(Close Window) 
Topic: Open Arrangement 


This is a simple question really. If you have a shuffled deck and you want to put it into Mem Deck order, what's the fastest way to do so? I'm not talking about setting it up during a performance or on the fly. I mean you're at home and you want to stack your deck. Is there a system for doing so other than just going one by one? I personally do it in groups of 3, but I feel like there's probably something faster that someone smarter has worked out. Anyone? 


I make 5 heaps on the table. 19, 1019, 2029, 3039, 4052 Then I sort each heap. 


Bob Farmer recently released his Tarodiction toolbox. That describes a method for this. Also Patrick Redford recently released his method (available to owners of his new book Temporarily out of order). There are threads for both here in Shuffled not stirred. 


It would be nice if we could collectively find an efficient solution to that challenge :) There are a lot of sorting algorithms which are very efficient with abstract data but fail miserably with physical objects (for obvious reasons.) Going from a shuffled deck to a stack order or to NDO is actually the same process; the numbering of the cards, 1 to 52, is the same, only the card values are different. Therefore you could use NDO and suits for your experimenting sessions. Like many people, I usually deal out four piles of 13 cards (113,1426,2739,4052); I then fan each pile and sort each sequence using no particular method. It’s not that bad, but there must be a way to improve on that. So, how about using an algorithm similar to the [url=https://en.wikipedia.org/wiki/Patience_sorting]Patience Sorting Algorithm[/url] ? [b]Method[/b] Hold the shuffled deck face up. [b]Phase 1: Dealing[/b] 1. Initially, there are no piles. The first card dealt forms a new pile consisting of the single card. 2. Each subsequent card is placed on the leftmost existing pile whose top card has a value greater than the new card's value, or to the right of all of the existing piles, thus forming a new pile. [b]Phase 2: Collecting[/b] 3. Pick up the smallest card. It will always be on top of the leftmost pile. 4. Keep picking up cards in order. The next card will always be on top of one of the piles. I have tried with 26 numbered cards and it’s pretty quick once you’re used to it. It takes some time to make the decisions without thinking. The issue I’ve found is that you sometimes need a lot of table top space. So what do you think? Is it possible to improve this sort/merge algorithm, or are there much better ones out there? 


Table space is indeed another parameter to take into account. If you have a large table you can directly deal the cards into their normal position if you draw a grid of 13x4 or 10x6 for example. You just deal the cards onto the grid and collect. Maybe if people are ready to try different versions they could also indicate their sorting time (after some sufficient training) ? 


Yes, the tabletop space issue is a big problem with the quicksort like method I posted above. A couple of times I had 14 packets. Too many for comfort. jmbulg is onto something there with the notion of fixed positions. I vaguely remembered a sort algorithm that uses a fixed number of slots and I went to my favourite website Stackoverflow (I'm a coder by profession) and I have found what I think is a workable method: A [b]radix[/b] sort. It's possible to sort a 52 deck of cards in two passes by always using 10 slots (2 rows of 5 slots, for example) on the first run and 6 on the second. I've tried and it's very quick. I'll post more about it later, but if you're in a hurry, navigate to this site: [url=https://stackoverflow.com/questions/3324715/efficientwaystosortadeckofactualcards]sorting an actual deck of cards[/url] and search for radix sort. 


Patrick Redford can get into Redford Stack from a shuffled deck in the process of doing a few tricks that almost automatically stack the deck for him. Ben 


Just depends on which stack your using. I use the "Redford" stack now after many many years of using mnemonica. I can get from a completely shuffled deck into "Redford" stack all in the hands using no table in less than 2 minutes using the Chinese shuffle. Very quick and convenient. This is all taught in Patrick Redfords new book "temporarily out of order" 


