Oh forgot to mention that I made up and used a special clear plastic clip that perfectly fits the width of the cards (poker size card) and is about an inch shorter than the card. I bought an all clear plastic picture frame from Dollar tree and used a hacksaw blade to cut it to size.

Claudio, neat idea about having pairs that are seven apart and change suits by one. This would eliminate the need for the Seven of Clubs special pip signal of upper or lower group. Or maybe I should say frees up the 7 of clubs pip to signal something else such as left or right pocket that the spectator put the target card in!

As it stands right now the Three Eights trick can be done with a borrowed shuffled deck. Of course your idea can only be done with the magicians specially prepared deck of pairs running through the entire deck and would almost certainly need a Svengali type setup (a regular deck there is a 50/50 chance that the spectator could cut in the middle of a pair so the card cut to could be the second card of the pair and the card following is not part of that pair but is the first card of the next pair).

This problem using pairs and a regular deck could be handled as follows: instead of a spectator cutting the deck to select the 2 cards, after a false shuffle and or false cuts the magician would hand the deck to the spectator and instruct them to start dealng cards face down on the table in a single pile and told to stop wherever they want and place that card face up on the table. When the magician sees that it is a low card he tells the spectator to also take the next card but if he sees that the face up card on the table is a high card the magician says that we need another card and then instructs the spectator to pick up the card that is on top of the discard pile and turn it face up too.

But this method using a non-svengali stacked regular deck could only be done once? Well the magician may get away with doing the trick again even if the pairs break differently than the first time by saying "let's make it more difficult, we will take the other card this time".

When dealing the cards face down onto a pile the low high pairs are being reversed so you can't just put the discard pile back on top of the deck because the first part of the deck would be high low pairs then the remainder of the deck would be the normal low high pairs.

If the discard pile was dealt face up then after the trick is over the discard pile could be turned face down on top of the remainder of the deck and everything's fine for doing the trick again but the bad thing about dealing face up is the spectators may see a pattern of low high. Another way could be to say to the spectator before they make their selections "think of any number from one to 10 ok now double it, now deal down that many cards and stop". Then say "now deal the next two cards to the side they are your chosen cards". This would ensure getting a matched pair.

But we still have the issue of returning the discard pile in the proper order to do the trick a second time.

Note: with a non-svengali deck we could NOT let the spectator cut the deck before doing the trick either time.

Hmmm, what if we have the spectator deal the cards face up but then put them one at a time face up on the bottom of the deck as they're dealing this would preserve the low high order and not give them a chance to see the faces of the card very long and then after the trick is over the magician just splits the deck where the cards are face to face and puts the bottom part back on top face down.

Perhaps using a Svengali deck like you said and just doing a simple two or three complete cuts and take the next two cards is the way to go (there is no discard pile so this keeps the pairs in the proper low high order).

The problem with cutting the deck or dealing and stopping is we are potentially limiting the selection of the chosen card and introducing magician stuff into the minds of the spectators taking away from the purity of the card selection. My original presentations I would spread the entire deck face up and tell the spectator to pull out any card you want. That way the audience knows there is absolutely no way a card is being forced.

Maybe we are gilding the lily but brainstorming is fun. So keep the ideas coming. Thanks.

In order to make it appear that the spectator has a choice not only for the target card but also for the three signal cards I have come up with the following method:
Ahead of time the magician has the following ten cards on the top of the deck:
9C 5S 6H 8S 6D 5H 8C 9D 3D 9D 3C
The order does not matter as long as there's two threes, two fives, two sixes, two eights, two nines. The remaining threes, 5's, 6's, 8's, 9's are scattered anywhere throughout the deck.

With the above 10 card stack on top of the deck I do a false shuffle and false cuts and then hand the deck to the spectator and tell them "we need some cards to be detectives say a number from one to 10, okay deal down to that number and flip the card face up, okay will use those" and then I proceed to have them pull out all four of those cards and use them as the three detectives and of course the one diamond card to represent the diamonds that the thief has stolen.

Note I may put the 9H in the 11th position just as a safety measure in case spectator says 10 and then can't count.

