important to us. I appreciate that. Thank you so much. Um,
okay. Other side. Thank you.
I have two questions. Um, what was the inspiration of breaking the barrier of the two worlds when people the players are able to cross it and what were the technical implementations
uh of the specific were they crossing the split screen?
Yes, when they're crossing the split screen.
Uh, so when they're crossing the split screen, uh, I mean, I don't know, you could probably ask T afterwards. He was the one who came up with it. Uh but I think it was a bit inspired by the initial idea of using the portal and things moving between the two worlds. So that was sort of the inspiration. Uh but it is actually quite fun. This segment of the level works differently than the segment uh previously where they're just

in like the two worlds. So in the first segment as I showed the players are like playing in their own worlds 5 km offset. But in the second part both players are actually playing in the sci-fi world. So the visual like copycats, they are in the fantasy world. And so how I designed that level, for example, when they're going into the water, I just created a swimming volume that is like aligned precisely with the split screen, but it is actually in the sci-fi world. Uh so so that was like and and then during that opening animation, we actually do a teleport during that cutscene where we swap the fantasy player position. Uh yeah, I hope that answers your question.
Thank you.
Thank you so much. Okay. Yeah. Hey, uh I'm thankful for for all this level design. Like I just finished split fiction last week and it was super great.
Oh, great timing. Thank you.
Yeah. I have a quick question on the narrative side in a sense for the final level in a sense like I when I was playing with my my girlfriend, right? On one side, we're super exciting for the final scene, but we're kind of confused

and having some thoughts on does it mean that because the two words are into one, right? often mean that they have a narrative implication that the two protagonists are merging or you know like their personality are they are getting closer to each other or something like that. Was that ever under consideration or is it more of a
I think for uh yeah we we did consider like different takes on how we would do like what would be the motivation for the characters in this segment and we decided to go with something very simple. So the narrative between the characters, we actually resolved that a bit more in the previous levels, the two levels that are right before this final level because it felt like there were so many cool things that we wanted to do gameplay-wise kind of. So we didn't want to like put too much pressure on the narrative there. So there it is basically just progress forwards and then when you get to the boss fight, beat the boss fight and and that's basically the narrative. But then like the emotion of it is that they have like uh now kind of connected as characters and they are more like they're not bickering and like arguing now they are

more united as a team. Uh but yeah so I hope that answers your question.
Yeah thank you.
Great question. Thank you.
Yes.
Uh hi incredible talk. Uh I just wanted to uh ask about pacing of the levels especially when you're talking about those Zelda doors. How does taking control away from players end up contributing to the natural pacing of gameplay? And how does it play into what you were talking about of like the less engaging moments of gameplay allowing for more exploration of uh the eyes?
Yeah. So I think uh I mean in this level it is very different from in other levels. Uh what we wanted to do like the experience of just looking at your friend's screen was a very strong experience that we wanted to like uh get out of this. Uh and that thing is like that's only with your eyes. It is like not happening in your hands. Uh, so I think the pacing was a lot of like alternating back and forth between that. So we try to like spread these cella cameras for example out throughout the level a bit uh to to get the natural

pacing back and forth of like doing the things with your hands playing and talking to each other and just like admiring the the level and the visuals of it.
Thank you very much.
Thank you. Great question.
Yes.
Thank you for the talk. I really like the game. I finished it like as soon as it released last year.
Oh, amazing. Thank you. Play testing was like a very important part of iterating in the game as mentioned in the talk. So I was wondering about your play test strategy. Yeah. Um to help improve the level like the timing, the frequency like when did you start arranging all these play tests?
Yeah. So so it uh we do a few different types of play tests like the core play tests are like we call them UXRs and they are happening like every two or three months pretty much and then we let people who have never played the game uh play. they play in pairs and it's like 20 pairs or something playing through the game recording themselves so we can watch the videos and see the reactions. Uh but of course like on a day-to-day level uh something that is very fun with working at Hastlife is that I can't play test any of the things I'm creating by

myself pretty much. So I'm always like doing play tests with other people and getting the responses. Uh and I'm also doing a lot of like trying to get fresh eyes even from the company. like when I've created a new puzzle, I have like this whole company of 80 people that I can test with. Uh so that is also like a great resource for us.
Thank you. Great question.
Thank you.
Yes.
Hello. Um it's amazing to see all the design considerations and iterations everything goes through, you know, to land on the final product. Something I was thinking about because you talked about how in split screen sharing information and communicating with each other is essentially a mechanic, but if you're also kind of encouraging screen pick peaking. How do you manage it so that one player isn't just kind of like steam rolling through the entire game by just seeing what he needs to see on the other player screen the entire time?
But that's a really great question. Uh so I think it's like also the two parts. One part of it is that uh as I showed it is really tricky to make the players look at each other's screens. So they

