Dammed
Finishing a glyph-based narrative game two cohorts built but nobody could play. University of Canterbury, 2025 – ongoing
Dammed is a narrative exploration game where the entire world is drawn in coloured glyphs, built so students can write interactive stories in Ink without wrestling with graphics. It began as a 2023 class project, picked up character art from a 2024 group, and by the time it reached me it had a good story, good art, and no way for anyone to actually play it.
This was a solo pass to fix exactly that: the UX and UI work that turns an inherited, half-finished prototype into something a player can pick up, save, and finish.
The problem was that nobody could finish it
What I inherited had all its value sealed off. There was no save, so the game demanded an uninterrupted multi-hour sitting. There was no objective tracking, so players forgot what they were doing and wandered. Nothing taught the controls, and the core mechanic, walking into a character to talk to them, was not something anyone guessed. Every playtest ended the same way: someone quit, politely, before the story went anywhere.
The interesting part was not the plot. Two cohorts had built something genuinely good and nobody could reach it. The job was to take the wall down without rebuilding the house.
I started from watching people, not the code
Think-aloud sessions with players who had never seen it, notes on every "wait, how do I…", then fix the most blocking thing and watch again. The failures repeated across everyone, which made prioritising easy: the same three walls stopped every tester, in the same order. That consistency is what turned a vague sense of "it's rough" into a ranked list of things to build.
Extend, don't overwrite
Two cohorts' work was already in the codebase, and the glyph renderer and the brown-and-cream identity were part of what made the game itself. Every fix had to sit inside the existing architecture and visual language rather than replace it, which is a different discipline from building from scratch: solving new problems without breaking work you didn't write.
Three walls, in the order they stopped people
Save. The single biggest blocker. Without it, the game was a multi-hour commitment or nothing. I added a pause menu on Escape, wrote saves to the browser's local storage, and surfaced a Resume option on the title screen so a saved game announces itself. Players started taking breaks and coming back, which is the whole game right there.
The pause menu. Save and resume are the difference between a game you can play and one you can only attempt.
Quests. People weren't lost because the game was hard. They were lost because they couldn't hold the objective in their head. I built a quest panel on Q that splits active from completed, tracks steps, and keeps the narrative reason for a task sitting right next to the task. Quest content lives in JSON, separate from the code, so the writing can change without anyone touching JavaScript.
Active and completed quests, with the story reason kept beside the objective instead of buried three dialogue screens back.
Reading the story. The hardest call. The 2024 art was full character illustration; the world is bare glyphs. Forcing them together would have cheapened both.
Keep them apart on purpose
Exploration stays glyphs; the art only appears in dialogue, in its own overlay. The switch itself signals the shift from moving to talking. The same overlay handles characters with no art, where the text simply reflows to fill the space, so the reader never sees a gap where a picture should be.
Art appears only in conversation, in its own overlay, so the glyph world and the illustration never fight for the same screen.
The same overlay with no art available. Text reflows to fill the space rather than leaving a hole.
What changed
These are observations from repeat think-aloud sessions with a small group, the same testers before and after, rather than hard metrics, but the shifts were consistent enough to act on. Players who used to bail in the first five or ten minutes were sitting down for half an hour to an hour. The constant "how do I…" questions dropped to the occasional one. People who physically couldn't return, because there was no save, started coming back across sessions. And instead of wandering, they moved at objectives. None of it was new content. It was the same game with the friction removed.