Iteration 1 started as most game projects do, with an idea. Four of us set out to build it: myself, Kay Blake, Brian Tromba, and Xander Hacking. We envisioned a whimsical roguelite dungeon shooter centered on a wizard wielding a gun and devised a story to support it. The story was straightforward and focused on a wizard who gets evicted from his tower by goblins, grabs his gun, and ventures out from a temporary home base in a quirky mini-mall to reclaim his tower.
Our team split into 4 roles, I would handle level design, Kay would work on art, Brian would create the AI, and Xander would focus on the character. We built rudimentary systems during this phase, but several core ideas persisted including randomized level generation, fast-paced movement, enemy variety, and a spellcasting mechanic.
Once we had a playable prototype, we pitched it and received encouraging feedback, enough to convince us there was an audience for this concept. However, life called for all of us, and we took a collective break before gearing for the real work.

Starting Iteration 2, the team regrouped with a fresh codebase and a shifted roster. The new lineup was: myself as designer, Kay Blake as artist, Lauren Sterling as programmer, Brian Tromba as AI programmer, and Xander Hacking as leadership and character controller programmer. Building from scratch meant reconsidering everything. The “wizard in a mini-mall” premise no longer fit our vision. We explored new directions such as fighting to recover stolen memories instead of a tower, deciding whether the primary weapon would be spells or the gun, or how to introduce level variety. For example, our early concept featured a world map centered around a colossal skeleton warrior where players would traverse his remains and enter dungeons around different appendages. While the narrative took a backseat to systems work, this visual grounded us for the time being.
Formalizing our commitment, we founded Grimoire Studios as our publishing LLC. I personally built the hub world and testing dungeon while the team tackled core systems such as spell casting, random generation, and movement. The AI work lagged behind, and community feedback on both the AI units and movement felt weak. I shifted focus to help stabilize these systems, taking on programming duties alongside design.
Shortly after, Brian had to step back from the project entirely. I took over as Lead AI programmer, and while trying to match my new AI to the movements state machine structure I discovered something. Examining the movement system's “Hierarchical” Finite State Machine, I found it was fundamentally broken. Around the same time, Xander encountered personal difficulties that forced him to step down from the project. He handed over programming responsibilities to me and leadership to Kay.
This meant I was now juggling level design, AI programming, and ongoing movement system work and was heavily stretched. However, this period coincided with community playtesting becoming more active, and their feedback was invaluable in guiding our next phase. After navigating these shifts and consolidating the codebase, the team took a month-long break to regroup.

Iteration 3 was our refinement phase, driven almost entirely by community feedback. With me split across multiple systems, I organized my work into focused sprints: first the character controller until the community felt satisfied with its responsiveness, then level design variety (adding Sky Islands, Scholar's Tower, Dungeon, and Sewers), then an intensive AI rebuild where I fixed the entire system and built unique enemy types for each zone. Between playtests and value tweaking, combat finally felt right.
It was also in iteration 3 that gameplay saw a major shift. Originally the story followed our wizard as he traveled dungeons from his mini-mall, trying to reclaim his tower. It then changed to him exploring dungeons across the land covered in the remains of the a giant. In this last change we meshed ideas and cut scope creep focusing on what we could plausibly achieve. Instead you would fight through waves in dungeons you explored, working your way through the different zones to make it to your tower. Only then could you face your nemesis spell sniper and reclaim your tower. While it didn’t feature a overly complex story, it was just the right of story involvement for our game to feel alive.
The final stretch was polish. I rebuilt most levels, retuned character values, optimized the AI across all zones, and refined the overall experience. This stretch involved heavy crunch time, but by the end, the game had come together. When it came time to publish, we made a decision. Rather than ship at the originally planned $5 price point, we decided to release for free on Steam. Despite our best efforts, the team agreed the product wasn’t at a level we would describe as marketable as a paid product. The community response was positive and the game landed at 88% positive across reviews. The community's positive reception vindicated our decision to release for free and our focus on quality over monetization paid off.