Constructing Analysis #1: Stuffed with Ambition!

I’m no stranger to crossword ambition. But there’s a difference between coming up with a cool concept, and making it into a publishable grid.

That’s why I want to talk about Tomas Spier’s Universal crossword from Monday, September 13, 2021: “It’s a Zoo Out There!” Because I thought that the constructor not only created a solid theme, but they even went the extra step by adding another layer of depth. But to pull off this ambitious plan, there needed to be compromise. So how exactly did Spiers negotiate with their 15×15 grid to bring it to fruition? 

The Puzzle in Question

The puzzle is centered around 50a. [Creature comforts? … and a hint to the words that bookend 19-, 24- and 44-across] – STUFFED ANIMALS (14)

The themers:

19a. [Valuable designation for a collector] LIMITED EDITION is a stuffed LION

24a. [Prepare to order a drink] BELLY UP TO THE BAR is a stuffed BEAR

44a. [Balancing daredevil] TIGHTROPE WALKER is a stuffed TIGER

As I was solving, I liked how the theme snuck up on me — it was very hard to guess the theme before the great reveal, which resulted in a nice “aha” moment!

Moreover, the animals aren’t just random: You might have noticed that they’re all part of a famous quotation from a famous movie (based on the book), The Wizard of Oz. 

But there’s a small problem – the animals are in the wrong order! It’s LIONs and TIGERs and BEARs! TIGERs and BEARs!

What do you think? Was this a simple oversight? Or perhaps, was there a more concrete, construction-related reason for it?

Solutions… And Drawbacks

Isn’t it silly how the constructor didn’t maintain the order? They should have just switched the second and third entries!

Well, in order to understand these decisions, I wanted to try reconstructing the puzzle from the bottom up. And though the switching solution seems simple, the issues arise almost immediately:

Oof. That I?W entry looks nasty. Same with U?F. Also, the middle-west section is already posing problems – THROB AND THUMB are probably the best options for 25-down, but what letter do I put next to that resulting H? 26-down as I AM ME? I AM HE? I’m not feeling it.

Okay, so maybe it’s not so easy as just a theme-entry transplant. How can we configure the black squares to avoid that T???B section? As I tinkered more with the grid, I realized that the difficulty of this puzzle’s construction lies in the length of the revealer STUFFED ANIMALS (14). The tiny black square needed to put a 14-letter word in a 15×15 grid creates so many limitations. Maybe for flexibility you’d want to put the first 14-letter word not in the fourth row, but in the third (a typical choice in many crosswords). But that would result in a two-letter word in the northeast corner – a crossword no-no.

The fifth row holds its own difficulties. By placing the first entry on the fifth row, all of the themers get much more smooshed as a result. In order not to touch, they’re forced to be in the 5th, 7th, 9th, and 11th rows. This creates additional restrictions: you’d have to include more words that pass through two, three, or even four themers, which would limit your word options. This is no small task, especially since the across entries span the entire grid.

One other creative option could be to put the themer in the third row anyways, but fill in the two spaces above it. I’ve noticed that constructors tend to have an aversion to this method, because it looks a little ugly – but to counteract that, they tend to use a staircase pattern in the corner:

That could be a pretty good way to go. However, the drawback is that those corners cause this grid to use a lot of black squares – 41, to be exact. (The NYT suggests 36 as the maximum for a 15×15, though Universal is a little more lax.) It’s not terrible, but less white space means less letters, which could make you feel like you’re shortchanging your solver a little.

You could, of course, remove some of those interior black squares. But that would exacerbate another feature: The word count. The example grid above contains 72 words, and removing more black squares will decrease that number further. Generally, less words often will result in longer words, which often need to intersect. This could result in a tougher solve, since there will be less clues to provide a foothold.

Again, that’s not inherently a deal-breaker. After all, the published version of the puzzle has 74 words, even though Universal has a maximum of 80. But lower word counts are often used in themeless puzzles – think of puzzles with stacks of three or four 10- or 11-letter words! – which are often possible because they have less restraints. For a puzzle like this with four large theme answers that need to be in the grid, you’re probably facing a tough battle.

