Strategy sweep
Every strategy below plays the life simulation four hundred times, on the same four hundred seeded lives, driven by the same function a human player will drive. This page exists so that the mechanics can fail here rather than in front of a classroom.
Running...
What this page is looking for.
Four ways a simulation can be broken without looking broken.
- Passivity wins. A player who never touches a control does as well as one who thinks. The simulation is then scenery.
- A dominant strategy. One approach beats every other on every outcome at once, so there is a right answer and therefore no decision. Checked as Pareto dominance across wellbeing, money, experiences, when you stop working and whether you run out, not by reading the top of a table. For a long time this page claimed that and the code underneath compared the top of the table to the middle of it on wellbeing alone. It does what it says from 2026-08-03.
- A lever that does nothing. Two strategies differing in exactly one decision produce the same distribution, which means that decision is decoration and should be cut rather than shipped.
- One deck, dealt twice. A single playthrough is eligible for nearly every card, so replaying deals the same year again. Measured as the share of the pool one life can draw from, plus a check that no card is gated so hard that nobody ever meets it.