Overview / Usage

Ginkgo is a single player 3D horror game set in a dreamlike space inspired by East Asian horror. Armed with a magic needle that can sew parts of the world together, solve puzzles and outsmart monsters as you explore this surreal world.

Playing as a young woman at her estranged mother's funeral, a chance sighting of your mother's ghost allows you to explore a dream world that shows you life from her perspective. Gain a greater understanding of why your mother, a famous seamstress who worked herself to the bone, acted the way she did and potentially reconnect after her death.

Features:

• Innovative needle and thread mechanics combined with unique 3D spatial puzzles
• A stylish, mysterious, and engrossing world modeled after East Asian psychological, spiritual, and supernatural horror elements
• A beautifully crafted and cohesive experience

This game is being developed under a year-long cycle as a USC Advanced Games Capstone Project. We work in week-long sprints, utilizing Agile/Scrum methodologies for task distribution. Our goal is to showcase a highly polished and cohesive experience in our debut at the USC Games Expo 2020.

The Team

• Creative Director: Joey Tan
• Production Lead & Narrative Lead: Milo Smiley
• Design Lead: Justin Lu
• Level Designer: Michelle Pina
• Usability Lead: Clerisse Cornejo
• Engineering Lead: Francesco Galletti
• Engineer: Christina Guan
• Engineer: Jeffrey Lei
• Engineer: Connor Aguilar
• Engineer: Katie Moulton
• Art Lead: Wayne Yip
• Art Director: Andrea Zhang
• Art Director: Nolan Lu
• Animator: Haley Tomaszewski
• 3D Artist: Jamie Liao
• Tech Artist: Sichen Liu
• 2D Artist: Quiana Dang
• Audio Lead: Kevin Ke
• Composer: Anthony Sabatino

The Problems

Like every other game development project, we have had to execute a fair amount of problem solving during production. Some of the largest issues we have resolved include:

Inconsistency with Cutting

While sewing has worked in our game since its prototype, cutting thread with the needle worked only part of the time in the game’s build while working all the time in editor.

Enemy Tuning

The central enemy of our game, a monstrous being called the Chase Yokai, was initially unbalanced and made the experience too difficult for playtesters to complete.

The Solutions

While these are specific solutions, each included group focused problem solving with communication as a priority, extensive research, and advice from the USC Games faculty as guidance.

Inconsistency with Cutting

To make sure it wasn’t an issue with our code, we began by isolating the code which specifically affected cutting in game to test each line via print and debug statements. While we found and resolved a few minor errors, the overall issue remained. Research showed us this could potentially be an issue with the way our editor built spline mesh collision, so we changed course and began a new search for creative ways to emulate cutting.

What we ended up on was a modification of existing code used by enemy AI that also cut thread. Instead of communicating with spline meshes, an invisible “pole” extends from the player character to check for sewn thread that the player is targeting. When the player cuts thread, the pole activates as a shredder and deletes it, mirroring the initial cutting motion almost perfectly.

Enemy Tuning

As a solution to this issue, the Chase Yokai’s experience was redesigned starting with its levels. Set dressing was added to the maze like levels with specific props and colors used to subtly guide players to the exit. This gave the impression of being stuck in a maze with a monstrous beast while allowing players to continue making informed actions which rewarded them for an attention to detail.

The Chase Yokai itself also underwent a minor behavior overhaul, its speed slowed significantly and its vision limited. Its gameplay had a new emphasis on stealth: instead of being chased by the Chase Yokai and having to slow it down, players instead had to pick their way through the level without being seen or heard. The quick interactions with the needle were changed to slower, more calculated movements which reinforced the stealth mechanic by allowing players to create red herrings, trip the Yokai, and even sew up windows and walls to hide from its gaze.

Playtest feedback with these changes showed players getting though these sequences successfully without losing any of the fear and tension from before.

Methodology / Approach

There are a couple key methodologies we have been following since the beginning of development that have contributed significantly to our production process. These include:

Player Centric Design

Ginkgo uses player centric design, or designing around a player’s experience and iterating based off player feedback. With players as our priority, the first thing design created was our player experience goals, which are constantly referred to when considering new levels or mechanics. Sticking to these goals throughout production not only allows us to create a unified, cohesive experience, but it gives us an unchanging point of reference to return to when puzzling over difficult design decisions.

RITE Testing

We have also been using Rapid Iterative Testing and Evaluation, or RITE testing in our playtests to for quicker iteration and tuning of mechanics. As opposed to conducting 6 to 8 playtests to see if a mechanic isn’t working and then deciding to fix it, we run about 1 to 2 players and address issues when they come up. This allows our team to jump onto issues faster, produces faster iteration, and gives us a playtesting format that meshes well with the limited scope of student game playtesting and the time constraints of college students.

Outside of these specific practices, we heavily prioritize communication down to the individual. We make sure every team member feels that not only does their work contribute an important part to a greater whole, but that their input is being heard in a welcoming environment. Production does individual check-ins after every milestone to hear what is/isn’t working, so we can continue crafting a production process that prioritizes the health and happiness of our devs.

Technologies Used

The technologies used to develop Ginkgo included:

• Unreal Engine
• Perforce Version Control
• Microsoft Visual Studio
• Autodesk Maya
• Blender
• Substance Painter
• Adobe Photoshop
• Logic Pro X

Other links

• Ginkgo Instagram
• Ginkgo Facebook
• Ginkgo Official Website

Collaborators