Note I really like to use the eights and the sixes more than the threes, fives, nines therefore I have strategically placed the eighths in the sixes and positions that I think might be more frequently named but any of those are okay to use.

Of course with this method you cannot spontaneously use a borrowed a deck. With a borrowed deck you would have to stick with the eights.

Oops, typo, I listed 9D twice (11 cards). Let's use:
9C 5S 6H 8S 6D 5H 8C 3D 9D 3C
on top.

Using pairs can be done impromptu (I have a few ideas and I’ll try to describe one tomorrow) but I wanted to address the handling should you decide to use a regular deck of cards which have been paired up.

Here’s the way I would present the trick. After a couple of false shuffles (a Charlier false shuffle is enough) and real cuts performed by the spectator, you instruct them to cut the pack once more, without completing the cut, and that the card(s) cut to will be used for the effect. They cut the deck. You pick up the bottom half and deal two cards face up. If they form a pair, you may decide to stop, if they don't, deal a 3rd card and you’ll have a pair.

Assuming you have 3 cards face up, you perform the effect with the 1st card (unpaired) and then repeat with the pair (nice progression that’ll throw people who think to know the principle behind the trick).

One way to add mystery to the effect is, when you have to deal three cards if the first two don’t form a pair, you’ll know that the 3rd one will pair up with the second card and you can therefore deal it face down. Do what you will with this.

Claudio, I think your pairs and dealing the third card is a brilliant solution to the split pairs.
If I understand it right the magician would arrange the three eights and have them given to the assistant who reveals the first card. The three eights are brought back to the magician who rearranges them a second time and has them given to the assistant and the assistant reveals the other two cards.
Very good methodology.
I think there are some big picture issues with the whole pairs approach:
1. it's a stack and if the spectator doesn't shuffle this takes away from the effect.
2. there will be several people in my more astute target audience that will detect a false shuffle and false cuts and then know that something is awry.
3. when the split pair situation (will happen about 50% of the time) occurs a third card is dealt then why bother sending the three eights a second time (theoretically the assistant could have the whole stack memorized and after determining the first card would know the next two cards and now it becomes an ESP thing).
4. cutting to select a card is more suspect than just spreading the deck face up and letting the spectator pick any card, the cut is too much magishing for my target audience. I want this to be a math mystery like Fitch Cheney and Rusduck's criteria and be able to present this as a challenge at math clubs etc.

I think your pairs idea would work very well for a typical lay audience and I really do like your solution oriented thinking!
Thanks much.

PS: a few posts back I mentioned doing a false cut and false shuffle for a ten card stack at the top of the deck but that was just a minor optional thing to select the three eights or three fives or three sixes etc. After thinking about it I will not do any false shuffles or false cuts for this trick.

Thanks, Glowball.

I understand where you’re coming from and that your audience is different from mine. I would never present a pure math effect to my audience, i.e. I would always occult the math (to my spectators, I’m a magician). However, the back-and-forth exchange can bring up some new ideas that could be useful to both styles of presentation. I think it already has, and I’ll share some further ideas towards that goal in a subsequent post.

To be honest, if I were to use the pair idea, I would use the principle either impromptu or full blown with a Svengali like deck. From my point of view, unless it’s a formal show, there’s nothing better than impromptu handlings as all you need is a deck of cards and you’re all set.

As I mentioned in a previous post, the Long/Short setup would be excellent for the effect as you can riffle and overhand shuffle such a deck with no fear of breaking the pairs* and you could actually use that deck for many other magic or mentalist effects.

The half-way house idea of using a regular deck with cards paired up is, in my opinion, too high a price too pay (onerous setup) for the pay-off so I won’t waste time to defend it as I would never use it (and I did not mention that option myself).

Now, for the impromptu use of the pairing principle. Here’s a possible handling: Have the deck shuffled by a spectator and get it back. While browsing the deck to remove 3 eights, notice the bottom card and bring its complement to the top. Table the Eights and give the deck a quick riffle shuffle keeping top and bottom cards in position. Spread the deck face down and ask the spectator to touch any card that you then upjog. You can either pick up the deck and perform Bill Simon’s Prophesy Move, or perform a Simon Aronson’s variation known as “Mark-a-Place” which is very deceptive (and is my preferred handling).

