motris ([info]motris) wrote,
@ 2006-06-26 23:15:00
Previous Entry  Add to memories!  Tell a Friend!  Next Entry
Kakuro Variant
So one of the comments I remembered reading post-qualifying test was that someone wished there was a magazine filled with just kakuro variations. I must admit that, unlike sudoku which can vary in many many ways to create a range of neat new puzzles, I think kakuro seems to not have as many routes available to it. That won't stop me from trying to offer some unique puzzles here over the coming months.



A lot of challenging variations already exist (I may not be using common names for these, but then I don't know that common names exist). Examples include "Zero Kakuro" where the use of 0 is permitted, "repeated digit kakuro" as in the Tuller/Rios books where each entry has a digit repeated twice in it (23 can be 9+9+5 or 8+8+7 or 7+7+9 but not 9+8+6), and "missing/incorrect sum kakuro" where there are either missing clues, or the clues are wrong by, say, one unit in either direction. Last week saw a good example of "plus/minus kakuro". There might be the possibility for a variation like "mod-10 kakuro" where the clues only reveal the ones-digit of the entry. I may try this idea sometime in the future.

One variant which occurred in a more complicated sudoku/kakuro form in the 2004 USPC was a "consecutive digit" puzzle where each and every instance of two digits vertically or horizontally adjacent that were consecutive would be marked.

This variant, particularly in the sudoku realm, is one of my favorites when it is "clean". Imagine a "Consecutive Kakuro" or "Consecutive Sudoku" where none of the digits ever touches a neighbor that is just one away. While it would still fit into the "Consecutive" category, it would be more accurately described as a "Non-consecutive" puzzle. The puzzle below is a Non-Consecutive Kakuro of exactly this type. For each row/column, the preceding number indicates the sum of that row/column. Only the digits 1 through 9 are used for each sum, and at most once in each sum. Further, no two adjacent digits can be consecutive numbers.

You will see that this adds some interesting constraints to kakuro (6 can no longer be a 3 digit clue as 1-2-3 has no entry that is valid; 30 must be 8-6-9-7 or 7-9-6-8; and a whole host of others for you to discover on your own). I tried to make the following puzzle have some easily solvable parts, some much harder parts, and some areas that are likely not possible until you solve others around them. In other words, I think it a fun, fair, and difficult challenge. It can be solved completely by logic alone and has a unique solution. Enjoy.






(Post a new comment)


[info]canadianpuzzler
2006-06-27 01:56 pm UTC (link)
As a constructor of variety sudoku (in addition to all the other wild and wonderful stuff you've seen from me ;-), and someone who enjoys the cross sums puzzle type in general, I'll be watching this puzzle series closely.

Craig K.

(Reply to this)


[info]lardarsegreg
2006-06-27 07:09 pm UTC (link)
Do you think that a non-consecutive killer could be interesting? I'd like to try one of those...

(Reply to this) (Thread)


[info]motris
2006-06-27 07:22 pm UTC (link)
Very interesting suggestion. I haven't seen one, but it makes a whole whole lot of sense, particularly given my kakuro writing experience last night.

(Reply to this) (Parent)(Thread)


[info]lardarsegreg
2006-06-27 07:39 pm UTC (link)
Well, I prefer killers to kakuro.

Of course, if you wanted to make kakuro interesting, you could always try a grid of hexagons. That could be fun...

(Reply to this) (Parent)


[info]lardarsegreg
2006-07-17 04:00 am UTC (link)
Small amount of closure for this idea:

http://sudoku.apinc.org/?p=124

(Reply to this) (Parent)(Thread)


[info]motris
2006-07-17 01:00 pm UTC (link)
thanks for the link. I actually saw it yesterday at sudoku.com's forum but have not gotten around to solving it.

(Reply to this) (Parent)

Thanks for the Puzzle
[info]gary3000
2006-06-28 02:16 am UTC (link)
Since I was the one who actually requested the Kakuro variants I figured I had better finish your puzzle before I posted again.

It was just as you described. Easy entries and then a couple of places where it gets a little harder. All in all a great challenge. I'm looking forward to your next puzzle.

I've also been watching Treasure Hunters. I really liked it when they revealed that the teams could have identified Mt. Rushmore by placing the reflective cylinder on the painting. That was a nice touch. Too bad none of the teams figured it out.

This week was less about the puzzles and more about putting the teams through a grueling workout with some nice scenery. My problems with the show are the endless product placement, the useless host and the feeling that the producers are taking a much greater hands on role than is being shown. It's still the only show I'm actually watching during the week (except for Deadwood on Sunday). I liked Hell's Kitchen in its first incarnation but I'm not particularly interested the second time around.

Fortunately I have the Venture Bros. DVD and the Criterion Collection edition of Dazed and Confused so I should be OK for the next few days.