OK, here’s the method. As from my previous post it’s based on [url=https://en.wikipedia.org/wiki/Radix_sort#Least_significant_digit_radix_sorts]The Radix Sort[/url]. For training purpose, It’ll be easier to number a deck of cards, in your stack order, on its face. You’ll be able to see better how the deck gets ordered. [b]Least significant digit radix sort[/b] Hold a shuffled deck, face up. Imagine you have 10 spaces (buckets, spots, etc.) on the table numbered from 0 to 9 where you’re going to build the card piles. [b]1st round: rightmost digit sort[/b] On the first round you will consider only the rightmost digit of you card position in the ordered memdeck and deal the card face up on its spot. So for instance: 2, 12, 22, 32, 42, and 52 are dealt to to spot 2, and 10, 20, 30, 40, 50 to spot 0. Once you’re through dealing the pack, you’ll have 10 piles having each 5 to 6 cards. Gather the deck from right to left, i.e. starting from the rightmost pile (spot 9) that you put on top of pile 8, then on pile 7 up to pile 0. [b]2nd round: leftmost digit[/b] Here you have to remember that cards numbered from 1 to 9, should be thought as numbered as 01 to 09 and therefore will go to pile 0. Hold the deck face up and visualize 6 spaces, numbered from 0 to 5 and deal the cards to their correct pile based on their leftmost digit. So, for example, 1 to 9 will go to pile 0, whereas 40, 41, 42, 43... and 49 will go to pile 4. Once again gather the deck from right to left and the deck is in memorized order with 1st card on top of the facedown deck. [b]A few hints:[/b] It'll take some time before you get used to the idea, but if you manage to deal the cards without thinking, the method is rather fast. If you’d rather visualize the piles ranging from 1 to 0 (think of them as 10), instead of 0 to 9, it’s fine, but when you gather make sure that the 0 pile is picked up last. It might be better, during the 1st round of dealing, to form 2 rows of 5 piles each, as it will be easier to find the relevant pile to put down a card. Something I have just realised is that after round 1, your deck is setup for the excellent Aronson effect [i]Histed Heisted[/i]. [b]Finally[/b] I’ve got some ideas to considerably speed up and limit errors during the dealing procedure and reduce the number of piles. I’ll post them if I manage to finalize them. 


Thank you for posting this Claudio! 


Sometimes an inthehandssort is useful. I use the following mainly as an inthehands sort for NDO. I've used it as well for Aronson, but it can be generalized to any stack: Run through deck upjogging all black cards. Pull out black cards to face of deck. Run through deck upjogging all spades and diamonds. Pull out this half to face of deck. Spread the bottom 13 spades and arrange in order with right hand as if arranging a bridge hand. Cut those 13 cards to the top of deck. Repeat with the next three suits. You are now in Bicycle NDO. To generalize for any stack: 1) upjog all cards within the ranges of 1426 and 4052, and cut to face of deck. 2) upjog all cards within the ranges of 2752 and cut to the face of the deck. 3) Spread 13 cards at a time and put in ascending order, then cut to back of deck. Repeat three more times. Hope this is useful to someone. 


I use two methods. Method One: If there's a table, I'll sort into piles 113, 1426, 2739, 4052. Then I fan each one to put them in order. This is similar to the first reply in this thread. Since I get into my stack from NDO, I also have an inthehands method, but it's slightly more efficient than Landmark's above post. Method Two: First, I like to begin with a (low) diamond on the face, and another diamond (any value) at the rear (i.e., the top of the deck). I'll explain why in a minute. Similar to Landmark, I upjog all of the black cards, HOWEVER, the spades are pushed higher than the clubs (the clubs are only upjogged halfway). Next I spread the cards into a kind of fan, and remove the spades one at a time* beginning with the king and ending with the ace, so they're in New Deck order (KA). I repeat with the clubs but start with the ace and end with the king (AK). Now I spread through the pack again, fully upjogging A7H, and midjogging 8K. I again spread and remove the cards and, starting with the ace, remove them one at a time. Now I'm left with the diamonds on the face, which I'll fan out and put in KA order. This is not quite New Deck Order because the spades and diamond suits are switched, but I need them switched for my particular stack (which is just four faros away). And the reason for having diamonds on the top and bottom is because I save the diamond suit for last, so they make for clamp the cards in place (e.g., if a spade were at the rear of the deck and I upjogged it, it could drop to the floor. Since a diamond is on top, it will keep the cards secure). Using this method I can get into stack order (without a table) in under two minutes (that's counting the four faros). Obviously it's not something that should be done in front of spectators. This method of halfjogging also suggests another solution: ujpog, midjog, and downjog, so that the suits can be sorted in one pass rather than two. I've experimented with it, but the sloppiness isn't worth the effort, especially since this is only done for myself. I'm sure someone could make it work with practice (this is essentially what Lennart Green does with a modified angle jog allowing for a quadseparation. Then again, the guy has hands like catcher mitts). *I said I remove the cards one at a time, but that's not necessarily the case. If the five and six are adjacent to one another (and in the proper order e.g., the five is on top of the six), I can pull them both out at once. 