are not generally doing it uh when we are not like explicitly wanting them to do that. Uh but it is also the case that like for example with the the pot they were shooting a lot of players do look at their front screen there and that also gives like another quality. So so there it's like both versions are fine. So there we really don't have to do anything really about it. Um I hope that answers your question.
Yeah. Awesome. Thank you so much.
Thank you. Yes.
More of a comment. Um my daughter and I love playing split fiction and it's one of our favorite game. It's one of the best design game I've played in a long time and I wholeheartedly agree on the co-op. It's you guys took it to another level.
Thank you so much. Thank you.
Yeah. And she bugs me about playing all the time which I love to. Uh and now we're playing it takes two.
Ah fun fun.
And uh and you guys were able to uh pack pretty much almost every genre of games in this game. People who haven't played it really should try it. From

sidescroller to actions to and I didn't get paid to talk about this. Um uh to flying 3D to racing to every single thing. It's amazing. It's really amazing.
Oh, thank you so much.
It's also a big test for relationships.
Yes.
Uh very much so. Be careful of that. Um the centipede.
I created a centipede.
Great. [laughter] Because I do have a small question on that. That one that that was the top for me in co-op. Yeah,
because it's really hard.
Yes,
you have to coordinate. So, what what was the idea behind how did you guys came up with that?
Uh, so with the centipede, yeah, I guess it might be confusing for people who haven't played it. Uh, but it is a segment where the players are tied together kind of and controlling each head of like a snakeish.
Yes.
Uh, and uh, and there in one of the segments you can like uh, swing by biting different points and you're like swinging on wheels and it's quite complex. Uh, and that is one of the absolute most challenging uh, segments

in the game. And I sometimes watch play tests where people were stuck on that for like 20 30 minutes, which was more than like the rest of the level. Uh, but and and it is like it is a bit divisive in that way. But I think also for the people like when they complete it, they get a very satisfying like experience, a really like they have really accomplished something, which is something that we usually don't do. Uh, but I think it was fun to have like at least one of those moments in the game. Uh, but yeah, it was just like I tried a lot of different things like spinning wheels in different directions and just play tested until it was fair enough and uh but also like very uh intense for the players.
But the idea that you're on a single creature together like that just sort of came out or
uh it was uh that one like for a lot of things that we're doing we get inspired by other games. Uh there was a game called like fogs that we looked a bit at and they were not really doing the same things like they were not doing the swinging for example but it was something that uh felt like uh we had

the idea of having a centipede before that and we didn't know what the centipede would be and then we just looked at a lot of different inspirations and uh I recommend you to play fogs as well like uh I I haven't played it myself actually I just watched videos of it but it seems to be a very fun game.
Thank you so much.
Thank you so much.
Yes. Next.
Um hi thank you for the talk. It was amazing. Uh thank you.
I've played it takes two twice and split fiction I think three times because there's no other game to play. Uh my question is u how do you I'm a visual designer so my question is about UX research. How do you come up with certain things and that the features that you talk about? And also how much time does it take you to do that iteration and research over and over again? And my final question is when is the next game releasing?
Uh I can't answer the final question. Sorry for that. Uh but it will be our best game yet. So uh you should keep your eyes out for that one. Uh uh but for the other thing like the process

when we're creating things, something that I'm often doing is like I'm sitting on the commuter train on my way to work and I'm sitting with my iPad and just sketching different ideas to just get a lot of ideas out. Uh and then when I get to work, I usually start prototyping them myself as quickly as possible. Uh and then I try to like I try to play test on a daily or like every other day. Like I try to play test a lot and just keep iterating. And I think something that allows us to do this very quickly is that we're very autonomous as designers as well. We're quite good at like programming and stuff. So we can create a lot of the features ourselves without putting it like on a task list for someone else who might not have the time at the moment.
Okay.
I hope that answers your question.
Yes. Thank you.
Thank you. I don't know how much more time do we have. Uh is there any? I don't know. No one is stopping us. So let's continue with questions. [laughter]
All right. Hi. Uh it was really really cool to see how you did the two worlds by simply having the fantasy world off to the right a bit.
Yeah.
Uh it reminded me of a project I did once about where it included time travel

