Creating Horror Roguelike Inspired by No, I'm Not a Human & Buckshot Roulette

The Story, Vision & Inspiration Behind I'm not a Psycho Invader

I'm not a Psycho Invader is a first-person horror roguelike set in a university you've seized. The roof is the only way out. Locked doors and barricades block the way up. Every hostage is valuable, dangerous, or a waste of time.

A run is one building. Floors are generated, the rooms and the people in them are randomized, and death sends you back to the ground floor. Nobody tells you who is who. Interrogate them, search their pockets, or gamble with a shot. Everything costs time, and panic makes it burn faster.

I was aiming for horror that comes from atmosphere and risky decisions rather than jump scares. No monster in the dark – just you, a shotgun, and a stranger.

The story arrives gradually, through calls and messages on the player's phone from an unknown contact. By the end, it becomes clear why you're there, who you're talking to, and how they know everything they do.

The main inspiration was No, I'm Not a Human, with its deduction mechanic of spotting the Visitors and the paranoia that builds up around it. Unfortunately, the developers there leaned more towards story, and I wanted something similar with a much heavier focus on gameplay. I've always liked how well such a simple concept as Russian roulette was turned into gameplay in Buckshot Roulette, so I took its item mechanic and combined it with the deduction mechanic. The result is a combination of one-sided Mafia and an escape room. Every hostage has a role, and to escape you have to find the character whose role fits the specific problem in front of you. To take down a security mesh, for example, you need to find the janitor – he's the one carrying bolt cutters.

Regarding uniqueness, I think the core mechanic of figuring out roles through interrogations and searches is unique in itself. But the real hook is the dark fantasy itself: seizing a university full of people you've taken hostage, and controlling their lives.

My vision for this game was to create a small, confined horror game that's oddly cozy and has its own quirks, where the player feels like a participant in a death game. A similar niche already exists in manga and anime: Future Diary, Darwin's Game, Danganronpa, and so on. That's exactly the potential audience for I'm not a Psycho Invader.

Development & Tools

I'm building the game in Unity, modeling the 3D assets in Blender, and making the characters in Photoshop. Since I work solo and I'm not an artist, the game's visuals are made up of very simple models with public domain textures. The visual style's cohesion rests entirely on a custom model shader and custom post-processing shaders like pixelation and CRT.

For comparison: the corridor with custom post-processing:

The corridor without custom post-processing:

Combining Ideas From No, I'm Not a Human & Buckshot Roulette

No, I'm Not a Human already has a fairly interesting mechanic for checking people, where you gradually gather information through the TV and the radio, but it only works for identifying monsters. That gave me the idea to use the item mechanic from Buckshot Roulette, which in the original game let you gain bonuses or information about the shells. I thought it would be interesting to add interrogations, where each of the hostage's answers narrows things down to a certain pool of roles, along with searches, during which you can find items that could only belong to specific roles. That way, with every action the player takes, they get more and more information about the hostage, and the range of possible roles gradually narrows. However, the game has a constraint in the form of time as a currency, so there's no way the player can spend ages checking everyone – at some point they'll have to start acting.

Gameplay Mechanics

The mechanics behind the obstacles and finding the right hostages are rather simple. Each floor has a set of obstacles, and as the player approaches one, a scripted intro plays. For example, a security mesh drops down and blocks the way, or a character flashes into view for a moment behind a barricade of desks. After that, the player gets a phone call from an unknown caller who hints at which roles are carrying the item they need, or which ones are the reason the obstacle is there – for instance, the desk barricade holds as long as the Negotiator is alive, so the player's objective is to kill him. From there, the player starts scouting the floor and searching for their targets.

Visual Style & Storytelling

Since I'm a programmer first and foremost, visuals were always the main stumbling block standing between me and actually making a game. That's why development started with the search for a visual style that would look good and that I could work in quickly and comfortably. I knew right away that in my case this was only achievable through shaders, so I kept experimenting with them until I arrived at this look and something finally clicked.

The storytelling side is still in early development; I can't say yet how it will end up working together with procedural generation. So it'll be a surprise for now.

Advice for Beginner Developers

When it comes to beginner game developers, I'd say that right now is probably the best time ever for small indie games. The barrier to entry is lower than it has ever been, and most games do find their audience. So here's the advice: the main thing is to just start, and to build something that sits at the intersection of what players are missing and what you personally enjoy.

You can find more information about the game on my X/Twitter, and you can support me by wishlisting the game on Steam (if you actually like it).

Max Soloha, Developer of I'm not a Psycho Invader