There’s a quote that’s often attributed to Albert Einstein which goes:

“If you can’t explain it simply, you don’t understand it well enough.”

Whether or not Einstein himself actually said this (it’s never been properly sourced, so it’s likely he didn’t), it’s still an insightful observation. It’s also one that yields a pretty powerful study tip when reversed:

**If you want to understand something well, try to explain it simply.**

By attempting to explain a concept in simple terms, you’ll quickly see where you have a good understanding of that concept. You’ll *also* be able to **instantly pinpoint your problem areas, **because they’ll be the areas where you either get stuck or where you end up resorting to using complex language and terminology.

This is the idea behind the **Feynman Technique.**

Named after the Nobel Prize-winning physicist Richard Feynman – who, in addition to being a brilliant scientist, was also called “The Great Explainer” for his ability to relay complex ideas to others in simple, intuitive ways – the Feynman Technique is a method for learning or reviewing a concept quickly by explaining it in **plain, simple language**.

In addition to helping you pinpoint those problem areas in the concept you’re trying to learn, the Feynman Technique gives you a quick, efficient way to shore up those areas using **targeted learning**. It’s a simple technique, but it’ll help you study much more efficiently once you put into action.

So how do you actually use it?

## How to Use the Feynman Technique

Since the root of this technique involves **explaining the concept, **you could execute it in a number of ways – including literally grabbing a friend and explaining to them what you’re learning. However, you don’t always have willing friends at hand, so here’s the simpler method that just involves a sheet of paper.

**Step 1:**Grab a sheet of paper and**write the name of the concept**at the top. You can use pretty much any concept or idea – even though the technique is named after Feynman, it’s not limited solely to math and science.**Step 2:**Explain the concept in your own words**as if you were teaching it to someone else**. Focus on using plain, simple language. Don’t limit your explanation to a simple definition or a broad overview; challenge yourself to work through an example or two as well to ensure you can put the concept into action.**Step 3:**Review your explanation and identify the areas where you**didn’t know something**or where you**feel your explanation is shaky**. Once you’ve pinpointed them, go back to the source material, your notes, or any examples you can find in order to shore up your understanding.**Step 4:**If there are any areas in your explanation where you’ve used lots of technical terms or complex language, challenge yourself to**re-write these sections in simpler terms**. Make sure your explanation could be understood by someone without the knowledge base you believe you already have.

That’s it!

## 3 Examples of the Feynman Technique in Action

As I mentioned earlier, simply defining a concept is only half the battle. If you want to explain is clearly, you have to **apply it** by working through examples.

In the spirit of eating my own dog food, I’ve included three examples of how you might use the Feynman Technique below.

### Example #1: The Pythagorean Theorem

We’ll start with a very simple example. The Pythagorean Theorem shows how you can find the length of any right triangle’s hypotenuse:

When I initially started writing this explanation, I simply wrote the sentence at the top and then added the formula.

However, note how the final page has a couple of additions:

- A small picture showing what a right triangle is
- An arrow clarifying the nature of C in the formula

This was my attempt to go back and further simplify the explanation. Even with a basic mathematical theorem like this one, there are still assumptions and terms that encompass ideas that you may not be 100% clear on. Challenge yourself to identify those things and define them.

### Example #2: Bayes’ Theorem

Since the Pythagorean Theorem is a pretty simple concept, I thought you might like to see an example using something more complex. **Bayes’ Theorem** – a concept used in probability theory and statistics – fit the bill nicely.

And here’s a page working through a specific example and using the formula:

These pages do a decent job of explaining Bayes’ Theorem at a very broad level, but I’ll be the first to admit that this is a topic that takes a *good long while *to truly grasp.

In fact, I had to spend three hours reading through A.I. researcher Eliezer Yudkowsky’s 15,000-word explanation of the theorem before it “clicked” in my brain, so definitely check that article out if you’re curious. You can also check out Arbital’s more recent guide, which is – by Yudkowsky’s own admission – much better and easier to follow.

### Example #3: The CSS Box Model

Here’s an example of how the Feynman Technique can be used to review a non-mathematical concept.

The CSS Box Model is a tool for representing the size of **HTML elements** (i.e. the code that makes up web pages just like the one you’re reading right now), as well as the spacing around them. I chose it as an example because it’s a concept that took me a long time to grasp back when I started learning how to build websites as a teenager.

To clarify that page’s general explanation, here’s an example of an element with specific height, width, margin, padding, and border values written in CSS code:

In addition to writing the code out, I thought it would be extra helpful to show exactly how each attribute affects the overall size of the element.

To a budding web developer, it might not be immediately obvious that, say, a padding value of **10px **actually increases the element’s width by **20px **overall (because the 10px is applied to each side).

If you happen to be curious about the Box Model and want to learn more, check out this guide.

