The New Sudoku Players' Forum

[ArkieTech]

Code: Select all

 *-----------*
 |5..|3..|7..|
 |2.7|4..|.5.|
 |.31|...|..2|
 |---+---+---|
 |.98|2..|..3|
 |...|8.3|...|
 |3..|..5|91.|
 |---+---+---|
 |6..|...|89.|
 |.1.|..8|2.5|
 |..9|..4|..6|
 *-----------*


Play/Print this puzzle online

[pjb]

Code: Select all

5      468    46     | 3      12    a129    | 7      68    a19     
2      68     7      | 4      18    a169    | 3      5     a19     
9      3      1      | 567    578   b67     | 46     468    2     
---------------------+----------------------+---------------------
47     9      8      | 2      147    167    | 5      467    3     
1      4567   456    | 8      9      3      | 46     2      47     
3      2467   246    | 67     47     5      | 9      1      8     
---------------------+----------------------+---------------------
6      2457   2345   | 1      2357  b27     | 8      9      47     
47     1      34     | 9      6      8      | 2      347    5     
8      257    9      | 57     2357   4      | 1      37     6     


The AUR of 19s at r12c69 can be used two ways:
1: Type 3: The 129/169 provide a naked pair of 26, and together with the 67 at r3c6 and 27 at r6c6 make a naked triple, => r4c6 <> 67; stte
2: Type 4: Because the 9s are alone in both row and column, r12c6 <> 1, => stte.

Phil

[pjb]

In addition to the about two, a third attacked with this AUR:

If 2 at r1c6 is false, then 6 at r2c6 is true: (6=7) r3c6 - (7=2) r7c6 - (2=6)[r12c6, r12c9] => r3c4, r4c6 <> 6; stte

Phil

[Leren]

Code: Select all

*--------------------------------------------------------------*
| 5     468   46     | 3    A12   B129    | 7     68   B19     |
| 2     68    7      | 4    A18   B169    | 3     5    B19     |
| 9     3     1      | 567   578   67     | 46    468   2      |
|--------------------+--------------------+--------------------|
| 47    9     8      | 2    a47-1 A167    | 5     467   3      |
| 1     4567  456    | 8     9     3      | 46    2     47     |
| 3     2467  246    | 67    47    5      | 9     1     8      |
|--------------------+--------------------+--------------------|
| 6     2457  2345   | 1     2357  27     | 8     9     47     |
| 47    1     34     | 9     6     8      | 2     347   5      |
| 8     257   9      | 57    2357  4      | 1     37    6      |
*--------------------------------------------------------------*

Reverse BUG digits 19:

If r4c5 = 1 => cells marked A <> 1 => cells marked B = 19 (DP) => r4c5 <> 1; stte

I could use some help with terminology and eureka notation for this move

Leren

[JC Van Hay]

Code: Select all

+-----------------+--------------------+------------------+
| 5   468    46   | 3      128   1269  | 7     468    19  |
| 2   8(6)   7    | 4      18    19(6) | 3     5      19  |
| 9   3      1    | 57(6)  578   7-6   | 4(6)  48-6   2   |
+-----------------+--------------------+------------------+
| 47  9      8    | 2      147   17(6) | 5     47(6)  3   |
| 1   4567   456  | 8      9     3     | 4(6)  2      47  |
| 3   247-6  246  | 7(6)   47    5     | 9     1      8   |
+-----------------+--------------------+------------------+
| 6   2457   2345 | 1      2357  27    | 8     9      47  |
| 47  1      34   | 9      6     8     | 2     347    5   |
| 8   257    9    | 57     2357  4     | 1     37     6   |
+-----------------+--------------------+------------------+

Jellyfish : Kite(6C7R4)+Kite(6R4C4)+Kite(6C4R2) :=> -6r3c68, -6r6c2;ste

[daj95376]

Code: Select all

*--------------------------------------------------------------*
| 5     468   46     | 3    A12   B129    | 7     68   B19     |
| 2     68    7      | 4    A18   B169    | 3     5    B19     |
| 9     3     1      | 567   578   67     | 46    468   2      |
|--------------------+--------------------+--------------------|
| 47    9     8      | 2    a47-1 A167    | 5     467   3      |
| 1     4567  456    | 8     9     3      | 46    2     47     |
| 3     2467  246    | 67    47    5      | 9     1     8      |
|--------------------+--------------------+--------------------|
| 6     2457  2345   | 1     2357  27     | 8     9     47     |
| 47    1     34     | 9     6     8      | 2     347   5      |
| 8     257   9      | 57    2357  4      | 1     37    6      |
*--------------------------------------------------------------*

Reverse BUG digits 19:

If r4c5 = 1 => cells marked A <> 1 => cells marked B = 19 (DP) => r4c5 <> 1; stte

I could use some help with terminology and eureka notation for this move

What I see is the X-Wing for <9> in [c69]. If r4c6<>1, then an X-Wing will exist for <1> in the same cells, and a DP will result. So, r4c6=1 follows.

[daj95376]

JC's eliminations from another perspective.

Code: Select all

 +--------------------------------------------------------------+
 |  5     468   46    |  3     12    129   |  7     68    19    |
 |  2    a68    7     |  4     18   b169   |  3     5     19    |
 |  9     3     1     | c567   578   67    | h46    468   2     |
 |--------------------+--------------------+--------------------|
 |  47    9     8     |  2     147  e167   |  5    f467   3     |
 |  1     4567  456   |  8     9     3     | g46    2     47    |
 |  3     2467  246   | d67    47    5     |  9     1     8     |
 |--------------------+--------------------+--------------------|
 |  6     2457  2345  |  1     2357  27    |  8     9     47    |
 |  47    1     34    |  9     6     8     |  2     347   5     |
 |  8     257   9     |  57    2357  4     |  1     37    6     |
 +--------------------------------------------------------------+
 # 64 eliminations remain

      |**** 2-String Kite ****|                              =>           r6c2<>6

 (6): r2c2 = r2c6 - r3c4 = r6c4 - r4c6 = r4c8 - r5c7 = r3c7  =>  r3c68<>6,r6c2<>6

                    |*** mutant Swordfish r4c47\r3b56 ****|  =>  r3c68<>6


The 2-String Kite elimination can be obtained via Multipe Colors, and the mutant Swordfish eliminations can be obtained through simple Colors.

