The New Sudoku Players' Forum

by **wapati** » Wed Oct 18, 2006 4:45 pm

If you like finny fish, you'll enjoy this one.

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

by **Carcul** » Wed Oct 18, 2006 9:58 pm

Wapati wrote:If you like finny fish, you'll enjoy this one.

Code: Select all: *--------------------------------------------------------------------* | 2 59 359 | 1357 4 1357 | 6 8 57 | | 3578 58 4 | 3578 6 2 | 9 57 1 | | 578 1 6 | 578 78 9 | 2 3 4 | |-----------------------+----------------------+---------------------| | 189 3 289 | 14789 1789 6 | 5 124 79 | | 1589 4589 7 | 2 189 138 | 38 14 6 | | 6 2489 289 | 134789 5 1378 | 378 124 379 | |-----------------------+----------------------+---------------------| | 58 6 1 | 578 3 578 | 4 9 2 | | 3589 7 23589 | 589 289 4 | 1 6 35 | | 4 259 2359 | 6 129 15 | 37 57 8 | *--------------------------------------------------------------------*

[r8c4](-9-[r46c4])-9-[r8c1]=9=[X-Wing: r45c15]-9-[r4c9](-7-[r6c9]=7=
=[r7c6]-7-[r6c6])-7-[r6c7]=7=[r9c7]-7-[r9c8]-5-[r9c6]-1-[r56c6],

and the cells r56c67 are all and only populated by “3,8”. So, r8c4 cannot be “9” and the puzzle is solved.

Carcul

by RW » Wed Oct 18, 2006 10:40 pm

carcul wrote:[r8c4](-9-[r46c4])-9-[r8c1]=9=[X-Wing: r45c15]-9-[r4c9](-7-[r6c9]=7=
=[r7c6]-7-[r6c6])-7-[r6c7]=7=[r9c7]-7-[r9c8]-5-[r9c6]-1-[r56c6]

Once again there's too much sand for my truck... "(-7-[r6c9]=7= =[r7c6]-7-..." :?:

You started such a nice chain there, why not keep it simple:

[r8c4](-9-[r46c4])-9-[r8c1]=9=[X-Wing: r45c15]-9-[r4c9](-7-[r1c9])-7-[r4c5]=7=[r3c5]-7-[r1c46]

=> r8c4<>9

RW

by **wapati** » Thu Oct 19, 2006 4:37 am

This one needs at least 4.

Chain gang..I dunno.

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

by **wapati** » Thu Oct 19, 2006 5:26 am

So way nasty for me.

This is about my limit, nice one.

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

I have a question, I am sure here is wrong. <sigh>

I was wondering what the pattern is, for "good puzzles" when you take the givens and add a correct not given and see whether it is now an "easy".

I didn't say that well.

I wonder whether most "good" puzzles have any, more than 1, lots of squares, that if filled in render the puzzle easy.

I'd like to know if there are more than one squares that may be critical.

by **wapati** » Thu Oct 19, 2006 5:53 am

WOW! This one is past my limit.

I sort-of know Sashimi X.

Hrrmmm. Here is a borrible. Happy Dancing!

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

by **ravel** » Thu Oct 19, 2006 8:43 am

wapati wrote:I have a question, I am sure here is wrong. <sigh>

I was wondering what the pattern is, for "good puzzles" when you take the givens and add a correct not given and see whether it is now an "easy".

You are talking about "backdoor" cells. It depends on what you define as "easy", what a backdoor is. With the basics i have implemented (box/line, tuples, x-wing, UR1) your puzzle has 15.
When you allow the Simple Sudoku techniques (box/line, tuples, x-,xy-wing, swordfish, multicoloring) only one puzzle is known, that does not have a single backdoor (nr 210 in my old list):

33: Ocean #10/M21/D21 (ER 9.8, gsfr 99650, suexr 977, RMS 8)
.....1..2.3..4..5.6..2.......5.....3.7..8..4.2.....9..9....4.......5..7...41..6..

by RW » Thu Oct 19, 2006 9:26 am

wapati wrote:This one needs at least 4.

Chain gang..I dunno.
Code: Select all
. . . | . 7 . | . 1 . . 7 . | 1 . . | . 5 6 3 . . | . . . | 8 . . --------------------- . . . | . 4 . | . . . . 3 . | 6 9 1 | . . 7 6 . 1 | . 3 . | . 4 . --------------------- . 2 9 | 3 . . | . . . . . 7 | . 8 . | . 2 . 1 . . | 2 . . | 4 . .

4 of what?

