Enigmatic Variations 1459
Winning by Ifor
Generated single letters anagram to JOHN NASH; extra words lead to PRINCE/TON and STOCK/HOLM. In STOCKHOLM, in the film A Beautiful Mind, NASH, but not LARDE, sees MARCEE, HERMAN and PARCHER.
Writing a setter’s blog is a new experience for me, so I’ve taken the liberty of describing my setting technique, and these sorts of puzzle, in general terms, as well as referencing the puzzle. I’ll happily answer any more specific queries, of course.
Barred cryptics are sometimes representative – the solver is required to modify the grid in some way to generate a picture. Often an endgame adaptation will be required so that the original image changes in some relevant way, as here.
After watching the film I was struck by the final scene, in which the camera cuts between what Nash and his wife see; it’s one of the rare points at which all three of his delusions appear together. Could this be utilised? The key issue was whether words of structure XXXPAR and CHERXXX, where XXX are real words and no more than six letters in total, to allow a reasonable grid height, could be found. One advantage accruing from years of setting and solving is that one acquires a vocabulary of all sorts of obscure words – so HEPAR and ISH came to mind quickly. I make no apologies for a two-letter answer; always approach these puzzles without preconceptions. HERMAN would be no bother managing (see what I did there?) but MARCEE pretty much forces CEEFAX; certainty and flexibility are, paradoxically, both helpful when building a grid. And so to construction of the grid’s right-hand-side, where again knowing the existence of WAES, MES and the relatively rare spelling of AMEERS helped considerably. Experience told me that by ignoring grid symmetry LARDE and NASH with their helpful letter combinations would be easy to accommodate.
At this point two possibilities arise – in the endgame should the solver be required to delete, or to add, the threesome? The second is the greater challenge and so much more to my liking. But how to indicate the requirement? Again, there are broadly two options. An instruction could be embedded in the clues (eg extra letters or misprints spelling out “fill the empty cells so that three names appear”) or placed overtly in the preamble, with the clues gimmicked in a way that leads to the theme. Prince / ton and Stock / holm had already occurred to me, and a dictionary check revealed that even the unlikely HOLM would yield the minimum of three synonyms necessary. Is this enough to indicate the theme? Maybe not if you’ve not seen the film, so Nash himself had better appear to prompt a Google search. The gimmick I chose is one I like – apart from anything else it allows an extra word which could well assist in giving the clues meaningful structures. It’s also thematic in the sense that the key letters “vanish”, like the extra words, from the clues.
I very much prefer gimmicked clues to form a defined subset rather than simply being “some of them”. (Just in passing – another of my principles is that no puzzle is ever the worse for some juvenile smut, whence 16a) My practice is to look at all the feasible groupings of clues, as usually some will produce the required numbers; this led to the 3/4/7/8/9 in the preamble. A check revealed that the 3s and 4s would work for JOHN NASH (if, say, all their answers actually contained an A or an N then the gimmick I chose would fail).
And so to cluewriting. My method is to start where the greatest difficulty is likely to arise, in this case (I thought) hiding a J-word in a clue, and finding somewhere to place HOLLY within a clue surface. I seldom spend more than an hour at any one time on the job, and am pleased to have one, or at best two, clues written within that period. And I usually have one in mind when out and about – whence the thousand-yard stare followed by frantic scribbling that friends and family are now expert at ignoring.
Well before this the preamble is written; it clarifies one’s thoughts. (When I first started setting someone very experienced told me to write the preamble first – nonsense, I thought at the time, but what good advice!) Usually a title will have suggested itself by this stage as well; I think many solvers would be surprised if they knew how much time goes into these aspects of a puzzle. More experienced solvers will recognise the double meaning of “confused person” and perhaps appreciate the apparently colloquial “how the other saw things”. At this point, incidentally, all the clues are in one large colour-coded spreadsheet, which is then copied and pruned to the (hopefully) definitive version. The final stage is to use manipulations in Excel to identify occasions where I’ve used the same word twice (regular Ifor solvers will know my over-fondness for “lost” as an archaism indicator, for example). Finally it’s all put together into the editor’s preferred format, sent to one or more fellow setters for testing, modified in the light of their invariably useful feedback – and then submitted. As I’ve often said, if solvers gain anything like the pleasure that I do from constructing puzzles then I’m well satisfied.
A full review of this puzzle can be seen over on fifteensquared.