Fill Quality Versus Theme Ambition

The most common rejection note in crossword construction is some version of the same sentence: good idea, rough fill. It means the constructor had a theme worth publishing and could not build a grid around it that a solver would enjoy.

This is not a failure of skill so much as a failure of pricing. Every theme has a cost, paid in the quality of the non-theme entries, and the constructor's job is to work out whether the theme is worth what it charges.

Where the cost comes from

Theme entries are long, they are fixed, and they usually have to be symmetrically placed. Each one removes an enormous amount of freedom from the grid. A 15-letter entry spanning row three does not just occupy fifteen cells; it dictates the third letter of fifteen different Down entries.

Add a fourth theme entry and the constraint compounds. Add a fifth, plus a revealer, plus circled squares, and you have a grid where the fill is not really being chosen any more, it is being surrendered to.

An ambitious theme in a rough grid is a promise the puzzle cannot keep. The solver only experiences the fill.

The three questions

Before defending a compromised grid, answer these honestly.

1. Would you enjoy solving this if you did not know the theme? Most solvers hit the fill before they see the pattern. If the first two minutes are unpleasant, the reveal has to be extraordinary to recover.
2. Is the worst entry defensible on its own? Not defensible as a consequence of the theme - defensible as an entry a solver could reasonably know. If your answer starts with the word because, it is not defensible.
3. Could a smaller theme carry the same idea? Four theme entries instead of six is usually a much better puzzle. Constructors treat entry count as a scoring system, and it is not.

What editors actually do

Editors read for the solver, not for the constructor. They know how much work went into placing six theme entries, and they will still reject the puzzle if the cost lands on the reader.

The practical threshold most of them use is roughly this: one weak entry is invisible, two is noticeable, three is a pattern, and a pattern of weak entries is a rejection regardless of how good the theme is. That is a much tighter budget than most new constructors assume, and it explains why so many submissions come back with the same note.

You can calibrate against published work. Reading a month of grids at XWord Info with an eye on the short entries rather than the theme is uncomfortable but instructive; so is the market and standards discussion the constructor community keeps at Cruciverb. The Times submission guidelines are unusually direct about the priority order: fresh theme, clean fill, and clean fill is not optional.

The three ways out

When the fill will not come clean, there are only three real moves, and constructors tend to try them in the wrong order.

Cut a theme entry. Cheapest and most effective. It hurts, and it almost always produces a better puzzle. Most published themes run four entries, not six, for exactly this reason.

Change the grid. Re-place the blocks, shift a theme entry by a row, break up a long stack. This is where a proper grid editor and a scored word list pays for itself, because you can test twenty block patterns in the time it takes to hand-fill one.

Change the theme entries themselves. A 13 and an 11 are far easier to place than two 15s. If the theme logic tolerates a different set of examples, take the shorter ones.

The uncomfortable version

There is a fourth option, which is to publish it anyway somewhere less strict, and it is not always wrong. Indie outlets and personal blogs have different standards, and a grid that is 80 percent excellent and 20 percent rough is still worth more than an idea that never left the notebook. Our guide to getting a first crossword published covers which markets tolerate what.

But be clear with yourself about which thing you are doing. Publishing a compromised grid because you like the theme is a legitimate choice. Believing the theme makes the fill acceptable is not, and solvers can always tell the difference.

If you want to see how commercial themed puzzles handle the same trade-off under much tighter vocabulary constraints, the daily archives at Daily Themed Crossword Answers and the syndicated grids collected at Thomas Joseph Crossword are a useful contrast to the broadsheet puzzles most construction advice is written about.