B A R E L Y B A D  W E B  S I T E  
   Links      Cute Clues   



Because this whole thing about crosswords goes on rather at length, you'll find several intersections on this crossword main page where you can either scroll straight on and miss something you might like or take a detourThe only purpose of the detours is to spread out your downloading time.  In the outlines at right the detours are marked with an ellipsis.



I. Introduction

I've worked an average of just over one crossword puzzle a day from 1972 to 2018.  During that time I've learned a thing or two about crosswords, and I've even gotten a few of my acquaintances hooked on them.  I want to hook you.

Plus which, since I first posted this essay about crosswords on this Web site I've engaged in a number of electronic conversations with a number of crossword constructors and other fans.  Without meaning to seem to draw a too-rapid conclusion, I can say that constructors of crosswords are more interesting to talk to than, say, your average doctor or lawyer or corporate chief.

In the twelve sections that follow I'll try to pass on to you some of what I've learned, both from my own experience and from that of others.

This monograph is written for everyone from the total novice to the thoroughly experienced player and even the thoroughly experienced constructor.  My hope is that in this article you will find some useful information, perhaps some slightly interesting information, and maybe even a dollop of entertaining information, no matter who you are.

Both as you're reading this and as you're working them, the most important thing to keep in mind is that crosswords are for fun and relaxation, and I hope you don't take them any more seriously than I do.  In fact, I prefer the phrase "playing a puzzle" to "working a puzzle" because that's the attitude I think you should have.

Top of Crossword main page

II. Why to play, and why not

A. Why to play crosswords

One good reason I play crosswords is that they're fun and relaxing.  They're a sort of hobby, a pleasant diversion.

Another reason is that they can be mildly amusing or even enlightening.  For example, in a New York Times puzzle that ran on March 12, 2003, by Myles Callum, the theme answer is a takeoff on a Groucho Marx quip, strung in perfect symmetry throughout the grid:


Here's another funny one, especially if you ever spent more than five years living less than three miles (exactly five El stops) from Addison AT Clark Street.  In this NYT puzzle of October 23, 2005, by David J. Kahn, I'll give the clues and ANSWERS together.

A man goes for a walk and FINDS A BOTTLE ON A BEACH.

When he pops the cork, a genie appears and says, "I SHALL GRANT YOU ONE WISH."

The man says, "I want to see PEACE IN THE MIDDLE EAST."

He then hands the genie A MAP OF THE AREA.

The genie studies it for a while and finally says, "This is impossible.  So I GRANT YOU ANOTHER WISH."

The man says, "I always wanted to see THE CUBS WIN A WORLD SERIES."

The genie replies, "LET ME SEE THAT MAP AGAIN."

(Needless to say, this joke made a lot more sense before the endearing Cubbies finally won a World Series, after over a hundred years, in a down-to-the-wire showdown that went into extra innings, with a rain delay beforehand, against the Cleveland Indians on November 2, 2016.  The Cubs came back from being down 1 to 3 by winning three straight.)

Crosswords encompass everything from the easy to the mind-taxingly difficult. The sense of satisfaction from completing a really hard clue is one every crossword fan will understand.

Yet another reason is that crosswords're educational.  I can guarantee you that every single person who knows me wants to be on my team in Trivial Pursuit, and the reason is that I've collected thousands of factoids in my brain from answering what I estimate to be closing in on nearly a million and a half crossword clues by now.

Of course, many of those million and a half answers are repeats.  For example, as every New York Times crossword puzzle fan will tell you, there's a limit to how many times you need to learn that Woodie Guthrie's son is named Arlo.  And of course many of those answers are not educational at all.  For example, I don't learn anything by giving the answer DROP to the clue "Fall."  But if only two percent of the clues I've answered have taught me anything new, that's still maybe 30,000 factoids, and that's way better than zero.  And it would be a mistake to think of those factoids as being sterile and isolated.  While it's true that in any given answer you can learn only a few facts, when you consider how many answers cover a particular topic over a long time, you can eventually put together quite a few facts about it.  For example, I have answered perhaps fifty clues about Galileo over the years, and in the process I've learned a lot about him.  I know when and where he was born and whom he worked for and what his accomplishments were and a fair amount about what happened to him throughout his life.  I've learned about the Catholic Church's reaction to his confirmation of Copernicus's discovery that the Earth revolves about the sun.  I even know his last name.

Here's another example.  I have learned that Henri Matisse was French, that he painted water media, that he was a leader of the Fauvist movement, that he did a piece titled "Le Bateau" (The Sailboat), that "Le Bateau" appeared at an exhibition at the Museum of Modern Art in 1961, that some MOMA curators hold Masters of Fine Arts degrees, and that Matisse's sailboat piece was hung upside-down at MOMA for 47 days.  And, believe it or not, I learned all that from one puzzle, the New York Times Sunday of November 23, 2008, by David J. Kahn.  That's a lot of information to cram into one crossword puzzle, and I'll bet it's more than you knew about Henri Matisse till just now.

You're probably wondering how a Matisse piece could get hung in one of the foremost art galleries in the world for 47 days upside-down.  Believe it or not, this puzzle contains a bonus that gives you the answer.  If you read the gray-colored squares they spell out SAILBOAT, and if you read the circled squares they spell out REFLECTION.  Matisse's "Le Bateau" depicts a sailboat and its reflection on the water, and the "words" SAILBOAT and REFLECTION in the answer grid are reflections of each other.  This is a shining (ha ha) example of the degree to which crossword puzzles approach true art.

