Setters' Advice Corner [to be archived 25/10]

I have one. Are we allowed to use imperatives to the solver which make the surface tight but add nothing to the cryptic? For example "Find" or "Put" at the start of a clue. Or does this make it unfair, as it could be mistaken for the definition?

My favourite bit of setting advice: "Easy clues for hard words, hard clues for easy words" i.e. if you have a obscure word (usually cos of GridFill Trauma™) the fairest way to clue it is with a 'plain sight letters' device e.g. hidden or acrostic (anagrams less reliable in this instance!); flip-side is that an everyday word can be disguised via a tricky combination device like synonym-reversed-then-deletion and/or a whimsical definition (e.g "joiner" for AND).

I have been instructed to post this here! Does anyone have to hand something that clearly explains same-sidedness? I am working with a group of new compilers and it's a concept that is tricky to define. 'Perceives taste and smell? (6)' = SENSES is the clue that prompted the discussion. So not the situation that often arises of a double def with both elements sharing a common root. Here, it was more a feeling that the clue was unsatisfactory/un-cryptic but how to explain quite why?

I have a pretty vague question which will probably be nigh-on impossible to answer, but what is the key to making a clue with a good surface reading? More specifically, how do each of you individual experienced setters personally go about it? Is it a question of looking within the word to find potentially related components, or is it purely a matter of experience? I will shamelessly plug my own 2 puzzles if anyone would be willing to check them out and give advice to help me and other new

As above what are the conventions on capitalising and de-capitalising words in a clue to make the surface read smoothly

When you delete a group of letters from a longer word (and you clue the shorter group as a word) do they have to actually exist AS that word within the longer word? For example if I clued the word ABOUT and then "but is out of it" to get AO, or must you remove them one by one in that case?

What is the general consensus on an infinitive verb cluing a base verb as a definition/charade? For example, a definition being "TO MASK" yielding COVER as the answer, rather than TO COVER. The same question goes for a charade / clue using substitution - if I had a clue structure that was something like DEF 'to mask' ..., could I fairly substitute 'TO MASK' with the word COVER? Something I've been wondering for a while, but never got the chance to ask!

Prior to the help of testers, how do you all go about deciding whether a definition is sufficiently close to your answer? I find that intuition, dictionary reference and the practice of ‘can I find a sentence where interchanging these terms bears the same meaning?’ Works for 95% of cases. But probably the most common tester criticism I get is in those edge cases where my assessment of fairness is too loose. How does one get ahead of this problem before handing off for feedback?

I'd be very grateful for pointers on writing Spoonerisms, particularly how to position the indicator and which way it works. When I'm writing or solving I'm not always confident which bit I should be Spoonerising. Thank you!

Another question I had - it’s not something I’ve done personally, but I have thought about it - if I had a clue in which its definition was a definition-by-example, and I had “DEF say” to mislead rather than “DEF, say” , this would be frowned on against, correct? In that the setter isn’t saying what they mean, and the grammar of a definition-by-example using “say” just doesn’t work without a comma as well?

What does the community think about modern references? As a young setter my occasional references to the things I enjoyed (horror media, memes) have been largely missed/disregarded as obscure if I just put it for surface, while people will complain if I made it a part of wordplay. What makes me jealous is that other setters almost always get away with their references (and even get praised lol)

Another question because we're already here: How can I make the grammar in a clue where I want to use an imperative indicator alongside other indicator(s) work? Example skeleton: [DEF] [IMPERATIVE ANAGRIND] [ANAGRIST] [CONTAINER-GRIND(?)] [CONTAINER-GRIST(?)] -> if I wanted this structure, what form would the containing indicator have to take to make it work grammatically with the imperative anagrind? I hope this makes sense, if not I can try to come up with an actual clue example to clarify

Can two or more pieces of a charade be treated as a unit for, say, a reversal? E.g. is "Most of the gin gets knocked back" an okay WP for NIGHT? Can the singular "gets" work on TH(e) + GIN?

A grid-setting question: how comfortable are you with using "underchecked" words in your grid? I might have the wrong word, but by "underchecked" I mean solutions with fewer than 50% of their letters checked by crossing words. When I see a grid w/ multiple entries of this sort, especially from experienced setters, I assume the setter's including a theme or is under some other special constraints. Such grids are harder for me as a solver, but their theme, Nina, etc can help to ease the challenge.

Another grid-setting question: how many unchecked initial letters can you have in a 15x15 grid? I was taught any more than 8 is unfair, but some of the sample grids on this site have twice that many, so there are obviously other schools of thought.

Just on the off-chance you didn't see this at the top the post 😁 UPDATE 18/10/26: this Topic will be closed for comments on Sunday 25/10/2026 as we are now encouraging setters to post their queries as individual New Topics. Thanks for your engagement in this thread, see you in the comments on the individual posts!