
Discover more from Marty's Magic Ruseletter
Corrupting the Classics: The Twenty-One Card Trick
Learn some alternative approaches to an ancient self-working card trick. Includes links to trick tutorials for "Trust Your Gut", "Unlucky for Some" and "Tiny Tantalizer".
Welcome to the first article in a regular column called Corrupting the Classics. I’ll re-interpret a classic magic trick in each post and provide one or more new ways of performing it. The title of this column, of course, begs the question: what makes a magic trick a classic? For me, the trick must meet the following three criteria:
First, it must be popular with magicians and audiences alike.
Like a classic car, it must be at least twenty-five years old. This longevity proves that the trick has stood the test of time. While this stipulation limits what tricks can appear in the column, it will encourage me to revisit older books, manuscripts and magic tricks (something I enjoy doing, anyway) and ignore the newest releases (something I should do more often).
It must have some historical importance to magic. For example, be connected to a magician of notable worth or represent a significant improvement or development in a particular plot.
Right, let’s start corrupting those classics!
This first article will take a deep dive into the much-maligned Twenty-One Card Trick. Also known as the Eleventh Card Trick or the Three Column Trick, this is perhaps the most well-known card trick among non-magicians. When performing, a member of my audience will invariably ask me if they can show me a trick. Nine times out of ten, they’ll deal out the familiar three columns of seven cards and proceed to perform a laboured version of the Twenty-One Card Trick (cue eye roll and suppressed yawn). 🙄🥱
In this post, I’ll discuss some of the inherent weaknesses of the trick and how we can remedy them. I’ll also share some tips for improving the method along with three fantastic variations of the Twenty-One Card Trick: “Trust Your Gut”, “Unlucky for Some”, and “Tiny Tantalizer”.
Even though the trick has a generally poor reputation among magicians, it remains a classic of card magic because of its strong effect—the revelation of a thought-of card—and its appealing and easy-to-remember method, which is entirely self-working. I think it is a fine piece of magic, so long as you spend some time developing a good presentation for it. Unfortunately, more often than not, when a layperson performs it, there is usually very little or no attempt at presentation, and the trick is delivered in a bare-bones manner. See the Numberphile video below for a typical layperson’s approach. The performance by Anastasia Chavez is charming in its own way, but, unfortunately, it exposes the method’s many inherent weaknesses (we’ll discuss these later on in this article).
So what does the Twenty-One Card Trick look like? In its most common incarnation, the magician takes twenty-one cards and asks someone to think of one of them. The cards are then gathered up and dealt into three seven-card columns. Next, the participant is asked to indicate which column contains their card. This process is repeated two more times. The magician then names the thought-of card or produces it in some impressive manner. The mathematical nature of the trick ensures that the selected card always ends up eleventh from the top of the assembled packet. For this reason, the trick can also be framed as a prediction.
As well as being arguably the most well-known card trick in the world, the Twenty-One Card Trick is also one of the oldest. It is well over four hundred years old at this point in history, so it definitely meets my criteria for being a classic magic trick.
The trick is usually passed down from generation to generation by word of mouth. Or, sometimes, it is discovered in the musty pages of an old magic book, such as Professor Hoffmann’s Modern Magic (1876) or Popular Card Tricks by Walter B. Gibson (1929). Unfortunately, I can’t remember when, where or how I first learned the trick (I don’t think it was from a book).

Here is an extract from Modern Magic1, Chapter III: Card Tricks With Ordinary Cards, And Not Requiring Sleight-of-Hand. Professor Hoffmann includes three easy methods of discovering a given card. The second method explains the mechanics of the Twenty-One Card Trick, although the description isn't particularly clear:
Second Method. — Deal the cards into three packs, face upwards, and request a spectator to note a card, and remember in which heap it is. When you have dealt twenty-one cards, throw the rest aside, those not being employed in the trick. Ask in which heap the chosen card is, and place that heap between the other two, and deal again as before. Again ask the question, place the heap indicated in the middle, and deal again a third time. Note particularly the fourth or middle card of each heap, as one or other of those three cards will be the card thought of. Ask, for the last time, in which heap the chosen card now is, when you may be certain that it was the card which you noted as being the middle card of that heap.
