Update 22 March

Here we are again, this time with an Independent daily puzzle from 2008.  My thanks to them for permission to republish.

I’ve spent part of this weekend being thankful I don’t get too many ideas that require a circular grid.  However, after a munging together of output from a grid-drawing program and the use of MS Paint, I have come up with something satisfactory, even if I didn’t relish entering 175 letters one at a time for the solution grid.  (NB if the puzzle ever appears, it’s unlikely to be before 2017, and neither editors nor solvers can expect to see it soon.)

I had a minor grumble from an editor earlier this week over my propensity to repeat the use of (say) s=second in two clues in a single puzzle.  If someone’s observing it, who am I to deny it?  Indeed, I think I know how it comes about.  The interesting part may be how it doesn’t get subsequently ‘corrected’.

At any given time I have up to a dozen puzzles being clued, and most advance incrementally, three or four clues at a time (which is why I was able to suggest a 2017 appearance above).  I almost never complete a set of clues at a single sitting – though there is one such puzzle coming up in The Independent shortly (I know, because I was sufficiently surprised to make a note of the fact).  The Independent daily and the Beelzebub puzzle are least likely to experience this, as there are tighter deadlines there.  But in other outlets I appear less often and I take a correspondingly looser approach (it’s astonishing how things still seem to pile up, though…).

It can be beneficial, as I will move from puzzle to puzzle, doing a couple of clues here, three there, and so on, which avoids my getting bogged down (the ‘Oh, no, not KOK-SAGHYZ again’ moment).  But I don’t generally look back at the body of clues already written – if I’m working from a notebook, I may not even have them listed since I’ll have noted the next ten (or whatever) candidates.  In such circumstances, if you’re trying to write the best or most apposite clue from the materials at hand, then it’s reasonably likely that you’ll end up using the same abbreviations or synonyms more than once.

But, of course, I do look back over the body of clues already written when I come to type up the puzzle for submission: my last task is to take the clues and write up a set of notes.  I rarely write notes as I go, because this last step means – well, if I can’t see how one of my own clues works after a period of time, I’d better write another one.  And at this point I do regularly spot (and adjust) repeated abbreviations – and also regularly miss them.

Why might that be?  I think it’s the desire to write the best (or most apposite) clue resurfacing.  If the two clues are both reasonably good then I’m more inclined to let the repetition pass (by which I mean fail to see it…).  If I can pick a decent-sized hole in one (you can always pick a hole in any clue) then I’m more likely to say ‘Oh, look, s=second again’.

Similarly I’m not so fussed about a hard-and-fast limit on the number of anagrams while the internal cry of  ‘[Expletive] – it’s another hidden!’ is a curiously enjoyable kickself moment in solving.  (Perhaps irrationally, I do worry about that point in the Down clues where you get a run of reasonably long words – I don’t like a run of anagrams there.)

The nature of the editor’s job requires that they work through a puzzle in a single session; some solvers also tackle puzzles that way (deadlines for competition puzzles must also encourage that line of thinking).  It’s likely that such an approach will draw attention to repeats. But the blogs show that many solve on a cut-and-come-again basis, often finding that that helps kickstart inspiration.  That’s much closer to the way I clue many puzzles, and I wonder whether solvers taking that line are more tolerant of  (or even notice) repeats.