The New Sudoku Players' Forum

[Yogi]

.4......5.....2.......8........5..8..3.7...........1..2.8...6.....4..72.3..5.....

17June19.png (10.28 KiB) Viewed 676 times

[Leren]

Code: Select all

*-----------------------------------------*
| 169 4     2   | 1369 37 179 | 8 a69 5   |
| 5   8     169 | 169  4  2   | 3  7  169 |
| 7   69-1  3   | 169  8  5   | 2  4  169 |
|---------------+-------------+-----------|
| 169 169   7   | 2    5  3   | 4  8  69  |
| 8   3    a69  | 7    1  4   | 5 b69 2   |
| 4   2     5   | 8    69 69  | 1  3  7   |
|---------------+-------------+-----------|
| 2   79    8   | 139  37 179 | 6  5  4   |
| 169 5     169 | 4    69 8   | 7  2  3   |
| 3   67    4   | 5    2  67  | 9  1  8   |
*-----------------------------------------*

Remote Pairs => - 1 r3c2. You can finish off with Some Skyscrapers and Kites followed by BUG +1, or for more Remote Pairs fun :

Code: Select all

*----------------------------------------*
| 1-69 4  2   | 1369 37 179 | 8 a69  5   |
| 5    8  169 | 169  4  2   | 3  7   169 |
| 7    69 3   | 169  8  5   | 2  4   169 |
|-------------+-------------+------------|
|b69   1  7   | 2    5  3   | 4  8  a69  |
| 8    3  69  | 7    1  4   | 5 b69  2   |
| 4    2  5   | 8    69 69  | 1  3   7   |
|-------------+-------------+------------|
| 2    79 8   | 139  37 179 | 6  5   4   |
| 169  5  169 | 4    69 8   | 7  2   3   |
| 3    67 4   | 5    2  67  | 9  1   8   |
*----------------------------------------*

Remote Pairs => - 69 r1c1, and then

Code: Select all

*---------------------------------------*
| 1   4   2  | 369 37 7-9 | 8 a69  5    |
| 5   8  b69 | 169 4  2   | 3  7   1-69 |
| 7  a69  3  | 169 8  5   | 2  4   69-1 |
|------------+------------+-------------|
| 69  1   7  | 2   5  3   | 4  8  a69   |
| 8   3   69 | 7   1  4   | 5  69  2    |
| 4   2   5  | 8   69 69  | 1  3   7    |
|------------+------------+-------------|
| 2   79  8  | 39  37 1   | 6  5   4    |
| 69  5   1  | 4   69 8   | 7  2   3    |
| 3   67  4  | 5   2  67  | 9  1   8    |
*---------------------------------------*

Remote Pairs => - 9 r1c6, -69 r2c9, -1 r3c9; stte

For some reason Hodoku required a forcing chain to solve this one. Is there some Hodoku setting I can change to avoid this ?

Leren

[SpAce]

1) Replace 169 with abc, and assign them in box 1:

Code: Select all

.-----------------.-----------------.-------------.
| *a    4     2   | 3abc  37   7abc | 8  abc  5   |
|  5    8    *b   | abc   4    2    | 3  7    abc |
|  7   *c     3   | abc   8    5    | 2  4    abc |
:-----------------+-----------------+-------------:
|  abc  abc   7   | 2     5    3    | 4  8    abc |
|  8    3     abc | 7     1    4    | 5  abc  2   |
|  4    2     5   | 8     abc  abc  | 1  3    7   |
:-----------------+-----------------+-------------:
|  2    7abc  8   | 3abc  37   7abc | 6  5    4   |
|  abc  5     abc | 4     abc  8    | 7  2    3   |
|  3    7abc  4   | 5     2    7abc | 9  1    8   |
'-----------------'-----------------'-------------'

2) Apply basics and get here:

Code: Select all

.-----------------.------------------.---------------.
| *a    4     2   | #b    #3   #7    |  8   #c    5  |
|  5    8    *b   | #c     4    2    |  3    7   #a  |
|  7   *c     3   | #a     8    5    |  2    4   #b  |
:-----------------+------------------+---------------:
| #b   #a     7   |  2     5    3    |  4    8   #c  |
|  8    3    #c   |  7    [1]   4    |  5    ab   2  |
|  4    2     5   |  8     ac  #b    | [1]   3    7  |
:-----------------+------------------+---------------:
|  2   #b     8   | #3    #7    ac   | [6]   5    4  |
| #c    5    #a   |  4    #b    8    |  7    2    3  |
|  3   #7     4   |  5     2    ac   | [9]  [1]   8  |
'-----------------'------------------'---------------'

3) Read results:

r67: b <> 1|6 -> b = 9 => +9 r2c3; stte

Edit: simplified the result step.

[SpAce]

For some reason Hodoku required a forcing chain to solve this one. Is there some Hodoku setting I can change to avoid this ?

No, because Hodoku is probably right that there's no easier conventional move available. If you claim your first step is such, you should explain it a bit better. As shown, it doesn't make much sense to me. What remote pair gives you -1 r3c2 ??? If it's some fancy move that neither I nor Hodoku understand, it needs better documentation.

Added. I'm guessing you meant:

Almost Remote Pair:

(1)r1c1 = RP(69)r1c18,r5c83 - (6|9=1)r2c3 => -1 r3c2

That's a neat move per se, but it's not a conventional move that you could simply call "Remote Pairs" like you did. Hodoku sees it as a net because it can't use complex almost-patterns in chains. That's also why it uses a different forcing chain by default (because it's simpler than a net, and thus the simplest move it knows here). So, Hodoku behaves just as expected.