This same effect will be produced with any number of cards, so long as such number is odd, and a multiple of three. The process and result will be the same, save that if fifteen cards are used each heap will consist of five cards, and the third card of each will be the middle one; if twenty-seven cards, each heap will consist of nine cards, and tint fifth will be the selected one, and so on.
Why Does the Twenty-One Card Trick Work?
There are only three rules that you need to remember to perform the trick successfully (this is, perhaps, why the Twenty-One Card Trick has become so popular):
That you deal the twenty-one cards face up from left to right, as you would in a game of cards. In other words, the first three cards should be dealt in a horizontal row. Then, each subsequent card should also be dealt from left to right, overlapping the card below it in an injogged position (watch the above video if you’re struggling to visualise what this looks like). In this way, you form three distinct columns of seven cards.
After a participant points to a particular column, that seven-card group must be sandwiched between the other two heaps.
Three rounds of dealing must be completed to ensure that the thought-of card ends up eleventh from the top of the packet.
Note: I prefer to turn the piles face down as I gather them up and then deal them face up again (stud style) rather than keeping the cards face up throughout the entire trick. Either way works, but you must remain consistent in your approach. Otherwise, the method might fail.
Like many self-working or semi-automatic card tricks, even when you know how to perform them, you often don’t understand why they work. Bridging this how-why knowledge gap is important. Attempting to understand why a trick works will help you structure and present it more effectively, resulting in a deeper level of deception. It encourages you to compare and cross-pollinate similar methods and processes, leading to new and exciting variations and discoveries. It also means you’re less likely to forget the correct sequence of events mid-performance, potentially saving you from embarrassing mistakes.
So, with this aim in mind, why does the Twenty-One Card Trick work? The trick relies on the Redistribution Principle and a covert process of elimination. After the first round of dealing, you have narrowed the thought-of selection to one of seven possibilities. Once the three piles are reassembled, the selected card lies eighth, ninth, tenth, eleventh, twelfth, thirteenth or fourteenth from the top of the face-down packet. After the second round of dealing, you’ve eliminated some more possibilities; the selection will be third, fourth or fifth in its column. This will put it in position ten, eleven or twelve once the cards are reassembled. When you deal the cards out again, the card will be fourth in the first, second or third column. Therefore, once you gather all the cards, the thought-of selection will always end up eleventh from the top of the packet or dead centre in the stack.
Twenty-One Card Trick Proof:
1 = 01 | 02 | 03 | 04 | 05 | 06 | 07 |
2 = 5B | 5C | 4A | 4B | 4C | 3A | 3B |
3 = 4C | 4C | 4B | 4B | 4C | 4A | 4A |
4 = 4B | 4B | 4B | 4B | 4B | 4B | 4B |
The table above illustrates why three rounds of dealing are required. This is a brute-force mathematical proof, also known as proof by exhaustion. It should give you the confidence that the trick will always work, providing you follow the correct sequence of actions. The number represents the location of the selected card from the top of each seven-card heap, and the letter indicates the pile—A on your left, B in the middle and C to your right. After three rounds of dealing, the card is always fourth from the top of its pile (see row 3 in the table). In the proof, the target location is “4B”, or the fourth card down from the top of pile B. Studying this table also reveals that, after a fourth round of dealing, the thought-of card will always end up in pile B (something that we’ll take advantage of in the trick “Trust Your Gut”).
Other Variations and Related Effects
There’s a similar category of trick that also uses the same redistribution principle in combination with a grid-based method of location. But, unlike their twenty-one-card cousin, these tricks are not self-working and require more mental effort to perform correctly. Instead, they involve locating the position of a selected card in the same way that you would locate a place on a map using coordinates.
Well-known examples include “Mutus Nomen Cocis Dedit” and the “Sixteen-Card Trick”. There are also pseudo-versions of the Twenty-One Card Trick that simulate the original effect using an unrelated method. The most well-known of these is probably Marlo’s “21 Card Trick Streamlined” from his book The Cardician, a trick designed explicitly as a form of “self-defence” against the traditional method.

