The New Sudoku Players' Forum

[mollwollfumble]

Attached is a link to "The Devil's Puzzle Book" that I created containing "ambiguous and nearly impossible puzzles".
http://freepages.misc.rootsweb.ancestry.com/~hallsofjamaica/The%20devils%20puzzle%20book.pdf

This puzzle book contains four Sudokus, three of which have been deliberately designed to defeat Sudoku-solving computer programs. Have fun, and don't give up too soon. (I could give you a hint to make the solution of all three much easier, but I won't).

[enxio27]

Your first and last sudoku grids are invalid puzzles. I'm not going to waste my time on the other two.

[m_b_metcalf]

Your first and last sudoku grids are invalid puzzles. I'm not going to waste my time on the other two.

In particular:

Code: Select all

Puzzle 5
 . . . . . . . . .
 3 . . . 9 1 4 . 6
 . . 7 3 . 4 2 . 5
 . 9 . 4 . 5 6 . 7
 7 4 . . 8 . . 5 1
 5 . 1 9 . 7 . 3 .
 2 . 5 7 . 9 1 . .
 1 . 6 2 3 . . . 8
 . . . . . . . . .
No. of givens =  35

brute found 0 solution(s) in    0.00 seconds.


Code: Select all


Puzzle 17
 . . . . . . . . 1
 . . . . . . . 4 8
 . . 7 . . 3 . . .
 . 6 . . 8 . . . .
 . . 2 . . . 7 . .
 . . . . 7 . . 9 .
 . . . 5 . . 6 . .
 4 8 . . . . . . .
 1 . . . . . . . .
No. of givens =  16

brute found 110552 solution(s) in    1.01 seconds. Last was:
  8  2  3  4  5  6  9  7  1
  6  5  1  7  9  2  3  4  8
  9  4  7  8  1  3  2  6  5
  7  6  4  2  8  9  1  5  3
  5  9  2  3  4  1  7  8  6
  3  1  8  6  7  5  4  9  2
  2  7  9  5  3  8  6  1  4
  4  8  6  1  2  7  5  3  9
  1  3  5  9  6  4  8  2  7

Common clues (18):
  .  .  .  .  .  .  .  7  1
  .  .  .  .  .  .  .  4  8
  .  .  7  .  .  3  .  .  .
  7  6  .  .  8  .  .  .  .
  .  .  2  .  .  .  7  .  .
  .  .  .  .  7  .  .  9  .
  .  .  .  5  .  .  6  .  .
  4  8  .  .  .  .  .  .  .
  1  .  .  .  .  .  .  .  .

Puzzle 17 has too few clues to be valid, and interpreted as an X-sudoku has no solution.

Mike Metcalf

[tarek]

Blow your trumpet until you fall on it .....

There is a good reason why these puzzles defeated logical sudoku solvers

Tarek

[m_b_metcalf]

Blow your trumpet until you fall on it .....

There is a good reason why these puzzles defeated logical sudoku solvers

Tarek

Quite! Of course, the puzzle with 17 clues can simply be checked against Gordon's data base, which I'm too lazy to do. At least the 18-clue puzzle represents a valid pattern:

Code: Select all

 1 . 2 . . . 3 . .
 . . . 4 . 5 . . .
 6 . . . . . . . .
 . 4 . . . . . 7 5
 2 . . . 3 . . . .
 . . . 8 . . . 9 .
 . 9 . 5 . . . . 7
 . . . . 2 . . 6 .
 . . . . . . . . .

and I suspect the particular value distribution he sets has been plucked from a data base.

Regards,

Mike

[daj95376]

Puzzle 5: (first Sudoku)

Code: Select all

 +-----------------------+
 | . . . | . . . | . . . |
 | 3 . . | . 9 1 | 4 . 6 |
 | . . 7 | 3 . 4 | 2 . 5 |
 |-------+-------+-------|
 | . 9 . | 4 . 5 | 6 . 7 |
 | 7 4 . | . 8 . | . 5 1 |
 | 5 . 1 | 9 . 7 | . 3 . |
 |-------+-------+-------|
 | 2 . 5 | 7 . 9 | 1 . . |
 | 1 . 6 | 2 3 # | . . 8 |
 | . . . | . . . | . . . |
 +-----------------------+


r8c6 is void of candidates because it sees all nine values for clues/givens.

If you remove clues/givens r2c6=1 and r8c1=1, then you have a valid puzzle that solves with Naked Singles.
Subsequent reductions in clues/givens are still possible.

If you remove clue/given r3c6=4, then you have a valid puzzle that solves with Naked Singles.
Subsequent reductions in clues/givens are still possible.

If you remove clue/given r7c6=9, then you have a valid puzzle that solves with Naked Singles.
Subsequent reductions in clues/givens are still possible.

Note: all three solutions are different. FWIW, all three solutions have these cells in common.

Code: Select all

 +-----------------------+
 | 6 . 4 | . 7 . | 3 1 9 |
 | 3 . 8 | . 9 . | 4 7 6 |
 | 9 1 7 | 3 . . | 2 8 5 |
 |-------+-------+-------|
 | 8 9 3 | 4 1 5 | 6 2 7 |
 | 7 4 2 | 6 8 3 | 9 5 1 |
 | 5 6 1 | 9 2 7 | 8 3 4 |
 |-------+-------+-------|
 | 2 8 5 | 7 . . | 1 . 3 |
 | . 7 6 | 2 3 . | 5 . 8 |
 | . 3 9 | . 5 . | 7 . 2 |
 +-----------------------+   # 8 solutions possible


Yuck !!!

