dailysudoku.com

[Ruud]

Here is a puzzle from the long forgotten top95.

[Mogulmeister]

I've no idea but as it's you I shall say Death Blossom somewhere ?

[keith]

My opinion: A very unpleasant puzzle, for those of us who are inveterate pencil & paper solvers.

Keith

[Ruud]

This is a very deceiving puzzle.

You need 2 naked pairs and 4 hidden pairs to get to this point:

[TKiel]

My hat is off to the pencil and paper people who slogged through this. Salute!

[Ruud]

Hint:

.------------------.------------------.------------------. 


| 4     8     7    | 3     12    12   | 56    9     56   | 


| 59    359   39   | 6     48    48   | 2     7     1    | 


| 1     2     6    | 57    9     57   | 3     8     4    | 


:------------------+------------------+------------------: 


| 7     34    5    | 89    348   489  | 1     6     2    | 


| 69    13469 349  | 2     1346  57   | 8     34    57   | 


| 28    1346  28   | 57    1346  14   | 57    34    9    | 


:------------------+------------------+------------------: 


| 58    45    1    | 48    7     6    | 9     2     3    | 


| 3     67    89   | 1     28    289  | 4     5     67   | 


| 269   4679  249  | 49    5     3    | 67    1     8    | 


'------------------'------------------'------------------'

[Mogulmeister]

67 can not exist in r9c2 - all the other pairs are locked in. Is this in someway bug related ?

[Ruud]

This move is called BUG-Lite.

A size 12 Deadly Pattern is very rare.

Ruud

[Mogulmeister]

Many thanks for sharing that.

[Hasel]

That was tough ,, but i did it .







Hasel

[Ruud]

Here is another One-Trick Pony.

A rather nice step cracks the puzzle, but there is a less elegant alternative that achieves the same result.

[Johan]

[Steve R]

Johan,

Yes, that’s my preferred solution too. If you have not seen it, may I recommend bennys’s original paper on almost locked sets? It is a model of clarity, effectively covering every known xy…etc wing in one fell swoop. (The endorsement does not extend to the rest of the thread, which adds little you will not be able to work out for yourself).

The question remains as to whether this approach satisfies Ruud’s criteria of elegance. I have a sneaking suspicion that he might be attracted by a uniqueness argument, even if, so to speak, it is in reverse.

Steve

[Johan]

[Mogulmeister]

I often assume on the extreme puzzles that where an ALS elimination is likely to occur one of the ALS's is usually a bivalue. This often works.

When this doesn't work, one can build out from there. A sort of reverse engineer. I look for the eliminands then go back and find the restricted common candidate between the sets which supports/denies the elimination.