In case you're further wondering about that Matisse piece, it's cut out of paper and here it is first upside-down then right-side up.


Update of November 23, 2009 (exactly one year after the crossword puzzle appeared): Many events or states or affairs that are merely coincidental are mistakenly referred to as ironic.  It so happens I lived in an apartment on 75th Terrace, then in 1994 I moved, and then I moved again, this time to a house also on 75th Terrace.  I mentioned this to someone at a party once, and someone else said, "Wow, I used to live on a 75th Terrace too."  The first person then said, "That's, like, really ironic."

No, it's not.  It's not the least bit ironic.  That's just coincidence.

(You can see another unexpected example of mere coincidence here, and, ironically, it also involves street names.  Kidding.)

And it's just coincidence that this update is published exactly one year after the puzzle was published, not irony.  Irony can involve coincidence, but it always involves more.

Anyway, today it was brought to my attention that the left image above, the one I said was oriented wrong at MOMA, is itself oriented wrong.  Now that's irony.  See the difference?

It should look like the image below, i.e., this is how "Le Bateau" was incorrectly hung in MOMA for 47 days, not the way I showed it above.

Sorry (and thanks to Kalvik for bringing this to my attention).

Also, in case you're wondering, my wrong hanging lasted not 47 days but 351.

(Also, to be honest, they're all pretty much the same to me.  Is that ironic or what?  I mean, what if that one squiggle of cloud or whatever it is had been shaped just slightly differently?  Or how do we know Matisse didn't slip with the scissors, look at it, and say, "Oh well, they'll never know.  They'll buy anything I do.  Hell, they'll buy it sideways if I call it 'Le Papillon' or 'Des Ailes D'ange'"?)

Well, anyway, possibly the best reason for playing crosswords is that they allow you to revel in the small pleasures of playing with the English language.  English, like any language, can be a toy to play with, like a kite or a computer, and I think it's good to play with it.  It seems to me that it's desirable to understand English, or whatever your language is, as well as possible, to strive always to get more fluent.

And not just because increased fluency in one's language increases clarity of communication.  It's also because it's in language that you think.

Much of your cognition results from how you think thoughts in your head, and when you do that -- when you reason things out -- you're doing so using language.  If language is the tool of mentation, it is a tool we should always strive to master more.

Yet another reason is that, according to some sources, playing crosswords or performing other such mental exercises can retard the onset and progress of senile dementia and Alzheimer's Disease.

B. Why not to play crosswords

They are time-consuming.  According to my crossword timer spreadsheet, I spend an average of about 6 days annually playing crosswords.


Top of Crossword main page

III. The rules of crosswords . . .

Crossword puzzles that appear in The New York Times follow certain rules, and I hereby propose to list some of them.  Some of them are of little interest, but many will likely help you play better if you know what they are, such as knowing that answers are never used more than once.  Some of what I call rules are really more like traditions that authors invariably observe (and that crossword editors enforce), such as that the capital of Italia is ROMA, not Rome.

Take this detour to learn about the rules of crosswords, then return here.


Top of Crossword main page

IV. How to play well . . .

In addition to the other information provided in this monograph, some of which should prove useful in playing well, in this section I list some specific tips, tricks and hints that I've found helpful in achieving the goals of finishing at all and finishing as quickly as possible.  Some of these suggestions will be obvious to the experienced player but not to the novice.  Others will seem like overkill to the novice but might interest the experienced player.

But first let me tell you why I think you should learn how to play well, by which I mean being both as accurate and fast as possible.

The two reasons you should be accurate are obvious: First, if you filled in a wrong letter you deserve to feel less satisfaction for having finished.  Second, and more important, presumably you got at least two answers wrong, which means you might have mis-educated yourself.

The two reasons you should be fast are less obvious.  One reason relates to the reason I gave above about why not to play crosswords, which is that they are time-consuming.  Clearly, if you can whip out a Monday puzzle in ten minutes instead of fifteen, that's five extra minutes you can use to get off the pot and on with your day.

The other reason to try to finish a puzzle fast is because it adds fun.

Harder. Since the mid-90s or so, I've wanted the New York Times' crossword puzzles to be harder.  As I said in a 1996 letter to Will Shortz, the editor of the puzzle, "I wish every day were Saturday."  (NYT crosswords get harder throughout the week: Mondays are the easiest, Saturdays the hardest.  Sundays are 96% bigger but tweaked to be about a Wednesday or Thursday level of difficulty.)

Because I've been playing crosswords for so long, I've gotten to the point where I find little satisfaction in merely finishing the Monday and Tuesday puzzles and some of the Wednesdays unless the theme is good.

Faster.  Consequently, in 1997 I added some challenge by timing myself, and you might want to consider doing so as well.  My best time so far, which was on a Monday, of course, is 4 minutes 48 seconds.  Experts can finish a Monday in 2 minutes, but I so rarely finish in under 6 minutes that I despair of breaking the 5-minute barrier ever again. 

Anyway, it's good to time yourself, to see how well you can think not just right but fast.

I use a clipboard to play the puzzles on, and I've wedged a stopwatch under the clip, so it's easy to time myself.  I usually finish the Mondays and Tuesdays in the range of 7 to 9 minutes.

