Plan for the average, die by the variance.
Margin of safety is deliberately building in room for being wrong. Engineers design a bridge to hold ten times its expected load; investors buy below their estimate of value; good plans assume the estimate is off. The model: never bet on everything going as expected, because it won't.
Most catastrophes aren't caused by bad averages โ they're caused by normal variance hitting a plan with no slack. The commute that's "usually 30 minutes," the project that's "about six weeks," the savings that cover "typical" expenses: all fine until the tail event arrives, and tail events always arrive eventually.
Two people each earn the same salary. One budgets 100% of it; the other budgets 70%. The same surprise car repair is a non-event for one and a debt spiral for the other. Nothing about the surprise differed โ only the margin did.
This is how Too Hard Pile trains it โ a real scenario where you commit to an answer before anything is revealed:
Quitting your job with three months saved
๐ก Runway is a guess dressed up as a number
Find one estimate you're currently relying on and ask: what happens if it's off by half? If the answer is "disaster," you've found where your margin belongs. Too Hard Pile's risk drills train you to see variance, ruin risk, and asymmetric bets in everyday choices.
Reading about mental models doesn't build the reflex โ reps do. Too Hard Pile schedules each drill back with spaced repetition until the model fires on its own, mid-decision, when you actually need it.
Short daily drills ยท spaced repetition ยท free to start, no signup to try
Try a drill โ no signup โ