Kevin Wald is probably the most prolific American constructor of variety cryptic crosswords, by far. Unfortunately, he is not as well known as some others, because his output is not shared widely. On occasion, you will find his puzzles in The Enigma, the publication of the National Puzzlers’ League (NPL). In the NPL, he is known as Ucaoimhu (pronounced “Oo-kuh-voo”), and universally admired for his creativity. Most of his puzzles are available on his website. In an e-mail conversation, he answered some of our questions about his work.
Your puzzles’ complexity is legendary. You combine layer upon layer of structure, in a way that is thematically coherent and entertaining for the solver. The resulting constraints must be quite a challenge. How do you meet it?
Generally, the constraints on the grid are separate from the constraints in the clues. So I typically start by just gridding, completely ignoring the “each clue contains…” stuff; I only have to worry about putting in a certain number of words with a specific property, or putting specific letters in specific squares, and so on. At that point, I find constraints very helpful, since they vastly reduce the number of possibilities I have to try. Of course, sometimes the constraints I’ve chosen make things impossible—but then I just loosen the constraints (or in extreme cases toss the whole idea out and try something else), and I don’t tell anyone about the original concept that I had to scrap.
After that comes clueing. In fact, the constraints on the clues might not even be fully decided until after the grid is finished. Say, if I know I want to put a message in but the puzzle idea doesn’t hinge on exactly how it’s phrased, I might settle on the phrasing, and even the technique used to hide it, after the grid is completed and I see how many clues I have to work with. Or if I wasn’t planning to do anything special with a subset of the clues, but then I get an idea later, I might just add that on.
Anyway, the clueing is generally the easier part for me; I can pretty much guarantee that any answer can be given a clue with one specific letter in one specific place, or with one specific letter substituted for one other specific letter, for example. Even more exotic constraints (writing one ambiguous clue that might clue two different specific words, say) will generally work out given enough time to try different possibilities. (And I can work on clues in my head while walking or waiting on line or watching a movie, which is not the case for gridding.)
It seems like you bring one or more topical cryptic to almost any event you attend. How do you manage to not run out of ideas?
Most of my puzzles are written “reactively”; there is some upcoming event, say, and that event itself inspires the puzzle. If it’s a party, does that party have a theme? If it’s a National Puzzlers’ League convention, what city is it in? If it’s a movie, what’s it about?