## Think Like a Child

One final tip: While you’re working through the Feynman Technique for any given concept, it can be useful to pretend that you’re explaining that concept to a child.

Doing this will boost your own understanding for one simple reason; in addition asking things like, *“Can I have another Oreo?” *and *“Can I go watch Dragon Ball Z now please?” *a kid is probably going ask…

“Why?”

While older people often become accustomed to taking things at face value, kids are naturally curious. They’re quick to point out their confusion.

If you teach a kid how the Pythagorean Theorem works and give him the formula for using it, there’s a good chance he’ll ask you:

“Why does that formula work? How can you know it’ll

alwayswork? Prove it, sucka!”

…and then you realize that the kid was actually Mr. T in disguise all along, and now your life depends on being able to explain a geometry concept. How did you even get here?

Seriously, though, this is a great mindset to adopt. Maybe you* do *know how the Pythagorean Theorem works, and maybe you can easily draw out the proof by rearrangement:

When it comes to other concepts, though, it’s likely that you’re relying on **assumptions, heuristics, **and other **black boxes **when it comes to certain details. So adopt a child-like mindset and challenge yourself to clearly explain the whole concept.

Once you’ve done that and worked through all the steps, you can further refine your knowledge of whatever you’re studying with other techniques, including:

Hope this helps!

*If you’re unable to see the video above, you can view it on YouTube.*

### Looking for More Study Tips?

If you enjoyed this article, you’ll also enjoy my **free** 100+ page book called *10 Steps to Earning Awesome Grades (While Studying Less).*

The book covers topics like:

- Defeating procrastination
- Getting more out of your classes
- Taking great notes
- Reading your textbooks more efficiently

…and several more. It also has a lot of recommendations for tools and other resources that can make your studying easier.

If you’d like a free copy of the book, let me know where I should send it:

I’ll also keep you updated about new posts and videos that come out on this blog (they’ll be just as good as this one or better) 🙂

### Video Notes

- Scott Young’s video – where I originally heard about this technique
- The Feynman Notebook Method – related concept on Cal Newport’s blog
*Surely You’re Joking, Mr. Feynman! –*a book of stories about Richard Feynman’s life, written by Feynman himself. I read this book last year and loved it.

Have something to say? Discuss this episode in the community!

*If you liked this video, subscribe on YouTube to stay updated and get notified when new ones are out!*

Just a little question- I have read about five to six chapters of the first volume of the Feynman lectures and I am thinking whether or not I am getting data about everything within that topic. Could it be that Feynman’s simplistic system is hiding information?

sir can you please tell me if we have to learn a new concept from book how should we read and remember it before using feynman technique

That’s really usefull! I’m horrible in Math and I’ve never understood Math so good like I got it after that post haha! Cheers mate!!!

Thank you i learnt about the Css Model box and i had covered CSS ON my own but didn’t get most concepts wanna be a software developer and this is a great help for me to study. Today is first day to discover your YouTube Channel am from EastAfrica a country called Kenya.

Awesome explanation of Baye’s theorum! If you want to understand the logic behind the superhero example, here’s how I think about it:

The important point is to realize that the odds of a superhero being a 6’2″ bearded guy is 5x more likely than the average person being a 6’2″ bearded guy (5% compared to 1%). Now we take this realization and apply it to our example.

We are trying to figure out what the odds are of our 6’2″ bearded guy being a superhero. It is ALREADY established that he is 6’2″ and bearded. So now we apply what we figured out before, that a superhero is 5x more likely to be 6’2″ and bearded than not. Since the odds of any random person being a superhero is .01%, the oddsof a 6’2″ bearded guy being a superhero is .05% (5 x .01%).

Please please tell me if my logic is faulty.

Am I the only one who thinks that his final conclusion in Baye’s is not satisfying?

Of the 700,000 superheroes, 35,000 are 6″2 bearded guys. (In 35,000 there is both secretly and Not secretly 6″2 bearded superheroes)

So I if we want to know how many 6″2 bearded is secretly a superhero, think we should mulitply it by the final answer, and now the answer is: “1,750” (out of 70,000,000 6″2 bearded guys, 1,750 can be expected to secretly be superheroes).

Someone, PLEASE read this. I’d love to know if I’m wrong or right. If I’m wrong, I need someone to explain/clarify it for me.

Thank you so much for this great article tho.

— from your friendly neighborhood, Spidey 😉

Here, the sets {secretly a superhero} and {a superhero} are equal. That is to say, if you are a superhero, you have a secret identity. Therefore, his calculations hold. You multiply P (S|B) with total bearded dudes, 70 million to get the superheros total no. within bearded dude community. This is the no. of superheroes with beards, cause the no. itself is absolute -equal to n ( B □ “intersection” S) ….(whereas probabilities are always ‘relative’).

