Shooting a Still Bird (Bitsy)
An adaptation of The Wet Nights at Beat Beach, an early version of a longer work that has since become Like a Mountain of Sleep. Retitled Shooting a Still Bird, the game places the player as an inebriated spring breaker navigating the final twenty minutes of a slasher film that never existed: trapped in a looping seaside manor by a narrator who built the trap and is watching them move through it. The player has no voice and no interiority. The narrator, the objects, and a cat have all of it.
Reflection
Part of what makes this interesting to me as a project is that Beat Beach functions here as a prior draft being reactivated — the game is in some sense an earlier, rawer version of the world that Like a Mountain of Sleep more fully develops. That layering feels appropriate to the work’s concerns: the text the player collects describes their own situation back to them in fragments they haven’t earned yet, drawn from a version of the story that predates the version they might encounter elsewhere. My work tends not toward the personally expressive — this is not a game about my interior life and I want the player to think mostly about what it means to move through a space designed against you. The feeling I want the player to take away is something like: I was being watched the whole time, and I was reading about it before I understood it. I want this narrative to be a bummer.
Johnson and Salter’s point about accessibility resonates here. The barrier to entry that engines like Unity and Unreal impose — requiring highly specialized skills or the time and resources to acquire them — is precisely what makes Bitsy interesting as a platform for this kind of work (Johnson and Salter 139). It removes that barrier without removing the capacity for meaningful critique but encourages the use of “simple” play. Their observation that indie games are “incredibly well-suited to conduct thoughtful critiques on society” because story and character drive the experience rather than console-controller mechanics feels directly applicable (140): Shooting a Still Bird has no mechanics to speak of. Movement is the only verb. That limitation is the argument.
What I find generative about their framing is the productive tension that surfaces around image-making in Bitsy specifically. The imagery in this game would be compelling if photorealistic: the manor rooms, the subterranean flower field, the staircase with false perspective. I generated a series of AI concept images during development — the coat of arms hallway, the painting room with water seeping from the frame, the shirt, the limestone circular chamber — not because they were directly usable in the engine, but because they helped me envision the production logic before I understood how Bitsy actually handles images. Working through that misunderstanding was itself productive. But the 16-bit constraint forced me to think carefully about how to use a two-dimensional space to create abstraction and disorientation rather than representation. This puts me in a different relationship to the pixel grid than some designers choose: rather than using it to approximate depth-of-field or high-definition environmental detail, I’m treating it as a formal constraint that does rhetorical work.
bee ulrich’s description of Bitsy games as poems rather than tomes captures something I keep returning to: “handcrafted spaces with so little headroom that every last pixel is placed with purpose.” The clearest example of this pressure is the sprite work. The game’s entry corridor is populated by disembodied arms — a direct citation of Cocteau, whose films use architectural limbs as a surrealist mechanism for animating otherwise inert space. To make this work interactively, each arm had to be built as an individual sprite, meaning I constructed dozens of distinct pixel figures: right arms, left arms, column variants, cat iterations, arm pairs for multiple corridor rooms. The interactivity that results is strange in the way Cocteau’s interactivity is strange — the arms gesture, deflect, speak — but building it required a granular, almost absurd level of repetitive pixel labor that the engine doesn’t abstract away. That felt right for a game about a trap.
The flower room operates on similar logic. Because the flowers are meant to be bioluminescent, I made them sprites rather than tiles, which allowed me to assign them dialogue and give the room a blue palette. The consequence is that the avatar and cat are likewise rendered in blue, since Bitsy’s color logic applies palette-wide. On reflection, I should adjust the tile colors to better homogenize the space — but the decision itself was productive, forcing me to think through how sprites, tiles, and dialogue points each carry different affordances and can be leveraged against each other to produce effects the engine doesn’t straightforwardly offer. I’m distributing the full text of the pamphlet across 26 of these flower sprites as multi-page dialog. Most are inaccessible to the player without walking through others. That felt right: partial vision, wounded body, text you can almost reach.
Bitsy is also genuinely low-code in a way I didn’t fully appreciate until I was inside it. I didn’t need to use AI to build this game. I’m still exploring Borksy for potential expansion, but the core work has been done within Bitsy’s native constraints, which feels appropriate to the project’s argument about what constraint can do.
This project has also sharpened my awareness of what it means to engage with games from an outside. My girlfriend played an early build and noted that it felt aesthetically reminiscent of Undertale, a game I know essentially nothing about beyond the fact that Dr. Salter has written about it. That comparison stuck with me, not because I can evaluate it, but because it suggested the work was legible within a tradition I hadn’t consciously drawn from. Exercises like this are leaving me increasingly aware that I want to make time to actually play games.
I plan to host the finished game on itch.io, though I consider this very much a proof of concept at this stage.