The New Sudoku Players' Forum

by **999_Springs** » Fri Feb 09, 2007 9:52 pm

How hard is this puzzle? Is it possible to solve it without using forcing chains?

Code: Select all: .5.|.29|... .29|..1|.5. ...|..6|4.. ---+---+--- 1.5|..8|.6. 4..|53.|... .9.|...|..4 ---+---+--- ...|...|..7 ..8|7..|..1 21.|..3|9..

(This puzzle was created by me so I am not looking to solve it, only to see how hard it can get.)

by **ArkieTech** » Sat Feb 10, 2007 12:08 am

999_Springs said

How hard is this puzzle? Is it possible to solve it without using forcing chains?

A good puzzle but not a hard one. Found a swordfish, xy-wing and a turbot

Did not need chains.

dan

by **daj95376** » Sat Feb 10, 2007 12:21 am

ArkieTech,

Please post solution not needing chains. TIA!

by **re'born** » Sat Feb 10, 2007 1:19 am

If one uses Unique Rectangles, it isn't too bad. After basic techniques:

Code: Select all: *-----------------------------------------------------------* | 368 5 4 | 38 2 9 | 17* 17* 368 | | 368 2 9 | 4 7 1 | 38 5 368 | | 378 378 1 | 38 5 6 | 4 29 29 | |-------------------+-------------------+-------------------| | 1 37 5 | 9 4 8 | 237 6 23 | | 4 678- 26 | 5 3 27 | 178* 1789* 89# | | 378 9 23 | 1 6 27 | 5 378- 4 | |-------------------+-------------------+-------------------| | 9 4 36 | 2 1 5 | 368 38 7 | | 5 36 8 | 7 9 4 | 236 23 1 | | 2 1 7 | 6 8 3 | 9 4 5 | *-----------------------------------------------------------*

the potential deadly pattern in r15c78, along with r5c9 implies r5c2<>8 (as well as r6c8<>8). After a few more singles, we get to:

Code: Select all: *-----------------------------------------------------------* | 36* 5 4 | 8 2 9 | 17 17 36* | | 36* 2 9 | 4 7 1 | 38 5 368* | | 7 8 1 | 3 5 6 | 4 29 29 | |-------------------+-------------------+-------------------| | 1 37 5 | 9 4 8 | 237 6 23 | | 4 67 26 | 5 3 27 | 178 1789 89 | | 8 9 23 | 1 6 27 | 5 37 4 | |-------------------+-------------------+-------------------| | 9 4 36 | 2 1 5 | 368 38 7 | | 5 36 8 | 7 9 4 | 236 23 1 | | 2 1 7 | 6 8 3 | 9 4 5 | *-----------------------------------------------------------*

The potential deadly pattern in r12c19 implies that r2c9=8, and the puzzle then solves with singles.

by **udosuk** » Sat Feb 10, 2007 9:59 am

rep'nA, your first UR doesn't need the help from r5c9:

Code: Select all: *-----------------------------------------------------------* | 368 5 4 | 38 2 9 |*17 *17 368 | | 368 2 9 | 4 7 1 | 38 5 368 | | 378 #378 1 | 38 5 6 | 4 29 29 | |-------------------+-------------------+-------------------| | 1 @37 5 | 9 4 8 | 237 6 23 | | 4 678 26 | 5 3 27 |-178 -1789 89 | |#378 9 @23 | 1 6 27 | 5 @378 4 | |-------------------+-------------------+-------------------| | 9 4 36 | 2 1 5 | 368 38 7 | | 5 @36 8 | 7 9 4 | 236 -23 1 | | 2 1 7 | 6 8 3 | 9 4 5 | *-----------------------------------------------------------*

Strong link of 1s in r5c78 => UR-type 4 => r5c78<>7
=> Naked triple of {189} in r5c789 => r6c1=8 => r3c2=8

And then there is a turbot fish of 3s:
Strong links in r48c2 & r6c38 => r8c8<>3

And the rest are all singles...

by **re'born** » Sat Feb 10, 2007 12:02 pm

udosuk wrote:rep'nA, your first UR doesn't need the help from r5c9

...and yet to get my exclusions that is exactly what you use

udosuk wrote:=> Naked triple of {189} in r5c789 => r6c1=8 => r3c2=8

I believe my UR was originally called a Type 3 UR, but the terminology police can correct me.

by **Carcul** » Mon Feb 12, 2007 11:06 am

999_Springs wrote:How hard is this puzzle? Is it possible to solve it without using forcing chains?