My good times for the Saturdays are around 30 minutes although I've done a few in under 20.  My average Sunday time is well under an hour, although if you normalize it from a 21 by 21 grid to a 15 by 15 it turns out to be 11% faster than the average Friday, and my fastest Sunday turns out to have been just under 11 minutes.

Here's a table showing how much faster or slower I have performed so far on average compared to other selected days:

Tuesday compared to Monday   20% worse
Wednesday compared to Tuesday   46% worse
Thursday compared to Wednesday   60% worse
Friday compared to Thursday   53% worse
Saturday compared to Friday   29% worse
Sunday* compared to Friday   21% better
Saturday compared to Monday!   452% worse

Your own times will vary, of course, but if you're like me what this means is that you should expect to hit the biggest wall going from Wednesday puzzles to Thursdays.  Hang in there.  Once you get to where you can finish the Fridays, it's almost all pure fun.

The data for the tables above and below came from my times tracker spreadsheet.

 04/97-04/05  M T W R F S N
Slowest 11:19 17:17 34:15 66:24 87:14 102:12 138:31
Average 7:21 8:48 12:50 20:34 31:33 40:34 26:04*
Std. Dev. 1:04 1:43 3:37 8:05 11:56 16:13 14:50
Fastest 5:01 5:52 6:43 8:44 12:23 12:58 21:26

*Normalized from a 21 X 21 to a 15 X 15.  The Sunday puzzle is 96% bigger.

Spreadsheet. In April of 1997 for this essay on crossword puzzles I wrote a small spreadsheet to track my daily times.  You may download it -- for free, of course -- to track your own.  All you do is enter your times and the spreadsheet automatically updates you, on a Monday through Sunday basis, with your three best times, your three worst, and your average, as well as the standard deviation of times for each day and some other statistics.  I myself record the times originally on a piece of paper in my clipboard immediately after I finish each puzzle, and every so often I transcribe the data into the spreadsheet.  I've tried to make the data-entry part as quick and painless and unscrewuppable as possible.  To learn more about this spreadsheet, which is written in Excel 2000 and no doubt can be read by or converted to other formats, just .

Update of May 2005: The spreadsheet now also contains a sheet that allows you to chart your times, to see whether you're getting faster or slower.  You can see and read a little about the chart of my times here.

Anyway, I regard finishing fast as being almost as desirable as finishing at all.

I think in formal speed-solving contests they let you leave cells blank or get letters wrong and then penalize you severely for it.  I've heard that the penalty is this: For each wrong cell they whomp you once upside the head with a warm walleye, and for each blank cell they use an outright carp.

Update of October 26, 1999  Here are the rules for scoring in the American Crossword Puzzle Tournament (ACPT) held and directed by Will Shortz each year in Stamford, Connecticut from its inception in 1978 till 2007, after which because it became so popular it moved to a larger hotel, in Brooklyn, New York:

  • 10 points are awarded for every correct word across and down.
  • 25 bonus points are awarded for each full minute the contestant finishes early.  [Note: Time allotments are usually 15 minutes for 15x, 25 minutes for 17x, 30 minutes for 19x, and 45 minutes for 21x)].  The bonus is reduced by 25 points for each letter that is omitted or entered incorrectly up to but not beyond the point the bonus returns to zero.
  • 150 bonus points are awarded for a complete and correct solution.

So, for example, if it's a 15 X 15 puzzle with 70 answers and you get them all right in 12 minutes, that'd be 700 points for the 70 correct answers, plus 75 bonus points for finishing three minutes under the limit, plus 150 additional bonus points for getting the complete, correct solution, for a total of 925 points.

(Just so you'll know, there are people who can solve a NYT crossword in two minutes.)

However, if there's just one cell wrong or blank in that same 12 minutes your score would be 680 points plus 50 bonus points for finishing three minutes early, or only 730 points total.  So, apparently it doesn't involve being struck with any species of fish whatsoever.

Update of July 2002  In a BOOK titled The New York Times Monday Through Friday Easy to Tough Crossword Puzzles, a collection of 50 NYT crosswords published by St. Martin's Press, the introduction by Will Shortz is based in part on the table of my solving times shown above.

shortz_intro.jpg (13,482 bytes) 137 X 124

Also, the URL given therein for my crossword essay was out of date.

It is barelybad.com/crossword.htm, the very page you are reading now.

As I said, some of the tips that follow might seem like overkill to the novice, and it's probably those very tips whose purpose is to promote speed, not accuracy.

I do hereby note a significant drawback to trying to play fast, which is that you can't take the time to wallow luxuriously and languidly in all the information to be found in your reference works.  If you don't feel pressured to finish fast, you may spend a few extra seconds or minutes or weeks reading about something that has caught your interest's eye.  Learning about the universe out there and in our heads, and the language we use to interpret them is, it seems to me, always desirable.

Take this detour to learn how to play crosswords well, then return here.


Top of Crossword main page

V. Relationships between clue and answer

It is insufficient to think of the answer as being a definition for its clue or vice versa.  Frequently one of those is the case, but there are other possibilities.  What's true in every case is that the answer and its clue form a relationship.  It's the relationship that you need to noodle out for each pair of answer and clue in order to complete the puzzle.

And in order to solve that equation, i.e., in order to identify that relationship where ANSWER = Clue, it's helpful to know which relationships can exist in crosswords.  If you get stuck on a seemingly easy clue-and-answer relationship, it's good to return to the advice above about keeping your mind open to all interpretations, and a good way to do that is to mentally scan through the possible relationships.

