The New Sudoku Players' Forum

by **Guest** » Mon May 30, 2005 11:47 pm

I can't make any progress on this one... Is there a trick to get me started? This is a tough one...

*42 *3* ***
**5 *** ***
*** 9** **8
1** 8** ***
**3 *4* 2**
*** **7 **6
7** **6 ***
*** *** 46*
*** *2* 59*

by **Guest** » Tue May 31, 2005 12:01 am

Minny wrote:I can't make any progress on this one... Is there a trick to get me started? This is a tough one...

*42 *3* ***
**5 *** ***
*** 9** **8
1** 8** ***
**3 *4* 2**
*** **7 **6
7** **6 ***
*** *** 46*
*** *2* 59*

The 3 and 2 pairing in the centre block should get you started.

by **Guest** » Tue May 31, 2005 12:24 am

Hmmm... Okay, I'll have to think about this one a little bit. I'll let you know if I get anywhere. Thanks...

by **Minny** » Tue May 31, 2005 1:18 am

I may be doing something stupid, but the 2,3 pairs don't really get me anywhere. It leaves me with 2?3?, 1?9?5? and 1?5?6? cell pairs in the center box, but I can seem to get anywhere with those in the other boxes. I do see 4 cells with 1?5?9?, but that doesn't seem to get me anywhere either. I don't see any X-wing, starfish, or other exotic combos that can help me elimiante candidates. Do I have to guess and see where that leads me?

by **Minny** » Tue May 31, 2005 2:04 am

Cool. I found 2?4? pairs in row 3... Stuff is starting to open up and I should be able to solve it now. Thanks.

by **Guest** » Thu Jun 02, 2005 4:10 am

the way i do is i look at each 9 square box and if there is a number in it, i look at the other two boxes on the same line and if there is two of a number you can easly work out where the third number for that line goes

by **jordyn44** » Thu Jun 02, 2005 4:11 am

the way i do is i look at each 9 square box and if there is a number in it, i look at the other two boxes on the same line and if there is two of a number you can easly work out where the third number for that line goes

by **Guest** » Thu Jun 02, 2005 7:49 pm

This strikes me as one of the most intimidating puzzles I've seen in a while. There are no numbers to be given off the top and there are a large number of difficult logical steps to go through:

Start with the 24 pairs in row 3
Then find the 3,6,8,9 quadruplet in column 1 (using the restricted 5s in the middle box)
Then 1,9 pair in row 7 (using restricted 8s in the bottom-right box)
Then 1,7 pair in column 9, then 5,9 pair in same column

AND THEN, even after all this and then having filled in a bunch of easy stuff, you end up with

*42 63* **5
**5 27* ***
*** 954 *28

19* 862 *5*
**3 145 2*9
52* 397 **6

759 416 **2
231 589 467
4** 723 591

I could easily solve this with trial and error - there are very few cells left and many of them have only pairs of possibilities.

But, surely, there is something out there I have missed.

Help!

by **scrose** » Thu Jun 02, 2005 8:05 pm

This puzzle cannot be solved using logic. Trial and error must be used. This can be verified using rubylips' solver. You need to take a guess at placing a 1 in either r3c2 or r3c7.

by **Animator** » Thu Jun 02, 2005 8:43 pm

Trial and error is required... so either it is not a Pappocom puzzle, or the poster made a mistake when copying it.

The key to solving it (using T&E) somehow is column 7 and box 6...

Placing 8 in r6c7 will make the grid invalid, as in, r5c8 will be left with no candidates...

by **jimichou** » Mon Jun 06, 2005 9:04 am

This one is tough. It took my solver 43 trials and 473 rollbacks till solved. I really wonder who can do a lucky guess with only very few trials.
BTW, my solver is here.
http://forum.enjoysudoku.com/viewtopic.php?t=450

by **Minny** » Mon Jun 06, 2005 11:48 am

You are right -- there doesn't seem to be a way to finish using simple logic rules. I am not sure where the puzzle came from -- it was circulating around my office -- but the numbers are correct and it looks to be a published puzzle. I don't enjoy a puzzle that ends in trial and error...

by **rafowell** » Tue Jul 19, 2005 6:38 am

It has been suggested the puzzle was transcribed improperly. In support of this - note that if we place a "6" at r2c2, the problem becomes symmetric, and also solvable without guessing.

Given the popularity of symmetric puzzles, and the unpopularity of puzzles requiring guessing, I would suggest that the original puzzle had a "6" at r2c2, but somewhere along the line it was omitted.

by **simes** » Tue Jul 19, 2005 6:47 am

scrose wrote:This puzzle cannot be solved using logic. Trial and error must be used.

Sorry, but it can be solved without T&E (and without the possible missing 6 at r2c2). It needs three forcing chains though, which I don't think rubylips has implemented.

by **Nick70** » Tue Jul 19, 2005 7:01 am

simes wrote:Sorry, but it can be solved without T&E (and without the possible missing 6 at r2c2). It needs three forcing chains though, which I don't think rubylips has implemented.

Three forcing chains? Not at all. You just need a single XY-Wing.

The New Sudoku Players' Forum

Almost Impossible --- What is the trick?

Almost Impossible --- What is the trick?

Re: Almost Impossible --- What is the trick?

you can do it

Help! Hard Puzzle

Minny