The New Sudoku Players' Forum

by **spirit_fantasy** » Fri Apr 07, 2006 10:19 pm

Code: Select all: 3 . 4 | . . . | . 1 . 5 . 7 | . 4 . | . 2 . . 9 8 | . . 3 | 6 4 . -------+-------+------ 9 3 1 | . 5 . | 2 6 4 4 8 2 | . . . | 5 7 1 7 5 6 | . 2 . | . 9 . -------+-------+------ 8 . 9 | 6 . 5 | 4 3 . . 7 3 | . 8 . | . 5 6 6 4 5 | . . . | . 8 . 3 26 4 | 25789 69 26789 | 789 1 5789 5 16 7 | 189 4 1689 | 389 2 389 12 9 8 | 1257 17 3 | 6 4 57 -------------+-------------+------------- 9 3 1 | 87 5 87 | 2 6 4 4 8 2 | 39 369 69 | 5 7 1 7 5 6 | 14 2 14 | 38 9 38 -------------+-------------+------------- 8 12 9 | 6 17 5 | 4 3 27 12 7 3 | 1249 8 1249 | 19 5 6 6 4 5 | 12379 39 1279 | 179 8 279

btw, how can i arrange it to be a better format?

by **Ruud** » Fri Apr 07, 2006 10:28 pm

Hi,

there is a beautiful Sashimi Swordfish with a maximum yield of 4 candidates in this puzzle:

The blue highlighted cells are the fish, the red cells are the cells where digit 1 can be eliminated as a candidate.

That does break the puzzle, but keep this picture in mind. It's very rare.

Can you post the original puzzle, please?

TIA, Ruud.

by **Sped** » Fri Apr 07, 2006 11:12 pm

I do not understand Sashimi Swordfish.

But.. if you color on ones you can exclude 3 or more of them right off the bat and it's all singles from there.

by **ronk** » Fri Apr 07, 2006 11:32 pm

Ruud wrote:there is a beautiful Sashimi Swordfish with a maximum yield of 4 candidates in this puzzle:
................
The blue highlighted cells are the fish, the red cells are the cells where digit 1 can be eliminated as a candidate.

Isn't there supposed to be a fin cell or two somewhere?

by **emm** » Fri Apr 07, 2006 11:33 pm

spirit_fantasy wrote:btw, how can i arrange it to be a better format?

Use the preview button to view the grid and align the columns so that they look right in the preview screen before you submit. They won't look aligned in the message screen.

Code: Select all: 3 26 4 | 25789 69 26789 | 789 1 5789 5 16 7 | 189 4 1689 | 389 2 389 12* 9 8 | 1257 17 3 | 6 4 57 -----------+-------------------+---------------- 9 3 1 | 87 5 87 | 2 6 4 4 8 2 | 39 369 69 | 5 7 1 7 5 6 | 14 2 14 | 38 9 38 -----------+-------------------+---------------- 8 12 9 | 6 17 5 | 4 3 27 12 7 3 | 1249 8 1249 | 19 5 6 6 4 5 | 12379 39 1279 | 179 8 279

Here's an XYchain linking bivalue cells from r8c1
1-r8c1-2-r7c2-1-r7c5-7-r3c5-1
A chain that starts and ends with the same number eliminates that number from all other cells that 'see' the start and end ie r3c1<>1

After that you need only singles to solve.

by **Ruud** » Sat Apr 08, 2006 12:24 am

ronk wrote:Isn't there supposed to be a fin cell or two somewhere?

Oh yes there is! Can't you see it? R7C5 is the fin.

If either R8C5 or R9C5 were candidates, it would not be a Sashimi.

Ruud.

by **Myth Jellies** » Sat Apr 08, 2006 12:24 am

Code: Select all: 3 26 4 | 25789 69 26789 | 789 1 5789 5 16 7 | 189 4 1689 | 389 2 389 *12 9 8 | 1257 *17 3 |*6 4 57 -----------+-------------------+---------------- 9 3 1 | 87 5 87 | 2 6 4 4 8 2 | 39 369 69 | 5 7 1 7 5 6 | 14 2 14 | 38 9 38 -----------+-------------------+---------------- 8 12 9 | 6 #17 5 | 4 3 27 *12 7 3 |-1249 *8 -1249 |*19 5 6 *6 4 5 |-12379 *39 -1279 |*179 8 279

The fin is in r7c5.

Code: Select all: 3 26 4 | 25789 69 26789 | 789 1 5789 5 *16 7 |*189 4 *1689 | 389 2 389 12 9 8 | 1257 17 3 | 6 4 57 -----------+-------------------+---------------- 9 3 1 | 87 5 87 | 2 6 4 4 8 2 | 39 369 69 | 5 7 1 7 *5 6 |*14 2 *14 | 38 9 38 -----------+-------------------+---------------- 8 *12 9 |*6 #17 *5 | 4 3 27 12 7 3 |-1249 8 -1249 | 19 5 6 6 4 5 |-12379 39 -1279 | 179 8 279

edit-some might see this sashimi swordfish in rows more easily.