Here's a YouTube video where at 1’40 you can see Allan Ackerman performing a slight variation of the Aronson move : On Your Right

In any case, once the selected card has been turned over, spread the deck face down again and encode whatever card has been selected. On the second round, you must justify the use of the two cards on the sides of the selection with some patter like “You’ve selected that card, but you could have touched any other card such as this one or that one. ” And you upjog the cards on the left and right of the selection. You then code either card of the pair and your assistant can emphatically announce the two cards.

What I find very subtle in this effect is that you actually force two cards without your assistant having any prior knowledge of them, so the effect is very dynamic.

* See Mark Lewis’ The Long and the Short of it for an excellent Svengali routine and technical details on different shuffles.

I think your method might not be direct enough and does look magishing

The simplest way is to use the Criss Cross Force which is surprisingly deceptive.

You could as well use the Prophecy Move (PM): Simply while browsing the deck to remove the jokers (there should be a joker in the deck) position one 8 at the bottom and one at the top. Ask the spectator to insert the joker anywhere in the deck and perform the PM. Ask them whether they want to use the left or right card.

There are many other forces that should be suitable and direct.

PS: Can you again briefly describe what you're using to hold the 3 Eights? I'm not sure I understand what the holder looks like. Thanks.

This is what I bought from the Dollar Tree store and cut it so I only kept the top part.

https://www.shoppopdisplays.com/11228/5-......der.html

The final result was about 2 in high and a 16th inch wider than a playing card.

This is a very durable plastic and I actually cut it with a band saw but it should be able to be cut with a hacksaw blade.

A key point about this holder is that the sides are open so this allows the magician to easily place the three card packet (the three eights) to either side so the three card packet will extend out one side a little bit and then the magician can run his hand vertically up and down a couple of times (thumb on one side and index finger on the other side) to true up the packet inside the holder but in reality is making the packet flush to one side leaving a small 1/16 visible gap on the other side for his assistant to notice.

The use of the gap is not necessary to reveal the target card, it is only necessary to reveal extra information such as which pocket (left or right) the spectator placed the target card and the diamonds.

The main purpose of the holder is so the magician's assistant will know which way is up and which way is down for the packet. A secondary purpose for using the holder is so the spectator won't casually change the order of the 3 eights.

I’ve made one holder with 3 card sleeves held together at one extremity by a rivet. The 3 sleeves can be fanned easily.

As I have expressed multiple times I prefer an impromptu handling without props, but yesterday (we met with a few friends) I started with an impromptu handling and finished with the prop (under the guise of not marking the cards in anyway) but using only 2 Eights to encode one* and then 2 cards at once (complement pair).

The con of using the prop I described is that it takes a bit longer to insert the cards, but there are a few advantages:

1. The assistant does not touch the cards which are fanned and shown to her by the spectator.
2. It’s quicker for the assistant to announce the card(s) as the key cards can be “read” at once without first being removed.
3. The orientation is obvious; because of the rivet the cards can only be displayed one way.
4. It’s possible to code any card with only 2 slots: either the 2 left ones or the 2 right ones, leaving either outside one empty. Leaving the middle slot empty would be too conspicuous.

If I can, I will construct the same 3-card holder but with the possibility to remove any of the 3 sleeves (though it’s possible to know which one has been removed) making the effect even stronger.

The routine was very well received.

* Two cards are necessary to code the value. By leaving the right sleeve (3rd) empty we convey (for instance) that the suit is Spades or Hearts according to the 2nd card orientation. By leaving the left sleeve (1st) empty we convey that the card is either Clubs or Diamonds based yet again on the 2nd card orientation.

PS: My assistant made up this mnemonics to remember the coding for 4/J, 5/Q & 6K: Four Jacks played HoCkey while Five Queens ate CaShew nuts with six Kings sipping CHampagne.

Some brilliant ideas in this thread. Thanks!