The Island

Welcome to the Island

This level is a third-person adventure focusing on composition and exploration. Take control of a lost robot who must find a way to refuel the train on the island, using it to reach the city far beyond.


  • Reusing a small space in creative ways to keep the scope of the level manageable.
  • To create a non-linear level that engages the player's desire to explore and discover.
  • To use well-crafted compositions and vistas to appeal to the player while also communicating vital information.
  • To include simple puzzles.


  • 3½ Weeks Fulltime
  • Unreal Engine 4
  • Maya (All assets except ground textures are self-made)
  • One-man project


This level primarily took notes from Rime, but I also looked at Breath of the Wild and Journey for design cues.


What is The Island about?

This level has two pillars: Visual Communication and Exploration.

The player should feel a desire to explore, and should be able to understand both their goal and how to navigate based on visual means alone. As such gameplay took a backseat in favour of focusing on the visual aspects and level design.

The Island can be split into two acts. The first act is a linear section that functions as a tutorial and sets up the player's goal. Everything the player needs to know to clear the game is presented here. Once they arrive on the northern island the second act begins. The player must find and clear two puzzle rooms to beat the level. Here the design is open, allowing for exploration and clearing the puzzles in any order.

The First Act

Communication through Composition

The Importance of Visual Communication

Everything around the player transmits information to them. Managing this information is key to creating a design that facilitates the player's experience and makes sure they obtain the necessary knowledge about the game in an engaging way with the least burden on their mental energy. This applies to everything from telling the story of the world to teaching gameplay mechanics.

Visual Communication is one aspect and applying it to a 3D space is complicated. Care must be taken that the imagery looks good and transmits the correct information from several angles. Creating 2D-like vistas is still possible, but to ensure they work the player's movement must be restricted - an inherent conflict in the design of this island, which is meant to be open. However the 3D space also opens up new possibilities, like relying on the players movement to create a sequence of imagery. In the intro section the player can follow the ridge of the cliffs to find a cave entrance, a 360 motion not available in 2D space.

In the end I chose to make the first act of this island linear, not only to make sure the player goes through the tutorial, but so that the vistas presenting the long-term and mid-term goals are guaranteed to be seen by the player from optimal angles.

The Bridge Landmark

The entire level is designed around one landmark: The Bridge.

This construct creates a leading line throughout the level, making it easy for the player to orient themselves.

Both the long-term goal (the city) and the mid-term goal (the train) are connected to it, ensuring the player can easily remember those locations. Care has been taken to weave the landmark into the level design, letting the player explore underneath the pillars. This helps make the bridge feel like integrated into the design instead of being window dressing and gives the player a sense of scale.


An example of how the flow of information was considered throughout the project. The starting area is deliberately sparse, so to not overload the player with information while they're learning the controls. Negative space in the form of the sky is used to invoke a sense of calm and help orient the player, as they follow the line of the cliffs their sight will land on the sole way forward. The crashed ship offers a sense of mystery and storytelling.

The City and the Bridge

After going through the tutorial cave - forcing the player into a contrasting dark confined space - they are then greeted with the reveal of their long term goal. From the city comes a bridge, the centrepiece that binds the level together.

Denial and Reward

The stone pillars offer a sense of rhythm to the player while  denying them a proper view of their mid-term goal. The player has to undertake a simple puzzle involving placing a rock into water so they can move across; gameplay variety, and a subtle way to make the player focus on something else than the narrative so they're offered a break between the reveals of the long-term and mid-term goal

At the end are a set of stairs, which when traversed will force the player's vision upward as they're greeted with...

The Level Goal

The sight of the train, the mid-term goal which will take them to the city. The city, the bridge and the train are all imagery with strong intrinsic connotations for the player.

By seeing them in succession the expectation is that the player can put together their goal - to find a way to ride the train to the city - purely through visual means.

The Second Act

The Art of Exploration

The Next Step

During the second act the level opens up considerably. The goal is to get the train moving, and to do that the player needs to load it with coal and water. There's a coal chute and a water bridge leading to the train, and by following them to their source the player will encounter two puzzle caverns that can be completed in any order.

Now that the level is far more open,

there are new design problems that must be solved.

Water means Death!

There are two ways the player can die. Falling from too great heights and touching water. Being a brittle robot they can't survive either.

Therefore water is used as a natural barrier, eliminating the need of invisible walls and allowing me to block off large areas of space without it feeling unnatural.

Multiple Paths

Once the player crosses the small bridge they can explore at their leisure. Above are some of the player paths I designed. Much effort went into trying to make the paths loop into each other, avoiding dead ends and making sure that once the player was done with one goal they wouldn't have to backtrack.

Problem: Desire Paths

One problem I encountered was the player desire paths.

For example, to get to the lower cavern I designed a path to go around a pillar. But while playing I found myself jumping down the cliff instead. While this isn't game-breaking, it felt immersion-breaking because as the player I could tell I wasn't going a route the designer had wanted.