Code: Select all: *---------------------------------------------------* | 368 5 4 | 38 2 9 | 17 17 368 | | 368 2 9 | 4 7 1 | 38 5 368 | | 378 378 1 | 38 5 6 | 4 29 29 | |----------------+----------------+-----------------| | 1 37 5 | 9 4 8 | 237 6 23 | | 4 678 26 | 5 3 27 | 178 1789 89 | | 378 9 23 | 1 6 27 | 5 378 4 | |----------------+----------------+-----------------| | 9 4 36 | 2 1 5 | 368 38 7 | | 5 36 8 | 7 9 4 | 236 23 1 | | 2 1 7 | 6 8 3 | 9 4 5 | *---------------------------------------------------*

[r2c7]-8-[r12c9]=8=[r12c1]-8-[r6c1]=8=[r6c8]-8-[r7c8]=8=[r7c7]-8-
-[r2c7], => r2c7<>8 solving the puzzle.

Carcul

by **ravel** » Mon Feb 12, 2007 1:04 pm

Carcul wrote:[r2c7]-8-[r12c9]=8=[r12c1]-8-[r6c1]=8=[r6c8]-8-[r7c8]=8=[r7c7]-8-
-[r2c7], => r2c7<>8 solving the puzzle.

I dont see the link [r12c9]=8=[r12c1], please explain.
In fact there is a nishio for 8 in r2c7, so the fishermen again are invited to identify the creature.

by **re'born** » Mon Feb 12, 2007 2:28 pm

ravel wrote:
Carcul wrote:[r2c7]-8-[r12c9]=8=[r12c1]-8-[r6c1]=8=[r6c8]-8-[r7c8]=8=[r7c7]-8-
-[r2c7], => r2c7<>8 solving the puzzle.
I dont see the link [r12c9]=8=[r12c1], please explain.

It follows easily from noting the potential deadly pattern in r12c19<36>. One of those four cells must be an 8.

Perhaps this is an even shorter solution using the same observation:

[r5c9]-8-[r12c9]=8=[r12c1]-8-[r6c1]=8=[r6c8]-8-[r5c9], => r5c9<>8, solving the puzzle.

by **ronk** » Mon Feb 12, 2007 2:42 pm

ravel wrote:In fact there is a nishio for 8 in r2c7, so the fishermen again are invited to identify the creature.

Based on just the 8s in Carcul's grid, the following is possible ...

Code: Select all: . . . | 8 . . | . . . . . . | . . . | 8 . . . 8 . | . . . | . . . ---------+----------+---------- . . . | . . 8 | . . . . . . | . . . | . . 8 8 . . | . . . | . . . ---------+----------+---------- . . . | . . . | . 8 . . . 8 | . . . | . . . . . . | . 8 . | . . .

... so the Nishio must be for different pencilmarks. Would you please post them?

by **ravel** » Mon Feb 12, 2007 3:14 pm

Thanks rep'nA and Ron, so i was both blind and wrong

by **999_Springs** » Tue Feb 13, 2007 12:57 pm

Does removing the 1 in r4c1 make any difference to the techniques used in solving the puzzle?

by **udosuk** » Tue Feb 13, 2007 1:20 pm

AFAIS, just a little bit of difference...

This is the sticking point without the clue at r4c1:

Code: Select all: *-----------------------------------------------------------* | 368 5 4 | 38 2 9 | 17 17 368 | | 368 2 9 | 4 7 1 | 38 5 368 | | 1378 378 13 | 38 5 6 | 4 29 29 | |-------------------+-------------------+-------------------| | 137 37 5 | 9 4 8 | 1237 6 23 | | 4 678 126 | 5 3 27 | 178 1789 89 | | 378 9 23 | 1 6 27 | 5 378 4 | |-------------------+-------------------+-------------------| | 9 4 36 | 2 1 5 | 368 38 7 | | 5 36 8 | 7 9 4 | 236 23 1 | | 2 1 7 | 6 8 3 | 9 4 5 | *-----------------------------------------------------------*

As you can see, all the URs and chains from rep'nA, Carcul still hold, but my UR-type 4 (making use of the strong link in r5c78) doesn't work anymore... (Thanks ravel for pointing that out...

)

Meanwhile ArkieTech still hasn't demonstrated to us his mysterious "swordfish, xy-wing and a turbot" solution... :?:

by **ravel** » Tue Feb 13, 2007 1:39 pm

udosuk wrote:As you can see, all the URs and chains from rep'nA, Carcul & me still hold...

To be more precise, the "Strong link of 1s in r5c78 => UR-type 4" doesn't exist here.

by **ravel** » Tue Feb 13, 2007 2:09 pm

999_Springs wrote:Does removing the 1 in r4c1 make any difference to the techniques used in solving the puzzle?

If you remove the 2 in r9c1 instead, you get a harder one:

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