How does the interviewer test your way of thinking? It is very simple to learn this way of thinking!

1. Two ways of thinking

In job interviews, questions like this are often examined:

How many Tesla cars are there in Beijing? 

How many pancakes can a pancake stand at the entrance of a certain alley sell in a year? 

How many product managers are there in Shenzhen? 

How many ping pong balls can fit in a bus? 

How many hairs does an average adult have?

This type of estimation problem is known as the Fermi problem, named after the scientist Fermi. Why is this question asked in an interview?

These kinds of questions can clearly distinguish between two types of people.

One is people with liberal arts thinking, who are good at admiration and fuzzy imagination, which mainly rely on people's first reaction and intuition, such as children who are born to see things to admire. 

The other is people with scientific thinking, who are good at solving specific problems through logical reasoning and analysis. This kind of thinking is not born in people, it needs long-term training.

9a458e3c5b8b8e3d836a8e74a8bcde20.jpeg

The company recruiters need people who have undergone long-term training, can do things well, and have rigorous logical reasoning and analysis capabilities. Therefore, Fermi questions can examine what kind of thinking a person has.

2. How to answer the Fermi question?

Ordinary people will feel confused when they get a question like the Fermi problem, and don't know how to solve it, so they simply guess a number based on their feelings.

This actually ignores the purpose of the interviewer's investigation. He doesn't want you to have a certain number, but wants to see your analytical thinking when facing the problem .

Therefore, you need to speak out your thoughts to prove that your way of thinking is scientific thinking, not a person who talks nonsense based on feeling.

To answer the Fermi question, we need to use the "logic tree analysis method" we talked about before, to disassemble a complex problem into sub-problems, and then solve them one by one .

3. Case

Someone once asked the scientist Fermi: "How many piano tuners are there in Chicago?"

585e7725da0a3c5c414301e0f5b7398e.jpeg

What is a piano tuner?

In order to maintain the accuracy of the piano sound, it needs to be checked regularly by a professional to adjust the inaccurate sound. A person who does this type of work is called a piano tuner. There is a part of very good-looking movies called "The Tuner".

We use the logic tree analysis method to disassemble this problem into sub-problems.

The number of piano tuners = the total working hours of all piano tuners in one year / the annual working hours of one tuner.

Therefore, we can decompose this problem into two sub-problems:

1) Total working hours of all piano tuners for 1 year

2) Annual working hours of a pianist

6362f5e31f4fe1e5be2ee71e1a804fc0.jpeg

"The total working hours of all piano tuners for one year" can be broken down into three sub-problems to solve:

1) How many pianos are there in Chicago?

2) How many times does the piano need to be tuned each year?

3) How long does it take to adjust once?

a5a8362e5627f70538b11a797155dcdf.jpeg

Now we solve these subproblems one by one.

1) First sub-question: How many pianos are there in Chicago?

I don't know this. Let's break it down again. First, you need to know how many people there are in Chicago, and second, you need to know a percentage of people who own pianos.

The population of Chicago can be checked online to be about 2.5 million.

What percentage of people have a piano? I don’t know, but we can guess that the piano is more expensive for ordinary families, and the piano occupies a large area and is inconvenient to keep at home, so we guess that the proportion of families owning a piano is 1%.

Why is it 1%, not 5%? Because 1% usually means the lowest probability, and some institutions have more pianos than individuals, such as conservatories, so let's guess a number again, which is about 2%.

With this data, a simple calculation shows that there are about 50,000 pianos in Chicago.

e004f17638b55dd92cd60d51a35d8c35.jpeg

2) The second sub-question: How many times does the piano need to be tuned each year?

Piano tuners are a rare industry, and there are definitely not many people. Pianos do not need to be tuned frequently like guitars, probably once a year. 3) The third sub-question: How long does it take to adjust once?

Probably about two hours.

cbcb3bc05bf6e61d665b0ee79c2e1a2a.jpeg

4) The fourth sub-question: How long does a tuner work each year?

There are four weeks of holidays in the United States each year, and there are about 50 weeks in a year. Working 5 days a week is calculated as 8 hours a day. Multiplying these three numbers, it can be calculated that a tuner works 2,000 hours a year.

But piano tuners have to run around, and they must spend time on the road, so about 20% of the time is subtracted, so the tuner works about 1,600 hours a year (2,000 hours-2,000 hours*20%).

3a9ddd9f5f4b281c926d16c54b0a6f10.jpeg

Now we summarize the four sub-questions. The total working time of all piano tuners in one year is the multiplication of the numbers of the three sub-problems, that is, 50,000 pianos need to be tuned once a year, and each tuning takes two hours, a total of 100,000 hours. Work 1600 hours.

We then divide the total working hours of all piano tuners in one year by the annual working hours of one tuner, and we get 62.5. Rounding it up, Fermi estimates there are about 63 tuners in Chicago.

2a0790e86e2c804b174610340b916293.jpeg

Is this answer correct? Later, Fermi found a list of tuners in Chicago, with a total of 83 people on it, and many names were repeated. You know, Fermi estimated it, and it is quite accurate.

4. Summary

The company uses "Fermi problems" to judge whether a person has logical thinking ability, and we can also consciously do "Fermi problems" to improve logical thinking ability in daily life.

The way to solve the "Fermi problem" is to use the logic tree analysis method to split a complex problem into small problems. Thereby turning a problem without thinking into a well-organized thinking.

Now leave a thought question, you can write your analysis ideas in the message area:

Estimate how many Tesla cars there are in Beijing?

a0fb1972f1e0437a1106dcc7cd46091a.jpeg ⬇️Click "Read the original text"

 Sign up for free Data analysis training camp

Guess you like

Origin blog.csdn.net/zhongyangzhong/article/details/129645692