The New Sudoku Players' Forum

[ArkieTech]

Code: Select all

 *-----------*
 |..3|.6.|4..|
 |.6.|8.7|.9.|
 |2..|.5.|..6|
 |---+---+---|
 |.4.|...|.2.|
 |7.2|...|9.4|
 |.5.|...|.3.|
 |---+---+---|
 |4..|.8.|..3|
 |.8.|3.5|.4.|
 |..5|.7.|1..|
 *-----------*


Play/Print this puzzle online

[SteveG48]

Code: Select all

 *-----------------------------------------------------------------------------*
 | 1589   d179     3       | 129     6       129     | 4       1578    2578    |
 |e15      6      e14      | 8      f1234    7       | 235     9       25      |
 | 2      d179     14789   | 149     5       1349    | 378     178     6       |
 *-------------------------+-------------------------+-------------------------|
 | 13689   4       1689    |a15679  a139    a13689   | 5678    2       1578    |
 | 7      c13      2       | 56     b3-1     68      | 9       568     4       |
 | 1689    5       1689    |a124679 a1249   a124689  | 678     3       178     |
 *-------------------------+-------------------------+-------------------------|
 | 4      c1279    1679    |b1269    8      b1269    | 2567    567     3       |
 | 169     8       1679    | 3      a129     5       | 267     4       279     |
 | 369     239     5       | 2469    7       2469    | 1       68      289     |
 *-----------------------------------------------------------------------------*


1r468c456 = 1r5c5&r7c46 - 1r57c2 = r13c2 - r2c13 = 1r2c5 => -1 r5c5 ; lclste

Or equivalent, if preferred, Kraken 1c2 => -1 r5c5 .

[Cenoman]

Yes, I prefer... It shows the role of empty rectangles "right-angle" weak links in B1 & B7

Code: Select all

 +-------------------------+---------------------------+-----------------------+
 |  1589   a179    3       |  129      6      129      |  4      1578   2578   |
 | b15      6     b14      |  8       c1234   7        |  235    9      25     |
 |  2      a179    14789   |  149      5      1349     |  378    178    6      |
 +-------------------------+---------------------------+-----------------------+
 |  13689   4      1689    |  15679    139    13689    |  5678   2      1578   |
 |  7      A13     2       |  56       3-1    68       |  9      568    4      |
 |  1689    5      1689    |  124679   1249   124689   |  678    3      178    |
 +-------------------------+---------------------------+-----------------------+
 |  4      x1279   1679    |  1269     8      1269     |  2567   567    3      |
 | y169     8     y1679    |  3       z129    5        |  267    4      279    |
 |  369     239    5       |  2469     7      2469     |  1      68     289    |
 +-------------------------+---------------------------+-----------------------+

Kraken column (1)r1357c2
(1)r13c2 - r2c13 = r2c5
(1)r5c2
(1)r7c2 - r8c13 = r8c5
=> -1 r5c5; lclste

Edit: corrected misuse of empty rectangles

[SpAce]

Nice solutions. I shouldn't say anything else, but can't help mentioning that there are no empty rectangles (or ERIs) on 1s in b1 and b7. ERIs are box-based (often grouped) strong links, but we don't have them here. Instead the Kraken 1c2 uses box-based weak links, while the related strong links are on the rows (and the column SIS). Thus a different pattern, similar to what's in a Grouped 2-String Kite.

PS. I think such useful box-based weak links are often more difficult to spot than ERIs, so I consider them more advanced -- hence nice solutions.

[Cenoman]

I have used a pattern name, ER, that I quite never use (the name, not the pattern...) In ER there are two words, rectangle and empty; unfortunately none of the possible rectangles is empty! Moreover, ER should be used only when a strong link is provided! I am all wrong. I focused on the "right-angle gear", too bad for me. I will be tarred and feathered!!
Edited the faulty post.

[SpAce]

I might have missed it myself if it wasn't for this recent discussion which made me think about the difference more. It's a pity that we apparently don't have a standard name for that situation, since it's a useful and nice (partial) pattern. It does work as a right-angle hinge like an ER(I), but as a weak link instead of a strong link, so perhaps something like "Inverse ER(I)" could work?

More importantly, I really liked the Kraken pattern here, no matter what the hinge parts are called. If generalized, it's perfectly symmetric:

Code: Select all

.-----------.-----------.-----------.
| 1   1   1 | .   .   . | .   .   . |
| 1   /   1 | /   1   / | /   /   / |
| 1   1   1 | .   .   . | .   .   . |
:-----------+-----------+-----------:
| .   /   . | .   .   . | .   .   . |
| .   1   . | .  -1   . | .   .   . |
| .   /   . | .   .   . | .   .   . |
:-----------+-----------+-----------:
| 1   1   1 | .   .   . | .   .   . |
| 1   /   1 | /   1   / | /   /   / |
| 1   1   1 | .   .   . | .   .   . |
'-----------'-----------'-----------'


Mutant Swordfish (Rank 1):

1:R28C2\r5c5b17 => -1 r5c5

Without the fin (r5c2) it would be a loop:

Code: Select all

.-----------.-----------.-----------.
|-1   1  -1 | .  -1   . | .   .   . |
| 1   /   1 | /   1   / | /   /   / |
|-1   1  -1 | .  -1   . | .   .   . |
:-----------+-----------+-----------:
| .   /   . | .  -1   . | .   .   . |
| .   /   . | .  -1   . | .   .   . |
| .   /   . | .  -1   . | .   .   . |
:-----------+-----------+-----------:
|-1   1  -1 | .  -1   . | .   .   . |
| 1   /   1 | /   1   / | /   /   / |
|-1   1  -1 | .  -1   . | .   .   . |
'-----------'-----------'-----------'

Mutant Swordfish (Rank 0):

1:R28C2\c5b17 => -1 c5, b1, b7

[SteveG48]

Wow. Fun discussion. I almost didn't post because I didn't have a singles solution.

Speaking of which.... anyone?