The VanShee
Project Overview
The project that helped me take my first steps as a game designer — and one where I was able to learn and grow enormously.
I took full ownership of 2 levels, handling everything from concept planning to blockout prototype creation, and even the monster faction concept planning and skill implementation for monsters placed in those levels. Beyond that, I planned a traversal action systemto create a dynamic movement experience unique to The VanShee's levels.
Although the project was unfortunately cancelled and never released, the nearly 3 years I spent here were not a failure but a tremendous opportunity for growth. The perspective I gained through experiences beyond level design — monster and system planning, as well as collaboration and communication with other departments — remains one of my most valuable assets to this day.
Key Responsibilities
·
Level Design Concept Planning & Blockout Production
Created 2 main open field levels at 1km × 1km scale
·
Traversal Action System Planning
Developed 3 traversal systems: climbing, zipline, and sliding
·
Monster Planning & Implementation
Created 30 monsters across 4 grades — normal, named, elite, and boss — for placement in City and Jungle levels
·
Mission Scripting
Mission design using Lua scripting
3 missions per level
·
Auxiliary Tool Planning for Workflow Efficiency
Level Design
Process
Concept Planning → Prototype Level Creation → Playtesting & Validation → Final Polish
Details
Common Design
- ·Designed with an ant-colony-style flow structure matching the open field direction, allowing free movement across all paths without restrictions
- ·Designed so that finding the optimal route to the objective during mission play itself becomes a measure of player skill
- ·Placed landmark elements at key destinations where bosses are located so players can perceive the goal direction
- ·Set minimum wall height to 6m to match character action characteristics, minimizing jump skill usage and camera collisions
City
Concept to prototype: 4 months- ·Based on the post-apocalyptic downtown setting, used the Jamsil district of Seoul as the foundation — leveraging the rarity of Seoul-set games as a key point
- ·Divided into two themes — "Overground" of collapsed buildings and rubble, and "Underground" dark and cramped — to transition atmosphere and play experience
Overground
- ·Wide open sightlines and spacious combat areas designed to enable hit-and-run tactics based on visual information
- ·Encouraged players to self-assess and choose their own difficulty
- ·Designed combat experience of gathering and sweeping clustered monsters using the wide combat space
Underground
- ·Narrow space, dark atmosphere, and complex paths created short sightlines and limited movement radius — providing spatial tension contrasting the Overground
- ·Designed combat experience of constantly reading the surrounding spatial structure to avoid being surrounded by monsters in confined space
- ·Designed movement experience penetrating building interiors to convey a sense of vertical level composition
Jungle
Concept to prototype: 3 months- ·Based on the setting of a refuge in a post-apocalyptic world, used Southeast Asian jungle as the motif for overall level work
- ·Aimed for a lower difficulty level compared to the complex City — minimized paths and complex indoor structures to create a more combat-focused design
- ·Reinforced environmental storytelling and atmosphere through ruined Buddhist ruins and Buddha statues
- ·Added two terrain types — "elevation differences" and "fall-off" — to differentiate monster placement and combat play experiences
- ·Used elevation terrain to increase long-range monster placement ratio, conversely guiding players to exploit plunge attacks against monsters
- ·In fall-off terrain, guided easier monster elimination through knockback skill combos while creating tension with the player's own fall risk
Key Learnings
- ·Experienced the full process from concept planning to prototype level creation to polish, and learned that level design is not merely about building maps — monster placement and combat experience design make up an enormous portion.
- ·Through collaboration with the background art team, I learned the importance of clearly communicating level design intent, and that bridging the different perspectives between planning and art has a major impact on the level's quality.
