Free Crosswords Online!

Tips for Setters
HOME
FAQ
My Puzzles
Guest Puzzles
Beginners' Puzzle
Yet more free puzzles!
Ximenean Clueing
Ximenean clueing revisited
Single letter indicators
Link Words
Cryptic Definitions
Tips for Setters
Surface Readings
&lit clues
Ninas
Which words can I use?
Sudoku vs crosswords
The Listener Crossword
Cheating?
Chambers Crossword Manual
Becoming a compiler
Feedback
Links
Compiling Software
PUBLISH HERE
Use of these puzzles
Bits and Pieces

I've written a few pieces about specific aspects of clueing. This one contains more general tips, and I intend to add to it from time to time. Like all advice, feel free to take it or ignore it as you wish!

Definitions first
Not every clue will be a masterpiece
Don’t be put off by comments on blogs
How many anagrams?
Capitalisation
Love Thine Editor

Definitions first

I learned long ago that it is a good idea to decide roughly how you’re going to define the answer before you start work on the wordplay. If you have several ideas, even better. There’s a good reason for this. The possibilities for defining a word are usually fairly limited, while the possibilities for wordplay are almost always far more extensive. So it’s much easier to make the wordplay fit the definition of a word than the other way round. Of course you can always tweak your definition, or change it altogether, if a better idea occurs while you’re developing the wordplay, but it’s a good idea to have a starting point. Once you’ve got that you can look at the structure of the word itself and see if it contains any elements which, in conjunction with the definition you’ve chosen, could make a coherent sentence.

Here’s an example using a clue I wrote for the FT and was particularly pleased with. I will point out that it’s very, very rare for me to look back on clues I wrote a while ago and think “Ooh, that was rather good, wasn’t it?” I usually do the opposite – I’ll see lots of things I don’t like and feel I could or should have done better. In this case, though, seeing this clue when the crossword was published brought back a pleasant memory of the sense of discovery I had when writing it.

I had to clue JALAPEÑO and right from the off it looked like a horror of a word to clue. I decided on “hot pepper” or the more cryptic “hot stuff” for the definition and moved on to the wordplay. I concluded quickly that it would have to be an anagram so I started to investigate the possibilities. The letters of POLE came out fairly quickly, and I hit on “dancing” as a possible anagram indicator. There’s certainly a connection between pole dancing and hot stuff, but what of the remaining letters, AAJN?

It didn’t take me long to see that these letters could spell out a girl’s name, Jana, and all that remained was to put the clue in order. That gave

Jana pole dancing is hot stuff (8)

which is pretty good, but I tweaked the definition to

Jana pole dancing is something really hot (8)

as I felt that this is more like something people might say.

I will admit to a vested interest in the clue, as the name Jana does have a certain significance for me and the image the clue suggested meant a cold shower or two. Personal involvement shouldn’t affect a clue, of course, and I would still be happy with it if I had none. I did get a comment on the relevant blog from someone who thought it disgraceful that I used an anagram for such an “obscure” word. I can only suppose his idea of a good meal is McDonald’s or KFC, and that’s why JALAPEÑO is obscure to him!

Not every clue will be a masterpiece

It would be great to produce a set of clues where every one is a perfect sentence with a subject, object and verb. Where every clue evokes a vivid image and is breathtaking in its originality and skilful use of technique. Where there is no need for single letter indications (first of, primarily) and not a cliché (EX = old flame, PO = river) in sight. Oh yes indeedy, it would be terrific.

Dream on. Such a puzzle doesn’t exist. And it probably never will. The finest setters of today, for example Anax, Monk and Paul, usually come pretty close, and I would hope that some of mine are sort of in the right ballpark. But the truth is, it is almost certainly not possible to achieve a set of masterpieces in one puzzle, unless you spend years writing one set of clues, and even then I’m not sure it can be done.

All words are clueable, but some more so than others. Those who comment on crosswords often don’t realise that it’s far more of an achievement to come up with a competent clue for an unfriendly word than to produce a good clue for a friendly one. In any one grid, there will be a mixture of both, however carefully you plan your gridfill. Take ENNUI and ETUI, which often force themselves into grids. These usually appear as hidden words, and there’s nothing wrong with that, though hiddens seldom are very interesting. There are obviously other possibilities, but the structure of these words means that you’re looking at “a bit of this and a bit of that” kind of clues, which, while perfectly sound, will never be paragons of cluemanship.