[quote]On Aug 16, 2017, Bobby Forbes wrote: Just depends on which stack your using. I use the "Redford" stack now after many many years of using mnemonica. I can get from a completely shuffled deck into "Redford" stack all in the hands using no table in less than 2 minutes using the Chinese shuffle. Very quick and convenient. This is all taught in Patrick Redfords new book "temporarily out of order" [/quote] I don't have any work by Redford, but what you describe sounds very much like a Bro. John Hamman concept explained in his effect [i]The Chinese Miracle[/i]. It uses a series (up to 6) of culls/outjogs, aka "chinese shuffles" to get from a shuffled deck to NDO, though obviously the concept can be applied to any stack. Is that the case? The question was about getting to stack openly. If you want to get to stack in a covert way (by skill or cover of trick(s)), there's a lot of material out there. I know that Doug McKenzie has a method to get from a shuffled deck to any stack in just 3 culls! He says that he could order up to 64 cards in 3 culls. That should give a clue as to the method. 


Keep track of four positions without a table? That's impressive. 


Yes, he is! #thumbsup 


As mentioned about the Bammo Tarodiction Toolbox, there's a fulldeck stack sort in that book that requires 8 piles dealt only twice. I think it's detailed in the second addendum.  Richard 


As one of the easter eggs to Temporarily Out of Order I shared a way to get into any full deck stack by dealing the deck out twice (that doesn't use any binary or markings). There are two methods I shared to do so. I also have some work with the Chinese Shuffle that's taught in the project as well. I hope this clarifies! 


For the best work on Radix sorting, get in touch with Tomas Blomberg. 


A few years back, Shoot Ogawa put out a DVD set: Busters. He can order a deck in hands in about 60 seconds. In a nutshell, the cards are upjogged by different amounts based on the value of the cards, then collected out of the deck in order. 


I usually sort the deck back into NDO, and then shuffle it into mnemonica 


Claudio kindly referred me to this thread in reply to a question I asked about setting up for Nick Trost's "Intuition." One way to set up would be to (1) first sort the second deck into NDO, and then (2) go through the first deck as a guide to set up the second deck. Since the second deck is already sorted (and perhaps laid out on the table), the work shouldn't be too bad. But still, I'm wondering if people have ideas about how to accomplish Step 2 efficiently. Thanks, Bob 


Bob, do you do mem work? But even if you don't, mark the back of a deck in memdeck order that you mean to study in the future. The marks must be visible only to you, of course, by that I mean not so obvious that laypeople would spot them. There are a lot of marking systems around. Mark the second deck in reverse order. From now on, you'll be able to setup both packets in preparation for the Nick Trost effect in a matter of minutes and without thinking, using a radix sort. The more you use this technique the faster it'll be. As well, as you suggested, keeping one deck in order would be a good idea. The advantage of a marked deck is that the NT effect can be performed entirely hands off, as you have the necessary information immediately. Plus, I believe a couple of books noted in this thread, will help you to set up the order under the guise of an effect. Anyway, this is one way, others will no doubt think of more. 