Code: Select all: *-----------------------------------------------------------------------------* | 24589 458 24568 | 4589 7 345689 | 239 1 249 | | 489 7 48 | 1 2 3489 | 39 5 6 | | 3 1 2456 | 459 56 4569 | 8 7 249 | |-------------------------+-------------------------+-------------------------| | 7 9 258 | 58 4 258 | 6 3 1 | | 245 3 245 | 6 9 1 | 25 8 7 | | 6 58 1 | 7 3 258 | 259 4 29 | |-------------------------+-------------------------+-------------------------| | 458 2 9 | 3 1 45 | 7 6 58 | | 45 456 7 | 459 8 4569 | 1 2 3 | | 1 568 3 | 2 56 7 | 4 9 58 | *-----------------------------------------------------------------------------*

[r4c3](-8-[r6c2]-5-[r1c2])-8-[r2c3]-4-[r1c2]-8-[r1c4]=8=[r4c4]-8-[r4c3]

=> r4c3<>8

Sorry if I spoiled your fun carcul...

RW

by **wapati** » Thu Oct 19, 2006 4:22 pm

RW wrote:4 of what?
[RW

I would need x, xy, xyz, swordish and ER/Turbot to complete it.

Your way looks shorter.

by RW » Thu Oct 19, 2006 4:29 pm

wapati wrote:I would need x, xy, xyz, swordish and ER/Turbot to complete it.

No, you wouldn't. ER + XY-wing is enough. You shouldn't trust a computer program to tell you what is required, it tells you what is required if you do the techniques in a specific order.

RW

by **Carcul** » Thu Oct 19, 2006 9:20 pm

Wapati wrote:This is about my limit, nice one.

Code: Select all: *-----------------------------------------------------* | 456 9 2 | 1456 456 16 | 7 8 3 | | 1 56 3 | 256 7 8 | 259 259 4 | | 7 458 48 | 9 245 3 | 1 25 6 | |-----------------+-----------------+-----------------| | 46 2 9 | 7 16 5 | 3 14 8 | | 8 1 46 | 26 3 269 | 2459 7 259 | | 3 7 5 | 8 129 4 | 6 129 29 | |-----------------+-----------------+-----------------| | 9 4568 468 | 3 2456 7 | 245 245 1 | | 2 456 1 | 456 8 69 | 459 3 7 | | 45 3 7 | 1245 2459 129 | 8 6 259 | *-----------------------------------------------------*

We have r5c7=9 or r9c9=5. But:

[r5c7]=9|5=[r9c9]-5-[r5c9]=5=[r5c7],

which implies r5c7<>2,4 and the puzzle is solved (using later a trivial ATILA step).

Carcul

by **Carcul** » Thu Oct 19, 2006 9:40 pm

Wapati wrote:WOW! This one is past my limit.(...) Hrrmmm. Here is a borrible. Happy Dancing!

Code: Select all: *---------------------------------------------------------------------* | 2345 2459 59 | 3489 245 345689 | 7 1469 1469 | | 34 8 6 | 7 1 349 | 5 2 49 | | 1 2459 7 | 49 245 4569 | 3 469 8 | |---------------------+------------------------+----------------------| | 468 46 2 | 5 3 7 | 1 4689 469 | | 9 3 1 | 48 6 48 | 2 5 7 | | 456 7 58 | 1 9 2 | 48 46 3 | |---------------------+------------------------+----------------------| | 258 125 4 | 39 7 39 | 6 18 15 | | 56 156 3 | 2 8 45 | 9 7 145 | | 7 59 589 | 6 45 1 | 48 3 2 | *---------------------------------------------------------------------*

[r6c3]-8-[r9c3]=8=[r7c1]=2=[r7c2]-2-[r1c2]-4-[r4c2]-6-[r24c9]
-4-[r8c9]=4=[r9c7]=8=[r6c7]-8-[r6c3],

and so r6c3<>8 solving the puzzle.

Carcul

by **wapati** » Fri Oct 20, 2006 5:12 am

RW wrote:
wapati wrote:I would need x, xy, xyz, swordish and ER/Turbot to complete it.

No, you wouldn't. ER + XY-wing is enough. You shouldn't trust a computer program to tell you what is required, it tells you what is required if you do the techniques in a specific order.

RW

Sheesh, next you will tell me that "windows" isn't safe on the Internet. :?:

by **wapati** » Fri Oct 20, 2006 5:18 am

This is hard for me, many methods required.

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

by RW » Fri Oct 20, 2006 6:03 am

wapati wrote:Sheesh, next you will tell me that "windows" isn't safe on the Internet.

Well, if you write two post that claims that it is, then maybe I would...

RW

The New Sudoku Players' Forum

Fun puzzle. So many methods to solve. No CHAINS required.