In this case I chose to make the desire path an actual route, adding rocks acting like steps so the player could go back that path. Time allowing, the original path would've been made more scenic and perhaps had an optional reward or puzzle

An alternate solution is to block the player in a natural way, which was done with the body of water on the hill. Originally this was a large empty space left useless after redesigning the map - by placing water it was repurposed as a small inaccessible landmark.

Reusing Space

One of my goals was to reuse space rather than creating new areas when I needed them. This led to the usage of several layers of elevation. Once example can be found in the pillars traversed during the tutorial section. The bottom half is later revisited when the player is looking for a power core needed for a puzzle.

In addition, by having the player return to familiar spaces they create a mental map of the island.


The Puzzle Caverns

What can the player do?

Since gameplay wasn't a priority it was left rather sparse. The player can pick up items, a feature that is used to move stones so they can create paths, and to move power cores that power elevators. There are two switches that activate a sequence. Water is a death hazard and can't be touched.


At the beginning of the game the player goes through a small tutorial section. Here they learn the limits of their movement such as jump heigh and jump length, that water is dangerous, and about the stones and power cores they can pick up and what uses they have.

Water Cave

One of the two Puzzle Caverns, marked as Cave D on the overview.

The player must find three power cores so they can raise a bridge to another ledge. Once they reach their goal they will find a lever, which starts a sequence lowering a bridge and delivering water to the train.

Coal Cave

The concept of using power cores returns, but this time they're limited in number and the player must reuse them. There's a coal cart they must transfer to a coal chute, after which a sequence plays where you can see it being delivered.


Problem: Fitting the Caverns into the Island

For gameplay meant to take a brief time to implement, the caverns themselves proved a much bigger design challenge. Since one of my goals were reusing space, I decided to implement caves inside the hills I'd created.

However this meant I had to create the caverns inside a constrained space and design the gameplay around that. Much time went into making sure no meshes were sticking out in the overworld, while at the same time allowing enough room within the caverns for interesting vistas and to make the gameplay and player paths clear.




Using Journey as my initial reference, I wanted to make a story-driven experience. The level started out linear and enclosed, but feeling stressed about the project I chose to do an open-world experience instead as I'm more used to it.

From the very start the bridge has been the central theme of the project. It has been a comforting guide, despite all the up and downs the ultimate goal of the game - to find a way to travel across the bridge - has always remained the same. At this stage however there was no train, instead an elevator you had to get running.


The blockout saw the island transformed from a pancake into having an elevation and the ravine running through it, an effort to make the island feel more dynamic and allow for re-use of space.

At this point I had a different mechanic involving shifting planes, but it proved too narrow and time consuming so it was scrapped.


This is where the project lost focus, and the borders between blockout, whitebox and gold overlapped. A large reason was my decision not to focus on gameplay which resulted in it being force-fit into the island after the layout was done.

The development of the island continued.

A revelation I had was how interconnected the areas wound up, a spiderweb of design choices. It was a balancing act, improving one area could mean encroaching on another and making it worse. Either I had to take that sacrifice, leave it, or take the more time consuming option of finding an alternate solution.


The stage where I spent too much time on the environment and less on the gameplay. However, as the visual aspect was my focus from the start, it was perhaps not as big a mistake as I thought at the time.

At this stage I began running through the island, making sure it looked and felt good.



This project taught me a lot about management. With covid forcing us all to stay at home, it has been a challenge to stay focused and gain feedback. The little moments in class where we would look over each others' work or discuss them during break, those don't exist, and if you don't make an effort to reach out your work will be developed in a vacuum.

That happened to this project. I am confident that had I not worked from home, I would be far more satisfied.

But it is what it is, and adaptation is part of our lives. I would've liked to make a better plan from the start, to focus more on the gameplay and the level design. Through my education one of my problems has been creating levels that are pretty, but play like a boring walking simulator. I've worked to amend this, and while I feel this project was a step up I feel I'm not quite at the level I wanted to achieve. Still, I must also recognize the steps I did take - incorporating a single gameplay mechanic (picking up items) and trying to use it in as many creative ways as possible, running through the level and trying to think of the ten-second rule, always varying the gameplay between scenic exploration, navigation such as jumping over water and pure puzzles. It's not good, but I feel I've started climbing the wall I've so far only scratched.

The other big lesson was working with visual communication. This is an area I find very fun, and when I realized all the areas I had created were interconnected into one big spiderweb of design it made me happy. I want to continue using that mind-set, of reusing space, using lots of verticality, of connecting several areas in various ways that make logical sense, and let it all flow in a way that the player creates a mental map and starts to feel at home.

All in all, I'm not satisfied, there are much to improve, but there has been a plethora of valuable lessons I wouldn't have learned if I hadn't undertaken this project (for example building caves in a constrained space, which was NOT planned at all and a total pain, but by the end I think I got somewhat good at it). I hope to improve in the future.