Memory Techniques: the Peg system (part one)

Part one

How would you like to memorize the first 20 digits of Pi in less than two minutes and never forget it again? I was asked that, almost 20 years ago, by one of my College Professors. We were talking, and I was explaining to him my interest (and frustration) with mnemonics and the Memory Palace in particular.

As he had told me, he had applied mnemonics in his mathemagical endeavors – a funny name for the quite impressive skill of performing apparently difficult mathematical calculations by head – and was quite adept at it. Back then, I only knew about the Method of Loci from a couple of books I had read, but had never been able to effectively use it.

Obviously, I said that would be amazing if I could do that. To which he replied: “repeat after me: ‘My turtle Pancho will, my love, pick up my new mover Ginger’. The sentence certainly sounded silly, but I was not taken aback – surely, I thought, the digits of Pi are somehow encrypted in this apparently absurd sentence.

I repeated the bizarre sentence a few times and, after much less than two minutes, it stuck – I don’t know why, but this sentence is weirdly memorable, despite its absurdity. The remainder of the two minutes, the professor spent writing down the Major System, a simple table that taught me how to decode the sentence and extract the first digits of Pi. At that moment I had to consult the table as I deciphered the sentence, but it was quite short and I believed I could memorize it in a few hours at most. There was no doubt that I had just learned over twenty digits of Pi in a few seconds. I was hooked.

Actually, memorizing 20 digits under two minutes (a feat that, back then, sounded almost like a miracle to me) is something I can do quite easily now – even under one minute is not a challenge. The key to that is using a combination of the Memory Palace (about which we talked at length in the previous articles) and another mnemonic technique – one of which I had never even heard of back then: a Peg System.

I believe that I can say, without exaggeration, that the Peg System was the one technique which had the greatest impact in my developing use of mnemonics. The Memory Palace, with its ancient history, certainly has a great deal of allure, but I have to admit that, before learning about pegs, it was more of a novelty, for which I had little practical use. By combining the two techniques I was finally able to effectively apply mnemonics in my daily life.

That is why, in the article about the Memory Palace, I suggested you should choose a few famous individuals to function as a connection between the location in your memory and the item you wanted to memorize. Those served as temporary pegs, so that you could test the technique without having to invest too much into something completely new to you.

But before we go into the Peg System, let me ask: did you try yourself with the grocery lists in the previous article? I looked over the comments as I am writing this, and noticed some pretty good results for first tries, so congratulations.

Using the Memory Palace, I memorized the 20-items list myself, and it took me 1:40 minutes, which I consider a decent time; but still a far cry from the time of high level competitions – to give you a baseline, the current world record for random words memorized under 5 minutes is held by Simon Reinhard with 125 words, and note that those are random words.

If your first attempt had a high time, do not worry, both practice and the application of the Peg System will reduce it considerably.

The goal of this article is that you develop a simple peg system that would allow you to memorize vast lists of number with ease – in the same way that the Memory Palace allowed us to memorize items in a grocery list.

While we will apply this same method to chess, starting with such a complex task would be counter-productive as a means to learning the basic principles. Instead, it is better to build up from some very simple applications.

As before, I will leave here a test that may give you a baseline to compare your results with and without the mnemonics. It is a list of 40 digits created by an online random number generator. It is considerably harder than the grocery list, as we tend to have a very hard time memorizing random numbers. Just as before, read the list bellow twice, cover it and count to ten out loud. Then write down in order all numbers you can remember.

7 9 0 1 4 7 2 2 6 0 1 9 5 1 4 0 6 7 6 9
0 5 2 6 3 3 5 8 9 2 2 9 2 4 1 8 9 1 2 0

Done? For now, let’s move on – we will return to this list after learning about the Peg System.  

So, what exactly is a ‘Peg System’? Invented by Henry Herdson, the word ‘peg’ is a reference to the pieces of wood at which people used – and some still use – to hang coats or hats.

As some of you have noticed, it may be extremely time consuming to create the images that you use to ‘fixate’ objects to your Memory Palace. While it diminishes with practice, there is always an unwanted delay when creating your ‘pegs’ on the fly. So, one of the purposes of a peg system is to drastically reduce this time delay by memorizing all your images beforehand, so that they are ‘at hand’ when needed: in a peg system, you create a permanent association between an image (the peg) and the object that you will memorize.

That may sound strange at first. If you already know the object, why on earth would you need to memorize it again?

Let me give you an example: imagine you enjoyed memorizing grocery lists. As you keep testing yourself with dozens of different lists, you notice that some products – such as bread – do tend to repeat across several lists (but not all).

Thus, you decide, wisely, that you can choose one specific image to always be associated with bread. Since I had Angelina Jolie as my peg for bread in the first list, I could, for example, decide that she would always stand for this particular item – a fact that I would have to memorize the old fashioned way. As a result, every time that bread appeared in a list, I would imagine Angelina in the corresponding place within my Palace (if bread was the third item, she would be on my third stop), with no need to explicitly imagine bread at all, as the information would already be encoded in the image by my previous memorization.

Notice how much time that would save. There would be no need to imagine the image interacting with the item (in this case, Angelina holding the bread, for example). Imagining the image interaction with the location within your palace is still necessary, though, as the location may vary from list to list – bread could be the first item in one list, the tenth in another or completely absent in a third.

So remember: always imagine your peg interacting with the location.