Hi Claudio, I don't do mem work (though of course eventually I want to do *everything* in magic :) ). I'll check into marking systems, and in particular the books in the other thread. Thanks, Bob 


Landmark, Your method is appealing and sounds easy (in principle  it will be good practice to upjog and entire half of the deck and pull out only those cads!). So I thought I'd start with your method before trying some of ideas, such as Claudio's that require more thought. But I have a question: How *does* one organize a bridge hand? I did a bit of web searching before asking this and I got lots of conflicting opinions and technical stuff. Would you mind briefly describing your own method for getting each suit into numerical order? Thanks, Bob 


It's nothing very special. Let's see if I can explain for Bicycle NDO. Suppose you are holding the deck in left hand, facing you with the 13 unsorted spades at the bottom. Spread the 13 cards out in a fan so that you can see all the pips. Now with your right hand, fingers pointing down, palm facing you, thumb closest to your body, pick out the KS from above; next scan the cards and pick out the QS on top of that, then the JS and so on up until the AS is on the face of the packet. Then just cut those to the top of the pack, and continue the same with the diamonds and the rest of the suits. 


OK, that makes sense. Thanks! 


Hi landmark, Just wanted you to know that your method worked really well for me. I thought you'd want to know that, yes, your method *was* useful to someone. Here's what I wrote on the thread in which I asked for an efficient way to arrange for Trost's Intuition: I'm reporting back, here. Thanks to everyone for their ideas. Here's what I did; I'm sure it isn't the most efficient method, but it seems plenty good enough for my purposes. I used landmark's technique to physically arrange one deck into new deck order. Then I spread the cards on the table. Then I spread the cards in the other, wellshuffled, deck, so that one spread was "above" the other, i. e., I the two spreads were parallel with one closer to me than the other. From there it was quite easy to arrange the NDO cards into a gradually growing deck that was in reverse order to the shuffled deck. I was pleased to discover that it was easy and natural to put the NDO deck into *reverse* order: I just went from the face of the shuffled spread toward the back, choose NDO cards that matched the shuffled cards, kept them faceup, and put each new card faceup card on top of the previously chosen cards. (I hope that made sense! Clarification on request.) I'm guessing the whole thing took me 1015 minutes, though I didn't time myself. But considering that this was my first time, and that I didn't follow all of landmark's instructions, I was happy with the result. If I can get the time down to five  or even ten  minutes, I'll be happy. That doesn't seem like too much time to spend preparing what's reputed to be a really effective trick. 


I favor inhands methods like the Radix Sort for getting a deck into NDO or cyclical stack. With "Card Setting" I can get the deck into NDO in just under two minutes See; Shoot Ogawa Busters 3 Shoot also teaches how to use Card Setting to put the deck into Si Stebbins order with two additional Faro Shuffles, all inhands. PressureFan 


I just separate my stack into first half second half then if I am using a table I fan each half and remove cards laying them on the table in order. I just timed myself and it took 1 min 50 seconds. If no table I separate into half stacks again and just run through each half placing cards on the face as I come to them. This way without the table took 2 min 20 seconds. Doing it this way one thing that helps to speed it up is look for the next two cards at the same time. If you find the second first just put it on the face then slip the first card under it when you come to it. This way you always find at least two cards with each run through of the half stack. 


I recommend to set up the deck always as if it were in front of the audience. Additional practice. If you can do it fast and casual in front of the audience, you're ready for a lot of great "impromptu" card magic. 


I use an idea, that belongs to Nikola, and I have read it in Tamariz’s book. This is done in front of the audience as part of your tricks. You !@#$e three cards to three spectators, that are the positions 50, 51 and 52 of the stack. The deck is shuffled by the spectators and you say you can name every single cards in the deck, except the three selections, but in a “random” order, and without repeat any card and even blindfolded. You give one third of a deck to 3 spectators, and they make a fan towards his faces and as the magician name the cards, are put on the table. This must be done in an increasing rhythm to cause the spectators can not follow you as a comic situation. If this is done correctly, is a great trick, and you can follow your routines. In the end, you reveal the three selections too... 