With unfriendly words, there’s no point in spending hours looking for the perfect clue, as it’s often hard enough finding an adequate one. You are better off moving on to the words which have more possibilities for an excellent clue. The same goes for three-letter words which are necessitated by some grids. Solvers don’t care about them very much and however good you are as a setter, you’re unlikely to come up with anything for ATE, EAT or TEA that is so much better than what’s been done before.

Unhelpful and unexciting words will always be unhelpful and unexciting, and as the man said, you can’t polish a turd.

I’d also say that there are times you have to cut your losses even with friendly words. For example, you may come up with what would be a topical, humorous and grammatically perfect sentence if it weren’t for an unaccounted for Z and a couple of Us. It’s tempting to go on and on tweaking it until your head spins, and sometimes this produces the result you want. It usually doesn’t though, and I would say that if you’ve spent more than a couple of hours trying to knock the clue into shape and it still won’t work, it’s best to abandon the idea and try something else.

To be quite frank, setters are not paid enough to spend the sort of time on a puzzle that it would take to produce a string of diamonds. There will be a few rhinestones in there, and that’s inevitable. It doesn’t matter as long as the diamonds sparkle and the rhinestones look decent enough. I know this might sound a bit trade unionish but I am not for a moment advocating laziness. I’m just being realistic and advising setters not to waste time trying to achieve the impossible.

I’ll finish this section by mentioning that I don’t always take my own advice. I’ve been known to spend a whole morning trying to make the unworkable work, and to add that extra pizazz to an unfriendly word like, well, pizazz. When I started writing tabloid level puzzles for an agency in the early noughties, I had to clue the word UNDO. I’d used up my quota of hiddens so the next obvious possibility was something like

Ruin a Parisian party (4)

I decided that was way too mundane and boring, so I spent far too long trying to find a really great, knock ’em out clue. I can’t remember what I finally arrived at, but I’m pretty sure it wasn’t any better than my original idea. A couple of weeks later, I came across a clue in an Araucaria puzzle which was along the lines of

Ruin a Parisian party (4)

If it was good enough for Araucaria it should have been good enough for me!

Don’t be put off by comments on blogs

This one is for those setters who have scaled the heights of Olympus and got on to the team of setters for one or more of the national papers. All of these papers’ puzzles – Times, Telegraph, Independent, FT and Guardian – are discussed on the relevant blog the day they appear (except prize puzzles, which are blogged when the solution is published in the paper). The format is much the same on all these sites: the blogger of the day writes an explanation of each clue in the puzzle, often with a few comments thrown in, and then the blog’s visitors can add comments. Nearly all of the bloggers and most of the commenters are fair-minded and polite, and if they have any criticisms they express them in a constructive and respectful way. It must be said, though, that some of what’s written on these blogs has caused annoyance and hurt to a few setters (and they’re just the ones I know about).

Of course, free speech must always prevail – and real free speech means that people have a right to say what they like, however offensive or “politically incorrect” it may be, so long it’s not defamatory. Also, setters have to recognise that if you put work of any sort in the public domain, you’ll attract a certain amount of negativity. It would be naïve, arrogant even, to expect nothing but a stream of glowing eulogies. Crossword setters are in a rather unusual position though, in that these blogs usually provide the only feedback you’ll get from the public. Authors and musicians, for example, have plenty of negative things written about them too, but if they’re selling lots of books or records and watching their bank balances grow they’re unlikely to be all that bothered. The people who made the Lord of the Rings films won’t have lost a second’s sleep over the obsessives who decided they didn’t like the films before they’d even seen them, in the light of the films’ phenomenal success. For us setters, though, it can be quite disheartening to spend hours working on a puzzle for the benefit of solvers’ enjoyment – we don’t do it for the money as there isn’t much – when the feedback is minimal and some of what you do get is negative.

Of course, just because feedback is critical it doesn’t necessarily mean it’s of no value. For example, if nobody can parse a certain clue in your puzzle, it’s almost certainly too hard, or there may even be an error in it. That goes for whole puzzles too – if the overwhelming tone is that the blog’s contributors found it too difficult, it’s worth bearing that in mind. Many of the people who contribute to these blogs are experienced solvers and if they have struggled with the puzzle, you can be sure that the average solver will not have got far at all. Even so, it would still be worth consulting your editor before you alter your style.