A. Equivalence. The commonest crossword relationship is that of simple equivalence, i.e., the answer and its clue are two expressions of the same thing.  For example, the clue is "Capital of the former Upper Volta" and the answer, as if you didn't know, is OUAGADOUGOU.

B. Commonality. In this relationship, each of the collection of elements of the clue is an example of the answer.  For example, the clue "Betta, tetra and opah" could properly be answered with FISH.

(Also note that in this example the answer FISH would be appropriate even if the clue had read "Betta, tetra or opah," since "fish" refers to both plural and singular.  And while I'm on this example, let me wedge in another tip, trick or hint: Remember that some present tense verbs use the same spelling for the past tense form, such as "put" and "read" and the dreaded "set.")

The less common commonality is the mirror of the relationship described above, in which the clue is the single element and the answer is the collection of elements.  To stick with the example above for a moment, if the clue were "Fish" it would not be a violation of the crossword puzzle rules if the answer were BETTA TETRA OR OPAH.

If you're thinking that this answer is unlikely to appear in isolation in a real puzzle, you'd be right.  But it certainly would not be out of character if it were a theme answer.

C. Set membership. A pair of mirror relationships is that of set membership.  Either the clue is a member of the answer set or the answer is a member of the clue set.

An example of the first instance is the answer ANIMAL for the clue "Insect."  An example of the second instance is the answer INSECT for the clue "Animal."

These are extremely simple examples, invented more for clarity than for how likely they are to be used in a real puzzle.  In fact, it would be unusual for the clue "Insect" to be given for the answer ANIMAL or vice versa.  The point is that neither pair would violate the rules of New York Times crosswords, so they are examples of two common relationships for which you need to keep a mental eye peeled.

D. Fill in the blank (FITB). This relationship hardly needs further explanation.  The relationship is that the answer is that collection of characters that completes the clue.  A self-referential example here would be the answer BLANK for the clue "Fill in the ___."

A slightly more complicated and typographically invisible sort of FITB arises in the form "Suffix with lob or mob" (answer: STER) or "Prefix meaning catholic" (answer: OLI).

E. Analogy.  In this relationship the answer can be the solution to a logical analogy presented in the clue.  For example the answer to the clue "Dog : Puppy :: Cat : ____" is KITTEN.

In the special notation used by logicians and mathematicians, the single colon means "is to," and the double colon means "as."  So, to take another example, "Area : map :: volume : ____" is read "Area is to map as volume is to What?"  (Hint: globe.)


Top of Crossword main page

VI. Errors? . . .

The number of outright errors and seriously questionable answers in New York Times crosswords is remarkably low.  In the one page you're reading at this moment I've probably made more errrors of one sort or another than in all the 3,652.5 daily New York Times crosswords published in the last ten years.

Still, errors and questionable answers do crop up, and it's helpful to know that.  When they do arise, the ideal result for you as the player is both to finish the puzzle according to the answer grid and to recognize the error.

How to complain.   In about 1996 I complained to Will Shortz, the editor of the New York Times puzzle, about the answer TIL as an acceptable substitute for "till," as in "until."  Neither "til" nor even " 'til" is a legitimate word in any dictionary I had consulted (despite the prevalent use of the second version by all sorts of uneducated and otherwise educated people, as though it were a contraction of "until"), so I wrote asking what authority allowed the answer TIL.  I cited a number of authoritative dictionaries that either do not list "til" at all or list it as an illiterate version of "till."  I even cited a few authors on usage agreeing with me, one of whom was the famous New York Times editor Theodore M. Bernstein.  In response to all that I received a post card in which Mr. Shortz pointed out, quite correctly, that exactly one of my references allowed as how " 'til" was regarded as a variant of "till."

That's all it takes, and so I was set straight about what's allowable as an answer to a NY Times crossword.  It is as it should be.

I also note that "til" has never appeared as a legitimate answer since I wrote that letter, although it did appear in the January 23, 1998, puzzle in answer to the clue "Up to."

But that's not exactly what the clue said.  In fact it said, "Up to, informally" (emphasis mine).  Yay for me.

FYI, "til" actually is a perfectly OK if somewhat obscure word meaning the seed of the sesame plant.

Anyway, because there are so few errors, it's satisfying in a puny sort of way to catch them.

Take this detour to learn about NYT crossword errors I think I've caught, then return here.


Top of Crossword main page

VII. Reference works . . .

There is a difference of opinion whether it's OK to use reference works to solve puzzles.  The people who think it's not OK explain their position with these three points:

  1. If you can get all the answers without having to look them up you'll feel more satisfaction for your accomplishment.
  2. If you can't do the puzzles without help then you shouldn't take full credit.
  3. It's cheating.

In reply I say:

  1. You may choose to feel more satisfaction if you like, although it seems to me a somewhat arbitrary standard.  As for me, if I can finish the puzzle even a single minute earlier by consulting a reference, I'm willing to pay the price of a loss of some additonal sense of satisfaction.  And I do agree that if person A and person B finished the Saturday NYT in the same amount of time but person A did it without references, person A is the more knowledgeable solver. 
  2. Here again, you may choose to set such an arbitrary standard, but my attitude is that by paying the price of knowing I didn't meet the no-references standard, I learned somethng I didn't know before.
  3. The third point is tautological and therefore specious.

