Pre-release multiplayer RPG · Late 2025 - present
Lunacia: The Original Dream
Leading a 3D online game that combines farming, home decoration, and PvE, while using AI to help a small core team handle a much larger product scope.
Context
In late 2025, I had an opportunity to pitch the next direction for Project Idle to the two founders of Sky Mavis: Trung Nguyen (CEO) and Tu Do (Masamune, Art Director). The pitch was approved, and the team received game asset support for a 3D version with deeper gameplay.
That was when Lunacia officially became a new project rather than simply a 3D version of the old game. Everything we had built and learned in Project Idle became the starting point, while Lunacia’s gameplay and systems were redesigned for a much larger scope.
From product direction to implementation
On Lunacia, I own both product and technical work: shaping the product, planning milestones, setting priorities and scope, then directly implementing key systems. The team has up to five people and works around milestones rather than running a formal Scrum process.
On the technical side, I work directly across Unity, Colyseus and Node.js, MongoDB, and Redis.
Product scope
The initial gameplay pillars were farming, home decoration, and PvE. Once we began connecting them, the economy and the number of required systems became far more complex than Project Idle. Product, art, and engineering work all grew with that scope.
I did not want to add people simply to chase the workload. At the same time, AI became a meaningful way for the small core team to maintain speed without scaling too quickly.
Current status
As of August 2026, Lunacia production is temporarily paused while I look for more funding and support. The previous runway ended before the game reached release, but I have not abandoned it. I am continuing to redesign the economy around seasons, clearer sinks, and a model that can scale more sustainably without pushing new or returning players too far behind.
Lunacia remains pre-release. It has not launched and has no live-operation or production user metrics to report. The redesigned economy is also still in the design stage and has not been validated in production.
Team
The team already knew how to work together
By the time we started Lunacia, the core team had worked together long enough that ownership, coordination, and decision-making were already clear. I was no longer spending much time on the staffing and operating problems we had faced during Project Idle.
The difficult work shifted toward technical barriers. Because Lunacia is an online game, we had to decide which state belongs on the server, how data stays synchronized and persistent, how reconnect works, and how to keep the game light enough for mobile.
From the first prototype to the latest public build
These are the development milestones I still have on video, from the first prototype to the point where the game began to take shape. The newest internal build is being redesigned for UI/UX and optimized for mobile, so it is not public yet.
Online-first architecture
Online gameplay took more effort than I expected
In an offline game, the client can handle many things on its own. Once Lunacia became an online game, nearly every state that affects gameplay or the economy had to be validated by the server. Trusting the client would eventually make cheating unavoidable.
In practice, a feature that costs one unit of effort offline can take three to five online. Beyond gameplay, we also need protocol design, validation, persistence, concurrency, reconnect handling, error recovery, and the client presentation that connects it all.
Playtest · 07/2026
30-40 people tried the build, but only one group finished the content
In July 2026, I invited approximately 30-40 people I knew to playtest the game, share feedback, and report bugs. The most encouraging result was a close group of 6-7 players who kept pushing through all available content and defeated the boss.
Most other players created a character, completed a few quests, and logged out. The build was still incomplete and the opening was not strong enough to retain them. I cannot call it a successful retention test, although the PvE content did keep one genuinely interested group engaged.
Load test
Approximately 300 CCU in the tested Town scenario
In the latest load test, the game server handled approximately 300 concurrent players on the Town map. I treat this only as the result of that specific scenario, not proof that the entire game is production-ready at 300 CCU.
Insight · Game Economy / System Design
Seasonal economy - how do you reset without making players leave?
I am exploring a soft-reset model with legacy, giving the economy room to recover without erasing the player's effort and history.
Insight · AI-assisted UI Pipeline / Design Systems
Figma → Unity with AI - from 2-3 days to 4-8 hours
I split the workflow into an archive cache, component mapping, query helpers, and validation gates so AI can implement UI faster while still following the design system.
Retrospective
The biggest lesson was not technical
The lesson was painful but clear: I did not hold the scope and MVP milestone tightly enough. We kept adding features because every one of them felt necessary, while failing to lock a smaller build that could reach players sooner. The scope grew, the runway shrank, and the budget ran out before the product was ready. If I did it again, I would set a smaller scope, make the release date non-negotiable, and validate with players before expanding.
Role & responsibilities
- Created the product vision and led the project through major changes in concept, gameplay, and technology.
- Owned the design and delivery of the persistent multiplayer world, progression and economy, inventory, crafting, equipment, combat, social features, mobile UX, and backend operations.
- Currently redesigning the economy around seasonal soft resets and legacy, with guardrails for inflation, supply and demand, sinks, catch-up, and open trading.
- Set the technical direction and directly implemented key Unity and backend systems.
Results
- Evolved Project Idle into a new 3D product with a broader scope and more layers of gameplay.
- Pitched the direction directly to Trung Nguyen (CEO) and Tu Do (Masamune, Art Director), the two founders of Sky Mavis, and received game asset support for the 3D version.
- Ran a closed playtest with approximately 30-40 people in July 2026; a group of 6-7 players continued through all available content and defeated the boss.
- Reached approximately 300 CCU on the Town map in the tested load scenario; this is not a production capacity claim for the whole game.
- The project remains pre-release, with no launch or live-operation metrics yet.