Making My First 15 x 15 Crossword

grid

I thought it might be interesting to record my thoughts about what it was like to compose my first crossword puzzle, which I made available as wij001.puz.

This commentary is in the form of 4 short sections about Theme, Grid Symmetry, Word Fill, and Clueing and Difficulty Level.

I've written so as not to give any spoilers outright, except for the most minor of words, but there are a few hefty hints.

Theme

This puzzle was suggested by the May 20 puzzle by Frederick T. Buhler in which HODGEPODGE was one of several theme entries with a H----P---- pattern. In a message to the Today's Forum, I suggested that relaxing the HP constraint might make for another interesting puzzle. Since that message listed several of the words that I eventually used in my theme, I figured I should add a "twist" to my theme lights so that the finished puzzle would be more entertaining for the Forum Cru who knew about the origin of my puzzle.

Of course, solvers who know nothing of the puzzle's origin might be perturbed by the "twist". After all, there is no obvious reason to switch my XabcdYabcd words into YabcdXabcd words and clue them using such cryptic indicators as "on the rocks", "after a tumble", "in arrangement", and "when upset". (Furthermore, such phrases are generally seen in cryptic puzzles as indicators to mix up all letters, not just transpose two chunks in a special way.) So this could be viewed as a weakness. It was really just intended to throw off the NYTwits who remembered my original post. Certainly a more conventional puzzle could be constructed with the same theme entries but without the "twist".

Also, I should add that my grandmother, who was a crossword maven, used to drink 16A Old-fashioned[s] on the rocks, so when that clue came to me, I couldn't resist using it in homage to her.

In elaborating on the theme, I decided I needed to select phrases according to some kind of coherent criteria. I ended up using only phrases with components ending in Y, all of which were adjectives (or could be used adjectivally). Furthermore, I decided that I wanted a variety of vowels, so I found theme entries that use U, I, O, and A, respectively, as the second and seventh letters. Although several other theme-word candidates were possible, I wanted to avoid repetition among the four vowels. Did solvers notice that variety? Did they care?

Grid Symmetry

Early on in composing this puzzle, I decided to try for a grid with 4-fold rotational symmetry. In other words, the grid looks the same from all four sides. Why the heck did I do that? I think I was inspired by Frederick Buhler's HODGEPODGE puzzle, which was an amazingly dense grid that also exhibits that kind of symmetry.

grid Buhler's grid

I knew that 4-fold symmetry would present special challenges, but I forged ahead. I quickly gave up on using Buhler's grid, because I could not fill it. (I began using HERKY-JERKY as an incipient theme phrase, and it proved very difficult to use.) Gradually I came to modify the grid, striving to make sure that no region could be isolated from the rest of the grid by the introduction of a single block.

grid My final grid

The requirement of 4-fold symmetry poses a challenge because any alteration to the grid requires 3 related alterations, not just 1 related alteration, as it would in a grid with 2-fold symmetry. This led to some "cheaters" (potentially unnecessary black squares) in my grid. Up until the very end of constructing the puzzle, 6A was going to start one square farther to the left, and be EMTS crossing with EQUAL. I was able to do similar fills at the left and right, but was stymied at the bottom.

Those of you who wondered about the clue 14D Superior char. can now see how that came about. Experienced constructors (and solvers) can tell at a glance that my final grid has at least four cheaters.

Word Fill

Filling the grid was fun and exasperating at the same time. I had to try many different arrangements of my theme words until I found four that would fit well with other words. Very early on I found that I could get 4D. Parley, 6D. Place for boots, and 44D. Contraband delicacy into the puzzle, which seemed like interesting and fresh words. The last one in particular was featured in a recent NYT article (not online). You can read more about why I thought this item might be considered contraband in this Wine Spectator article.

Soon, however, less ideal words reared their ugly heads. I think the first one was 22D. Resinous coating, which I ended up favoring over URETHRAL, which I suspected would be not well received except by Manny Nosowsky. I actually enjoyed being able to fit 24D. Solar path and 49D. Spinner into the puzzle, but was not thrilled by 23A. Voter, 39A. Trembling, 48A. Natural, in a way, and of course 68A. Ukraine river because they are uncommon and seem forced. I didn't think that 58D. Mahler-Werfel would pose much difficulty, but I learned that it did for some solvers. If the clue had been 58D. Mahler-Gropius-Werfel, I wonder if it would have been any easier, especially because of this popular and humorous Tom Lehrer song?

The long entry that I felt was most questionable was 52A. Supplier for a CD maker. I did not find this word in 10C or NI3, and had to see it in NI2 and the OED before I decided to go with it. I was prepared to clue this as "One who overcomes" until I found a more modern usage by doing some web research. Even though I was able to cook up a quasi-modern clue, the word itself is pretty uninteresting.

The little words provided an intriguing challenge in the clueing, but I found myself chagrined to have to put in the very tired ALA, ELO, ARI, ELBA, REDD, TYNE, A-LINE, and the ever-present ARENA into my puzzle. (I trust typing those in plain text doesn't spoil anyone's enjoyment!) I figured, however, that the other interesting words might offset those chestnuts. In retrospect, I see that this could lead to an impression of unevenness in the puzzle which I may strive to avoid in the future.

Clueing and Difficulty Level

Writing standard crossword clues is a new experience for me and proved challenging. Of course, early on I knew that I wanted to use some quasi-cryptic "?" clues for the four theme entries, which settled down pretty quickly.

For the less common words, I generally went with fairly straightforward definitions. The only exception was 24D. Solar path, which I could have clued more explicitly. Instead, I chose that clue hoping solvers might be tempted to enter ECLIPTIC into the 8-letter light. Unlikely, but who knows? [To learn more about the term, see this beautiful photograph and explanation.]

For shorter words, I often tried to see if I could find clues that had multiple senses (verb and noun, say, or adjective and noun), so that the correct answer word was only one of many avenues a solver might go down. Clues in this category include Exit, Part, Row, Medium, and Sound.

For fun, I threw in just a couple of punny clues, such as 6A. Wedding gift?, 42A. Bikini blast [link], and 59D. Arrived at home dirty? I hope these added a bit of fun.

I think that the impression of a puzzle's difficulty must stem not only from its words but also from its clueing, so I will see in the future how to get the right balance. I've also learned that striving to achieve 4-fold symmetry in a grid can backfire if it results in too many uncommon words. I will try to keep these lessons in mind if I am inspired by other puzzle ideas in the future.

puzzles by Will Johnston

NYT puzzle links

home

wij@world.std.com