Listen With Others

Are you sitting comfortably? Then we’ll begin

Archive for the ‘Setting Blogs’ Category

Listener No 4683, Diversions: A Setter’s Blog by Dysart

Posted by Listen With Others on 21 Nov 2021

I set this puzzle four years ago, before my puzzles on Vitruvian Man and Lamb to the Slaughter. The former took precedence because it was intended to coincide with an anniversary date; the latter because I was worried that another setter would beat me to it with the theme (I’d already had one Listener submission rejected at the last minute because of thematic duplication). My memory of its genesis and construction was a bit vague, but delving into the files on my computer has revealed a lot. I had already set an EV puzzle involving the conversion of the names of some Greek Gods to their Roman equivalents, but I wanted a puzzle with a more comprehensive set of Gods. One obvious set was the Olympians. Authoritative sources don’t completely agree on the composition of the set or the spellings, though all agree that there are 12 in the set. Encyclopaedia Britannica‘s list of 12 includes Hades and Hestia in preference to Demeter and Dionysus. The former pair eventually made it into the puzzle as an element of wordplay (HADES), and as a jumbled answer (HESTIA).

Brewer’s Dictionary of Phrase and Fable lists the dii majores under the heading, Gods. The more accurately transliterated Apollon is used in preference to Apollo, and Athena appears as Athene. Hestia is included in the major twelve, and Dionysus is listed below the main list.

The Greek Myths by Robert Graves features the twelve gods I used in my final grid, though it gives the earlier Greek form, Athene, rather than the later Athena. Graves explains how Hestia (one of the original 12) gave way to Dionysus, thus the gender balance was tipped in favour of males. [1]

Six of the original twelve were children of Cronus and Rhea (Demeter, Hades, Hera, Hestia, Poseidon and Zeus). The other six were children of Zeus.

I entertained various ideas for the depiction of the theme. At one point I considered anagrams of the names. All but two could be anagrammed, the two exceptions being Hephaestus and Dionysus. The weakness of this was that the theme might emerge too early; after getting two or three thematic entries the rest would be easy to get. Later I thought of an endgame in which the names would be revealed by changing some letters in the grid. Initially the deities were scattered haphazardly in the grid, and MOUNT OLYMPUS did not feature. Eventually I included a jumble of MOUNT OLYMPUS on the top row, though it was some time before I got a grid with most of the gods. Later I thought it would be far better to have MOUNT OLYMPUS on the bottom row and the gods above that, becoming progressively shorter, with ZEUS and HERA in the top two rows, thus forming the appearance of a mountain. This is where things got very tricky despite the latitude I had in the sequence of entries that could be modified to reveal the names, and in the entries on the bottom row. If the bottom row had one or two entries of at least five letters and one or two unchecked cells there were over sixty possibilities (though many with obscure words). The longest possible entry was UPON MY SOUL, which I liked but it created problems in the top row. I started with symmetrical grids and normal entries, but this proved impossible. I then switched to an asymmetrical grid, but that resulted in a low average entry length (the best I could achieve was 5.3, below the recommended minimum of 5.5). I was intending to have a hidden message from a clue gimmick in the across clues, e.g. REARRANGE LETTERS IN THE LAST ROW, but at some point I saw the letters S/R/U/OTTO in one of the trial grids. This gave me the idea that the downs could be jumbled, thus eliminating the need for a clue gimmick, and facilitating a symmetrical grid. The device I had chosen to reveal the gods would result in non-word entries anyway, so jumbles would not detract further from the final grid. I know some solvers dislike jumbled entries, but I eased things with generous checking – twelve of the down entries are fully checked. Jumbling wasn’t strictly thematic, but the gods of Greek mythology were a pretty anarchic and egotistical lot, creating a fair amount of chaos, so jumbling wasn’t altogether unjustified.

To complete the theme I wanted to incorporate features associated with the deities in the clues. This was fairly easy in most cases, and in some cases there are two or more thematic associations in the clues.­ The Listener website [] includes the associations in the detailed notes on the puzzle. Some of the enjoyment in setting the clues was looking for opportunities to work in thematic associations without detriment to the surface. Normally I eschew anagram clues when entries are jumbled, but I made an exception in two cases for thematic reasons (REMOUNTS and VAT DYE). After devising the theme, writing the clues is the best part of the whole setting process for me. Even after a puzzle has been test-solved I may spend more time trying to polish the clues. I think I was told once that Dimitry (John Grimshaw, former Listener editor) advised putting a puzzle aside after its initial completion, then looking at it again six months later with fresh eyes. It’s very good advice.