It takes some humility to admit defeat and if people say that they found a puzzle too hard, there is no reason to doubt their word. On the other hand, people who say that a puzzle’s too easy are rather less reliable critics. You only have the contributor’s word for it that your puzzle was such an insult to his intelligence that he solved it so quickly he finished before he started. Even if these comments are not from people showing off, they are inconsiderate. As I’ve said, many of the contributors to these sites are experienced solvers and just because they find a puzzle easy, it doesn’t mean that the majority of solvers will. I’m a fairly decent solver myself, after years of practice, and I sometimes hanker for the days when a Times puzzle would last me an afternoon. I now need at least ten of the things to occupy me for a two-hour flight, but I’m never going to suggest that the Times puzzle is far too easy and that they should make them harder on my account. It would be total selfishness.

As a setter, you can happily ignore comments of this sort, but I didn’t know that when my first FT puzzle appeared towards the end of 2008. This was my debut as a crossword setter in a national paper (other than two Listener puzzles). It should have been a day to be proud of and to some extent it was, not least because I got some lovely messages of congratulation from setters and solvers who knew me, but it was pretty much ruined when I looked at the “review” for the puzzle on the relevant blog. It was appalling in all senses: unfair and self-aggrandising as well as negative. The blogger made a point of saying it was far too easy, and it probably was quite easy as it was my first puzzle and so I’d played safe. Still, I’m fairly sure it wasn’t any easier than many which appear in the FT. I was the new boy and it seems the tradition of picking on new boys is no less rife than when I left school in 1980. In this case, being picked on meant some blogger showing how clever he is at the expense of trashing my work.

The comments worried me and I contacted my editor at the FT to ask if the puzzle really was OK. He told me to take no notice of the blogs, and to a large extent I agree with him. As I’ve said, I think there are some cases where negative comment can be useful to the setter – and you’d be a fool to ignore any praise that comes your way of course – but I strongly recommend that you don’t allow yourself to be influenced by the carpings of individuals if you don’t agree with them.

I should repeat that the majority of bloggers, like good reviewers in any field, show their knowledge of and enthusiasm for crosswords in what they write, and do not allow their personal prejudices to become apparent. One or two, though, display (often irrational) bias against certain setters, and others have an agenda which they press at every opportunity. One of the latter doesn’t like cryptic definition only clues. How do I know that? I know it because every time he comes across a CD clue in a puzzle he’s blogging, we get a rant about why setters shouldn’t use this sort of clue. Experienced setters will know to ignore it, but I fear that new setters may be tempted to abandon a perfectly decent CD clue to avoid one of these rants. My advice is, don’t let yourself be bullied into not using this or any other legitimate clueing technique just because some Internet bore doesn’t like it.

The same goes for the commenters. Most of them too are fair and supportive, but as always, there are a few who aren’t. There is a (mercifully small) handful who take their inadequacies – cruciverbal and otherwise – out on the setter with a stream of invective or sometimes a single word like “Horrible!” Usually they disappear for quite a while after they’ve said their bit, even if called to account by the decent posters, probably to let their brains cool down after such strenuous intellectual effort. It's debatable whether they qualify as trolls or not but you can safely ignore them, and nobody needs me to tell them that.

Some commenters who have caused upset to setters are regular, bona fide posters who obviously have at least some idea about clueing. They are the people who delight in finding fault. Again, they are in the minority, but these serial quibblers do get under some setters’ skin. At the time of writing, one well tiresome serial quibbler provoked a setter into a splendidly acerbic putdown, which given the circumstances was still admirably restrained. If it had been me I’d have ignored the SQ, but I still sent the setter a note of congratulation!

Actually, the SQs remind me of when I used to go to the opera when I lived near London for years. There was a group of regulars who’d been going since the première of Don Giovanni, and during the intervals they would loudly denounce every aspect of the performance they could possibly find fault with. I used to wonder why they bothered going to the opera at all. SQs are easy to spot: their posts often start with “Not too much to complain about today…” or something similar. To be fair they can be quite polite and some of their complaints are justifiable, but many others are petty and ill-informed, which makes them the cyber equivalent of the boy who cried wolf. After a flawless meal and impeccable service in a restaurant, your SQ will complain about the colour of the tablecloth. The restaurateur will ignore him, and you can too.