You cannot multiply P (S|B) to total no of bearded superheros, even if we assume P (S|B) is the probability of “selecting a superhero who is still in the closet, given I have selected the bearded dude community ” . This is because I am finding conditional probability given the event of selecting a bearded guy. So I always multiply with total no. of bearded guys. If you make secret superhero and superhero separate sets, the substitution in the formula changes.

Say, probability of a person being a superhero,secret or not,

P ($)=0.0002.

Given, P (B)= 0.01, P (S)=0.0001, P (B|$)=0.05.

Then, probability that, having selected a random bearded dude, you have selected a secret superhero is P (B|S).

P (S|B) = [ P (B|S)*P (S)] / P (B)…..Aaaand, in explaining, I have run across our gross misassumption . 🙂 You can never say that just because 5% of all superheroes are bearded, that 5% of all secret superheroes are bearded. If you do assume even distribution, you get the same answer, but, once again, this is a probability having known that you selected a random bearded guy.

Your multiplying with the no. of bearded superheroes would work if we solved for P (S| (B□$) ) = [P ((B□$)|S) * P (S) ] / P (B□$)…for equal distribution,

= P ( (B□$) □ S)/ P (S) * 0.0001 / 0.000005 = 0.025 / 0.05 = 1/2, which shows the even distribution, which makes you question why you used the formula in the first place….. P ( E|A) can be found by P (E□A) / P (A). It is when you don’t have numbers that we swap the probability of the intersection with truths we already know.

Okay, I think I have succeeded in confuzzling myself, so I will back out while I am ahead. Let me know if something doesn’t add up. I’m simply a kid who had a nice math teacher….

Cheerio!

This is great, thanks! I just watched your video and love the CSS explanation. I’m currently learning CSS and your explanation is really helpful.

I liked your video (although I don’t do the thumbs up thing). I would like to recommend to my students (I teach mathematic). Small problem – Bayes Theorem is P(A | B) = P(A and B) / P(B) not P(B | A) / P(B)

Glad you liked it!

I do believe the Bayes’ Theorem formula in the article is correct: P(A | B) = P(B | A) * P(A) / P(B). Here’s a good article on it: https://betterexplained.com/articles/an-intuitive-and-short-explanation-of-bayes-theorem/

Nick97 The statement is correct as shown. For example (e.g.) he might be Batman. He may not be Batman but Superman so you cannot conclude (i.e.) he is Batman.

I’m Batman

I think you meant to write “i.e.” instead of “e.g.” in the Bayes’ theorem example.

Great article, by the way!

Hmm, Thomas, you mentioned that you read Yudkowsky’s 15,000-word explanation before understanding Baye’s theorem hurh?

Herein lies a very interesting phenomenon. And also a potential irony/catch-22 situation.

Sometimes when we are reading new material, we come across theorems that appear very puzzling. To what extent are we dissatisfied? This can have great implications on our ability to learn.

I had a friend, who rarely did his homework, and he would tell us that when we encounter something that we do not understand in the book, we should go and research it. And read about it until you understand it fully. That’s what he does.

But I have another friend, he reads a lot of books and when he encounters something he does not know, he takes it at face value. And he is able to apply it. He does not necessarily need to understand it so thorougly. It is merely sufficient for him to familiarize himself with something. And the mother of familiarization is memorization, he tells me. So he would memorize the equation. And he does that very easily.

And when I ask him about the deeper explanation behind academic stuff, usually he does not know, and he sometimes says “why do you worry so much about irrelevant stuff?”

As far as neediness towards understanding is concerned, I think my second friend has it better. I would rather have less dissatisfaction when reading. If every new thing that I read causes me to be dissatisfied until I get to the bottom of it, it’s very hard to finish a big academic project.

But in the long term, I realize that my first friend’s attitude towards always wanting to find out more and not being satisfied with understanding it merely at face value is a good quality to have.

Therein lies the irony, I want to be like my second friend when I am reading, otherwise reading becomes very daunting. But in terms of long term outlook, I want to be like my first friend with his always asking questions attitude.

Thomas, have you come across any academic paper or article that explores this thing I observed in more depth?

Similarly, me and my friends noticed this problem during high school. We were always the curious kind, and refused to use equations or continue with other material until we fully understood /why/ something works the way it does.

We also got pretty terrible gradings that way, as fully understanding each concept was too costly for us in time – leaving little to fully learn the other 3/5 subjects in an exam.

So, we learned to move on and simply behave as you 2nd friend – it was necessary for us to get good grades in decent time.

You should definitely read “Surely you’re joking Mr. Feynman” – Ferynman talks a great deal about this exact issue you’re referring to and ‘why’ there’s an importance in understanding the ‘why’ so much (as mentioned in this article) – see the chapter about Richard teaching in Brazil.