Various titles occurred to me, but in the end the superficially misleading Diversions seemed best as the conjunction of di and versions reflected the different pantheons and the jumbling.

In view of the potential problems from the jumbled answers I asked several people to test-solve the puzzle. It’s always useful to get different perspectives anyway. My thanks to all the testers involved.

Once it was considered by the editors, the main problem was the length, resulting from a longish preamble and 48 clues, with most of the down clues taking up two or more lines. The editors managed to trim the preamble and several of the clues. The Listener editors normally try to preserve the basis of the original clue, avoiding complete rewrites as far as possible, something I very much appreciate. In this case there was the added complication of preserving the thematic references. I was able to add my bit by completely rewriting a few more clues to make them one-liners.

Some solvers commented in online forums that it might have been better to erase the non-thematic letters. It was a fair point, and I did consider it early on, though one undesirable effect of this would be that solvers could submit a correct solution without getting all the answers correct. My original intention was that the area should be shaded green with a curvy outline to resemble a mountain, as below, thus drawing attention away from the non-thematic letters. The first test solve encouraged this, but the final preamble I submitted was less specific to make it more concise.


[1]. Dionysus, whose mother was mortal, was more like a demigod. After achieving renown from a series of military conquests he was given a place among the twelve Olympians, Hestia, goddess of the hearth, resigning in his favour “to escape the jealous wranglings of her family” (The Greek Myths, Book 1 by Robert Graves). Who can blame her?


Posted in Setting Blogs | Leave a Comment »

Listener No 4682, Round Table Man: A Setters’ Blog by Banjaluka

Posted by Listen With Others on 14 Nov 2021

Round Table Man started out as a germ of an idea about 2 years ago. I had read the Larkin poem An Arundel Tomb and it struck me how it would be ideal as a theme for a crossword. At first the only idea I had was that the final completed grid would have a heart highlighted to symbolise its famous final line, “What will survive of us is love.” Prior to Round Table Man, I had never set a Listener crossword before and so had little idea of how to go about turning a promising idea into a Listener quality puzzle. I tried to form a grid that brought in some of the elements of the poem, but I never could get it to the point where I was satisfied with the outcome.

I am a keen but inconsistent solver – I will try one or two Listeners a month and probably finish just over half of them. I first came across the Listener crossword around 10 years ago, when I was asked to test solve a Listener by a new setter. That puzzle ended up as Listener 4182 Breach of Contract by Ron (now setting under ‘Eck).’Eck is an old friend, and we had a long history of enjoying crosswords together. We had often sat in a café/pub solving the Araucaria prize puzzle in the Guardian on a Saturday, and over the years he has slowly dragged me towards solving the tougher Listener and Magpie style puzzles.

When the pandemic locked us all down at home, ‘Eck suggested it might be fun to jointly set a puzzle. We tossed a few new ideas around, but nothing seemed to spark any excitement. I thought that maybe that the idea I had had rattling around for a few years might come into being if I had help from a more experienced setter.

An Arundel Tomb is a rich source of thematic material that could be incorporated. With an end point of a heart decided on, it was a question of what else should appear in the grid and how could we slowly let the solver in on the theme. It seemed right to have ‘The Earl’ and the ‘Countess’ to be lying side by side, as described in the poem. We discovered that while Larkin describes “It meets his left-hand gauntlet, still / Clasped empty in the other”, the actual tomb in Chichester Cathedral has his right-hand gauntlet clasped empty. We chose to not worry too much about which way round they lay, taking our cue from Larkin’s “error” and ended up placing the countess on the left of the grid and the earl on the right.

It was then a case of finding how much of the “little dogs”, “Latin names”, “endless altered people”, etc from the poem we could fit in and still describe the required heart at the end. Very quickly, ‘Eck came up with a grid that miraculously included all the elements you see in the final grid. It took a few iterations of trying various names that could be jumbled without their last letter, as well as finding recognisably Latin names to go around the base.

We worried about the balance of taking advantage of the poem and overdoing it. Reading Larkin’s poem again, I see even more lines or elements that on a different day could also have formed part of the crossword: “…their faces blurred…”, “…time has transfigured them into Untruth.”

Writing the clues was the fun bit. ‘Eck and I have different styles, but I found in the end we complemented one another quite well. I always wanted to give a variety of difficulty of clues. As someone who struggles with the harder puzzles, I always like there to be something solvable at a first pass so everyone can enter something in the grid to start them off. ‘Eck loves adding more and more layers of complexity. By sharing and collaborating, I hope we found a nice balance to the puzzle. We started by taking half the clues for a first run through and then we just went back and forth over WhatsApp, debating and improving one another’s clues for several weeks. We enjoyed sowing various red herrings in the puzzle. We had references to the Wizard of Oz in the clues and in the grid, as well as ‘Eck’s sneaky use of novel as the signal to anagram the puzzle’s title. We hoped these might have distracted people temporarily from finding the correct subject matter.