It can’t be said too often that we’re talking about a minority here. Most of the contributors make a positive contribution and the blogs are a good thing for crosswords, especially for new solvers. It may be that I’ve underestimated setters’ ability to develop a thick skin, but I don’t think so, judging from correspondence I’ve had with some setters. So I’ll finish by repeating the title of this section: don’t be put off by comments on blogs.

How many anagrams?

Where would the setter be without the anagram? After all, many words don’t break up neatly into separate components that can be exploited in any meaningful way. Often the only way to provide decent wordplay is to jumble up some or all of the letters in the answer, in order to link it to the definition with a sentence which makes some sort of sense.

There are, as I see it, three basic types of anagram clue. These are

  1. Full anagrams. This is where the answer is a direct anagram of another word or words, e.g.

  2. Strange green lady is very famous (9) LEGENDARY

  3. Nearly full anagrams. These involve an anagram of a word or words with one or two (often indirectly indicated) letters added or removed e.g.

  4. Writer of sonnets, very English eccentric (9) STEVENSON

    Pole missed fantastic clarinet concert (7) RECITAL

  5. Partial anagrams. Here only part of the wordplay is an anagram e.g.

  6. Drove badly? One’s exaggerated (8) OVERDONE

There are more complex forms of anagram clue, but we needn’t worry about them here as they seldom appear in daily papers.

Ximenes, in his book Ximenes on the Art of the Crossword, states:

The anagram clues I like best are of two sorts – those in which there is a really appropriate connection between the anagram and the definition, and those which come under my heading “& lit”.

I agree. “&lit” clues, in which the wordplay overlaps exactly with the definition, most often require an anagram. An example is the classic

I’m one involved with cost (9) ECONOMIST

To exemplify the first part of Ximenes’s statement, look at these two clues.

The new stadium is rebuilt for football team (4,3,6) WEST HAM UNITED

Carthorse is ordered for the players (9) ORCHESTRA

What a difference there is between the two clues! The first has a clear football connection and so is a very good clue, though it is so memorable that setters won’t be able to use it (or versions of it) very often. The second, apart from being such a cliché that any setter worth his or her salt would avoid it, is poor, even though it’s technically sound. I have spent my life listening to orchestras, and have yet to find any connection at all between a beast of burden and a group of musicians.

I would add that some anagram clues are noteworthy if the anagram indicator has some surface connection with the anagram fodder and so creates a smooth, meaningful surface reading. For example

Supply teacher makes some ground (7)

There’s no connection between TEACHER and HECTARE but the misleading use of “supply” (adverbial form of “supple”) goes nicely with “teacher” and the surface evokes memories of how classes often play up substitute teachers.

Good anagrams certainly have their place in puzzles as much as any other clue type, and crosswords would be much poorer without them. The problem with anagram clues is that unless the setter is careful, too many will appear in one puzzle. The result is that the puzzle becomes rather dull to solve, and it can look as if the setter isn’t really trying very hard, even if that’s not the case. So how many is too many?

Ximenes suggests a maximum of four full anagrams and a couple of partials in a 28-word puzzle. Alec Robins (who set as Custos for the Guardian), in his excellent but sadly out of print book Teach Yourself Crosswords, advises much the same. These books was first printed in 1966 and 1975, but this advice is generally adhered to today. It is advice I try to follow, and a great many broadsheet setters use a similar anagram count.

The only difference between then and now is that what I referred to earlier as nearly full anagrams are much more common these days. This sort of clue tends to be used when the definition can be connected nicely with an anagram, but there are one or two letters unaccounted for and these need to be added or removed. For the purposes of this discussion, these nearly full anagrams can be counted as full anagrams. In other words, your puzzle of 28-30 clues should contain no more than four full or nearly full anagrams, and two or three partials.

