Gem Heist

Balancing Flow, Feedback, and Stealth

Project Overview

Gem Heist is a Game Boy game where players need to sneak into a guarded camp to steal a precious gem. It’s a short but complete stealth experience focused on flow, clarity, and player intuition. The game was designed to create a satisfying quest within tight technical constraints and a minimal visual language.

Genre: Stealth Puzzle

Gameplay Length: 5 minutes

Design Focus:

Crafting a complete, well-paced quest experience with carefully structured gameplay flow.

What I Learned:

I learned how to simplify everything around the core mechanic to ensure it shines—clarity, pacing, and readability are more valuable than complexity.

Key Problems and Solutions

1. Reducing Map Size to Improve Gameplay Flow

The original map design spanned 10 map pieces, aiming to create a sense of vastness and narrative distance. However, playtesters found it too empty and slow for a Gameboy game.

Approach:

  • Reduced the total map pieces from 10 to 4

  • Shrunk each area to better suit handheld play

  • Removed an unnecessary woodland section

  • Added decorative elements to previously blank spaces

Result:

Players reported that the new map felt tight, focused, and well-balanced between story and gameplay. Exploration became more purposeful and less tedious.

2. Making Tree Interaction Intuitive

A critical puzzle involved a key hidden in a tree, but some players didn’t realize trees could be interacted with.

Approach:

  • Placed a tree directly in the player’s path to encourage early interaction

  • Reworded NPC instructions to give clearer, more direct hints

  • Maintained subtlety while increasing discoverability

Result:

Players now discover the mechanic naturally. Most interact with the tree immediately, while others explore and still succeed—matching the intended design for discovery and curiosity.

3. Implementing Line-of-Sight Without Vision Cones

The Game Boy's technical limitations made it difficult to implement realistic vision cones for enemies.

Approach:

  • Used simple rectangular collider zones for enemy sight

  • Redesigned areas to visually imply these zones through wall placement and obstacles

  • Ensured enemy vision felt readable through careful spatial design

Result:

The stealth mechanic was well received. Playtesters found it clear and intuitive, even without any on-screen indicators. This confirmed that a strong level layout could carry the mechanic on its own.

Design Process

Pitch

The project started with the idea that I want to implement an intact quest into the Game Boy platform. I used some quests in Grand Theft Auto V and Baldur’s Gate III for example. The quest in the pitch was designed based on experience flow, including 4 typical phases: Onboarding - Development - Climax - Resolution. I designed the quest with the classic idea of “Sneaking into a stronghold”, with the 4 phases were: Knowing the background - Getting into the stronghold - Deal with the most strictly guarded area to get the quest item - Safely get out.

Paper Prototype

Based on the questline, I added a bit of fantasy into the game, making it a side story regarding a guarded camp that happened in a large, magical world. I made a set of printed map pieces that can be “played” by acting. The map was made with Tiled, a GameBoy map editor. This version was tested with players and gathered feedback about the game’s map. The idea itself was confirmed practical.

Rough Demo

After I refined the map based on feedback, I implemented the map into GB Studio, the engine for GameBoy games. I also added the player’s character and its animations, letting the player walk around and explore the map. This prototype was used to know player’s behaviours when they are thrown into the map.

Alpha Versions

Based on the results of the Rough Prototype, the map was further improved and became even more intact and less blank. All game mechanics were implemented. Alpha version was tested multiple times and was refined based on each playtest session.

Final Game

The final game was showcased in the Intro to Game Design course. A naive player tried the game, finished the game in about 5 minutes, and commented that the game is exciting, fluent and intuitive.

Summary

In this project, I followed an iterative process centered on player feedback and rapid refinement. Initial design ideas prioritized narrative framing, but testing quickly revealed that pacing, clarity, and feedback were more critical to player experience. I continually cut unnecessary elements, restructured layouts, and reworded dialogue to sharpen the game’s core loop.

The design leaned heavily on spatial storytelling—guiding players through architecture, obstacles, and subtle cues instead of text or tutorials. Every change was grounded in how it would affect a player’s moment-to-moment understanding of what to do.