The radix idea described here is essentially a sort based on the decimal number system. You can do similar sorts in other bases, for example two or four. In a decimal based sort you need only two passes (because 10^2 > 52 cards in the deck). In a binary based sort you need six passes because 2^6 = 64 > 52 (and 2^5 = 32 < 52)  a little sloppiness with < and > rather than less than or equal to etc as appropriate but you can work it out. Six passes is a lot but it means you can do it in the hand with an angle separation or similar. Sorted into suits and in numerical order you can do this with four passes for numerical order and then one pass into red/black and finally one pass into the suits. The procedure is: 1st pass: separate the odds and evens. 2nd pass: separate A,2,5,6,9,10,K into one group and 3,4,7,8,J,Q into another. 3rd pass: separate A,2,3,4,9,10,J,Q into one group and 5,6,7,8,K into another. 4th pass: separate A,2,3,4,5,6,7,8 into one group and 9,10,J,Q,K into another. Then into colours and suits as described above. You need to be careful of the order you collect the two groups. If you do an angle separation you are reversing the order of the cards each pass. Then on the third pass you need to collect the two groups in the opposite order to how I have written so 5,6,7,8,K is the first group. If you can do something similar to an angle separation in the hands but separate into four groups on each pass then you can do this in two passes plus one pass for the suits. 1st pass: 1st group 1,5,9,K 2nd group 2,6,10 3rd group 3,7,J 4th group 4,8,Q 2nd pass: 1st group 1,2,3,4 2nd group 5,6,7,8 3rd group 9,10,J,Q 4th group K Then one pass for the suits. It should be obvious with a little experimentation the order you need to collect the cards after the first pass to achieve what you want after the second (and third) pass. 


Hi folks, For some reason I wasn't notified about the five posts prior to the one I'm writing now until today. So, a belated thanks to you all! Wayne, I'm intrigued by your method; easy to remember and it sounds pretty efficient to my untutored ear. One question: What is an "angle separation"? Thanks, Bob 


I think angle separation is the name of what Lennart Green does to separate the cards into two groups under the guise of just looking through the cards and putting them one at a time into your other hand and then strip them out into the two groups. Harry Lorayne has a similar method that he calls the Great Divide that he claims is better and often disputes Lennart Green origination. See this thread for example: https://www.themagiccafe.com/forums/viewtopic.php?topic=58372&forum=2 


Other thoughts: 1. You can use the angle separation or great divide methods as practice of those techniques 2. I described the method for sorting into suits but you can sort direction into a stack order by the same method. Using a binary sort you would sort i/ into two groups based on equivalence modulo 2 (odds and evens) ii/ into two groups based on equivalence modulo 4 (1 and 2 in one group and 3 and 4=0 in the other group) iii/ into two groups based on equivalence modulo 8 (1,2,3,4) and (0,5,6,7) in the two groups iv/ into two groups based on equivalence modulo 16 (18) and (0,915) v/ into two groups based on equivalence modulo 32 (116) and (0,1731) vi/ into two groups based on equivalence modulo 64 (132) and (3352) 3. If you can't do a 4separation in the hands then you could do it on the table. a/ If you are sorting into suits then two of these will get you into numeric order then you can do two binary sorts in the hands to sort into suits. 4. If you want new deck order then you can reverse the two appropriate suits at the end of this (or do something more complicated on the way where spades and diamonds were sorted in one order and hearts and clubs in the reverse order) 


For my current purposes (setting up for Nick Trost's Intuition by putting a deck in NDO into the reverse order of another, wellshuffled deck), using a table would be fine. I gather that angle separation/great divide are for use in the hands. On looking at your 4pass method, I'm wondering whether there's a pattern to the second pass? 