We were both genuinely pleased with the final effort and were happy that both our helpful test solvers as well as the Listener vetters seemed happy enough with the construction and did not want to change the basic premise. We had two major challenges come back. One was around length – the preamble was long given the number of extra elements we were including (clashes, extra letters, extra words etc) and we were pushing the overall maximum space limit. The vetters rewrote some of the longer clues and pared back a number of others. This was done very skilfully and both ‘Eck and I liked the edit. The second challenge from the vetters was removing some of the more risqué clue surfaces that were in our original version. These were also fair. The original clue for 23d should never appear in print!

A number of people have asked where the name Banjaluka comes from. When we were (a lot) younger, ‘Eck and I were looking at a large map of Europe and thought that Bosnia & Herzegovina’s second city would make a great name for a rock band. We decided then and there that one day we would form such a band. Sadly neither of us have learned to play any instruments, but when we needed a combined setter’s name this felt like the only appropriate choice.

Posted in Setting Blogs | 1 Comment »

Listener No 4681, 1 41 6:10: A Setter’s Blog by Colleague

Posted by Listen With Others on 7 Nov 2021

“Money is the root of all evil” is a song that I have known since a child. I happened to hear it on the wireless (!) a couple of years ago and wondered what the origin of the song was. I quickly discovered that it’s origin was in the Bible in 1 Timothy chapter 6 verse 10 as “For the love of money is the root of all evil”. I decided that it could be a suitable theme for a Listener crossword.

How to portray the theme in a grid was less obvious. I decided early on that a “floral” representation was the way to go – a “stem” and a “(square) root”. The stems were to be the EVIL part of the theme whilst the roots were to be the MONEY.

So I decided that the monies needed to be nine letters long each and that the middle of the 3×3 squares should be barred off. This meant that there needed to be 5 stems and roots because I wanted the barred off letters to spell MONEY.

It is surprisingly difficult to find 5 nine letter types of money with their middle letters being M+O+N+E+Y. To achieve this the first one had to be BAHAMIAN$. The last one had to be DANDYPRAT. I was aware that there was another spelling (dandiprat) but this ambiguity would be easily resolved by the fact that M+O+N+E+I does not make sense.

The 5 synonyms/definitions for types of “evil” were not much easier. Although the grid, because of it’s complexity, could not be totally symmetrical I wanted the “evil stems” to be symmetrical. The outside ones became 4 letters each (Harm and Bale), the next ones in were 7 letters (Disease and Immoral) whilst the middle one was 11 letters (Unfortunate). The main reason for choosing these lengths was that the 3×3 cells for the roots should not be touching. This made the grid 15 letters vertically which is practically the maximum size that The Times can accept.

So the grid fill continued along with the clue writing. I had decided that I did not want the use of extra words/extra letters/misprints etc which seemed all too prevalent at the time. My only exception to this was to be the inclusion of an extra word in two clues only (Mill and Ion). Together the resultant word had to be treated like the currencies in the grid – as a square root. Therefore, solvers would have been asked to write the treatment of this word below the grid – THOUSAND (being the square root of Million). I thought that this would have made a neat final task for solvers to tackle. I was, therefore, slightly disappointed that it had to be abandoned due to the grid size constraints.

I had based the grid on the modern version of the saying rather than the biblical one. (The title was the intro to the general subject – I was pleased that I managed to get Timothy into the grid as 41 across so that I could use the 41 as part of the title).


I understand that some solvers wanted the Biblical version to be the real theme with “the love of money” being somehow portrayed. The “roots” made a V-shape which some thought was supposed to be heart shaped. I did not consider this and, seeing the puzzle published the following week – 4682 Round Table Man, I’m glad that I didn’t! It was also deemed that I must be a Pink Floyd fan because (apparently!) on The Dark Side of the Moon album there is a track called “Money” with the words “Money, so they say, is the root of all evil today”. Sorry, I was not aware of it.

What a wonderful lot Listener solvers are!

Posted in Setting Blogs | Leave a Comment »

Listener No 4680, Manhunt: A Setter’s Blog by Shark

Posted by Listen With Others on 31 Oct 2021

Ideas for puzzles come from all sorts of places: an interesting word, phrase or definition; a recently read article, book or publication; an area of interest, hobby or past-time; or perhaps just divine inspiration. Once the theme is chosen, the tricky bit is formulating that concept into a crossword, particularly one that will be fun to solve, and having a finale that ultimately gives satisfaction to the solver (ensuring the effort was worth it).  