Considering Fill

I often find that creating themed crosswords can feel like a constant battle between theme and fill. You’ve got a killer theme, but you need legitimate intersecting words so that your solver can even uncover that theme. How many TBSPS of abbreviations, partials, obscure names, and other OLIO do I need to make my puzzle work? Are my ambitious theme entries worth that price?

Clean fill is a must in this competitive crossword-constructing landscape: especially for a puzzle like Universal’s, which maintains an “easy-medium difficulty level.” Theme might be the overarching movie plot, but the fill is the dialogue and the cinematography to keep you invested. Unfortunately in crosswords, we often can’t have it all, and we need to make choices. After all, we’re only human! (Sometimes humans with autofill.) 

Overall, I think the choices made in this puzzle are justifiable. I liked how entries are all about the same length, and span pretty much the full grid, which I found enjoyable to solve (and also pretty to look at!). The fill felt relatively fair too – I personally would try not to cross the proper names ENLAI and ENYA (as seen in the grid’s middle-west), but the N is pretty inferable. And with colorful themers and 58 letters of theme (far above the typical minimum of 45-50 squares), this grid must have required quite a bit of massaging to make it as good as it is.

But there’s a bigger question that’s worth asking… Is this trio of animals even worth keeping in the first place?

More Drastic Considerations

In their CrosswordFiend blog post, pannonica notes how the LIONs and TIGERs and BEARs reference, while famous, doesn’t really connect to anything. Besides the sequence being mixed up, the Oz reference doesn’t matter at all to the revealer, and there also aren’t any bonus entries (like OH MY!). And it’s hard to argue the point: I think if no sacrifices needed to be made, all of these suggestions would probably have resulted in a stronger puzzle. 

Herein lies the conundrum: As my attempts have made clear to me, simply fixing the sequence of the Universal puzzle would have been a massive ordeal – and including a bonus Wizard of Oz reference might have broken the CAraMEL’s back. But is the imperfect sequence of animals too much of a tease? Should the Universal constructor have scrapped these beasts for new and unrelated animals?

In my research, I stumbled upon another crossword, Bruce Stuphin and Doug Peterson’s LA Times crossword “Kiddie Taxidermy,” from Sunday, February 3, 2013.

The concept is similar: various creatures are “stuffed” to make common phrases, with the same STUFFED ANIMALS revealer. Stuphin and Peterson even use two of the same animals – LION and BEAR – albeit with different “stuffing.” But what struck me in the LAT puzzle was how the animals were connected. There’s no Oz reference (likely due to the size and number of themers), but there are still gentle guidelines: For one, HARE, MOLE, MULE, LION, etc. are all mammals. But unlike TIGER, they’re all exactly four letters, which also allows for a more “even” stuffing.

Ultimately, I’m not sure which is better. Yes, conceptually the published puzzle could be more complex. But at the same time, all the STUFFED ANIMALS are big, carnivorous mammals that you wouldn’t want to meet in the wild. And even though the sequence isn’t in order, it’s pretty straightforward to recognize that the animals are connected by the famous phrase. What outweighs what: The slight joy of recognizing the trio, or the slight annoyance of the trio not being in order? I think it depends who you ask, but I can imagine that a newer (dare I say, less jaded?) solver in particular might find less of the annoyance and more of the joy.

Final Thoughts

It’s up to each constructor how they want to balance their theme and fill. At the end of the day, I can’t fault the constructor for scrapping a tougher sequence in order to make the grid smoother. If something’s got to go, I’d rather it be the *slightly more accurate* Wizard of Oz reference.

But we can still daydream: What if fill didn’t matter in this grid? And what if we were to come up with my own overly-ambitious “aha!” moment? 

First, I’d try putting LIONs, TIGERs, and BEARs in order. And then, in the bottom right, as the last across-answer of the puzzle: an easter egg in the innocuous word TOmaTO, a final “stuffed” Oz-ian animal.

(For your own sanity, try not to think too much about a stuffed TOTO.)

Subscribe for alerts whenever I post. And follow me on Twitter.

Agree? Something I missed? Leave a comment!