I’m very wary of crossing proper names in my crossword puzzles. Many times, I’ve completely torn up an entire grid simply because of two proper names crossing. Obviously there’s a little nuance – e.g. countries, extremely famous people – but overall I’m pretty ruthless.
You might call it overkill; but I’d argue that these moments are informed from my own solving experience. I get frustrated when I’ve completed almost all of a puzzle except for a single elusive square – and it’s worse when, after checking, I have the feeling of “how was I ever supposed to get that?” Those moments happen most frequently with proper nouns, which can often have a “you-know-it-or-you-don’t” trivia vibe. So when I wear my constructing hat, I want to avoid this same feeling for my solvers.
But two recent puzzles – by Ian Livengood and Rafael Musa – caused me to wonder if I’m being unnecessarily cutthroat towards my work. And as I thought about it more, I realized that perhaps not all “difficult crossings” are equally, well, difficult.
With that hypothesis: How might we rethink these moments? Are there ways to mitigate the effects of these “difficult crossings” in grids? And with this whole business of not crossing proper names: Am I just being overdramatic?
Spoilers (click for solving link):
New York Times Midi Crossword: “Green Day” on Tuesday, March 17, 2026 by Ian Livengood
Boswords 2026 Spring League: Week #1 Puzzle by Rafael Musa
Maybe I am being a little overdramatic about difficult crossings. Or at the very least, I do think there are tools and nuances for thinking about potentially problematic areas.
How many tools/nuances? At least 3, I’d estimate, with the first one being this:
1) Using Different Knowledge Pools
This is a fundamental trait of crossword puzzles: Because every square is usually part of 2 clues, a solver essentially has 2 resources to fill in any given square (i.e. the Across answer and the Down answer). So if you don’t exactly know how to spell an answer, for example, you’re hoping/praying that the crossing word will help.
As a solver, one of the worst feelings for me is when a single gap in knowledge makes it impossible to progress. For instance, I have basically zero knowledge about baseball. This means that if there’s a puzzle where (god forbid) two non-Babe Ruth baseball player surnames cross, I’m pretty screwed.
So as a constructor, it’s worth mitigating these possible instances by intentionally avoiding situations where two crossing words essentially “ask the same question,” i.e. draw from such a similar knowledge pool where you likely either know both or know neither. Instead, two crossing answers should ideally complement each other and offer a broader entry point.
I notice these “complementing” proper nouns in Ian Livengood’s “Green Day” NYT Midi puzzle, particularly in the upper-left corner. The section contains ULTA (clued as the beauty brand) crossing LOGAN (clued as the Boston airport) and AESOP (of the fables), both of which cross MAYO CLINIC (clued as the hospital; the theme also helps due to the first part being an Irish locale).
To be honest: If I had constructed this section, I’m not sure I would have kept it, just on account of the pure density of proper nouns (which account for nearly half of the entries in this section).
But I do think there’s justification. Not only are the entries all pretty prolific, but I think it’s important how they draw on different knowledge bases: beauty/business, literature/history, healthcare, airports. In my mind, this would be more of a problem area if it had ULTA crossing AVEENO, or LOGAN (as currently clued) crossing O’HARE.
The Fireball spec sheet encourages clues that are “a mix of all topics of interest to a cultured person,” which I think is a pretty nice summation of what many crosswords aim to do. With that in mind, I think this puzzle fits the bill pretty well in terms of its breadth of topics.
2) Lending a Hand with Creative Clues
Rafael Musa’s Week 1 puzzle in last week’s Boswords Spring Themeless League had an unassuming moment that I think was handled with some subtle thoughtfulness. The spot in question has MAHAL (clued as [Mumtaz __, empress with a namesake mausoleum]) and PATEL ([Kidman co-star in “Lion”]), crossing at the ‘L’.
Names are a big sticking point for me when I’m constructing. There’s so much variation for both given names and surnames, and I feel like solvers can’t (and perhaps shouldn’t) assume that a clued name of a person they don’t know is “recognizable” or “traditional.” So although in the Musa puzzle you could argue that PATE_ is probably PATEL (i.e. Dev Patel), my opinion is that it’s not a sure thing (perhaps even less so for knowing that Empress Mumtaz MAHA_ should be Mumtaz MAHAL).