Angle Separation is a complete outandout ripoff of my THE GREAT DIVIDE. I don't "claim," Wayne  IT IS. 


[quote]On Jul 17, 2019, WayneBurrows wrote: The radix idea described here is essentially a sort based on the decimal number system. You can do similar sorts in other bases, for example two or four. In a decimal based sort you need only two passes (because 10^2 > 52 cards in the deck). In a binary based sort you need six passes because 2^6 = 64 > 52 (and 2^5 = 32 < 52)  a little sloppiness with < and > rather than less than or equal to etc as appropriate but you can work it out. Six passes is a lot but it means you can do it in the hand with an angle separation or similar. Sorted into suits and in numerical order you can do this with four passes for numerical order and then one pass into red/black and finally one pass into the suits. The procedure is: 1st pass: separate the odds and evens. 2nd pass: separate A,2,5,6,9,10,K into one group and 3,4,7,8,J,Q into another. 3rd pass: separate A,2,3,4,9,10,J,Q into one group and 5,6,7,8,K into another. 4th pass: separate A,2,3,4,5,6,7,8 into one group and 9,10,J,Q,K into another. Then into colours and suits as described above. You need to be careful of the order you collect the two groups. If you do an angle separation you are reversing the order of the cards each pass. Then on the third pass you need to collect the two groups in the opposite order to how I have written so 5,6,7,8,K is the first group. If you can do something similar to an angle separation in the hands but separate into four groups on each pass then you can do this in two passes plus one pass for the suits. 1st pass: 1st group 1,5,9,K 2nd group 2,6,10 3rd group 3,7,J 4th group 4,8,Q 2nd pass: 1st group 1,2,3,4 2nd group 5,6,7,8 3rd group 9,10,J,Q 4th group K Then one pass for the suits. It should be obvious with a little experimentation the order you need to collect the cards after the first pass to achieve what you want after the second (and third) pass. [/quote] Just brilliant. For anyone out there interested in inhand separations, check out Karl Hein's ideas on this. He's got a beautiful routine called 3 degrees of separation. Highly recommended. P.S: As the alreadyenormous amount of MDs grow, they all seem to be designed to be NDOfriendly, so I am getting more and more used to going through a NDO sorting process followed by the getting into stack. It is quite handy, and a great way of practicing. 


Harry, I said you claimed it is better. I am sorry I do not know. I have seen Lennart Green's separation taught. I have not seen yours. About whether Lennart Green's separation is original I said that you dispute this. I have read what you say about this but I have not heard Lennart Green's take on what he has done and what inspired him. Really though separating cards into two (or more) groups through a form of jogging is not really groundbreaking technique that is worth arguing over in my view. 


[quote]On Jul 22, 2019, Alex Carantoña wrote: Just brilliant. For anyone out there interested in inhand separations, check out Karl Hein's ideas on this. He's got a beautiful routine called 3 degrees of separation. Highly recommended. P.S: As the alreadyenormous amount of MDs grow, they all seem to be designed to be NDOfriendly, so I am getting more and more used to going through a NDO sorting process followed by the getting into stack. It is quite handy, and a great way of practicing. [/quote] Thanks Alex. I will check out your recommendations. 


Obvious stealing/rip off is worth arguing over in MY view, Wayne. You might want to learn my THE GREAT DIVIDE before you even discuss this. You might even do some research and read my stuff about Mr. Lennart coming to every lecture I did in Sweden, Finland, Norway  when I was demonstrating/teaching/selling my THE GREAT DIVIDE  before he published his ripoff. You just might. 


[quote]On Jun 3, 2014, Harry Lorayne wrote: Agreed. Lennart is one of the best cardmen around today, with plenty of contributions  he doesn't need to steal. [/quote] Source: https://www.themagiccafe.com/forums/viewtopic.php?topic=58372&forum=2&start=320 