Making the World’s Most Difficult Puzzles: Interview with Puzzle Master Jack Krijnen

Jack Krijnen was born in 1952 in the Netherlands, where he has spent his entire life. It was while studying mathematics at the Eindhoven University of Technology when, in his own words, he, “got infected by the puzzle virus.”

Krijnen explains, “Pieter van Delft and Jack Botermans are to blame for that; in 1978 they published the book Spelen met puzzles, and that was what triggered me to make my first puzzle: Van der Poel's 18-piece burr. All I used was a small saw and an old chisel that I had found in my father's shed. Still, every time I look at it, I reminisce.”

From there on Krijnen says it was a small step to his first design effort; in 1980 he finished the design of the Cube. “It's kind of special,” he says, “as it requires 9 moves, all with the same piece, to free that piece.”

For the next 20 years he says the puzzle virus lied dormant. “I designed and made only one or two puzzles,” he says, adding, “Working, pleasing my wife and raising two children, a boy and a girl, took most of my time – not to regret that, by the way. But with the children growing up, the amount of puzzle time increased and from 2000 on it became a real hobby, or, according to my wife, addiction. At the moment, there is not a day that I'm not one way or another occupied by puzzles. It can simply be talking about puzzles via mail or social media. But also making puzzles I like for myself to solve, or making small batches of my own designs to sell or give away.”

Me: For someone like me, who doesn't know much about making puzzles, can you describe the process of making one. How does an idea occur? How do you turn an idea into something physical? What processes are involved?

Krijnen: You cannot simply sit down and say to yourself: Well, let's design a new puzzle tonight. It always starts with an idea. Sometimes that's all there is to it. I got the idea for the Wine-rack while I was assembling a wine-rack; I made it from the remaining parts, the same day. This must have been somewhere in the 80's; much later, in 2008, I discussed the design with John Devost and came up with a modification, the Pinhole Cube. Totally different appearance, but the same puzzle.

Sometimes it involves more inspiration than perspiration, such as with my elephant design. My daughter's favourite animal is the elephant, so I wanted to design one for her as a puzzle. After I figured out the locking sequence, the rest of the design was merely a crafting challenge: how to give it a nice look. The name of the puzzle is not surprisingly: Mary's Pet.

Me: Can you explain more about the technical aspects of design?

Krijnen: I have spent much time on high level designs, especially of 18 piece burrs. In 2002 Goh Pit Khiam held the record with his Burrloon. It took 33 moves to free the first piece. For me this was a fascinating achievement, and I wondered how I could improve on this. The concept I finally made to work was based on the Piston Puzzle of Peter Marineau. This was just a 6 piece burr, but it had a very ingenious to-and-fro sub-sequence of moves. Tipperary has a sub-sequence of 6 moves that subsequently have to be taken back. This is repeated 3 times. It interferes with a 7 move key sequences, resulting in a level of 43. In all, I spent 6 months with an average of 2 hours daily to reach a design with a unique solution realising the intended solution pattern.

Nowadays a level of 43 for the 18 piece burr is just a fact in the history of this form factor. I have been in competition and cooperation with Alfons Eyckmans. The record now is a level of 166 reached in the joint design of Supernova. I do not expect this to be improved upon in the coming years; I performed extensive computer analyses with a program I wrote.

Speaking of computers: the free availability of Burrtools as a solution as well as a design tool for interlocking puzzles has given a boost to the number of puzzles with high levels. Looking back, I value my Tipperary design higher than Supernova. It is the difference between an idea coming to life and just reaching a higher level.

Me: So what happens in the workshop?

Krijnen: A lot is said already in books, publications, blogs and forums. On YouTube one can find video's about practically every aspect of wood crafting, including the tools to use and how to use them. So, this won't be more than a small introduction on the subject.

It starts with finding a woodshop and choosing your wood. I visit a shop not too far away a couple of times a year to fill up my wood supply. They sell different kinds of wood by small boards, typically 80x15x2 cm. That's quite ideal for my purposes, single copies or small batches of puzzles. Usually I work with common and easy-to-use woods like maple and mahogany. Because puzzles sometimes require three or more contrasting wood colors I make sure to have other kinds on stock also, like walnut, zebrano, padauk, bubinga, and more. More woods may contribute to an appealing appearance too.

For making rectilinear burrs, for instance the traditional 18 piece shape, really square rods are essential for a good result. I make them the standard way, saw them with a saw table from the boards and smooth the sides with a planer. I size the pieces with a small table saw using a homemade shed. The same shed is used to remove the notches from the pieces. Most puzzle makers use additional parts and glue for inner corners, but I prefer the use of a small router and completion with a chisel. As far as I know, only Maurice Vigouroux works the same way. Some puzzles require a routing pattern on the edges of the pieces to enforce a unique solution.

