Does NumWar's Core Loop Hold Up?
Hypothesis
Yes. Rock-paper-scissors gives fair tension, knocking down the tower turns subtraction into a visible thrill, the three-way choice provides strategic tension, and short, comeback-friendly rounds lower the barrier to “one more round.”
Key metric
Spontaneous “one more round” rate R ≥ 50%
The question
Can a minimal game that wraps addition / comparison inside a "rock-paper-scissors → knock down the tower" battle loop get children to play repeatedly on their own, and spontaneously count, compare, add, and subtract along the way?
The hypothesis
Yes. Because:
- Rock-paper-scissors gives a zero-cost fair start and tension;
- "Knocking down the other's tower" turns subtraction into a visible, satisfying action;
- The three-way choice of attack / build / defend provides just-enough strategic tension;
- Short rounds + comebacks lower the barrier to "one more round."
Experiment design
- Use NumWar's physical rules (playing cards / dice + blocks).
- Sample: ≥ 5 children.
- Observe and record — no guiding, no teaching.
- Record round length, friction points, and the verbatim brightest moments.
Key metric
Spontaneous "one more round" rate R ≥ 50% — the hypothesis holds.
Results
To be filled in. After collecting ≥ 5 test records, summarize R, the distribution of round lengths, the top 3 friction points, and the verbatim brightest moments.
Reflection
To be filled in. Write it down win or lose — we record "what we learned about how children understand the world," not just the wins.
This is momomori's first experiment template. Every later ability branch reuses this structure for its core validation (EXP-002, EXP-003…).