I won’t be discussing these variations in this article. Still, you may well be interested in them. In that case, I recommend you read A Brief Analysis of the Twenty-One Card Trick and Related Effects by my friend Justin Higham, a very knowledgeable card conjurer and former student of Ed Marlo. In fact, you should read his article regardless because it provides an excellent overview of the principle at the heart of the Twenty-One Card Trick (it also helped me immensely when researching this article).
Problems with the Twenty-One Card Trick
The biggest problem with the Twenty-One Card Trick is the repetitious dealing. Many magicians insist that this makes the trick dull—I disagree. If you have something interesting to say during the repeated rounds of dealing, and you justify them with an engaging presentation in some way, I don’t see why the trick is destined to be dull. Boring tricks don’t bore audiences. Boring performers do—harsh but true!
Some of the most well-regarded card tricks involve a lot of dealing. For example, “Out of This World” by Paul Curry is considered by many magicians to be the best card trick ever invented, and it involves all fifty-two cards of the pack being dealt to the table! Even Dai Vernon thought it was the best trick to perform for laypeople, and Roberto Giobbi considers it to be an almost perfect card trick.
However, the three rounds of dealing do telegraph that the method is mathematical, especially if someone in your audience is a mathematician or a computer scientist, for example. After all, the Twenty-One Card Trick is, in essence, one big sorting algorithm. But again, this is where a solid presentation comes to your rescue by justifying the repetitive dealing.
The next biggest issue is that the number of cards used tips the method because the trick is so well known. There’s a straightforward solution to this issue: use a different number of cards. Most people who know how to perform the trick are under the misapprehension that it only works with twenty-one cards. This is because, while they know how to perform the trick, they don’t understand why it works. As a result, the underlying mathematical principle remains elusive even after repeat performances. In fact, the trick works with any number of cards, so long as it is odd and a multiple of three (something Professor Hoffman mentioned in Modern Magic). You can even perform it with fifty-one cards (the thought-of selection ends up ninth in the indicated pile or twenty-sixth from the top/bottom of the pack).
The other big giveaway that you’re performing the Twenty-One Card Trick is the face-up column-based layout of the cards. But, again, this is easy to circumvent. Rather than dealing the cards into face-up columns, you can deal them into three face-down piles. This has the added benefit of keeping the faces of the cards hidden from the performer throughout. This feature of the trick—that you never look at the faces of the cards—actually strengthens the presentation. The downside is that it does slow things down a little, and you need a way to justify searching for the selected card in this way.
Disguising the Method
If we’re going to stick with the basic method, then it is wise to find a way to disguise it. There are four changes we can make to do this:
Use a different number of cards. The trick can be performed with any odd number of cards that is a multiple of three. The more cards you use, the more choice a person has when selecting a card. I think this makes the trick more impressive. However, it also lengthens the time the trick takes to complete because there are more cards to deal out each time.
Eliminate the distinctive face-up column-based layout. Instead, deal the cards into three face-down piles and get your participant to guess which of the three piles contains their thought-of card.
Develop a compelling presentation for the trick to justify the repetitious dealing.
Change the revelation. Adjust the revelation to add an extra layer of deception.
You don’t have to make all four of these changes. This clever handling by my friend, and underground cardician, Justin Higham, provides an excellent way to camouflage the usual redistribution procedure simply by changing the way you present the effect.
Justified Twenty-One Card Trick
First, remove twenty-one cards from a shuffled pack and fan them out with their faces towards a spectator. Next, have her remember one of the cards, then shuffle them again. Finally, deal the cards face up into three seven-card piles in the usual manner.
Ask your participant to point to the column containing their card. Then say, “Would you please concentrate on just the colour of your card? Don’t worry about the suit or value. Simply think of the colour.” Feign concentration for a few moments as they do this, then say, “Okay, I now know the colour of your card!”
Let the impossibility of this statement sink in, and then gather up the three piles, putting the indicated pile in between the other two as per the standard method.
