# 37% Rule Created: [[2023_08_19]] 15:40 Tags: The 37% Rule, also known as the Secretary Problem or Marriage Problem, is a famous problem in probability theory and statistics. It's about optimal stopping: when you should stop looking and make a choice, given that you can't go back to previous options. Here's how the problem is typically framed: Imagine you're interviewing a series of secretaries for a job. You can interview each one by one, but once you pass on a candidate, you can't go back to them. The challenge is to determine the optimal strategy to maximize your chance of selecting the best secretary from the pool. **Here's how the 37% Rule works:** 1. **Divide your search into two phases**: - The "look" phase where you're just getting a sense of the candidates. - The "leap" phase where you're ready to make a selection. 2. **Spend the first 37% of your search (the candidates or time) just looking**. Don't commit to any option during this time, regardless of how good they seem. This is about establishing a benchmark. 3. **After that 37% period, leap at the next best opportunity**. So, in the context of the secretary problem, you would hire the next secretary that's better than all the previous ones you've seen in the first 37%. Mathematically, using this strategy gives you a roughly 37% chance of choosing the best option, which is the best probability you can achieve given the constraints of the problem. **Applications in real life**: The 37% Rule doesn't just apply to hiring secretaries. It can be used in various scenarios where you have a limited set of options and limited information, and once an option is passed over, it can't be revisited. For example: - **House hunting**: If you're looking at a certain number of houses over a set period of time, consider spending the first 37% of that time to just explore, and then be ready to commit to the next house that meets or exceeds all the others you've seen. - **Dating**: If you consider dating a certain number of people over a period of your life before settling down, the 37% Rule suggests you spend about 37% of that time getting to know what you want in a partner, and then commit when you find someone who's better than all the previous partners. It's important to note that while the 37% Rule provides a mathematically optimal strategy for these types of problems, real life is more complex and unpredictable. The rule is more of a guideline or a heuristic, not a definitive strategy for all decision-making scenarios. However, it does offer a fascinating insight into how we can approach problems of uncertainty and decision-making. ## References - [[Algorithms to Live By - Brian Christian|Algorithms to Live By]]