New York Times, Saturday, September 5, 2026

When I construct a crossword, one of my goals is to submit a high-quality original clue for each entry. If this sounds insane to you, I admit I felt pretty crazy myself when I read the proofs for this puzzle. Of the 72 clues I submitted, only 18 were kept verbatim, while 38 were changed all the way to a different cluing angle. So much of my effort wasted! Is the universe (i.e. the editorial team) sending me a sign to do something differently?

I had three hot takes:

1. I have terrible judgment about what constitutes a good clue, so the more effort I invest, the more reliably I create clues not worth publishing.
2. The editors are actually the ones with terrible judgment.
3. Originality in a crossword clue is valueless.

Unlike my hot takes, reality is nuanced and complicated. Accurate analysis does not consist of parceling out blame in correct proportions among three bitter contenders; rather, it is correctly interpreting events within a larger context. Possibly there is nothing for which to assign blame, because nothing went wrong.

Take, for example, the entry AT SEA. I submitted the clue {Out of one's depth} after verifying that Crosserville's database doesn't have "depth" in any of its 600+ instances of AT SEA clues. It was edited to {Befuddled}, which was the most common, clocking in at 45 prior appearances. The latter is so popular precisely because it is high-quality in multiple ways: it's concise; it's a close match without being a laydown; it's a fun word in its own right; it's comprehensible to almost all solvers; etc.

Typically, an original clue will have flaws that tried-and-true clues do not have. In this case, cluing an idiom with an idiom risks that either or both will be misconstrued as literal; the two figurative meanings don't exactly match each other (confused vs. inadequate to the situation); it's longer; it's less familiar; etc. Are the defects worth the possibility of creating an "aha moment" even for solvers who have encountered AT SEA over and over again? It depends.

Furthermore, at my stage of involvement, I can't know everything it depends on. The editorial team has additional information and additional constraints: maybe a test solver was offended by something; maybe the fact-checker debunked the premise; maybe something duplicates an entry in another puzzle in their pipeline; maybe a quadrant of the puzzle is shaping up to be too difficult and needs a foothold; maybe the character count has room to spare or is running particularly high; etc. All kinds of tradeoffs are best managed closer to publication.

My submission is a contribution to a team process rather than an attempt to single-handedly create a finished product. The editors have many jobs to do; I embrace the role of giving them creative raw material with which to do their jobs. Yes, I could have given them the best recycled clue I could find for each entry, but they can easily recycle clues themselves, case by case, if it is called for by all the considerations they have to balance. By providing a fresh clue for each entry, I gave them options they couldn't have gotten anywhere else.

In process terms, I can understand that even my clues that the editors cut were a positive contribution. It is reasonable for originality to play a larger role earlier in the process and a smaller role later on. I hope and trust that our team effort produced a better product than it would have if I had tried to make the editors superfluous. If I had gone out of my way to preempt everything that they do, I might well have left them with less work to do, but they would also have had less to work with.

Surely there is more than one right way to approach cluing, but I expect I will continue trying to submit a high-quality original clue for each entry. I'm content to be a cog in the machine, as opposed to trying to minimize the number of editorial changes that will happen after a puzzle passes out of my hands.