The term "reference" as used here has different meanings.  We can think of them as lying on a scale from not-really-cheating-much-at-all on up to full-blast, no-doubt-about-it cheating.

At the good end of the scale, I think it's positively desirable to use references such as dictionaries and atlases, for two reasons.  One is that the more you use them the more efficient you get at using them, which is good because it saves time.  I know France is on page 28 of my atlas, and I often open my unabridged dictionary to within a few pages of the word I want.  I'm willing to pay the price of believing I cheated a little in order to save time not having to look up France in the table of contents.

A better reason is that if you do go ahead and use your dictionaria and atli and so on, such as how to pretend-pluralize words in Latin, you're bound to learn something.  As you're rummaging around looking for whatever five-letter word you need, you can't help stumbling right smack into facts you didn't initially care about, facts you might find useful in understanding the universe of your life.

Moving supposedly further up the scale from good cheating to bad, I really don't see any significant difference between using paper references and using the Internet as a substitute.  Learning to use the Internet efficiently to get information is desirable, and it can save a significant amount of time.  And the Internet, with its hyperlinks to anywhere, makes it so much quicker to follow up on whatever piques your interest.  Furthermore, the Internet has a lot more information than any collection of paper references you have, unless your name is Library of Congress.

Yet nearer to full-on cheating is simply asking someone who knows the answer.  I had no idea that in hockey there's a deke or that in baseball there's an A-Rod, but I knew my brother would, so I asked him and got my answers.  I didn't learn how to use my references more efficiently (unless you count knowing my brother's telephone number, which I'll give you if you have any sports questions), but at least I learned the explanations (decoy, Alex Rodriguez).

Perhaps even further along the cheating scale is using a reference, whether online or on paper, that simply shows you all the possibilities for a given pattern of known and unknown letters.  If the clue is "Plenipotentiary" and the answer LEGATE turns out to fit your puzzle, and if you don't know what either word means, you have nominally finished the puzzle but you haven't learned anything.  I'd rather spend a few extra seconds using a more informative reference and learning something than finishing faster but remaining ignorant.

The extreme end of the cheating scale is simply to ask online.  Several sites purport to answer general-knowledge questions, with varying degrees of accuracy but in some cases within a minute, so you could simply log in to one and type, "What German industrial city on the Ruhr River starts with ES and is five letters long?"  (Hint: Essen.)  I used ChaCha.com for this purpose hundreds of times till it suddenly got unreliable.  Yes, I was cheating, and no, I don't feel bad about it.

As I say, I think the "no-references" standard is arbitrary.

I would not finish maybe a fifth of the puzzles if I eschewed my references.  As I see it, you will surely learn a lot more if you do use them, and I truly do consider learning to be a really good reason to play crosswords, no matter how many people disagree with me, which is a lot.

Not only do you learn about the relationship you looked up, you can also fortuitously learn a bunch of other stuff along the way.  For instance, just this morning (yes, this very morning) I was looking up the word "cline" in a dictionary (Rhu2) when I noticed the headword on the other page was "cliticize," which at first I thought was the Japanese pronunciation of "criticize," which of course drew my attention, so I looked up "cliticize" and discovered its handy adjectival form, "clitic," the definition of which you might want to look up yourself, following which you will probably be a lot disappointed but a little educated.

If you're an experienced solver (or constructor) you won't care much which references I use because you'll have gradually gathered your own collection, but if you're a novice you might care a little.

Take this detour  to learn about the reference works I've chosen, then return here.


Top of Crossword main page

VIII. Theme answers . . .

One of the funnest features of New York Times crossword puzzles is the theme answers.  Not all of them have a theme, and they can't help seeming a bit pedestrian or repetitive every once in a great while, but if you've read this far then often you'll think they're remarkably clever and entertaining.  Crossword constructors play with language at two levels.

Take this detour to learn about some outstanding NYT crossword themes, then return here.


Top of Crossword main page

IX. Try writing your own . . .

If you would like to learn how to play crossword puzzles better, the single best method is to play them a lot.  The most difficult method (not counting having to read this really long essay), but a surprisingly educational one, is to try to write your own.

There are two benefits to trying to write your own crossword puzzle.  One is that you will have a more detailed appreciation of how difficult it is to write one at all, not to mention how much more difficult it is to write a good one.

The other benefit is that you will become a better, faster player.  Once you have sat in the author's seat for a while, you can understand better how certain answers get the clues they do.  You'll ask, "If I were the author of this puzzle and this is the answer I need to complete this part of the grid, what clue can I give?"

Take this detour to learn how to write a crossword, then return here.


Top of Crossword main page

X. Hardest puzzle . . .

One particular crossword has generated a lot of interest, the puzzle I declared is the hardest one ever published by The New York Times.  Also, it contains what has been for me the most baffling clue-and-answer relationship ever.

Take this detour to learn a lot more about the hardest NYT crossword, then return here.



Top of Crossword main page

XI. Miscellaneous thoughts . . .

Take this detour to read a few more miscellaneous thoughts about crosswords, then return here.


Top of Crossword main page

XII. Conclusion

From playing crosswords I have enjoyed many thousands of small doses of pleasure.  If you are an experienced player, you know what I mean.  If you are a tyro, consider the benefits as I see them.

First, I've learned a lot.  Second, I always revel in playing with and presumably getting better at using my language, which is the tool, the translating mechanism, I use to think thoughts and to communicate with other humans.   And third, I've spent all those hours relaxing.

