Pre-production, research & planning
Starting new projects can be daunting, but I've always found that breaking it down into questions have helped me define and better understand the scope of the project. Questions like:
-
What's my time frame?
-
What do I want to show with this project? What's the goal?
-
What would the mechanics/features be?
-
Which games are their that's the most similar to what I want to create?
Analyzing similar games for mechanics or style reference helps me get inspired and come up with new ideas. I often have a gameplay playthrough running in the background while planning my experience or gathering reference materials in PureRef.
The next step is to be honest with yourself about time estimation and time-consuming sinkholes, such as scripting, player control, and UI. Relevant questions I’d ask include:
-
Can I script this myself? If not, how long would it take to learn?
-
Are there templates that already offer something similar to what I want?
-
Is this feature essential, or do I just want it because it's ‘cool’?
Lastly, I like to seek a second opinion of what I've gathered so far.
As we all know, in the rush of inspiration, it’s easy to develop tunnel vision, but fresh perspectives can introduce new ideas, uncover potential pitfalls, and provide valuable reality checks. I’m fortunate to have had talented and passionate teachers and co-students whose insightful feedback have helped me redefine my idéas.
During the blockout and playtesting phases, I like to grab screenshots and do paintovers to help visualize my ideas for the environment. In this case, I wanted the lower lands to have a "Mirkwood" feel, with overgrown roots and trees.
Another significant change was the Wishing Tree puzzle. The original idea for what was then called the "Cave Puzzle" was inspired by some puzzle designs in God of War, where the player needs to throw their axe at certain shrines or pots to proceed. I wanted to build on this idea but use my bow instead. After blocking out the cave room, I came up with the concept of shooting down "hollow wishes," which were made through selfish acts and angered the Norse gods.
In this project, I've made numerous changes, driven both by feedback and by new ideas that emerged as I visualized them through paintovers. One significant change was redesigning the Raider's Camp area, reducing its size by half. This adjustment improved the level’s pacing, flow, and balance between the two areas.
With a solid paper design as a foundation, I can save time by importing it into the engine and tracing the larger shapes of the map. In this project, I’ve primarily used Unreal Engine’s Poly Extrude tool to block out large areas with simple geometry. This allows me to quickly iterate on different sections while continuously testing from the player’s perspective.
During this process, I focus on key areas of the map first, such as the start and end points, reveals, and points of interest (POIs). Once I'm satisfied with the overall layout and pacing, I start adding essential elements like rough enemy placement, basic geometry for cover, ledges, boundaries, and the general positioning of buildings and structures.
At this stage, I also start to recognize what’s working and what isn’t. Perhaps a reveal isn’t as impactful as I initially envisioned, or the level flow doesn’t work as intended. The more issues I can identify and resolve early in the process, the more time and frustration I’ll save later in the project.
Next step in the processes of creating levels would be to set up a player gym, where I can test mertics like jump distances, crouch heights, climbing & vaulting.
I've found that establishing consistent metrics early helps streamline the processes of creating and managing your asset bank for climbales, hard and soft covers, interactables etc. Having consistent metrics also strengthens your visual language to the player, building upon the importance of affordances as I've mentioned in Return to ashfall. Questions I'd ask myself when deciding on metrics would be:
-
How tall would crouch and full covers be?
-
How mush bufferzone do I need to not confuse the player?
-
How far can my character jump?
Another feature of the player gym is of course to develop and test you're features & mechanics. I've decided to create a bow and arrow system & and a more advanced
AI behaviour tree which can react to noice events, run EQS to flank the player, and react differently depending on where you'd hit them with arrows.
I belive player gyms are key to establish rules and guidlines for yourself and your team, making you are more effective designer.
Metrics
A showcase of how I use the player gym to decide on metrics for vaulting and obstacles that can be crouched under.
Prototyping features
I've used the player gym to bug test features like hiding from the AI perception in bushes.
Assetgym
II do enjoy the process of creating my own assets when applicable. To often I find myself relying too much on asset packs for environments but when I want to realize a very specific vision of a area, I'll happily to jump into blender and do a low poly version of the asset. Working in group projects, I've found that creating placeholder versions of what I want, testing the metrics of it and later sending it to the artists have streamlined the creating on asset lists and metrics, removing the potential confusion of scale and intentions of assets.
I've found that drawing traditional top-down maps helps me visualize player flow, key areas, and how to pace the distance between points of interest, story beats, and enemies. However, not everything can be effectively visualized in a top-down view due to the limitations of 2D; some aspects simply need to be felt out during the blockout stage of production. That said, a major advantage of top-down maps is the ability to quickly iterate on them or use them to convey a general idea during a presentation.