It’s therefore wise not to use up your quota in the first seven clues of the puzzle! It’s not a bad idea to look through the words in the grid and decide which words, if any, look as if they are going to have to be anagrams, so you know how many you have left. For the other words I look first of all to see is there is a really good anagram that I can’t resist, and if not, I’ll use an anagram as a last resort. Partials are, I think, best left for intractable fragments of words that mean nothing and won’t break down easily any other way. For example, INTO A for words ending in -ATION.

Are these rules inviolable? Of course not. There are times you may have to increase the anagram count slightly, for example if the inclusion of a theme has left you with a higher number of intractable words than usual. I am writing this piece with broadsheet level puzzles in mind, and of course things are very different if you’re writing easy puzzles for magazines or local papers. These call for the most basic clues, and your options for abbreviations, indicators etc. are very limited. You’ll have to rely far more on anagrams in these puzzles than you would at broadsheet level. I’d still try not to use too many, though.

A final aside about certain anagram indicators. There are a lot of related anagram indicators based on the idea of “mad” – such as crazy, cuckoo, out to lunch, nuts, bats, bananas, loony and of course mad itself. Recently there have been objections from the usual suspects that this mocks people with mental illness. Poppycock! All of these terms have secondary meanings which have nothing to do with mental illness and simply mean “out of sorts” in one way or another – I’m nuts about her, he drives me crazy, my parents got mad etc. I would bet my entire CD collection that not one single setter has ever used one of these words in a clue and thought “Ha! That’ll give people a really good laugh at the expense of those in psychiatric care!”

Crosswords are intended for everyone, so of course taste and decency should be observed. But it would be a great shame for setters to lose these indicators, many of which offer great opportunities for misdirection, from their arsenal because some killjoys have nothing better to do than pretend to take offence on behalf of others. My advice is to exploit the richness of the English language for all it’s worth unless you’re specifically told not to, and if that happens you should protest vigorously!

Capitalisation

This one isn’t really a general tip, but it isn't a broad enough topic to merit an article of its own. There are two issues concerning capitalisation which arise in crosswords.

1. Missing capitals. Consider this clue for CLINTON:

Customer ignores third of items on bill? (7)

Is it fair to decapitalise Bill here? Ximenes says not, as does anyone who shares his view that cryptic grammar is important. That includes me. The “libertarians” will say that by taking this view I am being pedantic and anal retentive; what’s more, they will accuse me of trying to take the fun out of crosswords. Well, I certainly am anal retentive and pedantic when it comes to capitalisation generally. I do not take much notice of online postings or emails which omit capital letters for proper nouns or the pronoun I. I’m not talking about the occasional failure to hit the shift key here: we all do that from time to time, and spellcheckers won’t pick it up if the non-capitalised word means something in its own right. I’m referring to people who can’t be bothered to put capitals in at all and yet expect us to take what they write seriously. I don’t mind it so much in text messages – writing texts on many phones is a fiddly and laborious process and most texts are hastily written communications to be read once and forgotten about. Still, I always capitalise and punctuate in any texts I write, and may well be the only person on this planet who bothers to do this. But that’s all beside the point.

In crosswords, the definition should be accurate. It can be misleadingly worded, as in the old standbys like flower (river), driver (golfer), or perhaps Nice chap (French male name). All of these obey Afrit’s injunction: You need not mean what you say, but you must say what you mean. Deliberately misspelling a word does not say what you mean, and although calling this a misspelling may be overstating the case a bit, that’s what omitting a capital letter amounts to.

As for me trying to spoil the fun of crosswords, I deny it. In fact, I’m trying to do the exact opposite. Good crossword clues are all about clever, but fair, misdirection; when the penny drops I get that satisfying feeling of “why didn’t I see that before?” Clues in which I have to second-guess the setter because he or she has withheld important information (yes I know, that does sound a bit pompous) lead to me feeling that I’ve been cheated and wasted my time. That’s anything but fun! If I were the only one saying this it probably would be fair to accuse me of being a killjoy, but since the majority of crossword editors, all competent setters and, judging by comments on forums, a great many solvers too would be grossly underwhelmed by the CLINTON clue above, there must be a heck of a lot of us spoilsports around or, just perhaps, we’ve got a point.

Of course that doesn’t mean that setters can’t exploit proper names that mean something in lower case too. One way round the capitalisation issue is to use a homophone indicator, e.g.