For finishing the pieces I've long had a preference for oil; recently I'm using wax finish too.

It is more of a challenge if the shape of (parts of) the puzzle is less straightforward. I've mentioned Mary's Pet already: how to shape tusks, trunk, tail?

Some puzzles come with a cage or frame, glued together from several parts of wood. You have to ensure the joints are strong enough; no one likes the construction to break when first strength is applied. So usually just gluing together is not sufficient. Other ways to add strength to glued pieces are the use of dowels or splines.

The frame of Power Tower is built from 9 parts, each made in way similar to making the pieces of rectilinear burrs; next the frame is glued together and made into the rounded shape with a band sander.

Even more of a challenge – and fun – are polyhedral dissections. Stewart Coffin is the well-known name here. I have made several of his designs following his instructions for the basic building shapes as tetrahedral blocks, rhombic pyramid blocks and so on. Special jigs and sawing under different angles are required for those.

As I said, this only touches on the subject and is far from complete. I have made puzzle boxes with a wooden inlay; you have to find out the best way to do that. In books and forums you can learn about adequate techniques. But in the end it comes down to choosing the way that suits you best – mostly after experimenting and the spending of time, and wood. The satisfaction: did I really do that?! This is the reward!

Me: What for you is the most impressive puzzle you've ever come across?

Krijnen: That's Bill Cutler's Binary Burr. It won a first price at the IPP puzzle design competition in 2003. Bill created an interlocking equivalent of the famous Chinese ring puzzle, where each extra ring doubles the number of moves to solve the puzzle. I never thought of even trying this; it was like attempting maybe not the impossible, but definitely the unthinkable.

Krijnen: There is a difference between a high number of moves and difficulty of solving. It is easy to design a 2 piece interlocking puzzle with for instance 1000 moves to solve (though crafting would take some time). However, it would not be much of a disassemble challenge. It would only be boring.

Me: Is there a puzzle you struggled with, or just could not complete?

Krijnen: I certainly like the challenge of mastering complexity. In 2003 I designed Tipperary; as I told you before it took me 6 months with an average of 2 hours daily. At the start I had no idea whether I would be able to complete the task I committed myself to.

In the autumn of 2013 Goh Pit Khiam invited me to join his research into n-ary burrs. These are burrs with a number (say: m) of key pieces; each key piece can be in n states. Solving the burr requires the passing of every possible combination of states of the key pieces. So, the level of the puzzle is proportionally to n^m. Bill's Binary Burr is an example with n=2 and m=6.

Pit Khiam was looking for extendable (in both n and m) designs, and had come up with some attractive specimen already. I noticed they all required extra synchronizing pieces to enforce the n-ary sequence. I wondered if it would be possible to do without the additional pieces, and achieve the synchronization with the n-ary sequence with the key pieces themselves. For two months I tried, with no success. I let it rest for over a month. I was about informing Pit Khiam I was going to cease my attempts, but gave it a last try with another setting, 3-ary instead of 2-ary. To my own surprise I found a rudimentary working design within hours! The next week was hectic, we spammed each other with numerous improvements, and finished the Power Tower design. I didn't get much sleep that week...

Me: Can a laymen create a very difficult burr?

Krijnen: Nowadays, with the free availability of Burrtools, it is possible to define a high level and difficult burr without much design experience. It is merely a matter of trial and error, and have the program do the work. However, it will seldom lead to a really appealing or elegant puzzle.

Me: I guess as you make them, you are great at actually solving puzzles?

Krijnen: I'm not great at solving; except for puzzles requiring a systematic and logical approach.

Me: Some of your puzzles must drive some people crazy – I can barely do the easiest ones! – but there must be lots of benefits to people solving puzzles. Can you suggest why these puzzles might help people in an educational way?

Krijnen: You can practice thinking in shapes, it enforces the spatial visualization ability. Exercise understanding of symmetrical shapes, train your deductive capabilities. And in the end, there's always the fun a good puzzle offers.

Me: Lastly, the younger generation don't seem as interested in thing such as these beautiful interlocking puzzles? It seems a pity, as so much work, craft, creativity goes into these puzzles? Do you have anything to add concerning this?

Krijnen: I do not recognize this. I'm a regular attendant of the Dutch Cube Day, and of several forums. It's true that the really young, under 20, are a minority, but the puzzle world is far from a grey community and spreads over a wide range of age. Coming to think of it, age was always of minor importance.

A Special thank you to Jack for his time from all those at SiamMandalay.

By James Farrell,

If you are in any way involved with interlocking puzzles James would like to hear from you. Contact: jamesaustinfarrell19@gmail.com Send James an email.

May 14, 2015 —