Deal the cards face up into three columns of seven as before, and again have her indicate which column contains her card. Then say, “This time, I would like you to concentrate on the suit of your card. Don't worry about the colour or the value. Just visualise the shape of the suit. Imagine that you were drawing that shape on a piece of paper…” Then, after a suitable pause, say, “Okay, I now know the suit of your card!”
Gather up the three piles again, ensuring that the indicated pile is sandwiched between the other two.
Deal the cards face up into three columns as before, and have your helper indicate the necessary column, and as soon as they do so, remember the central card in this column (the selection). Then say, “Finally, can you please concentrate on the numerical value of your card? Don't think of the colour or the suit, just the number.” After a few moments of concentration, say, “Okay: I now know the value of your card!”
At this point, you are telling the truth: you really do know the value, along with the colour and suit. So scoop up all the cards and place them aside, then look at the spectator and say, “The colour of your card is... The suit is… And the value is…!”
If you only change one thing about the standard handling of the Twenty-One Card Trick, use this fantastic idea by Justin Higham. It makes a world of difference to this trick because it fully justifies the three repeat deals. Thanks, Justin!
Here’s a small addition to Justin’s idea: you could use a marked deck in combination with this mind-reading presentation and perform the trick without looking at the faces of the cards yourself. This would greatly enhance the mental aspect of the trick and help obscure the method.
Trust Your Gut
Now that we’ve discussed various ways to improve the trick's standard handling, let’s look at a more involved variation. “Trust Your Gut” is, perhaps, my favourite way to present the Twenty-One Card Trick. In it, the trick is reframed as an “experiment” to measure a person’s innate sense of intuition. This presentation justifies the repetitive dealing procedure, helping hide the method's mathematical nature.
I’ve also paired the Redistribution Principle with Roy Baker’s excellent PATEO Force. Combining these two methods makes the revelation of the thought-of card more impressive because it involves you and your participant making multiple decisions together. Introducing another layer of methodology also makes the trick more difficult to deconstruct. To justify the collaborative process of elimination, I talk about combining my “superior powers of intuition” with those of my volunteer to help locate the mentally selected card.
It is common for people to use some kind of Magician’s Choice (Equivoque) with The Twenty-One Card Trick. You’ll notice that Anastasia uses a simple Equivoque in the Numberphile video above. The issue with how most people do this, Anastasia included, is that they use an inconsistent process of elimination. It will still fool many people, but the inconsistencies will make some people suspicious; they might even guess that you’re controlling the outcome through the way you interpret their choices. The PATEO Force avoids such difficulties and provides a more consistent way to guide your helper to find their thought-of card (with a bit of help from you, of course).
You can learn “Trust Your Gut” by reading this secret page on my blog:
I hope you give it a go. It is an excellent trick to perform with a borrowed deck in a casual social situation when you have more time on your hands.
Twenty-Seven Card Trick—A Better Option?
If your intention is to disguise the method, then I think using twenty-seven cards may be a better option than using twenty-one. To do this, start with a standard pack of fifty-two cards and add a single Joker. This will allow you to legitimately claim that you’re using half the deck to speed things up. Next, divide the cards into two equal piles and ask a spectator to choose one. This not only eliminates the need to count out a specific number of cards but also provides the spectator with a broader selection to choose from, making the trick more impressive.
Here’s the proof by exhaustion for the Twenty-Seven Card Trick:
Twenty-Seven Card Trick Proof:
1 = 01 | 02 | 03 | 04 | 05 | 06 | 07 | 08 | 09 |
2 = 6A | 6B | 6C | 5A | 5B | 5C | 4A | 4B | 4C |
3 = 5C | 5C | 5C | 5B | 5B | 5B | 5A | 5A | 5A |
4 = 5B | 5B | 5B | 5B | 5B | 5B | 5B | 5B | 5B |
The other significant difference with this method is that the chosen card ends up being precisely fourteen cards from the top of the pack. This unique feature makes it ideal for a delightful prediction effect called "Unlucky for Some". If you're interested in learning this trick, click the button below:
In many respects, “Unlucky for Some” is similar to “Trust Your Gut”, but it uses a different revelation (a prediction rather than a location). Instead of testing someone’s innate sense of intuition, in this trick, you measure a person’s natural level of luck.