Customer ignores third of items on bill, we’re told? (7)

I’ve done this a few times in some of my earlier efforts but in recent years I’ve come to think it’s rather lame. A far better solution is to rewrite the clue so that the proper noun comes at the start of the sentence or after a full stop. Some clues lend themselves more easily to this than others, and it took a bit of thinking to rework my example. A possible way to put Bill at the start would be

Bill for one customer doesn’t include tip for service? That’s acceptable (7)

You may say it’s a bit wordy compared to the original and you’ll get no argument from me on that score, but it still misleads the solver into thinking of a restaurant tab rather than a person called Bill, and the necessary indication of definition by example (“for one” in this case) is quite neatly disguised by the surface. “Tip” is defined in Chambers as a small piece forming an end and the furthest part (among other things) but if you’re one of those who see it as an indicator for first rather than last letters, “at the end” after “service” is a viable alternative. The ON is clued by its meaning in the phrase “that’s just not on.”

In wordplay elements, as well as definitions of the answer, capitalisation should never be ignored. Look at this:

Criminal wants it with cream? (6)

This is a very loose reworking of a clue that appeared in a national newspaper not long before writing this. The idea is that you combine IT with BAND, a reference to the 60s rock group Cream. The rather open-ended definition, along with a reference to a band that finished over 40 years ago, makes things pretty tricky, but would probably pass muster if it weren’t for the omission of the one vital piece of information we need – the capital C. I think this is about as unfair as it gets, don’t you?

2. False capitals. Here’s Bill again in another clue:

Charlie intervenes in a matter for Bill (7)

C in A COUNT gives ACCOUNT, which is a bill. But is it a Bill?

Imagine someone writing a note to a friend: “Took the car in for a service and got a Huge Bill!” The friend may perhaps conclude that the car was serviced by a morbidly obese mechanic called William, but I think it’s unlikely. The meaning is clear. It’s not uncommon to capitalise for effect, especially in eye-catching headlines (Aliens Abducted my Hamster!), and also to emphasise a person’s importance (the Queen, the President). Capitalisation occurs in religious contexts (the Redeemer, the Saviour) and also in expressions that don’t involve people (the Grand Plan, the Big One etc.). And don’t forget that all nouns are capitalised in the German language.

The general consensus from those who care about and understand cryptic grammar is that false capitalisation is a minor sin, if indeed it is a sin at all. Ximenes only really sanctioned it for nouns, but these days it’s accepted for pretty well any part of speech except perhaps prepositions and conjunctions, and there is nothing to be gained from giving those words false capitals anyway. Pronouns are capitalised when referring to God (He, Him, His etc.) and it would be a great shame if we couldn’t have, by analogy,

Doctor Who harboured good and evil (6)

This rather nice cryptic definition clue is one that another setter originally sent to me to ask my opinion on the false capital. It is part of a very fine puzzle now on this site (here). The following example shows that false capitalisation can be used in wordplay elements too. Back to our old friend Bill again:

Rely on Bill’s relation (7)

It’s quite common for Bill to turn up as AC in clues, as in this second clue for ACCOUNT. As an example of a capitalised verb, Port Said has turned up occasionally in the Times. It's not a reference to the Egyptian city – it indicates a homophone of a port somewhere else in the world. I can’t remember which, though I am pretty sure it wasn’t Dubrovnik or Dnipropetrovsk.

I hope that I have managed to show clearly why there is a huge difference between omitting capitals and putting extra capitals in. There’s one other factor which mitigates – if mitigation is needed – the use of false capitals, and it arises from the tangential effects of using them. Capital letters stand out – that’s which the whole point of them. A capital letter in the middle of a clue grabs your attention and solvers soon learn that words which look like proper nouns on the surface – especially names – as often as not are to be interpreted with the capital ignored. This means that false capitals can actually make clues easier, and one really would have to be an anal retentive killjoy to complain about that!

Love Thine Editor

All the broadsheets, and some other major publications, have a crossword editor. But what does a crossword editor actually do?

“Takes out all my best clues, that’s what,” may be the answer from some disgruntled setter who’s just had a favourite clue disallowed, but that’s hardly fair. A crossword editor has numerous tasks to perform – far more than a literary or features editor, for example – and I’ll start by outlining these.