You too can enjoy all those small and large pleasures, meted out in small daily doses.  If you are a complete novice you will likely find your first several crosswords to be impossibly difficult; you simply will not be able to finish no matter how long you try.  And that's OK; it's to be expected.

But as you gain experience you will start to catch on.   Do hang in there.  You'll find after a couple weeks or months -- depending on how good you are with language to begin with and how many facts you already know and how good you are with references -- that you can finish some of the Monday New York Times puzzles.  I can almost guarantee you you'll remember the first time you finish one by yourself.

After a while longer you'll be finishing the Mondays most of the time and some of the Tuesdays.  Within six months or a year you'll be looking forward to the Sundays, because you'll enjoy how the big Sunday puzzles give the authors room to develop their themes more deeply, and by then you will have caught fully on to what to expect from a NY Times crossword.  You'll be crashing through those clues in no time, writing or typing in lots of them as fast as you can write or type, slowing down only to ponder over the tough ones, and taking in all the pleasure of the Cute Clues.

And maybe a year after that you'll actually be looking forward to the puzzles that are -- pound for pound -- the toughest of all, the fearsome and unpredictable Saturdays.  At that point you will have become permanently hooked.

And in the meantime you should not hesitate to see whether anyone you know plays crosswords or wants to.

Crosswords can definitely be a group activity, whether it's a group of ten co-workers in an office e-mailing or texting one another about the clues or a group of two people holding the puzzle between them and relaxing together in intimate contact.

I'm not sure I went down all the detours . . .   Take me back to the top crossword outline again.

Outline of
Crossword Puzzles  

I.  Introduction
II. Why and why not to play
     A. Why to play
     B. Why not to play
III.  "Rules" of crosswords . . .
     A. Rules for the grid
     B. Rules for clues, answers
IV. How to play well . . .
     . . .
     Cute Clues
V. Clue/Answer relationships
     A. Equivalence
     B. Set membership
     C. Fill in the blank
     D. Commonality
     E. Analogy
VI. Errors? . . .
     . . .
     Will Shortz replies
     New errors?
VII. Reference works . . .
VIII. Theme answers . . .
IX. Try writing your own . . .
X. Hardest Puzzle . . .
XI. Miscellaneous thoughts . . .
XII. Conclusion
Crossword links
Crossword Glossary


Garson Hampfield
video of July 29, 2008

Just go there and watch.


Crossword Glossary

It might be helpful to understand certain terms used by crossword constructors and their editors, so here's a glossary.  It's in alphabetical order.

American-style  American-style crosswords and British-style crosswords are both crosswords in a general sense, but they are significantly different and immediately distinguishable.  British crossword grids permit a lower ratio of whites to blacks, and they permit unchecked letters, which makes the constructor's job much easier, which it needs to be because another difference is that the all the clues are of the cryptic variety.
answer  Just what it sounds like, the answer to a clue, sometimes called an entry.  Note that yet another term is word, even though that "word" might comprise two or more literal words, for example, POPTHEQUESTION.

(Note that the convention among constructors, which I follow in this essay, is to render clues in double quotes and answers in all caps, viz., "Clue" = ANSWER.

cheater  In terms of grid layout, a cheater square is one (and its corresponding symmetrically opposite counterpart) that could have been a white but was set to black just to make the constructor's job easier, e.g., turning four five-letter words into four four-letter words.  Cheater squares are only venial sins, and they are frowned upon more by constructors themselves than by the playing public, 99% of whom don't notice they could have been lights.
checked  A white square in the grid that is crossed by an answer in the perpendicular direction is said to be checked.  Also called keyed.

Unchecked squares, sometimes called unches, are forbidden in all mainstream American-style crosswords.  One effect of this rule is to make the constrctor's job more difficult, but another is that the player gets two shots at a tough cell, not just one.

constructor  One half of the writing team, the originator of the puzzle, called a setter or a complier in Britain.  See editor.
corner  a particular area of the grid, even if not in one of the four actual corners, that is more or less bounded by any combination of borders and black squares and answers that cannot easily be changed such as theme answers.  A constructor might say, "I'm stuck in a tough corner."  For that matter, a player might say the same thing.
crosser  An answer that crosses another in the perpendicular direction.  For many years, until the editorship of Will Shortz starting in 1993, NYT puzzles did not eschew and sometimes even enouraged white squares in which both crossers were ungettable without resort to reference materials.

Now a rule followed by NYT puzzles and others is to avoid at almost all costs a white square in which both the Across and the Down answers are obscure or otherwise ungettable.  If you've placed the answer HHT ("Genetic blood vessel disorder: Abbr.") as an answer in your grid, you'll want to make sure all three crossers are easy.

Often an ungettable white square will arise where two proper nouns cross.  Because proper nouns can be spelled so many unpredictable ways, you might have to research one or both names to squish out the right letter.