Thirty-Three Card Trick
You can use thirty-three cards instead of twenty-one. There are a few advantages and disadvantages associated with using this number of cards.
Firstly, you can construct a Piquet Pack from a regular deck of playing cards. Piquet (pronounced “PK” in English) is a two-person trick-taking card game from the early 16th century. It was so popular that it became France’s national game. However, more recently, it has fallen into obscurity. Nevertheless, David Parlett, a games scholar and historian, still thinks it is a game worth playing:
“This classic game of relatively great antiquity, though still one of the most skill-rewarding card games for two, is now played only by aficionados and connoisseurs. Originating around 1500, its decline from about the end of World War One may be ascribed to the popularity of Gin Rummy and other lowbrow games that are easier to learn and faster to play.”2
— David Parlett
Most importantly, Piquet uses a thrity-two card pack: the Ace, King, Queen, Jack, Ten, Nine, Eight, and Seven in each suit. If you add a single Joker to this setup, you have the thirty-three cards needed for the method to work. Strictly speaking, Jokers were not added to the thirty-two-card pack. In fact, they were not included in any decks until the mid-1850s, well after the invention of Piquet. But we can use our artistic licence and pretend they were introduced earlier if needed.
With this smaller deck, it is quite possible to perform a wide range of card magic. In fact, the reduced number of cards makes some tricks and sleight-of-hand techniques easier. Using a Piquet Pack also gives you an interesting talking point to incorporate into your presentations. For example, many of the world’s first card conjurers, such as Professor Pinetti (1750 to 1800), Johann Nepomuk Hofzinser (1805 to 1871) and Jean-Eugène Robert-Houdin (1805 to 1871), all performed card tricks with a Piquet Pack. If you’d like to learn more about Professor Pinetti, I recently published a detailed article on the flamboyant Italian conjurer:
Here’s the proof by exhaustion for the Thirty-Three Card Trick:
Thirty-Three Card Trick Proof:
1 = 01 | 02 | 03 | 04 | 05 | 06 | 07 | 08 | 09 | 10 | 11 |
2 = 8C | 7A | 7B | 7C | 6A | 6B | 6C | 7A | 7B | 7C | 4A |
3 = 5A | 6C | 6C | 6C | 6B | 6B | 6B | 6C | 6C | 6C | 7C |
4 = 6A | 6B | 6B | 6B | 6B | 6B | 6B | 6B | 6B | 6B | 6C |
5 = 6B | 6B | 6B | 6B | 6B | 6B | 6B | 6B | 6B | 6B | 6B |
By using a thirty-three-card pack, you don’t have to count out a specific number of cards as you do with the Twenty-One Card Trick. This means that you can give the entire pack to a spectator, have them shuffle the cards, think of one of them and then give the cards another shuffle. You can then perform the trick without ever touching the cards. You can also talk your participant through the Jay Ose False Cut between each deal to further throw them off the scent.
However, the table above exposes a problem with this particular method. After three rounds of dealing, there is still a possibility that the card is not in the expected final position (sixth from the top of its eleven-card pile or seventeenth from the top of the reassembled pack). See row 3, positions 5A and 7C. This will only happen when the thought-of card is the top or bottom card of its eleven-card pile after the first round of dealing. You can easily fix this issue by performing a fourth round of dealing.
However, I prefer to change the selection process as follows: instruct your spectator to deal the cards into three piles, select one and look through the cards, picking one of them. In this way, the first deal is incorporated into the selection process. You can then reassemble the piles, putting the one containing the selected card between the other two, and continue as usual (with three rounds of dealing). This will allow you to use thirty-three cards with the same presentation from “Trust Your Gut” or “Unlucky for Some”. It is helpful to point out that this idea can be used with most variations of the Twenty-One Card Trick.
Another way to deal with this issue is to keep the cards face down during the trick (I usually do this anyway). After having a spectator think of any card in the entire pack, deal the cards into three piles. Ask your participant to guess which pile contains their card. Once the correct pile has been located, ask them to shuffle (or cut) the cards, ensuring that their selection does not end up on the top or bottom of the packet. You ask them to do this to apparently prevent you from secretly glimpsing the card’s identity. In truth, this subtlety stops the card from ending up in either of the two problematic positions.