Now, you could create a peg for all the common items on grocery lists (milk, coffee, etc.), but I personally would consider it a poor investment of your time. Unless you actually plan to compete in memorizing grocery lists (a competition that does not exist as of yet), this set of pegs would be of little use. For myself, I prefer to create pegs for information that is part of a limited set. Things like the moves in the algorithms to solving a Rubik’s cube or chess pieces or the cards in a deck – you may remember that video in the previous article where Jonas Von Essen, the 2013 and 2014 world memory champion, says that his peg for the ace of clubs is ‘Kate from Lost’, the famous TV series.

I find that a deck of cards is the easiest way to explain the different levels of complexity that a peg system may have. Since a deck of cards is composed of 52 cards (13 of each suit), the simplest peg system that could be applied to memorize the order of a full deck would include 17 images – at least, that is the simplest system of which I can think of.

In this simplest Peg System, the images would be divided into one for the rank of the card and one for each suit. That means that there would be one image for ‘ace’, one for ‘king’, and so on, until the deuce – for a total of 13 images. That is, of course, not enough as you would need to memorize both rank and suit, so you would need another four images for them (one for spades, one for hearts, one for diamonds and one for clubs).

Creating a peg system this way, every card would need to be represented by two images. You could choose, for example, the image for ‘ace’ is your favorite basketball player and that the image for ‘clubs’ is a tree. Therefore, if the ace of clubs is the first card in the deck (remember that when memorizing single decks, one is actually memorizing the order of the cards in that particular shuffle), then you would enter your single deck memorization palace, go to the first stop point and use the techniques described in the previous article to create an image that included both pegs – you could, for example, place the player dunking in a hoop attached to a tree.

While that solution requires minimum effort, it is also quite ineffective – remember rule #1: the greater the groundwork, the better the results. Anyone could memorize the 17 images in at most a couple of hours, but, I highly doubt that anyone could use such a system to achieve a decent time.

A much more efficient (and effortful) approach would be to associate one image to one card, for a total of 52 images. In principle, the criterion you choose to create the associations is irrelevant. As long as you remember that the ace of clubs is, for example, Kate from Lost, the system will work efficiently. The problem here is that, without some guiding principles, the groundwork may become overwhelming and impede your progress. That is why I found that it is much better to apply a few tricks and techniques to facilitate the peg creation process – I will explain those soon, but there is one more level of complexity that a peg system may reach.

The most complex application of the Peg System (of which I am aware) is called the PAO (Person, Action, Object) system. In the PAO system each item (in our example, each card) is associated with a total of three pegs – a person, a verb and an object. The practitioner would group cards in sets of three and form a sentence (often an absurd one) using the Person associated with the first card in the set, the verb associated with the second and the object with the third.

Let me give you an example. I am memorizing a shuffled deck and the first three cards are the king of hearts, the ace of clubs and the ten of diamonds. If my pegs for those three cards are, respectively, ‘Julius Caesar throws a die’; ‘Kate swims in the sea’ and ‘Bill Gates makes money’ – then my sentence would be ‘Julius Caesar swims in money’. I would store this image in the first stop point in my Memory Palace and move to the second set of three cards which I would place at the second stop point in my Palace.

By applying PAO, I could store a whole deck with only 18 images (17 sets of three, plus one solo image for the last odd card). That is, of course, if I decided that using a PAO system to memorize a single deck was the most efficient system for me – it is important to keep in mind that the most advanced system is not necessarily the one that will satisfy the specific needs for the task at hand.

At this point you already know a lot about Peg Systems, but unless you are planning to become a competitor in memory championships, memorizing cards is probably not something about which you feel passionately. What about something more interesting, such as chess?

In his record breaking blindfold simul, Timur Gareyev used a variant of the PAO system. As he explained to me:

"For example 1.d4 my move is Princess Diana, which the image already belongs to several destinations along my memory palace since I know I will be white and play the move on those particular boards. Now imagine games 13, 25, and 37 where I have Princess Diana at the Fountain, Campus Cafeteria, and Inauguration Speech Plaza. I receive responses 1...Nf6, 1...d5, and 1... d5 accordingly. Board 13 I have Princess Diana throwing a horseshoe into the fountain (Knight move represented by a horseshoe) Board 25 & 37 I add Diana holding her hands in namaste (symbolizing a symmetric response)."

Notice that this PAO peg system followed the same principles that I just described for memorizing cards except that chess moves come in pairs and not triads, so the images only contained a Person and an Action (so maybe it should be called a PA system).

Throwing a horseshoe was Timur’s peg for the 1…Nf6 knight response.

As far as I know, Timur Gareyev does not practice mnemonics for other purposes, but notice that, if he wanted to apply himself to memorizing decks of cards, he would need to create a different set of pegs, for this purpose. That is the versatility of Peg Systems, you can apply them to a large variety of problems. In the next (and final) article, we will delve into how mnemonics can be applied specifically to chess, and will discuss a little more about Timur’s use of the Memory Palace and Peg systems.

Another advantage of pegs is that they facilitate long term memorization. The last task to which I created a number of pegs was speedcubing (the practice of solving a Rubik’s cube in the least amount of time you can). Unlike the most basic solutions, which require a dozen or so algorithms (most of which are very simple), speed cubing requires tens of complex algorithms with many moves each.

To help me memorize, I created one peg for each possible move of the cube and used them to store the algorithms in a long term palace. Unfortunately, the necessity of decoding this information makes accessing it while solving the cube, time inefficient; but, having it stored in a palace gives me free access to the information, which I can review at any time (in the shower, as I ride an elevator or as I wait in a line).

That process slowly converts the codified images into decodified memory, which I remember just like I remember my phone number. But for that I must consistently visit my palace and review and decodify the images therein stored.

 And, if I ever find myself blanking, I can go to my palace on the spot and get the necessary algorithm – that will add a few seconds to my time, but is better than just failing to solve the cube.

Continued in part two