[blue]

There's a "telling" answers section.

[daj95376]

There's a "telling" answers section.

His answers section contains two solutions:

(1) He appears to have moved r7c6=9 to r8c6=9 for his first solution.
(2) He appears to have moved r3c6=4 to r8c6=4 for his second solution.


I found three solutions.

Code: Select all

# daj derived puzzle 1:  remove r2c6=1 and r8c1=1 from givens  =>  r8c6=1

 +-----------------------+      +-----------------------+
 | . . . | . . . | . . . |      | 6 2 4 | 5 7 8 | 3 1 9 |
 | 3 . . | . 9 . | 4 . 6 |      | 3 5 8 | 1 9 2 | 4 7 6 |
 | . . 7 | 3 . 4 | 2 . 5 |      | 9 1 7 | 3 6 4 | 2 8 5 |
 |-------+-------+-------|      |-------+-------+-------|
 | . 9 . | 4 . 5 | 6 . 7 |      | 8 9 3 | 4 1 5 | 6 2 7 |
 | 7 4 . | . 8 . | . 5 1 |      | 7 4 2 | 6 8 3 | 9 5 1 |
 | 5 . 1 | 9 . 7 | . 3 . |      | 5 6 1 | 9 2 7 | 8 3 4 |
 |-------+-------+-------|      |-------+-------+-------|
 | 2 . 5 | 7 . 9 | 1 . . |      | 2 8 5 | 7 4 9 | 1 6 3 |
 | . . 6 | 2 3 . | . . 8 |      | 4 7 6 | 2 3 1 | 5 9 8 |
 | . . . | . . . | . . . |      | 1 3 9 | 8 5 6 | 7 4 2 |
 +-----------------------+      +-----------------------+


Code: Select all

# daj derived puzzle 2:  remove r3c6=4 from givens  =>  r8c6=4

 +-----------------------+      +-----------------------+
 | . . . | . . . | . . . |      | 6 5 4 | 8 7 2 | 3 1 9 |
 | 3 . . | . 9 1 | 4 . 6 |      | 3 2 8 | 5 9 1 | 4 7 6 |
 | . . 7 | 3 . . | 2 . 5 |      | 9 1 7 | 3 4 6 | 2 8 5 |
 |-------+-------+-------|      |-------+-------+-------|
 | . 9 . | 4 . 5 | 6 . 7 |      | 8 9 3 | 4 1 5 | 6 2 7 |
 | 7 4 . | . 8 . | . 5 1 |      | 7 4 2 | 6 8 3 | 9 5 1 |
 | 5 . 1 | 9 . 7 | . 3 . |      | 5 6 1 | 9 2 7 | 8 3 4 |
 |-------+-------+-------|      |-------+-------+-------|
 | 2 . 5 | 7 . 9 | 1 . . |      | 2 8 5 | 7 6 9 | 1 4 3 |
 | 1 . 6 | 2 3 . | . . 8 |      | 1 7 6 | 2 3 4 | 5 9 8 |
 | . . . | . . . | . . . |      | 4 3 9 | 1 5 8 | 7 6 2 |
 +-----------------------+      +-----------------------+

Notice potential <49> UR [r78c68] prevented by clue/given r7c6=9


Code: Select all

# daj derived puzzle 3:  remove r7c6=9 from givens  =>  r8c6=9

 +-----------------------+      +-----------------------+
 | . . . | . . . | . . . |      | 6 5 4 | 8 7 2 | 3 1 9 |
 | 3 . . | . 9 1 | 4 . 6 |      | 3 2 8 | 5 9 1 | 4 7 6 |
 | . . 7 | 3 . 4 | 2 . 5 |      | 9 1 7 | 3 6 4 | 2 8 5 |
 |-------+-------+-------|      |-------+-------+-------|
 | . 9 . | 4 . 5 | 6 . 7 |      | 8 9 3 | 4 1 5 | 6 2 7 |
 | 7 4 . | . 8 . | . 5 1 |      | 7 4 2 | 6 8 3 | 9 5 1 |
 | 5 . 1 | 9 . 7 | . 3 . |      | 5 6 1 | 9 2 7 | 8 3 4 |
 |-------+-------+-------|      |-------+-------+-------|
 | 2 . 5 | 7 . . | 1 . . |      | 2 8 5 | 7 4 6 | 1 9 3 |
 | 1 . 6 | 2 3 . | . . 8 |      | 1 7 6 | 2 3 9 | 5 4 8 |
 | . . . | . . . | . . . |      | 4 3 9 | 1 5 8 | 7 6 2 |
 +-----------------------+      +-----------------------+

Notice potential <46> UR [r37c56] prevented by clue/given r3c6=4


[Pat]

His answers section contains two solutions:

(1) He appears to have moved r7c6=9 to r8c6=9 for his first solution.
(2) He appears to have moved r3c6=4 to r8c6=4 for his second solution.

some ingenious players have developed this method of
"solving" by moving the givens.

the rule against moving (or changing) the given cells
is often un-stated.