The New Sudoku Players' Forum

[rjamil]

Hi,

While trying to solve April 7, 2018 puzzle, after several basic and one known skyscraper move, found strange Skyscraper move in Hodoku as follows:

Skyscraper 2.png (15.98 KiB) Viewed 1114 times

Code: Select all

6.8..34.74.3......927..438......78.5..1.2.9..5.64..7..8341...79........81.58.....
:0400:1:6.8..34.74.3......927..438......78.5..1.2.9..5.64..7..8341...79........81.58.....:514 515 118 524 525 226 526 926 128 248 966 169 284 584 684 984 585 685 985 286 287 288 695 995:127::
.-------------.------------------.-------------------.
| 6   15   8  | 29    19     3   | 4     259    7    |
| 4   15   3  | 2679  16789  168 | 1256  2569   126  |
| 9   2    7  | 56    156    4   | 3     8      16   |
:-------------+------------------+-------------------:
| 23  49   29 | 369   1369   7   | 8     1346   5    |
| 37  478  1  | 356   2      568 | 9     346    346  |
| 5   89   6  | 4     1389   18  | 7     123    23   |
:-------------+------------------+-------------------:
| 8   3    4  | 1     56     256 | 256   7      9    |
| 27  679  29 | 37    347    569 | 156   13456  8    |
| 1   679  5  | 8     347    269 | 26    2346   2346 |
'-------------'------------------'-------------------'

.-------------.------------------.----------------------.
| 6   15   8  | 29    19     3   |  4      259     7    |
| 4   15   3  | 2679  16789  168 | *-1256  2569   *126  |
| 9   2    7  | 56    156    4   |  3      8      *16   |
:-------------+------------------+----------------------:
| 23  49   29 | 369   1369   7   |  8      1346    5    |
| 37  478  1  | 356   2      568 |  9      346     346  |
| 5   89   6  | 4     1389   18  |  7      123     23   |
:-------------+------------------+----------------------:
| 8   3    4  | 1     56     256 |  256    7       9    |
| 27  679  29 | 37    347    569 | *156    13456   8    |
| 1   679  5  | 8     347    269 |  26     2346    2346 |
'-------------'------------------'----------------------'
Skyscraper: 1 in r3c9,r8c7 (connected by r2c79) => r2c7<>1

Will someone please elaborate the above mentioned move (in exemplar form) for easy understanding the logic behind the same.

However, the above mentioned move is seems to me as Locked Candidate 2 (Claiming) in b3 move as well.

R. Jamil

[champagne]

very strange in fact.

In your pm, digit '1' is locked in band 1 for column c9.
This is enough to clear 1r2c7 .

if you omit this, you can apply the described move as well, but this is not what a solver would do unless a flag skips the first clearing

[StrmCkr]

Code: Select all

Sashimi X-Wing: 1 c29 r12 fr3c9 => r2c7<>1

this one also looks odd but does work when you look at it as a box-box alignment with R1c9 blank and the fin on r3c9

Code: Select all

Skyscraper: 1 in r3c9,r8c7 (connected by r2c79) => r2c7<>1

degenerative skyscraper as one of the digits is already solved/missing.

see the explanation in previous post with the additional note { a catch all fish solver will find degenerative cases}

[SpAce]

A self-cannibalistic Skyscraper? I guess it sort of makes sense if you see it like this:

(1)r8c7 = r2c7 - r2c9 = (1)r3c9 => -1 r2c7

That's standard Skyscraper logic, all right, even though most would have a hard time seeing r23c9 as two nodes and forgetting the effect of r2c7 on both. It works because standard AICs don't have memory.

PS. Claiming is of course the simpler and preferable logic here, but it's always instructive to see things from another perspective! A good example of that.

[eleven]

Just a programmer's problem. No one who looks at the puzzle, would take it serious.