The New Sudoku Players' Forum

[eniacap]

In their Bank Holiday puzzle supplement the Times published some Super Fiendish puzzles which normally give me no problems. In this case No29 puzzle I had to guess as it wouldn't succumb to xy-wing, swordfish etc., techniques.
Has anyone come across Super Fiendish puzzles which require guessing before?
Thanks

[tarek]

Hi eniacap,

If you could post this puzzle here, or preferably under "Help with particular puzzles", then we could have a better understanding of the situation.

Click on this link for tips regarding posting puzzles.

tarek

[Glyn]

eniacap I have moved this topic to the appropriate section as it relates to a specific puzzle.

Here is the Times Super Fiendish No 29 from August 23rd supplement.

Code: Select all

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

Singles and locked candidates should get you to here.

Code: Select all

.------------------.------------------.------------------.
| 2     468   3    | 68-4  5     6a7A | 9     4A7   1    |
| 1     5     467  | 4a6   2     9    | 37-4  8     37-4 |
| 47    48    9    | 348   1     37   | 2     5     6    |
:------------------+------------------+------------------:
| 467   1     47   | 5     369   8    | 347   34679 2    |
| 467   3     5    | 1     69    2    | 8     4679  479  |
| 9     2     8    | 7     36    4    | 13    136   5    |
:------------------+------------------+------------------:
| 3     49    1    | 2     7     5    | 6     49    8    |
| 5     7     46   | 9     8     136  | 134   2     34   |
| 8     69    2    | 36    4     136  | 5     1379  379  |
'------------------'------------------'------------------'

There is an X-wing for 7 in r24c37 => r2c9,r4c18<>7, but it still leaves a little work to do.

More powerful is the xy-wing based on r1c6 shown above. => r1c4,r2c79<>4. Singles to follow.

The Times Super Fiendish is usually not too horrible, the Telegraph Diabolical can sometimes be more demanding.

[ArkieTech]

eniacap I have moved this topic to the appropriate section as it relates to a specific puzzle.

Here is the Times Super Fiendish No 29 from August 23rd supplement.

Code: Select all

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

Singles and locked candidates should get you to here.

Code: Select all

.------------------.------------------.------------------.
| 2     468   3    | 68-4  5     6a7A | 9     4A7   1    |
| 1     5     467  | 4a6   2     9    | 37-4  8     37-4 |
| 47    48    9    | 348   1     37   | 2     5     6    |
:------------------+------------------+------------------:
| 467   1     47   | 5     369   8    | 347   34679 2    |
| 467   3     5    | 1     69    2    | 8     4679  479  |
| 9     2     8    | 7     36    4    | 13    136   5    |
:------------------+------------------+------------------:
| 3     49    1    | 2     7     5    | 6     49    8    |
| 5     7     46   | 9     8     136  | 134   2     34   |
| 8     69    2    | 36    4     136  | 5     1379  379  |
'------------------'------------------'------------------'

There is an X-wing for 7 in r24c37 => r2c9,r4c18<>7, but it still leaves a little work to do.

More powerful is the xy-wing based on r1c6 shown above. => r1c4,r2c79<>4. Singles to follow.

The Times Super Fiendish is usually not too horrible, the Telegraph Diabolical can sometimes be more demanding.

There is a nice swordfish swimming about (4)

dan

[Glyn]

I'm afraid I can't see a Swordfish in 4, Dan where is it hiding.

[ronk]

I'm afraid I can't see a Swordfish in 4, Dan where is it hiding.

The "x-wing for 7 in r24c37" and a naked pair step must happen first.

[Glyn]

Thanks Ronk that 4 in r5c9 had to go first. So we can avoid the xy-wing following that route.

[eniacap]

eniacap I have moved this topic to the appropriate section as it relates to a specific puzzle.

Here is the Times Super Fiendish No 29 from August 23rd supplement.

Code: Select all

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

Singles and locked candidates should get you to here.

Code: Select all

.------------------.------------------.------------------.
| 2     468   3    | 68-4  5     6a7A | 9     4A7   1    |
| 1     5     467  | 4a6   2     9    | 37-4  8     37-4 |
| 47    48    9    | 348   1     37   | 2     5     6    |
:------------------+------------------+------------------:
| 467   1     47   | 5     369   8    | 347   34679 2    |
| 467   3     5    | 1     69    2    | 8     4679  479  |
| 9     2     8    | 7     36    4    | 13    136   5    |
:------------------+------------------+------------------:
| 3     49    1    | 2     7     5    | 6     49    8    |
| 5     7     46   | 9     8     136  | 134   2     34   |
| 8     69    2    | 36    4     136  | 5     1379  379  |
'------------------'------------------'------------------'

There is an X-wing for 7 in r24c37 => r2c9,r4c18<>7, but it still leaves a little work to do.

More powerful is the xy-wing based on r1c6 shown above. => r1c4,r2c79<>4. Singles to follow.

The Times Super Fiendish is usually not too horrible, the Telegraph Diabolical can sometimes be more demanding.

Thanks Glyn. I did spot the x-wing for 7 but not the xy-wing, hence my difficulty. All is clear now.

[Pat]

Thanks Glyn. I did spot the x-wing for 7 but not the xy-wing, hence my difficulty. All is clear now.

puzzles in The Times may need Swordfish,
but never xy-wing.

as mentioned above,
the present puzzle can be solved with a Swordfish.

[tarek]

A better puzzle design would make it difficlt to bypass a swordfish without using extreme techniques.

The Pure Swordfish collection has some nice examples.

tarek

[eniacap]

Having pondered a while over the swordfish mentioned I conclude it must be at rows 2,4 and 8 and columns 3,7 and 9.
The reason for my pondering is that I assumed each of the 3 rows/columns in a swordfish must have 2 candidates as in an x-wing. This is always the pattern shown in examples in tutorials. However in this instance 4 out of the six have 3 candidates.

[ArkieTech]

Having pondered a while over the swordfish mentioned I conclude it must be at rows 2,4 and 8 and columns 3,7 and 9.
The reason for my pondering is that I assumed each of the 3 rows/columns in a swordfish must have 2 candidates as in an x-wing. This is always the pattern shown in examples in tutorials. However in this instance 4 out of the six have 3 candidates.

Good Pondering The 3 columns have 3 or less giving you some removals on the rows 2 and 4.

[Pat]

hi eniacap,

a Swordfish is a "hidden pattern" of 18 cells (marked "/")
the 9 cells marked "X"
are those where the digit may occur

Code: Select all

../ ... /./
**X *** X*X
../ ... /./

**X *** X*X
../ ... /./
../ ... /./

../ ... /./
**X *** X*X
../ ... /./



[eniacap]

Thanks Pat. I can see the nine but clicking on the link to the "Hidden 18" brought me to a thread which was totally beyond my comprehension I'm afraid.

The point I was trying to make above was that all tutorials I have seen on the internet use examples of swordfish which have 3 column or rows which each contain 2 candidates and this is easy for a beginner to spot. Each candidate can be true or false and one of them the required number, thereby eliminating all others - just like an x-wing. It is when there are 3 candidates in a row or column (as above) when it becomes more difficult to understand.

[Bigtone53]

eniacap

A swordfish requires at least 2 appearances in only 3 rows (columns). Most puzzles seem to work on the 2 in each principle or perhaps a 3 somewhere. The Times SuperFiendish however typically works on what is basically the more obvious version of the swordfish, namely exactly 3 contenders in exactly 3 columns (rows), ie a special case of the above. I guess that most pencil and paper solvers would find such threesomes easier to spot.

Good luck