and the way that I did it was I had the layers stacked. But the way that I was able to hide those layers was that I had it so high that it was out of the render view of the players. I couldn't see the world above them. Yeah.
I noticed that with yours they were a bit they share the same skybox. They both had the sci-fi skybox. Uh and I figured that would be really hard to do. Um, so I noticed that in the fantasy world you had to block out the sky by placing the environment. Were there other tricks you used so that you kind of were able to hide that if you really could see past it, you would see the sky from the fantasy?
Yes. So something that we actually did is on your screen you can only see your world. So the other world is actually not rendered. But then like the sky is like a common layer that both uh both levels can see. Uh but yeah, I think the other biggest thing was the use of post-process volumes because that like creates just an overall different feeling like how the lens behaves and all of that. Uh that makes it quite easy to just make the worlds a bit different even though we're having the same skybox. Um yeah, I hope that answers

your question.
Yeah, that was great. Thank you.
Thank you.
Yes.
Hi. Um I'm artist uh at Blizzard Entertainment D4. Uh I'm super amazed by you know how much uh puzzle solving you're doing in like the level design. I think my question is more towards to the production time. Um so when you're designing or redesigning uh the puzzle if you encounter problem is it something like production give you you know x amount of time and saying hey you have you know a months to you know uh finish this puzzle or is it something uh as a designer you like to ask hey I want you know uh three months to working on like this problem like who's like you know designing and sorting out like the production time you know to solve one level
I think I get the question So uh with the production cycle um generally we're working like two or three months on a level and then we're switching to another set of levels. So everyone is like working kind of on two different levels. Uh and then often it is like I mean it we iterate very late. We cut a

segment uh we cut a segment that is even like it is on the physical disc but it's cut if you download the game. So we iterate like into the very last minutes. Uh, and and it's like with with the I I think it benefits the game a lot because it allows us to really like um make the game play as good as it can be. But often it's like the later it gets into production, the more likely it is that we won't like completely redesign things. So I think it's just an organic uh thing kind of
for sure. So if you uh do see the problem you know uh your presentation you throw like some of the ideas you know uh during the presentation. So, so production is like you have to give you like extra time to continue working on like this problem, right?
Um I uh I I mean it's I Yeah, we I don't know. Can Can you say that?
Oh, I'm saying um so during your presentation there's a few times like

you throw like the entire design into the garbage like the paper uh garbage thing. So is it something like uh you actually needed extra time to redesign that specific uh puzzle?
I mean I don't think we have it that specific at Hitelight. We have it more like organic like uh we have a certain time to work on the level before we switch on to the next level and we want to get the things playable. So it's kind of up to us as designers to decide like do I want to put my time into redesigning this puzzle or do I want to put it elsewhere where it like gives us more. I hope that answers your question.
Question. Thank you so much.
Thank you so much. Yes.
So, yeah.
Okay. One or two more questions.
So, I have a question. How do you do the reveals? Like, how do you play test them and make sure they're hitting correctly? Even though the people you just play test with are now they now know the reveal, so they're now
contaminated. So, I think like uh uh some things are very easy to just test with people more like gamy segments like uh um and puzzles and stuff like that.

the things that are more like the reveals, those we can only play test when we have these big like uh play test sessions every three months. So there we really put a lot of effort into like uh getting the experience as good as it can be so that we have time to iterate on it because that iteration time is naturally a bit slower I guess.
Uh but yeah, thanks for your question.
Yes.
Hi. Hi. Um first of all, your talk has been my favorite at GDC. It was great. Um
Oh, thank you. That makes me so happy.
Thank you. Uh, so I'm a senior. I'm majoring in game design. And in between like making my own games and looking for internships, what made you want to get into games? And what advice would you have for people who are looking to break into the industry?
Uh, great question. So, uh, it was actually when I was in high school, I was deciding between like going to architect school and like games. And then they just had like a presentation uh where they where like the presentation for the two programs. The architect one was really boring and the

[laughter]
And I I I I love architecture, but I really wouldn't want to work as an architect. Uh but but the games one, they were just like playing a video of like a Google office with a slide indoors, and I was like, "Ah, this is what I want to do." Uh I guess uh but then I just I mean I I I love uh making games especially, so I I put a lot of time in my spare time to like design Counter-Strike levels and do different things like that. Uh but I went to TGA, which is a Swedish game school. uh and though they have like a built-in internship program kind of uh and through that I got the opportunity to have my internship at Hazelite uh and then like the internship worked out and I got to stay. So uh I don't know it's kind of difficult because every journey is very specific so it's it's tricky to like give advice from just my my experience but that's at least how it was for me. I think that was all of the questions we had time for. But uh uh I can go somewhere if someone has more questions. Uh
yeah, go out in the hallway and do a

wrap-up or or something like Thank you so much. [applause]