But top-down maps aren't the only things I like to visualize at this stage. I also sketch puzzle designs, concept art for points of interest, or general mood pieces during pre-production, especially in group projects, to help present an idea or convey the feeling I want within a specific level or area
While working on the paper design for The Long Journey, my goal was to divide the latter part of the map into two distinct sections, with key puzzles located on either side of the river. The river serves as a datum, tying the layout together and helping the player orient themselves throughout the map. The challenge would be to create effective one ways, pushing the player around the map. As well as creating interesting locations that played well in both directions.
About the project
The goal of this project was to showcase my level design skills and how I approach backtracking, allowing the player to choose their own path and the order in which to complete tasks. I also focused on guiding the player through interesting locations and atmospheres to enhance the feeling of a grand journey.
My second goal with this project was to document and showcase my creative process — from the early decisions I made to the iterative changes I implemented based on feedback and observations.
Narrative
I'm a huge fan of medieval / fantasy world building in games so I wanted to make a experience that centered around norse mythology.
-"You play as Eirik, a hunter from a village that's been raided by an enemy viking faction, who's traveling with the remaining survivors in search for safer lands.
The road on which you travel is blocked by an ancient pilgrim gate and it's up to you as the player to find the missing zeals needed to open the gate."
Scope
The idea behind the level was to create a semi-open design with branching paths, allowing the player to approach two puzzles from any direction.
With a five-week timeline, I decided early on to use the Game Animations Sample Project, as it already includes a fully implemented traversal system and motion matching for animations. This saved me time, as I didn’t have to build my own system from scratch.
I also wanted both the start and end of the experience to be at the beginning of the level. This challenged me to design a layout that allowed the player to backtrack in both directions, ensuring they wouldn’t miss any content while still encouraging exploration through pickups.
Additionally, I explored more advanced technical scripting in this project, creating a custom bow-and-arrow system and further developing the AI combat and reaction logic, including noise detection and an environmental query system.
As a foundation, I used and customized the Game Animations Sample Project, which is free to download on FAB and serves as an excellent starting point for anyone looking to create a third-person game.
Assets
-
Mixamo.com for custom animations
-
Stylized Birch Forest by Sandvector Studios (foliage)
-
Dreamscape: Stylized Environment Tower by Polyart Studios (foliage)
-
Elevenlabs.io for AI generated voice lines
All other assets where created by me in Blender or Unreal Modeling Tool
The Long Journey
A third-person semi-open adventure experience
Team Size
Solo Project
Development Time
5 weeks / half time
Tools
Unreal Engine 5.5
Game Animation Sample Project
Blender 4.0
Focus
Level Design & Backtracking
Creative Process
Puzzle Design
Narrative / World Building
Scripts
Advanced AI Behaviour
Bow & Arrow Combat
2 puzzle systems
Scripted Cinematic Sequence
Overview
Top downs & paper design
.png)
Setting up the player gym
The Creative Process
Rough Blockout
Iterations & redesings
Playtesting & observations
Getting feedback from external playtest is the bread & butter for all level designers. Since you as the designer know how you want it to play, and usually only play as intended. By asking people to playtest your levels and just observing with pen and piece of paper, chances are that they'll find bugs and exploits you'd never think about, or you'll realize that things that you tough where obvious was actually not.
Reflections
Designing this level was a challenging but rewarding process, especially when it came to creating open-ended and branching paths.
I wanted the player to have multiple ways to navigate the environment, but this quickly became complex. Certain one-way routes didn’t function as intended, and some paths ended up being so obscure that players would rarely, if ever, choose them.
Given more time, I would have refined the AI behavior further. In the video, I chose not to show the AI attacking the player, and there’s a good reason for that. When the AI reaches the player, an Environmental Query System triggers, causing the AI to circle around the player. However, the attack animations I sourced were not up to standard, and as a result, the AI’s attacks would almost never connect.
Another technical challenge I faced involved the arrow trajectory system. The blueprint relies on two line traces to calculate the velocity from the arrow’s tip toward an aiming point derived from the center of the player’s camera. As long as the arrow has a valid target, the system correctly applies force and gravity. However, when aiming into the distance where the aim offset lacks a valid endpoint, a random force factor is introduced to the calculation. This unpredictability makes it nearly impossible to aim slightly above enemies’ heads with any consistency.
I’m also not entirely satisfied with how the shrines are used to teleport the player back to the start of the level. My original vision involved creating engaging backtracking routes throughout the map, but in hindsight, this proved to be an overambitious scope. While the final implementation works, there’s plenty of room for improvement. For instance, the falling bridge still looks like a placeholder.
That said, I’m genuinely proud of what I’ve accomplished with this piece. It was a valuable learning experience, and I’m excited to apply these lessons to future projects.
Full Gameplay Walkthrough