First of all, he will correct typos or factual errors, such as a clue that defines Istanbul as a capital city. (I use the masculine pronoun because the great majority of crossword editors are men at the time of writing.) Let’s be honest here – we all goof from time to time, and a second pair of eyes looking at the puzzle afresh can save the setter a fair bit of embarrassment.

That’s the basis of any editorial work, of course, but the crossword editor’s responsibilities go far beyond that. He has to check for errors in the cryptic part of the clues – for example anagrams that don’t work or stray letters – and not only that, he needs to check that each clue is fair and solvable. Something that’s obvious to the setter may be far too obscure or convoluted for even the best solvers; what’s more, there may be an issue of technique involved, such as a duff container-and-contents indicator or the wrong part of speech.

Then, of course, the crossword editor has to make decisions on grounds of taste and decency. For example, he may reject a clue like

The plane crashed – it’s a Jumbo (8)

if there has been a recent air disaster, and quite understandably so. I avoid clues like this anyway, as it would be just my luck that it would appear in the paper on the day that a 747 went down.

Another job the crossword editor has to do is schedule each day’s puzzle so that there is an overall balance of easy and hard puzzles, with not too many of one or the other in succession, and to decide which crosswords should be used as prize puzzles.

I act as crossword editor for this site when I publish guest puzzles, and so I perform all the duties listed above except the last. My job is considerably easier because I only publish puzzles occasionally, whereas the crossword editor of a newspaper has to prepare six a week (Sunday puzzles too, in some cases). That’s why, to misquote Mark Twain, I shall always try to do right and be good so that God will not make me the crossword editor of a newspaper. Frankly, it sounds too much like hard work!

Now that we’ve discussed the onerous multitasking a crossword editor has to undertake, let’s return to our disgruntled setter from the start of this section. If I’m honest, I’ve grumbled to myself a few times when I’ve been asked to redo a clue in whole or in part, but that soon passes and far more often than not I realise that the editor is right after all. I have been known to argue when a clue of mine is queried, but only very seldom, and I am convinced that this is the best way to do things. There’s nothing wrong with defending the occasional clue if you feel really strongly about it, but if you do it too often, relations between you and your editor may break down to the point he regrets having you on his team of setters. By accepting you as a setter for his paper, he has put his faith in your abilities, and it is only courteous to repay the compliment by trusting his decisions in turn.

A good editor will always consult with the setter before making any changes. The only exception is if he notices an error at the last minute and there isn’t time to discuss corrections with you. Resentment would soon breed if every time you looked at the published version of a puzzle you thought “Hey! I didn’t write that!” – and all the more so if the changes actually weaken the clues. Both the editors I work for, at the FT and Independent, almost always consult with me about any alterations to be made and I respect them for that; and as far as I know this practice is observed by all crossword editors at broadsheet level (and rightly so).

Contrast that with the agency I used to work for, producing easy tabloid-level puzzles. I did quite enjoy writing these puzzles and I was grateful for the experience, but the editing there was a shambles, with clues (and sometimes answers) being changed for no apparent reason and without any consultation with me. I’ll admit that when I first started with them I did sometimes pretend I was writing for the broadsheets and put in some clues which were probably too advanced, but much of what was substituted was rubbish. Even when I gave them what they wanted – the sort of thing that makes the Sun puzzle look like the Listener – the meddling went on, and one particular howler was a clue for EGOIST. I can’t remember my original clue, but the substitution was

I am the most important person in his life (6)

You what? Even then I didn’t bother to argue; I churned out puzzles, took the money, and thanked my lucky stars that the puzzles appeared without identifying the setter. It is well-nigh impossible to have a decent working relationship with a crossword editor in a situation like that, but luckily it is inconceivable that such incompetence would occur in the sort of publications that carry quality crosswords, so I repeat my exhortation to respect your editor’s decisions in good grace and only query them if you really feel you have to.

One other point, and I know this from my experience editing puzzles for this site. There is no quicker way to annoy a crossword editor than to send in constant revisions once he’s prepared your puzzles for publication. Obviously it’s OK if you correct an error that neither of you spotted before, but if you send what amounts to a rough draft followed by a series of updates as you think of better clues, his patience will soon run out.

So, love thine editor. He’s the boss, after all!