Elysian Siege is a game developed in Unreal Engine 5 by a team of 20 ESAT students over the course of nine months. It's a 2.5D platformer focused on player movement, where the main challenge lies in mastering the character's abilities and advancing as quickly as possible through different rooms.
As one of the Game Designers on this project, in the following paragraphs I'll be sharing how the design process for this game unfolded.
We could summarize this analysis in a single sentence: Elysian Siege is the Metroidvania that never was. But the real question is - what changes led us down that path?
At the start, our premise was to develop a side-scrolling platformer, but with 3D art. The main design pillar throughout the project was to focus on movement, downplaying combat - treating it more as an obstacle than a mechanically challenging system.
With that in mind, we envisioned a slow-to-medium-paced game intended to offer a reflective, thoughtful experience. The goal was to achieve fluid player movement through responsive controls and mechanics centered around mobility.
Initially, the core mechanic was time-slowing. Levels and other elements were designed around this ability, so that overcoming challenges would rely on analyzing the environment. However, after early research and playtests, we realized the slow pacing might not fit the kind of experience we wanted.
A faster rhythm began to emerge - one focused more on exploration and interaction. That shift brought us to our main inspirations: Metroidvanias.
Most relevant references
It was the beginning of a big change. With the altered pace, our unique mechanic came into question. The project had to reinvent itself - and we had to adapt alongside it.
Due to production constraints and evolving needs, we were forced to alter the proposed mechanics, which drastically changed the game's pace. The time-slowing concept and environmental interactions were removed, replaced by more traditional platforming mechanics.
This led to the introduction of dash and glide, which added extra mobility but significantly sped up the overall rhythm of gameplay.
Given the project's tight timeline, we were working simultaneously on documentation, level design, and mechanics development. These major changes inevitably affected level design. We managed to partially adapt the existing levels to the new gameplay rhythm, but this introduced one of our biggest recurring problems: player guidance.
Originally, exploration was our main focus, so explicit guidance wasn't a priority. Later, as the game evolved, we modified levels to follow a more linear path, though not all of them could be reworked due to time constraints.
In the end, we implemented a guiding element that I personally consider quite effective - a subtle environmental effect that naturally integrates into the scene while still being eye-catching enough to draw the player's attention. Its movement subtly leads the player toward key points of interest.
Despite all these shifts, we managed to mitigate most issues thanks to polished player navigation mechanics and strong visual and audio feedback, which helped reinforce immersion, visual clarity, and the sense that every action had an immediate consequence.
Level 1. Final dressing iteration.