Patternless Crosswords

[Omniscient]

I know there's a pretty large number of crossword fans, and authors, on the Dope and there's been some time spent sharing online outlets for the puzzles, but I haven't heard if anyone's up for the patternless, or diagramless, varieties.

The Chicago Sun-Times always has one which I enjoy doing, but sadly there's no online version out there that I know of. Personally I think it's more fun than the New York Times crossword.

[K364]

I always do the weekly diagramless... I find that the extra element of logic required to figure out the pattern is challenging and very satisfying when you solve it. When I first started these, it took me the better part of the week and sometimes I didn't finish. Now, after several years I usually finish one off in an hour or two.

[melondeca]

I love crossword puzzles and I would be interested in seeing a patternless puzzle. I don't think I have ever seen one.

[Sarahfeena]

Originally Posted by melondeca

I love crossword puzzles and I would be interested in seeing a patternless puzzle. I don't think I have ever seen one.

If you want to try a diagramless puzzle out, you can find good ones in Dell's Crossword magazines...they usually include a few diagramless in each issue.

[Fern Forest]

If I'm given the starting square I'll usually do pretty good but without one it seems like you'd need to writehalf the answers by the clues before you could figure out where to start.

Any help you can give on that?

The ones where it's not in the top line particularly baffle me.

[K364]

Originally Posted by Fern Forest

If I'm given the starting square I'll usually do pretty good but without one it seems like you'd need to writehalf the answers by the clues before you could figure out where to start.

Any help you can give on that?

The ones where it's not in the top line particularly baffle me.

Yes, that's certainly the hardest part. BTW, the first clue HAS to be in the top line, but probably not in the first square of the top line. Diagramless designers have some sneaky tricks to make the first half anything but easy. Because of standard crossword symmetry, the second half is much easier (and a bit anti-climactic).

What you usually do is solve in "clumps"... in other words you find sections of the crossword, but do not know how they fit together. As you work down, the sections start to link up. To annoy you, the designers may start new sections as you work down, so things may get more complex instead of simpler.

If you can't get the clues in the first sections, it is almost impossible to get started.

There are a number of other "tricks" about diagramless... see this older post

[Fern Forest]

Originally Posted by K364

Yes, that's certainly the hardest part. BTW, the first clue HAS to be in the top line

Cool, thanks for the links.

I guess it is there somewhere. The one I recalled was a pattern like an X only the upper left and lower right didn't reach to the edge so 1. was in the upper right.


What's silly is I love carefully shading in all the squares I know to be filled. As it so happens I do mazes the same way, carefully filling in all the dead ends. I guess it's no surprise that Paint By Numbers (not paint by number) are my favorite puzzles.

[Omniscient]

Originally Posted by Fern Forest

If I'm given the starting square I'll usually do pretty good but without one it seems like you'd need to writehalf the answers by the clues before you could figure out where to start.

Any help you can give on that?

The ones where it's not in the top line particularly baffle me.

It's not really that difficult if you ask me. All you need to do is look at the numbering pattern for the clues.

twickster gives a good explanation in the thread linked in the previous reply here:

Look at the ACROSS words to get the length of your first word. For instance, if the Across words are 1, 6, 9, and 13 Across that means 1 Across is five letters (with the numbers 1, 2, 3, 4, and 5 going in those boxes for the Down words), 6 Across is three letters (6, 7, 8), 9 Across is four letters (9, 10, 11, 12).

Thats essentially the crux of it. You basically know what the length of the first answers are based on the numbering. Tougher puzzles trick you by not starting in the upper left corner but instead of starting one or two spaces over, but the numbering of the down clues usually give that away. Ditto for puzzles having blanks between answers of more than one square.

Also, once you've learned what the symmetry is, solving those top few rows give you the layout for the bottom few rows. I tend to work the puzzle this way and finish the clues in the middle of the grid last.

Incidentally, I never write answers to clues off to the side, if I can't place them in the grid I don;t answer them. It's tantamount to cheating for me, sorta like Sudoku puzzles where you're required to guess and use process of elimination instead of logic.

[kaylasdad99]

Originally Posted by Omniscient

Incidentally, I never write answers to clues off to the side, if I can't place them in the grid I don;t answer them. It's tantamount to cheating for me, sorta like Sudoku puzzles where you're required to guess and use process of elimination instead of logic.

Anything to boost the challenge level, eh?

Do you do them on a typewriter, too?

[Omniscient]

Originally Posted by kaylasdad99

Anything to boost the challenge level, eh?

Do you do them on a typewriter, too?

No, but if I ever see a half done one in pencil and not using all caps I get all twitchy and anxious.

[Trunk]

The New York Times rotates a diagramless into it's supplementary puzzle section every month or so.

I find them enjoyable and challenging, and it took me a couple completely failed attempts before I ever got past the first line. Once I figured out some tricks, like the twickster quote above, I started rolling.

Here's a question for you all: in the NYT one, they make you turn the page to find out which square it starts in. I always thought, "that's funny. why do I need to turn the page?" But from reading this thread. . .can I assume that some people do them without knowing which square they start in? Is that just an additional challenge to people? It never occured to me to do that.

