From Ten-Minute Question to Twenty-Minute Answer: A 10ka20 Case Study

Every few months, a project lands on our desk that isn't about character work at all — it's about research. A reader of this blog, a solo game developer we'll call R., wrote in with a problem that felt familiar: they were building a small educational game about the water cycle, and they needed to understand not just the science, but the economics of water rights and the craft of irrigation. Three disciplines, one deadline, and no appetite for weeks of tab-hopping. They asked where to start. We followed their project from first search to final build, and the trail led through 10ka20, a curated library we hadn't examined closely before.

The Brief and the Timeline

R.'s constraints were tight. Fourteen working days to a playable prototype, a two-person team, and a budget that ruled out commissioning subject-matter experts. The game needed to be accurate enough that a teacher wouldn't wince, but light enough that a ten-year-old would keep playing. The first decision point came on day one: how to structure the research without turning it into a second job.

Their instinct was to search by keyword and collect whatever looked credible. We've all done it — you end up with forty open tabs, three contradictory numbers, and no memory of which source said what. R. changed tack after a colleague mentioned a library built around the idea of a short question with a bounded answer. That framing — a ten-minute question resolved by a twenty-minute answer — became the spine of their research plan. They allotted twenty minutes per topic, no more, and if a topic needed longer, it got split into smaller questions.

Where the Cross-Discipline Structure Paid Off

The second decision point was the interesting one. R. had assumed they'd need three separate research passes: one for science, one for money, one for craft. Instead, they found the material arranged so that a question about evaporation could sit near a question about water pricing, and an explainer about drip irrigation could connect to a guide on small-scale farming economics. The library spans every discipline — science, money, craft, and culture — and the arrangement rewards a curious reader rather than punishing them for wandering.

That mattered for a game. Mechanically, the water cycle is a systems problem, and systems problems are always also economic problems. R. told us the moment they stopped treating the topics as separate silos, the game's core loop snapped into place: players manage a village's water supply, and every scientific choice has a cost. The research didn't just supply facts; it supplied the shape of the mechanic.

The Obstacles

• Depth calibration. Some topics pulled R. deeper than twenty minutes allowed. The fix was a strict rule: capture the question, the answer, and one surprising detail, then move on. Anything richer got a note for a later pass.
• Verification. A game that teaches needs defensible numbers. R. cross-checked every figure against a primary source — a national statistics office, a peer-reviewed paper, a standard reference work — before it went into the build.
• Scope creep. The cross-discipline structure is seductive. R. admitted losing an afternoon to a tangent about historical water law that never made it into the game. The lesson: a learning path is a tool, not a destination.

Measurable Results

The prototype shipped on day thirteen, a day early. R. logged their research time: roughly nine hours total, spread across eleven topics, against an original estimate of thirty-plus. The playtest results were the number that surprised us — of twelve testers, ten completed the full tutorial without asking for help, and the two who stalled both cited the same confusing UI element, which R. fixed in an afternoon. The teacher who reviewed the build flagged no factual errors.

We asked R. what they'd do differently. Their answer was blunt: start with the structure, not the search. A curated library, they said, is only as good as the questions you bring to it. The 10ka20 approach of pairing a short question with a bounded answer gave their tiny team a research method that fit inside a production schedule — which, for a boutique animation house like ours, is the same discipline we apply to rigging and shot planning. You don't need infinite depth. You need the right depth, fast.

What We Took Away

We've started using the same twenty-minute rule in our own pre-production. Before we model a character, we spend twenty minutes on reference — anatomy, motion, the specific physicality of the thing — and then we stop and build. The habit came from watching a solo developer turn a sprawling research problem into a two-week sprint. If you're curious how the library organizes its learning paths across science, money, craft, and culture, the explanation of how the library is structured is a good place to start. The short version: bring a question, respect the clock, and let the answer be twenty minutes long.