But what improves this crossing for me is using additional trivia – i.e. the tidbit “with a namesake mausoleum,” which references the Taj MAHAL. I like how it offers an extra reference point for solving, using the realm of well-known world architecture. The reference ultimately adds increased feelings of confidence/”eureka” for the solver: Even though it took me a little while to make the connection to the Taj MAHAL, I at least felt relatively sure of the correctness after my realization.
One of my favorite strategies for massaging difficult crossings is through incorporating wordplay in clues, which I’ve written more extensively about in a previous post. Maybe your clue uses a hidden word (e.g. [An Iowa city found in “tamest”] for AMES). Or maybe you even essentially reveal one problematic letter of an acronym (e.g. [Scannable merchandise acronym whose 1st letter stands for “Universal”] to essentially reveal the ‘U’ in UPC). Regardless of the exact method, I think it’s a nice idea to blunt possible solver despair.
All in all, I like how these tools can help confirm a correct answer. Musa’s clue in particular, I feel, maintains a relatively high level of difficulty; the improvement is that the solver is more likely to get the little “aha” feeling where they know they’re correct. You don’t necessarily have to make the puzzle “easier” – but if the solver enters the right answer, shouldn’t they feel that satisfaction?
3) Coming to Peace with the Nature of Crosswords
The last tool for dealing with difficult crossings is understanding – specifically about the relationship between the constructor and solver, and how that affects your crossing decisions.
Yeah, yeah, lofty words. But what does that actually entail?
First, I think it comes from keeping the specific audience in mind. Although I write lots of beginner-friendly puzzles, not all of my puzzles should be easy – or are meant to be easy. So if I know my puzzle is meant for more experienced solvers (maybe it’s a late-week submission, for example), perhaps I can accept different crossings: They’re probably more familiar with crossword mainstays like baseball’s ALOU brothers, so maybe it’s not such a big deal to cross that entry with our canine friend ASTA – even if I would avoid that crossing for a puzzle catered more towards beginners (and especially younger solvers!).
I think this flexibility comes from a place of familiarity, and also a little respect, and also some pragmatism. For instance, the NYT midi relies on the USA-specific locations LOGAN and MAYO CLINIC, which I think are intentionally catering to an American solving audience. On a separate note, it can be annoying when you’re overly babied as a solver. The excitement comes from the challenge of the unknown – being able to flex your vast knowledge in that big brain of yours. More experienced solvers simply have more tools to solve with, and are likely more forgiving of an easier proper-name crossing if it enables some otherwise banger fill.
Second, I think there has to be some acceptance: Not everyone will be able to solve your crossword. That’s just a fact. Inevitably, no matter how you try, people will fail to complete the puzzle. Maybe the solver doesn’t know the word you thought was general knowledge; maybe the Gen-Alpha kid at the speed-solving competition loses in the finals because they don’t know who OPRAH is. (It feels possible, right?)
I’m no stranger to this frustration. There are times when I’m solving a supposedly “easy” crossword and get stuck on one single square. And yes, it boils me to no end. But also, isn’t that the risk I take as a solver? That failure is possible? Because if not, maybe it’s worth switching to word searches. (Usually this threat of word searches is enough to jolt me out of whatever rage I was in.)
All in all, I’m trying to view difficult crossings more contextually, and realistically. I can’t guarantee that everyone will succeed at solving my puzzle, and people will inevitably be frustrated. The most I can do is try to achieve balance as best I can, with the understanding that it will never be perfect for everyone, but hopefully it’s good for a decent amount of people.
Final Thoughts
At the heart of this post, I think, is the tug-of-war of solving a puzzle, that tension between too easy and too hard. And while it’s certainly not an exact science, I do think you can use certain techniques and decisions to pull towards one side or another, or at least change how it’s easy/hard.
But ultimately, the best you can do is think about it for yourself. For me, I think my pickiness has resulted in work I’m proud of. But by being a bit more mindful when I evaluate certain crossings, and not simply rejecting proper-name crossings en-masse, I think I’ll be able to achieve some more ambitious grids that certain solvers will find increased joy in.
Anyways, that’s all from me! In the meantime, I’ve got a couple of upcoming crosswords that I’m excited about: the Monday, March 23 Wall Street Journal crossword, the Tuesday, March 24 Puzzmo crossword, and then I’ll be writing the Puzzmo mini for the month of April as the mini-constructor-in-residence. Happy puzzling!
Do you have additional ideas? Things I missed? Feel free to comment your thoughts!
And click here for a related post from last September, titled “A Little Help Here?”