Terms used by constructors for these problematic squares are amusing -- blind crossing, dead spot, knothole, and my favorite, coined by Barry Tunick in 1983, dirty double-crosser.
crosswordese  A type of answer constructors try to avoid is called crosswordese, which is an answer chosen primarily because it has such commonly used letters even though it's not spoken or written often.  Words with lots of vowels and common consonants such as N and R are useful for filling in a difficult part of the grid, but a few of those words are both obscure and tough to clue except too literally.  Crosswordese answers are, if you think about it, always short.  Examples are ERN, REN, ERNE and of course NENE, all of which are animals and three of which are birds.
cruciverbalist  One who constructs crossword puzzles, or one who plays them.  Cruci- is the root for the term "cross," as in crucifix, and verbal, of course, just means "word."
cryptic  A style of crossword less popular in the U.S. than in Britain, in which the clues always carry two meanings.  These can be fiendishly difficult until you master the special rules for how clues may be formed.  For more information go HERE.
daily commuter  A type of puzzle designed to appeal to daily commuters such as subway riders and train passengers, and anyone else who prefers the following characteristics:

1. It does not get harder throughout the week, which means it takes the same amount of time every day.  2. It is is easy, easier than a NYT Monday, which means many people can finish it without references. As an example, the clue in a March 23, 2010, Tribune Media Services (TMS) puzzle for the answer BTU is "British thermal unit: abbr."  3. It has no theme.  4. Daily commuter puzzles, sometimes just called commuters, are not known for their sparkling and witty clues and answers; some are pretty much entirely generated by a computer program that is deadly accurate but also deadly straightforward.

Just so I could write this paragraph for you I recently (October to November 2010) played 24 consecutive such dailies and recorded my times.  The average was 5 minutes 50 seconds, which means my average NYT Monday of 7:21 takes 27% longer, a whopping minute and a half more, which no matter how much faster or slower you are than I am means these 24 daily commuters were really easy compared to NYT Mondays.  To confirm they don't get harder throughout the week, the range of those 24 TMS times is from 4:56 to 6:31, which is only 1 minute 35 seconds, with a standard deviation of less than 27 seconds.

By contrast, the range of the averages for my NYT Mondays (7:21) through Saturdays (40:34) is over 33 minutes with a standard deviation of more than 13 minutes.

If you're so good a player that you regard these puzzles as a waste of time, try solving using only the Down clues.  I started doing this off and on around August of 2009 with these daily commuters, and I recently finished one in 5:52, which is 20% faster than my average Monday NYT time using all the clues. 

dark, light  Alternative terms for the black and white squares in the grid.
diagramless  A special type of crossword in which you must fill in not only the answers but also the little numbers in the upper-left corners of the white squares.  Furthermore, you must figure out where all the dark squares go and blacken them.  See here for more information.
difficulty  The degree of difficulty of a crossword puzzle, just as you would expect.  NYT crosswords and almost all other high-quality franchises start off easy on Monday and progress to really hard by Saturday.  The gradations from one day to the next are remarkably consistent, at least according to my own recorded times for several years.  So, a constructor might wonder, "Is this clue too difficult for an 'early in the week puzzle'?" or "Is this answer too difficult for a 'Thursday'?"

If you give it some thought you'll realize this is a brilliant marketing technique.  It means no matter what your level of experience, there'll be a puzzle just right for you.

editor  The other half of the writing team (see constructor above).  Crosswords are submitted by constructors to editors, who not only reject many but also edit the ones they accept.  The editor's role can be surprisingly large.  It is not uncommon for a good editor to change in part or even in whole 75% of the clues in a puzzle, and sometimes even the answers, and sometimes even the grid itself.  The principal reason is to moderate the clues so they match the desired degree of difficulty for that day of the week.  For example, a clue for a Saturday might be "Caesar's male pigeon," for Wednesday "Pioneer Pinta passenger," and for Monday "Ohio capital."

NYT crosswords name not only the editor but also the constructor, a practice I think should be implemented in all crosswords.
fill  Any answer that is not a theme answer.  Most of most crossword puzzles is fill, but that's not to say it is unimportant.  Good editors look for fill that is lively and interesting and fresh and entertaining and cute.  And yes, educational.
finger  A line of blacks pushing into the grid from an edge.  Constructors talk about shifting and lengthening fingers
FITB  An initialism for fill-in-the-blank.  See partial, below.
gimme  An answer that any player will get easily, e.g., "Lincoln's nickname" in three letters.  A constructor might make sure to use a gimme as a crosser for an otherwise ungettable answer.
grid  The rectilinear arrangement of black and white squares into which the verbal answers are written by the player, with some answers reading across from the left right and the others from the top down.  Many rules apply to the layout of grids, probably more than you've noticed if you're a novice player.
in the language  A phrase, a term of art, denoting whether an answer, or perhaps part or all of a clue, is well-enough known to appear in a crossword.  Here I'm not referring to merely obscure names such as Lespinasse or Lessines but rather to terms such as DIY and jump the shark and on the down-low.  If these terms are unknown to you then, for you, they are not "in the language," but that doesn't mean they aren't in the language for inclusion in a crossword.  Lots of terms and phrases you might not have heard are in the language, but typically you've heard of most of them or should have.

The term used by constructors for an answer that is not "in the language" is GREEN PAINT.  If you stumble into a discussion about crossword construction and the term GREEN PAINT is used, now you'll know what they're talking about.