by **Sped** » Sat Apr 08, 2006 12:38 am

Code: Select all: 3 26 4 | 25789 69 26789 | 789 1 5789 5 16 7 | 189 4 1689 | 389 2 389 12 9 8 | 1257 17 3 | 6 4 57 -----------+-------------------+---------------- 9 3 1 | 87 5 87 | 2 6 4 4 8 2 | 39 369 69 | 5 7 1 7 5 6 | 14 2 14 | 38 9 38 -----------+-------------------+---------------- 8 12 9 | 6 17 5 | 4 3 27 12 7 3 | 1249 8 1249 | 19 5 6 6 4 5 | 12379 39 1279 | 179 8 279

Here's another XY chain. It's not as good as emm's because it's longer and only leads to one exclusion (the 9 at r1c7), which doesn't quite solve the puzzle:

9-(r1c5)-6-(r1c2)-2-(r3c1)-1-(r3c5)-7-(r7c5)-1-(r7c2)-2-(r8c1)-1-(r8c7)-9

I'm telling you folks.. there are XY chains everywhere. This is an under utilized technique.

by **ronk** » Sat Apr 08, 2006 1:25 am

Ruud wrote:
ronk wrote:Isn't there supposed to be a fin cell or two somewhere?

Oh yes there is! Can't you see it? R7C5 is the fin.

I didn't look close enough, as I remembered your software (or you) coloring the fin cell(s) a third color. So much for the memory, I guess.

by **emm** » Sat Apr 08, 2006 3:38 am

Sped wrote:there are XY chains everywhere. This is an under utilized technique.

Sped - I shortened the chain - I thought you might have got there before me.

I’m pleased that you have shone the light on XY chains.
I find them easier than nice loops since you don't have to worry about strong / weak / continous / discontinuous etc

spirit_fantasy - I hope you don't feel swamped by all these replies. Here's another one.

1s are obviously the target candidate in this puzzle. Here’s my other favourite – strong links eliminate these *

Code: Select all: . . . | . . . | . . . . 1 . | . . . | . . . * | . | . 1 . | . . . ---|---+---|---+------ . | . | . | . | . . . . | . | . | . | . . . . | . | . | . | . . . ---|---+---|---+------ . 1 . | . 1 . | . . . . . . | . . . | . . . . . . | . . . | . . . . . . | . . . | . . . . . . | . . . | . . . 1 . . | . * . | . . . -|-----+-------+------ | . . | . . . | . . . | . . | . . . | . . . | . . | . . . | . . . -|-----+-------+------ | 1---|---1 . | . . . 1 . . | . . . | . . . . . . | . . . | . . . . . . | . . . | . . . . . . | . . . | . . . 1 . . | . 1 . | . . . -|-----+---|---+------ | . . | . | . | . . . | . . | . | . | . . . | . . | . | . | . . . -|-----+---|---+------ | * . | . 1 . | . . . 1 . . | . . . | . . . . . . | . . . | . . .

by **Kent** » Sat Apr 08, 2006 4:21 am

Code: Select all: 3 26 4 | 25789 69 26789 | 789 1 5789 5 16 7 | 189 4 1689 | 389 2 389 12 9 8 | 1257 17* 3 | 6 4 57 -----------+-------------------+---------------- 9 3 1 | 87 5 87 | 2 6 4 4 8 2 | 39 369 69 | 5 7 1 7 5 6 | 14 2 14 | 38 9 38 -----------+-------------------+---------------- 8 12 9 | 6 17 5 | 4 3 27 12 7 3 | 1249 8 1249 | 19 5 6 6 4 5 | 12379 39 1279 | 179 8 279

Can I say that the start of the chain as r3c1 and the ending of the chain is r7c5 therefore eliminating 1 at r3c5??

by **emm** » Sat Apr 08, 2006 4:36 am

Yes. 1-r3c1-2-r8c1-1-r7c2-2-r7c9-7-r7c5-1

by **spirit_fantasy** » Sat Apr 08, 2006 5:51 am

Ruud wrote:
Can you post the original puzzle, please?

TIA, Ruud.

i reli hv no idea wt u guys talkin abt....probably im jus a newbie...but anyway thx fo solving it...

this is the original puzzle..

Code: Select all: . . . | . . . | . 1 . 5 . . | . 4 . | . 2 . . 9 8 | . . 3 | 6 . . -------+-------+------ . 3 . | . 5 . | 2 . 4 4 8 . | . . . | . 7 1 7 . 6 | . 2 . | . 9 . -------+-------+------ . . 9 | 6 . . | 4 3 . . 7 . | . 8 . | . . 6 . 4 . | . . . | . . .

by **Sped** » Sat Apr 08, 2006 12:37 pm

spirit_fantasy wrote:
i reli hv no idea wt u guys talkin abt....probably im jus a newbie...but anyway thx fo solving it...

Heh. We got kind of carried away with the mutant fish and the chains. But that's what happens when you post an interesting puzzle.

The easiest to understand method that cracks this puzzle, I think, is colors. A good explanation of colors (a very useful technique) can be found here, about 80% of the way down the page:

http://angusj.com/sudoku/hints.php

by **ronk** » Sat Apr 08, 2006 12:54 pm

Sped wrote:The easiest to understand method that cracks this puzzle, I think, is colors.

I agree ... and think you should illustrate your earlier statement ... "if you color on ones you can exclude 3 or more of them right off the bat and it's all singles from there."

The New Sudoku Players' Forum

really need help for this one!!!

really need help for this one!!!