Stress test the complete system. Version 0.3 must loop cleanly for 5 minutes.
Today is not a building day. Today is a playing day. You play your game silently for 5 minutes. No code changes during play. You observe what works and what doesn't. Write down what you notice. Then fix only critical bugs. Then play again.
Loop Test: Explore → Encounter → Win/Lose → Risk increases → Bank or Respawn → Repeat.
Requirements:
- Risk visibly increases over time in the danger zone
- Loss resets unbanked XP and respawns to Camp
- Banking feels safe — risk resets, XP is secured
- Tier system doesn't exist yet (that's Week 4)
Loop Test: Accept Job → Buy Materials → Build → Earn Karma → Risk debt → Upgrade → Repeat.
Requirements:
- Debt pressure is visible and increasing
- Bankruptcy resets correctly when debt exceeds -50
- Decision between safe/small vs risky/large jobs is clear
Does the loop function without explanation? Can someone who hasn't seen it before understand what to do?
No new features. Only bug fixes from today's play session. Resist the urge to add "just one more thing."
Fixing bugs during play instead of writing them down — play first, fix later. Observation and coding are separate modes.
Adding new features instead of fixing existing ones — today is about stability, not novelty. New features break what already works.
Not playing the full 5 minutes — the whole point is sustained play. Short tests miss the problems that emerge over time.
This is a discipline exercise. Watch silently while they play. Do not help. Do not suggest. Write your own notes. After they play, compare notes. The question is not "Is it perfect?" but "Does the loop work?" This teaches the difference between a prototype and a polished product.
Both games have working 3D world, movement, event triggering, win/lose resolution, risk escalation, banking/debt tension, and persistence. The core loop is playable.