Sign in Subscribe

The Gnomon's Guide to Constructors

The Gnomon is interested in publishing high-quality, American-style variety cryptics every other month.

Rates, rights and submission

We’re hopeful to run puzzles of very different sizes and scales, and so our pay rates will scale with size of puzzle:  For the first year, The Gnomon will pay a base amount of $5 per clue, with a possibility of some additional payment for particularly intricate constructions or cluing constraints.  We’re hopeful to increase this rate if subscriptions can support it after the first year.

The Gnomon will retain exclusive rights to the puzzle for one year after its first publication, and perpetual rights to republish after that:  You’re welcome to include your puzzle on your blog or in your book, but only one year after its first appearance here.  The Gnomon retains the rights to run your puzzle again in the future, most likely in a compiled puzzle book, but possibly also rerunning in special issues or similar.

Puzzles can be submitted directly to me (Andy Yingst) at gnomoncryptics@gmail.com as a .puz, .ipuz, .pdf, or a google sheet.  If submitting a pdf, I’ll also need a text or doc file with the clues and any instructions. All constructors are encouraged to submit a grid before cluing, or to submit a concept for discussion before gridding, especially so that a pay rate can be agreed on early in the process.

What we’re looking for in a puzzle

Shenanigans-

We’re looking for variety cryptics here, which means cryptics with some extra rules governing the solving of the clues or the entry of the answers.  This can be clues with extra letters or words that must be removed to parse, solutions that must be modified according to some rule that must be discovered to fit in the grid, instructions that lead to an image in the full grid, some entries being unclued and related to a theme, or just strangely shaped grids.  Anything is fair game, from something as simple as an alphabet jigsaw, to as complicated as something with multiple layers of extraction giving additional instructions on how to modify the grid once filled.
Theme puzzles (puzzles with no extra rules, only a strong theme appearing in the clues and/or fill) may be considered, but probably only if every single clue is on theme.

Fill-

A typical puzzle should have well-known fill, with an expected maximum of about three unknown entries for a modern, educated solver.  More complicated variety shenanigans can excuse more obscure fill, with the admittedly ambiguous warning that the payoff should be worth the strain.  The setter should be aware of which fills are more likely to be unknown, and either make sure they’re very well-checked in the grid, or have an easier-than-typical clue.

Checking-

Even for grids with no strange rules going on, every word in every Gnomon grid should have at least half of its letters checked (crossed) by another word or possibly by some other variety method.  For most variety puzzles however, the presence of some additional difficulty in either solving the clues or in placing answers in the grid will increase the demanded ratio to 60% or more ideally, ⅔.  Many outlets forbid double-unches (two consecutive uncrossed letters in the same word.)   While not ideal, The Gnomon doesn’t have a strict ban on them, especially in a word that’s ⅔ checked.  Triple-unches are forbidden.

Cluing Restrictions-

In a puzzle of about 32 clues there should typically be no more than:

  • 5 full anagrams
  • 2 hiddens
  • 3 total variations-of-hiddens (acrostics, alternation clues, etc, including hiddens.  So using 2 hiddens in a 30-clue puzzle means max one alternation, or max one acrostic).  

For the above, these numbers increase faster than linearly however… a 50 clue puzzle could reasonably have 9 anagrams and 6 hidden variants, as simpler clues help solvers get through longer puzzles with less fatigue.

  • A maximum of two splitsies per puzzle–that is a maximum of two clues where a word must be divided into pieces to parse, as in maybe “massive” to indicate MIVE or “halfback” for BA.  Words using a hyphen don’t count towards this limit.
  • A maximum of two noun-part indicators per puzzle.  This includes for example, “industry leader” for I, or “downy wings” for DY.  Note that something like “redhead” for R counts against both your noun-part limit and your splitsy count.  (I think noun-part is completely fair, but it irks enough solvers that we should use restraint.)  Note that something like “industry’s leader” or “leader of industry” don’t count as noun-part in this sense; using grammar to indicate possession doesn’t bother the objectors.  Also note that something like “first grade” for G, while totally fine for an Australian cryptic, is way too far outside the norm for an American one and is forbidden in The Gnomon.
  • A maximum of one partial acrostic.  (Taking the first letter of one word is fine; taking first letters of many words to make the entire answer is fine; using initials to get more than one but not all letters of the answer is the restriction here.)  
  • Partial alternation (taking alternating letters to give some of the answer but not all of it) is forbidden.

Shared roots–

One of the defining characteristics of an American cryptic as opposed to a UK one is ensuring that there is no etymological link between the wordplay and the definition (or between the multiple definitions.)  While we don’t have an absolute ban on shared roots here, any clue with a shared root should either be distant or non-obvious enough as to not attract attention, or the clue should have something very fun going on to distract, or possibly some combination.  This probably requires examples to discuss clearly:

> Anxious spring prior to the end of May (5)
[JUMP+_Y]
This clue absolutely wouldn’t fly in the Gnomon.  The words jump as in ‘to spring’ and the adjective jumpy are very obviously related, and this isn’t so much “wordplay” as it is “adding a suffix”.

> Discover (possibly) a comedian (4)
[CARD ddef]
If you look them up, these two senses of the word “card” are related and this is a shared root, but the concept that connects them seems to be so far outside of English as spoken today that it could get a pass.  (Though note, this clue I just threw together as an example doesn’t really have an exciting surface or anything that fun to it, so the shared root might be enough to tip the scales to throw this back for a rewrite.)

> This is possibly a sign of the onset of labor (and possibly isn’t) (11)
[CONTRACTION, ddef]
Here the shared root is blatant:  A contraction of the uterine muscles and a contraction of two words are both definitely two senses of ‘shrinking’ in the exact same way.  The thing that saves this clue though is the fun hidden definition on ‘possibly isn’t’, both hiding the fact that ‘isn’t’ is a noun in the cryptic reading, while also serving as a true fact about labor contractions.  Another clue that transcends its shared root from a recent Guardian:

> Beauty is in the eye of the beholder (5)
[LOOKS, ddef]
Again, the shared root between LOOKS the noun (“beauty”) and LOOKS the verb (“appears” = “is, in the eye of the beholder”) is blatant, but wow… the fact that the surface is a well known adage that technically parses is amazing.

Note that even in situations like these last two, with a blatant shared root in a clue that’s otherwise fun or cool enough to make up for it is probably still only going to fly once or twice per puzzle.