top of page
Search

Jizo No Akumu

  • Writer: aka1819
    aka1819
  • Feb 20, 2024
  • 4 min read

Updated: Apr 25


A Japanese psychological stealth horror:Play as Jizo, exploring a nightmarish manor after her grandmother’s funeral, uncovering family secrets, and evading a demon.

Engine: Unreal Engine 5 | Team Size: 30 | Role: Creative Director, Producer, Systems Designer, Level Designer | Genre: 3D Psychological Horror

OVERVIEW

Jizo no Akumu is a 3D psychological horror game set in 1980s Osaka, Japan. I served as Creative Director, Producer, and Lead Designer across the project's full development cycle.


As Creative Director and Producer, I led a 30-person cross-disciplinary team through pre-production, production, and demo delivery managing timelines, aligning creative vision, and coordinating across design, art, engineering, and narrative teams to ship a 15-minute playable demo on schedule. My focus was bridging atmospheric psychological horror with modular technical systems in Unreal Engine 5, ensuring that narrative triggers remained performant while maintaining the high psychological tension necessary for the genre.


THE CHALLENGE:


Managing a large team while developing an original game raises unique production problems. Three challenges emerged early:


  • Team alignment: With 30 contributors working remotely, design intent didn't always translate clearly to other disciplines.

  • Feature scope: Early production included more systems than the team could realistically deliver within the timeline.

  • Design communication: Decisions needed to be documented and distributed quickly to avoid blocking parallel workstreams.

  • Systemic Tension: Ensuring the mechanics didn't disrupt the psychological pacing or allow players to bypass the environmental storytelling and intended scares.

A screenshot of our Trello board, showing color-coded tasks and teams. Organizing team with clear milestones accountability for a 30-person tea
A screenshot of our Trello board, showing color-coded tasks and teams. Organizing team with clear milestones accountability for a 30-person tea


HOW IT WAS RESOLVED?

Drawing on references from Silent Hill, Fatal Frame, and The Medium, I focused on realistic scope and sustainable production workflows that honored the psychological horror tone we were building toward.


  • Created standardized design documentation for cross-team reference, giving all disciplines a shared source of truth for tone, mechanics, and design intent

  • Held regular check-ins to align teams on current direction and surface blockers early

  • Established weekly syncs that gave all disciplines visibility into project status without requiring constant async communication


Silent Hill's use of environmental dread, Fatal Frame's tension through limited player agency, and The Medium's dual-reality mechanic all directly informed how we approached the astral projection system and spatial design of the shrine areas. These references helped the team develop a shared visual and mechanical language early in production, which reduced misalignment between disciplines as development progressed.


This approach reduced friction by making creative decisions explicit and reference-grounded rather than assumed.

  • A screenshot of the art team communication channel  on discord showing a positive check-in conversation
    A screenshot of the art team communication channel on discord showing a positive check-in conversation

DESIGNING THE CORE MECHANIC: ASTRAL PROJECTION:


Design Intent


The astral projection mechanic allows the player to separate from their physical body and explore the environment in a spirit form. The goals were:


  • Create dual-body puzzle potential

  • Enable low-risk exploration moments

  • Give players a new lens for reading environments without hand-holding


Results


  • Players understood the core interaction without explicit instruction during playtesting

  • The mechanic gave players multiple approaches to solving environmental puzzles

  • Exploration felt driven by curiosity rather than external pressure

  • Design documentation formalized the mechanic's rules for the broader team, reducing implementation inconsistencies


After iteration, players demonstrated a clear understanding of the mechanic within the first few minutes of each demo session.


Jizo's physical body is left behind in front of the door in order to search for clues, and he returns to the body once a solution is found.
Jizo's physical body is left behind in front of the door in order to search for clues, and he returns to the body once a solution is found.

PUZZLE AND LEVEL DESIGN


Design Problem


The game's shrine area was initially too dark and ambiguous. Players lost their sense of direction because there was insufficient contrast between readable navigation paths and background elements.


Design Changes


To improve spatial readability:


  • Introduced environmental cues lanterns, architectural details, and interactive objects to guide the player's eye toward the critical path

  • Adjusted lighting contrast to emphasize navigation routes without breaking atmospheric tone

  • Shortened distances between key puzzle progression points to reduce disorientation



SHRINE PUZZLE DESIGN


The original shrine puzzle created confusion around building spatial intuition and collision-precise platforming.


I revised the puzzle by:


  • Reducing the number of required steps to reach the solution

  • Extending the action window on key interactions to reduce frustration

  • Aligning the puzzle logic with the game's narrative context so solutions felt motivated by the world rather than arbitrary


This made the shrine feel intuitive while preserving meaningful difficulty.


The Shrine Puzzle: the player previously used the astral projection puzzle to find the key to the puzzle in order to unlock the door.
The Shrine Puzzle: the player previously used the astral projection puzzle to find the key to the puzzle in order to unlock the door.

The Technical Pivot: Astral Projection & Vulnerability


During early playtests, the Astral Projection mechanic felt detached from the core psychological loop; players were using it as a safe scouting tool, which inadvertently lowered the sense of dread. To fix this, I re-scripted the transition logic in UE5 to leave the physical body vulnerable during projection. While players are encouraged to find 'safe zones' before transitioning,


I implemented a negative feedback loop where remaining in spirit form for too long increases the likelihood of the physical body being targeted. This forced a high-stakes temporal calculationplayers are not just searching for safety, but racing against their own physical exposure. This adjustment successfully re-integrated the mechanic into the game's psychological pillars by ensuring that total safety is never guaranteed.


WHAT I WOULD IMPROVE


I would implement earlier vertical slice milestones to reduce scope creep across the team. Defining a hard content lock earlier in production would have freed up more time for polish and iteration on the core mechanics rather than managing late-stage feature additions. I would also formalize the design documentation process at the start of pre-production rather than building it reactively as communication gaps emerged.


CONCLUSION:


Jizo no Akumu taught me that directing a team is itself a design problem. The systems that keep a team aligned are just as important as the systems players interact with.


This project strengthened my ability to:


  • Led cross-disciplinary teams through production hurdles by translating creative vision into stable, hands-on technical solutions that preserve the integrity of the psychological horror experience.

  • Design and implement original game mechanics from concept through playable demo

  • Build production infrastructure documentation, syncs, scope management that supports creative work rather than slowing it down


See the complete 15-minute demo below, including the astral projection mechanic, shrine puzzle sequence, and environmental storytelling in action.









 
 
 

Comments


bottom of page