In fact, the internet has made it easy for you – someone’s posted it here http://v.cx/2010/04/feynman-brazil-education

Excerpt:

“After a lot of investigation, I finally figured out that the students had memorized everything, but they didn’t know what anything meant. When they heard “light that is reflected from a medium with an index,” they didn’t know that it meant a material such as water. They didn’t know that the “direction of the light” is the direction in which you see something when you’re looking at it, and so on. Everything was entirely memorized, yet nothing had been translated into meaningful words. So if I asked, “What is Brewster’s Angle?” I’m going into the computer with the right keywords. But if I say, “Look at the water,” nothing happens – they don’t have anything under “Look at the water”!”

The problem here is you’re studying to pass exams, not to truly understand the problem/concept. You need to really understand the concept to do anything of value with it. It’s not entirely your fault, education systems are designed for you to pass or fail exams, not to understand or measure the material to a point of true value – or so Feyman said 50 odd years ago (and still applies today).

Can you find the solution to the candle stick problem if you don’t know what a box is?

Can you speak Japanese fluently if you do not understand what the words mean, or in what context they speak it, and learn Japanese by imitating every phonetic sound and scribble?

Problem solving is very… trollish in this regard. A lot of stuff you thought was trash will provide solutions in the long run. That’s where “diffusive thinking” come from, as in A Mind for Numbers.

Obsessiveness is also a feature of all star athletes (and I suspect geniuses, i.e. top experts), as said in The Power of Habit.

Favoring speed over fundamental correctness will cost you effectiveness in actuality, as well as more time as you try to fix the problem. This latter concept is well-supported in coaches, trainers, and masters world-wide, but you can find this well featured in Unleash the Warrior Within, a SEAL story, and The Little Book of Talent.

On the other hand, true, the company does what the CEO measures, and if you ace those tests you’ll meet more geniuses and highly motivated people. Plus, according to Algorithms to Live by, time limits adds a natural constraint against complexity (usually correlated with high resource consumption, lack of prioritization, and over- extrapolation), and the simplest answers are usually the right ones.

So, what game are you really playing? What’s your bigger picture? In other words, what metric aka value criterion are you using to gauge yourself? Do you prioritize the fire you should put out not, or long-term productivity? Or, do you nail both, but only in the field of professional proficiency, as you abandon your social life and sound sleep for the next ten years?

To end this, I leave you with this:

“Weakness and ignorance are not barriers to survival; arrogance is.”

-Death’s End by Cixin Liu

Patricia,

That is a well-stated summary of an almost universally-unstated dilemma in education. I think a satisfactory, insightful response would require an essay, if not a textbook.

But I would offer the thought that education and learning require different strategies.

Education has become more of a vehicle for acquiring socially-empowering bona fides and credentialing (especially in the Humanities and Social Sciences, or those involving extensive qualitative efforts) and less about deep learning. If your strategy is to disrupt your own learning or the institution’s with deeper questions and critical thinking (especially if it detracts from the prevailing sloganism) you will be attempting to exercise a noble instinct in an environment that will only harm, injure, and personally attack you for your intellectual good deeds. The academy has become a very hostile space for those who question the prevailing political hegemony. And most of the “learning” acquired in these disciplines is about conforming to the prevailing political hegemony. So stick to practices that optimize your credentialing outcomes. That would exclude your first friend’s approach. A notable exception to this might be a university education in the hard sciences.

For true learning, on the other hand, or for those engaged in hard science majors that depend upon great quantitative efforts, your first friend has the much more intelligent strategy. The deep understanding of relationships in collateral knowledge will empower his or her life’s work. Career advancement and making a genuine difference are much more tightly correlated to this than other “identity” factors. Also your first friend will already be more proficient in considering, questioning, learning, and applying collateral relationships, all of which are hallmarks of truly brilliant work.

Best.

Hahaha, I am sure, in the process of creating these notes, you begin to realize why this technique does not come intuitively for most learners. This business of simplifying the thing out, especially in the Bayes theorem that you wrote, seems like a big waste of time at first. You are writing out in long English sentences something that you know already. And not only once, you got to write it out several times.

I am glad there are people like Feynman and you who point us to things that are counter-intuitive. Things that, at first glance, appear like a complete waste of time. But upon further observation, actually help to solve important problems.

Great post! especially the last example is very useful. Thomas which website would you recommend for Html and CSS?

Dumbest explination to bayes therom everyone does not observe denominator set in bayes therom its n(b)

Do you have a video on studying/taking online or distance education courses? I’m going to be taking some for the first time but I don’t know where to start

Thanks for sharing such a useful post like this. Please keep up your good work to guide people.

I was struggled a lot to learn this concept but after visiting this post that became simple for me. Thank you for sharing this post.