(DIY means do-it-yourself, jump the shark refers to a particular EPISODE of the television show "Happy Days," and on the down-low is still imperfectly defined.)

novelty puzzle  Every so often a puzzle will contain an unusual element that expresses its theme, such as a rebus.  You'll find several ingenious examples of novelty, or specialty, puzzles here.  It is said that NYT novelty puzzles appear most often on Thursdays.
NYT  An abbreviation for The New York Times, the newspaper that has, arguably more than any other, set the standards for American-style crosswords.  Among the other highly respected puzzles are those in The Los Angeles Times (LAT), edited by Rich Norris; The New York Sun (NYS), no longer in publication as of September 30, 2008, edited by Peter Gordon; The Wall Street Journal (WSJ), edited by Mike Shenk; and The Washington Post (WP), edited by Fred Piscop.
pangrammatic  Denoting a grid in which all of the letters of the alphabet appear.  These are rarer than you might
think, what with there being little use for Z and Q and K and J, but sometimes constructors will tweak a couple of answers just to achieve a pangram.  Ninety-nine percent of pangrams go unnoticed by 98% of the most experienced players and 99.99% of the rest of us, so they are basically an inside joke among constructors and their editors, like a Boy Scout badge only your troop will ever know about.
partial  An answer that does not easily stand by itself.  An example is the answer INHOT, which cannot be clued in any way except as part of a longer phrase.  For example, you can clue it using the FITB clue " ___ water," so the answer INHOT is the moiety in the idiomatic phrase in hot water, i.e., it is a partial.

Another way to clue a partial is to use it as part of one or more idiomatic phrases.  The answer AGEOF cannot be clued in any way except as a partial, which might be clued like this: "It can precede innocence or consent."

According to the rules of The NYT, partials may be no longer than five letters (but there are almost always exceptions to rules).

player  The solver of the puzzle, the paying customer, the end user, the idealized person to whom any given crossword puzzle is directed
rebus  A type of novelty puzzle in which a square is to be filled in with a picture or more than one letter.  For more information go here.
revealer  In theme puzzles there is sometimes a clue and its answer that tell the player something about that theme.

For example, the clue at 45-Across is "What 20-Across and 36-Across are" and the answer to the clue at 45-Across is TWO TENTS.

As another example, the clue "Apt alternative title for this puzzle" gives the answer OMARKHAYYAM.

If you're trying to finish fast you'll want to solve the crossers of a revealer right away so you can learn something about the theme, which will probably move you along quicker, but which might also deaden the thrill of learning what the theme is without looking ahead

The moment when you figure out what the theme is is always pleasurable.  You can spot most revealers pretty easily by scanning the clues.

Revealers can be located anywhere in the grid, but they most often are a long theme answer or they're in the exact middle (8th row down and centered) or they're the last Across answer of the grid (15th row and right-most).

stack  A series of two or more long answers one on top of the other, thus forming a stack; also,
the same thing vertically, though they're rarer.  An 11-letter double stack is tough to write the Down answers for, so you can imagine how tough it is to write a 15-letter triple.  For an example that's just a little bit tougher go here.
symmetry  All mainstream crossword grids are symmetrical in some way.  More than 98% of NYT grids exhibit 180-degree symmetry,
meaning that if you rotate the grid 180 degrees in a clockwise (or counter-clockwise) direction, the darks and lights will be in the same locations. For more information go here.
theme answer  For puzzles that have themes, a theme answer is any entry that relates to that theme.  For more information go here.  Compare fill.
themeless  Any puzzle that has no theme answers (see immediately above).  NYT themelesses (which is a real word if you're a constructor or an editor) appear quite often on Fridays and even oftener on Saturdays.  Themeless NYT crosswords are challenging not only because they appear at the end of the week but also because they do not offer a theme you can grasp onto to apply to other answers.
wide-open  A wide-open grid is one in which there are few black cells, where there are big swaths of white cells without interruption.  The more wide-open is the grid, the more difficult is the puzzle to write, and probably to solve, so you'll see grids getting more wide-open as the week progresses from easy to hard.  If you're a novice player, it's a good idea to get in the habit of observing each new grid to see how wide-open it is.
word count  The number of answers (not necessarily literally the number of words) in the grid.  It is desirable for the word count to be low, other things being equal.  Why?

Think of the word count as the inverse of the average answer length.  The higher the word count, the shorter is each answer, and short answers are typically less interesting than long answers.  The supply of short answers that haven't been used a lot has been almost depleted, so new clues have to be invented for much-used words.

The maximum allowable word count for a NYT 15 X 15 grid is 78 unless it's a themeless, in which case it's 72.  (If you want to see an extreme example of achieving a low word count go here.)  The maximum for a 21 X 21, which is 96% bigger, is 140 words.

That said, note that constructors can be ingenious at creating cute clues for seemingly dull answers.  If you scan through the list of Cute Clues starting here you'll see lots of four- and even three-letter answers.





























Click  ?  at the top of any page to see all the nifty navigation aids, including Search Site and Site Map.








Outline of  
Crossword Puzzles  

I.  Introduction
II. Why and why not to play
     A. Why to play
     B. Why not to play
III.  "Rules" of crosswords . . .
     A. Rules for the grid
     B. Rules for clues, answers
IV. How to play well . . .
     . . .
     Cute Clues
V. Clue/Answer relationships
     A. Equivalence
     B. Set membership
     C. Fill in the blank
     D. Commonality
     E. Analogy
VI. Errors? . . .
     . . .
     Will Shortz replies
     New errors?
VII. Reference works . . .
VIII. Theme answers . . .
IX. Try writing your own . . .
X. Hardest Puzzle . . .
XI. Miscellaneous thoughts . . .
XII. Conclusion

Crossword links
Crossword glossary


B A R E L Y B A D  W E B  S I T E  
   Links      Cute Clues