The New Sudoku Players' Forum

[whitefork]

this one

..5..47.2
3..1.....
....8.36.
...75.6..
.9.....7.
..7.62...
.46.3....
.....1...
9.38..5..

The program says is invalid - not solvable by logic alone. Have I made a typo in entering it? I can't seem to see one.

Thanks

[emm]

If it's a Pappocom puzzle it should be symmetrical. You're missing a clue from row 8 box 9.

[whitefork]

Well well. It does say "By Pappacom" and the Globe printed it with an empty cell at (8,9). Putting a 3 in there does make it valid.

Yet the puzzle seems to have a unique solution with the empty cell.

[rubylips]

The printed puzzle has three solutions. Logic alone takes us so far:

Code: Select all

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

    8   6    5 |  3  9  4 |    7   1    2
    3   2   49 |  1  7  6 |  489   5   48
   14   7  149 |  2  8  5 |    3   6   49
---------------+----------+--------------
   24   3   14 |  7  5  8 |    6  29  149
    6   9   28 |  4  1  3 |   28   7    5
  145  58    7 |  9  6  2 |   14  38   13
---------------+----------+--------------
   27   4    6 |  5  3  9 |   18  28  178
   57  58   28 |  6  4  1 |   29  39   37
    9   1    3 |  8  2  7 |    5   4    6


Unfortunately, it's not possible to solve for r8c9, the cell identified by em as the missing cell. Since two of the three solutions have a 7 in this cell and just one has a 3, clearly Wayne intended for a 3 to go into that cell, as surmised by whitefork.

Code: Select all

I.

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

II.

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

III.

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


The puzzle isn't totally straightforward even with the 3 in place. I'll leave it as an exercise.

Code: Select all

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

Straightforward to here:

    18     6     5 |  3   9  4 |     7     18    2
     3    28  2489 |  1   7  6 |   489      5  489
   147    17   149 |  2   8  5 |     3      6  149
-------------------+-----------+------------------
   124   123   124 |  7   5  8 |     6  12349  149
     6     9    28 |  4   1  3 |    28      7    5
  1458  1358     7 |  9   6  2 |   148   1348  148
-------------------+-----------+------------------
   128     4     6 |  5   3  9 |   128    128    7
  2578  2578    28 |  6  24  1 |  2489   2489    3
     9    12     3 |  8  24  7 |     5    124    6


[whitefork]

A misprint most likely. Rubylips, thank you for pointing out the alternate solutions. This is the first time I've run into this situation in published puzzles. Does this happen sometimes?

I've been doing these for a few months now but I'm new to this community. Thanks for responding.

[Pappocom]

Hello all. There is evidently a problem with the way this puzzle is printed. I am getting in touch with the Globe to see how they managed to get it wrong.

If only papers would print the puzzles from the PDF files I supply them with there would never be a problem. It's when they try to transcribe from my PDF file, using it as a data source, that problems arise.

- Wayne

[gtmtnbiker]

Hello all. There is evidently a problem with the way this puzzle is printed. I am getting in touch with the Globe to see how they managed to get it wrong.

If only papers would print the puzzles from the PDF files I supply them with there would never be a problem. It's when they try to transcribe from my PDF file, using it as a data source, that problems arise.

- Wayne

What is the end resolution of this misprint? Was the only error in the puzzle in r8c9? Or are there other errors?

--Bill

[Animator]

Indeed. r8c9 = 3. That's the only change.

(Note, if you don't believe me then you can check it on the Solutions page (http://www.sudoku.com/solutions.php). Just select Boston Globe and enter 54723. (The blue numbers are the clues))

[Pappocom]

I confirm that Animator has it right.

The Boston Globe printed their apology the next day.

- Wayne.

[lbass]

Thanks for the clarification. I had arrived at a different solution than the one published in the Globe also, and was surprised, as I thought there was only one unique solution to a puzzle.

I never really paid much attention to the symmetry in the puzzles. Is symmetry somehow a requirement of a "legitimate" puzzle or is it a feature of Pappacom's sudoku's in particular? Could the lack of symmetry be used as a flag for an incorrectly published sudoku?

-LB

[Animator]

It's a feature of Pappocom's sudoku's.

Further reference: Three features of all Pappocom puzzles