[twickster]

Originally Posted by Trunk

Here's a question for you all: in the NYT one, they make you turn the page to find out which square it starts in. I always thought, "that's funny. why do I need to turn the page?" But from reading this thread. . .can I assume that some people do them without knowing which square they start in? Is that just an additional challenge to people? It never occured to me to do that.

My dad used to do them on graph paper, so no, you don't need a starting hint for those. It's traditional to provide the hint on a different page for that reason.

[Omniscient]

Originally Posted by Trunk

Here's a question for you all: in the NYT one, they make you turn the page to find out which square it starts in. I always thought, "that's funny. why do I need to turn the page?" But from reading this thread. . .can I assume that some people do them without knowing which square they start in? Is that just an additional challenge to people? It never occured to me to do that.

The Chicago Sun-Times one never gives you the starting square, only a blank grid and clues, and I do it whenever I have a paper handy without any issue. As noted above, finding the starting spot really isn't very challenging.

[Fern Forest]

Originally Posted by Omniscient

The Chicago Sun-Times one never gives you the starting square, only a blank grid and clues, and I do it whenever I have a paper handy without any issue. As noted above, finding the starting spot really isn't very challenging.

I guess it depends. If it's 15 across on the top and 1 through 13 are all down answers and 1, 5 and 10 are across then those I can get.

But sometimes they get really tricky. Like that lopsided X I mentioned. If I recall in that one the upper left piece started in the 3rd row. So across and down went like this

A: 1, 5, 6, 10
D: 1, 2, 3, 4, 6, 7, 8, 9

There may have been a bow too so the first across touched neather the right edge or left edge which would make the clues different.

But I pretty much look up the starting square automatically since I started. Discouraged in the beginning I guess. Tomorrow I'll try not to. I keep my puzzle books at work for my breaks.

[Omniscient]

Originally Posted by Fern Forest

I guess it depends. If it's 15 across on the top and 1 through 13 are all down answers and 1, 5 and 10 are across then those I can get.

But sometimes they get really tricky. Like that lopsided X I mentioned. If I recall in that one the upper left piece started in the 3rd row. So across and down went like this

A: 1, 5, 6, 10
D: 1, 2, 3, 4, 6, 7, 8, 9

There may have been a bow too so the first across touched neather the right edge or left edge which would make the clues different.

But I pretty much look up the starting square automatically since I started. Discouraged in the beginning I guess. Tomorrow I'll try not to. I keep my puzzle books at work for my breaks.

I'm having trouble picturing this, but the numbering you have look like it's impossible. there's no way for 5 and 6 to both be acrosses. It defies the rules of crosswords.

[K364]

It could look like this:

xxx1234xxx
xxx5   xxx
x67    89x
xt       x

or

xxx1234xxx
xxx5   xxx
678    9xxx
t       xxx

or

xxx1234xxx
xxx5   xxx
xx6    789x
xxt       x

or

xxx1234xxxxxxxxxxxx
xxx5   xxx6789x
xxxt   xxxe   x

[Voyager]

A quickl google revealed a Dell book of diagramless puzzles here. There are also 2 NYT diagramless books, both of which I have done.

In the good old days the Times ran two diagramless puzzles, without the grid.

For grids, not all puzzles have symmetry. My favorites make pictures - these often don't. The nastiest ones have the first answer on the right side of the grid, and the second one row down on the left. The pattern looks standard for the first few clues, but then you run into a long answer for an even clue.

I figure out for sure where the first answers go when you get to a row that totally fills the grid from left to right. Then you can start inserting things. I always do newspaper puzzzles with extra paper - in a book I do look at the first clue just to keep from carrying a pad along with the book.

[Fern Forest]

Originally Posted by Omniscient

I'm having trouble picturing this, but the numbering you have look like it's impossible. there's no way for 5 and 6 to both be acrosses. It defies the rules of crosswords.

K364 almost got it, it was like this

XXXXXXXXXXX1234
XXXXXXXXXXX5
XX6789XXXXX10
XX11 XXXXXX12

[Fern Forest]

Given a diagramless with symmetry, could you find the pattern with just the clue numbers? It seems like you should be able to.

I worked on it today but maybe I'm not good enough of a crossword puzzle solver, or the clues were too hard for me, to get enough to start filling things in.

[pulykamell]

Originally Posted by Omniscient

I'm having trouble picturing this, but the numbering you have look like it's impossible. there's no way for 5 and 6 to both be acrosses. It defies the rules of crosswords.

It's possible, but it's not like the patternless crosswords in the Sun-Times. I love the patternless there (as opposed to the patternlesses in GAMES magazine, which are all sorts of weird and don't follow normal crossword conventions.) They're not terribly tricky, they follow standard crossword symmetry and, overall, are pretty easy. On a Monday, I can generally solve all three crosswords (the NY Times, patternless, and Sun-Times crossword) in the Sun-Times on the El ride to work (about 35-45 minutes). The patternless is usually easier, I find, than the NY Times puzzle. I rarely solve the Times Thursday and Friday puzzles, but I can almost always get the patternless.

They look difficult to the uninitiated, but patternless crosswords are fairly straightforward once you do several of them and understand crossword construction conventions (diagonal symmetry, no words of under three letters, how numbering works, etc....)