After you’ve tested your participant’s luck/intuition/psychic ability, the mentally-selected card will be the seventeenth card from the top (or bottom) of the assembled pack. You will now demonstrate your magical powers. First, divide the cards into three piles, A, B and C. The thought-of card will be the central card of the centre pile (B). Next, eliminate piles A and C. Then perform a Down-Under Deal on the remaining pile (B). The card left in your hand after the Down-Under Deal will be the chosen card.
I’m developing a complete card magic act that uses a Piquet Pack. One of the tricks in it will be a variation of “Trust Your Gut” with thirty-three cards.
Nine Card Trick
Of course, you can perform the trick with less than twenty-one cards. The trick is still functional when using only nine cards. Here’s the relevant proof by exhaustion:
Nine-Card Trick Proof:
1 = 01 | 02 | 03 |
2 = 2A | 2B | 2C |
3 = 2B | 2B | 2B |
4 = 2B | 2B | 2B |
As you can see, it only takes two deals for the card to be in a known location in the assembled packet (fifth from the top/bottom). This positioning lends itself to a couple of fun revelations of the thought-of card.
Firstly, you can spell/deal the word “M-A-G-I-C” and turn over the last card dealt to reveal the chosen card. Or, you can reassemble the piles after the second round of dealing so that the pile with the selection is on top of the reconstructed packet (thus, the chosen card is second from the top). To reveal the card, perform a Down-Under Deal.
You can also perform a streamlined version of “Trust Your Gut” with nine cards. The method is, essentially, the same as the one I describe using twenty-one cards. However, during the second phase of the trick, you only need to make four eliminations in total during the PATEO Force.
However, the small number of cards might make the method more transparent to those members of your audience who are logical or lateral thinkers. In fact, running through the process with nine cards, one of them face up, is actually an excellent way to understand how the Redistribution Principle works. I haven’t performed these variations enough to know whether this is a genuine concern, but it is worth considering when selecting a method.
Fifteen Card Trick
And finally, you can perform the trick with fifteen cards. This number represents a “happy medium” between nine cards (the minimum required) and twenty-one. Here’s the relevant proof by exhaustion:
Fifteen Card Trick Proof:
1 = 01 | 02 | 03 | 04 | 05 |
2 = 4C | 3A | 3B | 3C | 2A |
3 = 3C | 3B | 3B | 3B | 3A |
4 = 3B | 3B | 3B | 3B | 3B |
This handling automatically positions the mentally selected card in the correct location for a so-called “tantalizer deal”. This makes it a perfect method for performing “The Tantaziler” from The Royal Road to Card Magic when you’re short on time and don’t want to deal with a full pack of fifty-two cards. In fact, I’ve written up a detailed explanation of this trick, which I call “Tiny Tantalizer”, for those of you that might not be familiar with the original trick. You can learn it for free on my blog:
Another way to enhance the presentation of the Nine-Card Trick is to refer to the three piles as three hands from a game of five-card stud, one of the earliest forms of the card game stud poker (that originated during the American Civil War). While this game is less popular these days, it could provide an interesting historical context that you could incorporate into your performance. In addition, you could build a fun and engaging gambling-themed presentation around this context, which could help to distract from the repetitive dealing procedure and keep your audience entertained. So why not add a touch of American history to your performance and take your audience on a journey back in time to the Wild West?
I trust that you found this brief exploration of the Twenty-One Card Trick engaging and informative and that it has piqued your interest in experimenting with the Redistribution Principle. Despite its reputation among magicians, this method can be surprisingly effective when executed properly, and I have personally used it to great success in my performances. So don't underestimate the power of this simple yet elegant method. Feel free to incorporate some of these ideas into your own repertoire with the confidence that they do indeed fool people.
Yours Magically,
Marty
Professor Hoffmann, Modern Magic, 43-44.
David Parlett, “Piquet,” Historic Card Games, https://www.parlettgames.uk/histocs/piquet.html.