When I came across the two-needle telegraph, the diagonal and vertical lines that could be used in the grid as a code, the backstory of how the murderer was caught, the interesting incomplete alphabet and the fact we have generally all heard of Wheatstone from the Playfair cipher, made this concept too good to put aside.

I started by creating the size of the grid. I wanted the height of the grid to be the usual 12 or 13 with each row decoding to a letter. GARB OF A KWAKER was the obvious choice from the telegram and the encoded letters could be adjusted to create a reasonable crossword of rows. I then decided to use the letters split by diagonals to lead the solver to the theme. This was restricted to 27 letters and so A MURDER HAS GUST BEEN COMMITTED AT seemed apt. It then seemed logical to have SALT HILL hidden within the grid to complete the above phrase.

At this point the grid was already very constrained, but after quite a few hours of adjustments, I got the presented grid. I tried my hardest to get TAWELL hidden and to be highlighted as a finale, especially as the theme is based partly on him. It is a manhunt and so surely he should appear somewhere. Unfortunately, I couldn’t. However, on balance I felt it was worth keeping him in disguise just to round off the puzzle, even if it may have appeared a potentially unnecessary added extra at the end. I never know how much of the grid people spot in terms of Easter Eggs, but for those that may have missed it, the grid uses all the letters of the cipher … effectively a pangram in code (there is no J, Q or Z).

I really wanted this puzzle to go to the Listener, but the nagging doubt the editors might mention the Internet being required to solve it, was in the back of my mind. I therefore searched my local library and found two books on the subject, one about Tawell and the other about the telegraph, so I had my retort waiting. And so it turned out that I did indeed have to play that card, but there was still a reservation by the editors that the only lead-in to the theme was the beginning of the telegraph and so there had to be another way to help solvers. I then decided to use the five normal entry clues to assist. The downside being that some will now get the theme too early, but this did not seem to detract from the overall solve and I am over the moon by the positive feedback sent in by solvers.

Shark puzzles do tend to be (unintentionally) at the harder end of the spectrum, so this appears to be one puzzle where the hard work has indeed paid off. Glad you enjoyed it.


Posted in Setting Blogs | Leave a Comment »

Listener No 4679, Basic Fact: A Setter’s Blog by Opsimath

Posted by Listen With Others on 24 Oct 2021

I’ll start by thanking the editors for taking the time to look over my submission and improving so many clues for me. I won’t reveal here the number of minutes (sic) they each said it took to solve this. Nor should I mention that one of them suggested it was “short in the thematic department”.

Well, I can do with the 200 quid, so I’ll take the money and run — and on no account will I look for any feed-back online. Therein lies the danger of a nervous breakdown. We setters can be sensitive souls. Bear in mind the rate of pay is about 50 cents an hour.

Manufacturing one of these puzzles and getting it into print involves various layers of pleasure: there’s a great moment when an idea comes which has that vital element of “symmetry”. Then, I find using Qxw to produce a valid grid is the most enormous fun. Constructing the clues is bearable now and then when one stumbles on a good clue. Nerve has to be summoned to send in the final submission. There’s a brief acknowledgement from “Vetter 1” and then silence until the moment you receive notification: “it’s ready for publication. Please solve the clues back into the grid to check.” (At that point you find you can’t solve your own puzzle, and that’s embarrassing. Just trust the editors to have got it right.)

The real pleasure comes when the setter receives that package of feed-back from St Albans. Unlike the Internet — anonymous, 50% poisonous — solvers who want a prize must include their name and address, and are unlikely to be damningly negative. After my first Listener though, someone did comment “Not Really Listener Standard” [Yes, I know your address and I’ve lurked outside on Google Maps Street View! The fact that you had two errors in your grid gave me no pleasure.]

Of the setting process for “Basic Fact” I remember very little, but my files show that “BASIC 1” dates back to January 2017, very early in my career as a setter. Versions 2, 21, 301 and 401 came soon after. None of them much resemble the published offering, so somewhere in between I lost track of versions 500 through 901. But I can see that, right from the start, I felt obliged to use PLUSHLY, COEQUALS and UNEQUALS, none of which I was very happy with.

I do know that I had Chalicea look at one earlier version, and out of respect to her I made sure to fit PREMIER CRU into my later efforts. I’d like to thank Jason Watkiss and Jon Penman for encouragement and suggestions for the version I finally submitted.

Finally, I like to think that anyone who complains the “equation” is “too obscure” would only be embarrassing themselves. Thankfully, “too easy” is not a phrase the Editors seem to use.

Posted in Setting Blogs | Leave a Comment »