(Reply to this) (Thread)

Re: Thanks for the Puzzle
[info]motris
2006-06-28 04:16 am UTC (link)
The problems with HK include the show not really changing since the last season, the contestants being way way less interesting and capable and quite noticeably so, and the "bleeping" risotto becoming more tedious than fun when Gordon uses the same put-downs time and again. I guess the only moment I've enjoyed (it lasted just a single second) was when they took the always awkward looking Jean Philippe and had him doing a silly string fight with the little kids during the "kindergarten lunch service". Unfortunately, it was way too brief a moment for what it was worth.

I'm glad to hear someone has successfully done the above kakuro challenge (although I trust my own test-solving). The next puzzle should be fun and maybe a little new, even if somewhat obvious as a variant. It may take some time to make it as clean as I want though so wait a couple days. I also have some other ideas in mind but I'm glad to get suggestions from people too for what they would like to see.

(Reply to this) (Parent)


[info]byronosaurusrex
2006-06-28 05:32 am UTC (link)
A very clever puzzle type, and well-executed; I took my first shot at it yesterday before the other comments went up, and started out in the upper middle portion, filled that out partly and decided that I would have the best chance of breaking out in the lower right...but then, I kept getting double sevens in the part of the line of 45 that goes into that section.

I had half a mind to comment that it seemed broken, but my better instincts kicked in and I decided to take the long way around, working from the top down and filling in the corner last. The logic was more unwieldy at times than was probably necessary as a result, but I got what I was looking for...the working corner fill I had overlooked.

I always have to do things the hard way, don't I? ;-)

P.S.: Congratulations!

(Reply to this) (Thread)


[info]motris
2006-06-28 02:07 pm UTC (link)
(spoilers as I partially describe solution paths)

The bottom right 11 cells are very easily solved without conflict so I'm not sure what your problem was. The lower left is much harder but can be gotten if you start at the intersection of 9 down and 20 across. It takes a couple logical deductions even after that and is a little tougher than I wanted but you can also completely solve those 11 cells independently. This makes the bottom middle trivial. You can also get the bottom middle without the bottom left if you just use constraints on the 29 down column. In this case, the 22 across at the bottom must have 9 or 5 in the middle, but putting a 5 in the middle makes the rest of the 29 unfillable. So, put the 9 in there. Then the position of the 5 is forced, and the 7 from the lower right gives the 8 and 7 placements as well. Again - there is some subtle new logic that is available in this type of puzzle and I tried, in my harder sections, to overuse this logic so just "kakuro-solving" skills would not give enough headway.

The top left and top middle are the other two breakout points. It seems you used them and worked clockwise through the puzzle, which is definitely a possibility.

(Reply to this) (Parent)(Thread)


[info]motris
2006-06-28 03:06 pm UTC (link)
Actually, I see pretty readily that the bottom middle doesn't need anything else to solve. My comment about the 8 and 7 placements is silly when you consider placing the 9 really forces the rest.

(Reply to this) (Parent)

Thanks!
(Anonymous)
2006-06-28 11:49 pm UTC (link)
Just wanted to thank you for a very enjoyable puzzle.. I also look forward to kakuro perhaps getting a similar kind of 'variation' treatment that sudoku has gotten since its popularity explosion. The non-consecutive rule adds a nice twist and some interesting new techniques, this is a really elegantly designed puzzle. Thanks again!

Andy

p.s. Congrats on the USPC, and on your US State sudoku, it's a work of art!

p.p.s. In case you hadn't seen it, djape's Daily Sudoku of June 26 is a nice non-consecutive sudoku - I'm a bit stuck on it actually. If it's not still featured on the "Daily Sudoku" page, it's "20060626_..." in his archive.

(Reply to this) (Thread)

Re: Thanks!
[info]motris
2006-06-29 03:48 am UTC (link)
Thank you for your comments. My goal as a puzzle writer and solver is always to find new challenges that, while close to familiar ground, force one to rethink the ways to solve them. It is the discovery process with a new puzzle/puzzle type that is the most rewarding part of solving puzzles to me. Yes, I've done hundreds (more likely thousands) of regular sudoku and kakuro but it is the extra variety that keeps them stimulating. I'm glad that you got this kind of experience with this "non-consecutive kakuro." For the moment, I'm trying to write up some other new ideas, but I may revisit this "non-consecutive summation" puzzle notion in a kakuro or killer sudoku form in the future.

Also, thanks for pointing out djape's recent puzzle - I stop by his site every week or so as it has lots of good offerings but hadn't seen this daily yet. I'll print it out and give it a try at home when I have some spare time.

(Reply to this) (Parent)(Thread)

Re: Thanks!
(Anonymous)
2009-05-17 08:57 pm UTC (link)
the middle left part is not quite unique, what i mean is there is 2 solutions that will both solve, still this puzzle is very fun!

(Reply to this) (Parent)


Create an Account
Forgot your login or password?
Login w/ OpenID
English • Español • Deutsch • Русский…