37 Percent Rule Calculator

Look without committing for the first 37 %, then take the first option better than everything you have seen. Over an open time window your chance of landing the best one is 36.8 % — and it stays 36.8 % however long the window is. Knowing the exact number of candidates does slightly better.

This is the secretary problem, the classic optimal-stopping model: candidates arrive one at a time, each must be accepted or rejected on the spot, and none can be recalled later. The strategy that maximises your chance of picking the single best one is to look without committing through the first 1/e of the sequence — about 36.79 percent, usually rounded to 37 — and then take the first candidate who beats everyone seen so far. The number this calculator returns is that chance of success. In the time model it is the constant 1/e itself, 36.8 percent, however long the window: someone searching from 18 to 35 switches from looking to leaping at 24.25 years, roughly 24 years and three months. With a known number of candidates it can be computed exactly, and small pools beat the limit — 43.3 percent for five candidates, 39.8 for ten, 38.4 for twenty, 37.1 for a hundred, falling back toward 1/e as the pool grows. The caveat that matters most is how narrow the goal is. The rule maximises the probability of landing the absolute best candidate, not average satisfaction and not minimal regret. It also assumes you only ever learn whether someone ranks above or below those already seen, and that the two phases stay strictly separate.