Traversal Action Planning
Process
Design Document → Feature Testing → Final Polish
Details
Common Design
- ·Created 3 traversal actions: climbing, sliding, and zipline
- ·Given the game's high combat frequency, planned to provide clear rest periods and minimize movement tedium in the game's repetitive play loop
- ·Since this is not an action-adventure game but a combat-core game, having traversal actions active at all times would interfere with combat. Implemented as volumes placed in levels rather than as character abilities, to avoid disrupting combat
- ·Structured traversal actions as shortcut movements — guiding players to discover optimal routes through traversal on repeated missions
Climbing
- ·Created for vertical movement purposes
- ·Two types based on climbing actor distance: standard movement, and movement via direction + jump button input
- ·Placed at intervals throughout climbing sequences requiring input, conveying rhythm and the fun of controls
Sliding
- ·Created to forcibly push players in one direction toward boss rooms or specific arenas
- ·Made controllable during sliding, conveying the fun of dodging walls or traps while moving
Zipline
- ·Designed for use when crossing long distances
- ·Camera control enabled during zipline traversal to allow looking at the surrounding scenery, providing a moment of rest
Key Learnings
- ·Through close collaboration with animation and programming teams, I learned that different departments have different perspectives, and that there is a great deal for planning to check — and that going forward I need to look broadly from other departments' viewpoints as well.
- ·I had not given much thought to camera framing and transitions during the planning stage, but through actual testing and feature development I realized how important this was. I learned that not only functional design but also the directorial perspective must be seriously considered.
Combat Design (Monsters)
Process
Concept Planning → Skill Implementation → Playtesting → Final Polish
Details
Concept & Skill Planning
- ·Planned overall appearance, weapon concepts, and skill mechanisms
- ·Planned monster appearances and weapons appropriate to each level environment to be clearly recognizable and predictable
- ·Key focus: Can this monster provide sufficient play fun when placed alone?
Skill Implementation
- ·Skill implementation work using Lua scripting
- ·Created diverse monster skills including melee, single projectile, multi-projectile, homing, and buff types
- ·Animation and hitbox adjustment using Unreal Engine montages
Playtesting
- ·Validated that even prototype monsters without art resources provide sufficient mechanically intended fun
- ·Repeated cycles of testing and revision
Final Polish
- ·Assembly work after art resource handoff
- ·Verified animation and hitbox alignment and completed final quality
Key Learnings
- ·Through repeated monster design iterations, I learned what is most important for making good monsters.
- ·Monster concept must connect with the level environment for immersion to come alive.
- ·Monster concept and skills are deeply connected to level design.
- ·Monsters must match the background atmosphere of the level for immersion to work.
- ·Even the same monster delivers vastly different play experiences depending on placement.
- ·Attacks must be readable from motion and weapon form alone for players to feel the combat is fair.
- ·A well-designed monster is one that is threatening in both solo and grouped placement, while still giving the player sufficient fun.
Mission Design
Process
Based on Mission Settings → Scripting → Playtesting → Final Polish
Details
·Actor placement work aligned with scenario settings from the lore team
·Full mission flow setup from start to end using Lua scripting
·Scripted various mission objective types such as reaching a destination, defeating a boss, and eliminating a set number of monsters
Key Learnings
- ·Through collaboration with the lore team, I experienced the communication process of proposing and coordinating directions that could enhance play experience fun.
- ·Difficulty and experience vary by mission, and to stably control this mission flow I learned I must carefully consider all variables and unexpected situations a user might attempt from the user's perspective.
Auxiliary Tool Planning
Details
NavLink Proxy Generator
- ·Planned a large-scale NavLink generation tool using volumes to reduce the cost of placing individual NavLink proxies
- ·An auxiliary tool enabling easy NavLink placement by setting the generation count and NavLink direction within the volume range
Climbing Actor Data Extractor
- ·Identified that dozens to hundreds of climbing actors are placed per level, and validating all of them incurred significant costs
- ·Extracted climbing actor data to a CSV file, enabling rapid identification of problematic areas through data and minimizing validation cost
Key Learnings
- ·Discovered that play feedback or art resource replacement in level placement work increased the need for actor repositioning, leading to a high incidence of human error.
- ·Learned that auxiliary tools that streamline validation costs can improve both work efficiency and quality simultaneously.
Images
City
Jungle
