What is the 37% rule?
In: 0
The rule is the solution to the [Secretary Problem](https://en.wikipedia.org/wiki/Secretary_problem).
The problem is this: you’re interviewing applicants for a secretary position. You have a certain number of applicants, say 100. After interviewing a candidate you can either hire them or let them go, but once you let them go you can’t decide to interview the again or hire them, even if they were better than every other applicant.
The problem is to find an optimal method of hiring the best applicant. As it turns out, from a probability perspective, the best approach is to let first 37% of the candidates go (after interviewing them of course), and then keep interviewing and hire the first candidate who is better than all previous candidates. Statistically speaking, this gives you the best chance of hiring the best candidate.
The secretary problem would be best solved by having a properly designed interview with an actual mark grid and several interviewers, then setting a target score to hit in order for your secretary to be hired EG. First person to score 80/100 is deemed to be good enough and then the process stops, whether it’s after one candidate or 98.
The 37% is an odd solution which I’m sure would never be applied in the same sort of real-life context.
Latest Answers