[Leren]

The way I explain the first move is as follows. r1c8 and r5c3 have the same 69 parity and they can see all of the 6's and 9's in Box 1 except for r3c2.

If r1c8 and r5c3 are 6, r3c2 = 6; If r1c8 and r5c3 are 9, r3c2 = 9. Thus r3c2 is reduced to 69 and has the same parity as r1c8 and r5c3.

Note that the target cell (r3c2 here) could have more than 3 digits in other situations. So I see this as an extension to the method of Remote Pairs, where you can (1) make eliminations using two RP cells of the same parity and (2) create a new RP cell of the same parity, which could be used in later RP moves.

Leren

[SpAce]

The way I explain the first move is as follows. r1c8 and r5c3 have the same 69 parity and they can see all of the 6's and 9's in Box 1 except for r3c2.

If r1c8 and r5c3 are 6, r3c2 = 6; If r1c8 and r5c3 are 9, r3c2 = 9. Thus r3c2 is reduced to 69 and has the same parity as r1c8 and r5c3.

That logic works but doesn't make it a Remote Pair, imho. Remote Pair is a simple, well-known, and easily recognizable pattern which contains two all-conjugate X-Chains in parallel (that could be used separately as well). I think at least eleven has used group extensions with it too, but that's about as much additional complexity I would accept under that name (even then I'd rather call it "Grouped Remote Pair"). What you're describing is quite a bit more complex because it requires net logic, can't be broken into two independent chains, and has a completely different elimination logic. It obviously has elements of Remote Pairs but I can't see any stretch of logic that would put it in the same category.

Note that the target cell (r3c2 here) could have more than 3 digits in other situations. So I see this as an extension to the method of Remote Pairs, where you can (1) make eliminations using two RP cells of the same parity and (2) create a new RP cell of the same parity, which could be used in later RP moves.

The keyword being extension which you didn't mention with your move. If you call your move Remote Pair without any other qualifiers, I would expect to see exactly that and not something else -- especially if you use the normal Remote Pair right after it. Your grid markings weren't exactly helpful in deciphering your logic either, as only 3/6 relevant cells were marked, and you used letters to mark parities (pretty confusing considering their normal use -- I'd suggest ' and " for parity markers).

That being said, a neat move anyway. My only beef is with the naming and the grid markings.

PS. Besides the Almost Remote Pair I mentioned earlier, it could be seen as a Bivalue Oddagon (or whatever it's called):

Code: Select all

.--------------------.---------------.-------------.
| *69+1  4      2    | 1369  37  179 | 8  *69  5   |
|  5     8     *69+1 | 169   4   2   | 3   7   169 |
|  7     69-1   3    | 169   8   5   | 2   4   169 |
:--------------------+---------------+-------------:
|  169   169    7    | 2     5   3   | 4   8   69  |
|  8     3     *69   | 7     1   4   | 5  *69  2   |
|  4     2      5    | 8     69  69  | 1   3   7   |
:--------------------+---------------+-------------:
|  2     79     8    | 139   37  179 | 6   5   4   |
|  169   5      169  | 4     69  8   | 7   2   3   |
|  3     67     4    | 5     2   67  | 9   1   8   |
'--------------------'---------------'-------------'

5-link Bivalue Oddagon r1c18,r5c83,r2c3 using internals:

(1)b1p16 => -1 r3c2

I think that would be my preferred way to express that elimination. If I remember correctly, Steve once suggested calling it Remote Pair Oddagon (which I would gladly accept).

(Note that if we used box1 externals it would be a closer match to your logic: (6=9)r3c2 => -1 r3c2.)

[Cenoman]

Code: Select all

 +---------------------+--------------------+------------------+
 |GBd169   4     2     | C1369  C37  D179   |  8 GBd69   5     |
 |   5     8   Aa169z  |  169    4    2     |  3    7    169   |
 |   7     169   3     |  169    8    5     |  2    4    169   |
 +---------------------+--------------------+------------------+
 | Gd169   169   7     |  2      5    3     |  4    8    69    |
 |   8     3    B6-9   |  7      1    4     |  5 GBd69   2     |
 |   4     2     5     |  8      69   69    |  1    3    7     |
 +---------------------+--------------------+------------------+
 |   2     79    8     |  139    37   179   |  6    5    4     |
 | Fc169   5    b169   |  4     E69   8     |  7    2    3     |
 |   3     67    4     |  5      2   E67    |  9    1    8     |
 +---------------------+--------------------+------------------+

Kraken cell (169)r2c3 & Almost skyscrapers (9)c18 & (6)r15
(1)r2c3 - r8c3 = (1-9)r8c1 = [SS(9)r4c1*=r1c1-r1c8=r5c8]
(6)r2c3 - [(6)r5c3=r5c8-r1c8*=r1c1] = (6-3)r1c4 = (3-7)r1c5 = r1c6 - (76=9)b8p59 - r8c1 = [SS(9)r4c1*=r1c1-r1c8=r5c8]
(9)r2c3
=> -9 r5c3; ste

TM 10x10

Hidden Text: Show

Code: Select all

9r5c8 9r1c8
9r4c1 9r1c1 9r8c1
            1r8c1 1r8c3
9r2c3             1r2c3 6r2c3
                        6r5c3 6r2c8
                        6r1c1 6r1c8 6r1c4
                                    3r1c4 3r1c5
                                          7r1c5 7r1c6
                                                7r9c6 6r9c6
            9r8c5                                     6r8c5

=> -9 r5c3; ste