My solver found the eliminations through its template logic, but my Fish program couldn't find all of the eliminations as a single fish pattern. So, I went on a hunt for unification logic. I found my results prior to seeing JC's post.

[David P Bird]

Reverse BUG digits 19:
If r4c5 = 1 => cells marked A <> 1 => cells marked B = 19 (DP) => r4c5 <> 1; stte

I could use some help with terminology and eureka notation for this move

I would notate it like this:

(19)ReverseBug:r5,c1,r3,c3,r9,c7,r6,c8,r7,c4,r8,c2,r4 => r4c5 <> 1

The path covers all the known cells for 1 & 9 in every intermediate row & column, so the loop can't be completed in r4c5 to leave r12c69 uncovered.

Only using rows and columns keeps it simple, but there may be more than one loop to identify.

[aran]

Code: Select all

*--------------------------------------------------------------*
| 5     468   46     | 3    A12   B129    | 7     68   B19     |
| 2     68    7      | 4    A18   B169    | 3     5    B19     |
| 9     3     1      | 567   578   67     | 46    468   2      |
|--------------------+--------------------+--------------------|
| 47    9     8      | 2    a47-1 A167    | 5     467   3      |
| 1     4567  456    | 8     9     3      | 46    2     47     |
| 3     2467  246    | 67    47    5      | 9     1     8      |
|--------------------+--------------------+--------------------|
| 6     2457  2345   | 1     2357  27     | 8     9     47     |
| 47    1     34     | 9     6     8      | 2     347   5      |
| 8     257   9      | 57    2357  4      | 1     37    6      |
*--------------------------------------------------------------*

Reverse BUG digits 19:

If r4c5 = 1 => cells marked A <> 1 => cells marked B = 19 (DP) => r4c5 <> 1; stte

I could use some help with terminology and eureka notation for this move

Leren

In theory, either UR or not UR
but UR not an option so
UR19r12c69=1r12c5 : =><1>r4c5

[ArkieTech]

In addition to the about two, a third attacked with this AUR:

If 2 at r1c6 is false, then 6 at r2c6 is true: (6=7) r3c6 - (7=2) r7c6 - (2=6)[r12c6, r12c9] => r3c4, r4c6 <> 6; stte

Phil

An aur xy-wing

(6=7) r3c6 - (7=2) r7c6 - (2=6)19r12c69

[Marty R.]

Just the Plain Jane UR on 19. R4c6=1 either from internal or external analysis.

[Leren]

David P Bird wrote:

I would notate it like this:

(19)ReverseBug:r5,c1,r3,c3,r9,c7,r6,c8,r7,c4,r8,c2,r4 => r4c5 <> 1

The path covers all the known cells for 1 & 9 in every intermediate row & column, so the loop can't be completed in r4c5 to leave r12c69 uncovered.

Only using rows and columns keeps it simple, but there may be more than one loop to identify.

David, that's fairly close to how I actually found this move.

Perhaps I could write it as

(19)ReverseBug: r3456789 / c1234578 / b1456789 => r4c5 <> 1

ie the number of rows, columns and boxes covered by given/solved 1's or 9's is 7 in each case (equal to the number of given/solved 9's and one more than
the given/solved 1's). The 1 can then be eliminated from a cell that is located in these rows/columns/boxes where the 1 is not given/solved (so would leave
the number of rows/columns/boxes unchanged but make the number of given/solved 1's and 9's equal.

Of course this move is really meant to be used when the number of rows/columns/boxes is small (and the corresponding DP would be more complex and harder to spot)
but it was interesting to see that the method still works with so many rows/columns/boxes.

Leren

[David P Bird]

Leren, I did realise that you were making a rather an academic point using a Reverse-GUG construct when there was a glaring UR available.

How best to notate an eliminations depends on whether you are trying to keep notes for yourself or are trying to convey the logic to another reader - who may possibly be a novice. Weighing up our alternative suggestions, I feel that readers should be able to trace a path more easily than digest a fish construct. However, I admit that this still leaves them the job of confirming for themselves that two cells are covered in every box visited.

Although I've ranted on before about making the notations we post reader-friendly, the only reaction has been sarcastic comments, so I've given up. Consequently you can hardy be criticised if you do what everyone else here does, and choose whatever method you personally prefer.

[Marty R.]

Leren, I did realise that you were making a rather an academic point using a Reverse-GUG construct when there was a glaring UR available.

How best to notate an eliminations depends on whether you are trying to keep notes for yourself or are trying to convey the logic to another reader - who may possibly be a novice. Weighing up our alternative suggestions, I feel that readers should be able to trace a path more easily than digest a fish construct. However, I admit that this still leaves them the job of confirming for themselves that two cells are covered in every box visited.

Although I've ranted on before about making the notations we post reader-friendly, the only reaction has been sarcastic comments, so I've given up. Consequently you can hardy be criticised if you do what everyone else here does, and choose whatever method you personally prefer.

Keep on ranting. Reader-friendliness is right up there with motherhood and apple pie. I'm a notation novice and experienced player, albeit less skilled than most who post here. I like things kept as simple as possible. I don't want to see easy technique A described as more complicated technique B and don't want to see tricked-up notation just for its own sake.