The Lost Feed

🔬Weird Science

Inside the Curious Math of Choosing the Best Candidate

Ever wonder how to pick the best person from a group when you only get one shot? Discover the surprising math behind making perfect choices, like hiring a secretary.

1 views·5 min read·Jul 20, 2026
The Secretary Problem

Imagine you have a big decision to make. Maybe you are hiring for an important job, or looking for a new apartment. You know you have a certain number of choices, but you can only look at them one by one.

Here is the tricky part: once you pass on a choice, you cannot go back. You must decide on the spot if the current option is the best you will ever see. How do you make sure you pick the absolute top choice, knowing you only get one chance?

The One-Shot Decision Dilemma

This exact problem has puzzled smart people for decades. It is often called the Secretary Problem, but you can think of it as any situation where you must pick the best from a sequence of options. The rules are always the same: you know how many options there are in total, you see them one by one, and you must accept or reject each one immediately.

Once you reject an option, it is gone forever. If you reject all options, you are left with nothing. The goal is to maximize your chance of picking the single best option out of the entire group.

Imagining the "Secretary Problem"

Let's use the classic example to make this clearer. Suppose you need to hire a new assistant. You have 10 candidates lined up for interviews. You interview them one by one, and after each interview, you must decide if that person is the best one for the job. If you say no, you cannot call them back later.

If you say yes, the search stops, and you hire that person. Your mission is to pick the single best candidate out of all ten. This is a tough spot because you do not know what future candidates might be like when you are looking at the first few.

Why Simple Strategies Fail

Some simple ways of thinking about this problem do not work well. For example, if you just pick the very first person you interview, your chances of getting the best one are very low. What if the first person is just okay, but the next one is amazing?

On the other hand, if you wait until the very last candidate, you might have passed up many great options. You are then forced to pick the last person, even if they are not very good, because you have no other choice left. Neither of these extreme approaches gives you the best odds.

"The challenge is balancing the desire to gather information with the need to make a timely commitment."

The Surprising "37% Rule" Emerges

It turns out there is a mathematically proven strategy that gives you the best chance of picking the absolute best option. This strategy is often called the 37% Rule, or the 1/e rule (where 'e' is a special number in math, about 2.718). It sounds strange, but it works surprisingly well.

The core idea is to use a certain number of initial options just to learn what a good option looks like. You do not pick any of these first options. You just observe them to set a benchmark for quality.

How the 37% Rule

Works in Practice

Here is how you apply the 37% Rule: First, figure out how many total options you have. Let us say you have 10 candidates for the job. You would then calculate 37% of that number. For 10 candidates, 37% is about 3.7, so you would round down to 3.

This means you should interview the first 3 candidates and reject all of them, no matter how good they seem. You are simply using these first 3 to get an idea of the quality range. Remember the best one you saw among these first 3.

After you have rejected the first 37% (or 3 in our example), you then move to the next step. You will now choose the *very first candidate

  • you interview who is better than anyone you saw in that initial 37% group. If you reach the very last candidate and have not found anyone better, you simply pick the last one.

Beyond Hiring: Where Else This Math Applies

This clever strategy is not just for hiring. It can be used in many parts of life where you face similar one-shot decisions. Think about searching for an apartment: you see a few, learn what is available, then pick the first one that beats your early observations.

It can also apply to finding a parking spot. You might drive past a few empty spots early on to see if better ones are available closer to your destination. Then you take the first one that seems good enough. Even in online dating, some people use a similar approach by observing profiles before making a first contact.

Finding the Best Apartment

When apartment hunting, you might have a list of 10 places to visit. According to the 37% rule, you would visit the first 3 or 4 apartments and just take notes. Do not commit to any of them. After that, the first apartment you see that is better than all the previous ones is the one you should seriously consider.

This approach helps you avoid settling too early for an average place. It also stops you from waiting too long and missing out on all the good options. It balances the need to explore with the need to decide.

The

Limits and Nuances of the Strategy

While the 37% Rule is powerful, it does have some conditions. It works best when you know the total number of options upfront. It also assumes that each candidate's quality is independent and that you can accurately rank them against each other.

Life is often messier than a math problem. Sometimes you do not know how many total options you will get. Other times, the options might not arrive in a truly random order. However, even with these real-world challenges, the core idea of observing before committing remains very useful.

This simple mathematical insight shows that sometimes, the best way to make a big decision is not to overthink every single choice. Instead, you can use a clever two-step process: learn from the beginning, then act decisively when a truly superior option appears. It is a reminder that even complex choices can be made clearer with a bit of forgotten math.

How does this make you feel?

Comments

0/2000

Loading comments...