dailysudoku.com

[Asellus]

Kathy,

Here's the "horizontal" Skyscraper on <2>, discussed above, in your grid:

[Marty R.]

[nataraj]

Thank you, Asellus and Tracy for the kind explanations. I can see whence the term "color wing" and have also grasped the subtle differences between kite, skyscraper and turbot fish.

Not bad for a start into a new day ...

Regarding solving techniques, I do not believe in using the source of the puzzle as valid information in deciding which steps to try. Unique rectangles are regarded by some as a dubious technique because of the additional piece of information that the puzzle has a unique solution. Well I don't have a problem with that. But "knowing" that The Times super fiendish can always be solved by xwing or swordfish and "knowing" that Dailysudoku.com can always be solved by xy(z) (or whatever) wing and choosing one's route accordingly is going too far for my taste.

When the terrain gets rough, I tend do do a quick search for obvious x-, w- and xy-wings before trying the more time-consuming color wing, xyz, xy chain and medusa coloring as last resort.

I have a nagging feeling that in the end it is more a question of random chance which key move I see first than one or the other technique being superior (as we have learned many times that there are quite a few parallel paths to solution).

[duffy]

I reached Peter's grid copied from his initial post as follows:
+-------------+-----------+------------+
| 2 7 5 | 19 19 6 | 4 8 3 |
| 6 8 9 | 5 4 3 | 2 1 7 |
| 1 3 4 | 2 8 7 | 6 9 5 |
+-------------+-----------+------------+
| 48 29 3 | 7 69 45 | 58 26 1 |
| 5 1 6 | 3 2 8 | 9 7 4 |
| 7 49 28 | 49 56 1 | 3 56 28 |
+-------------+-----------+------------+
| 9 45 7 | 8 35 2 | 1 345 6 |
| 348 245 128 | 6 135 9 | 7 2345 28 |
| 348 6 128 | 14 7 45 | 58 2345 9 |
+-------------+-----------+------------+
I got stuck for a long time on the fascinating pairs in rows 4 and 6: there are three naked triples in 4, and three pairs in 6, but I could not make this delicious finding pay off. After a guess, I uncovered a different FINNED X-WING--I think I have this terminology right--than the one apparently referred to elsewhere: that is, 2's in r4c28, r8c28. The puzzle fell out when I noted, using this wing, that a 2 in r4c8 would dissallow any 2 in box 9, so the 6 in r4c8 breaks the puzzle.
I am describing all this because I could not remember much about a finned x-wing, which leads to a question: Is the finned x-wing a "standard" pattern one should look for?

Don D.

[duffy]

1) Correction: There are two sets of naked triples in row 4. (There is no conjugate hidden pair.)
2) Marty and Tracy: My memory indicates that you both have talked about finned x-wings before, so a word from one of you would be appreciated.
Thanks,

dd.

[Marty R.]

Duffy, this is a writeup I did on the Finned X-Wing.

[Glassman]

Marty — Grateful thanks for your explanation of a technique new to me, but not needed here, as I solved this 19 Sept 2007 puzzle using the xy-wing on pivot r9c6 and the skyscraper on 2s (a lovely and logically perfect technique only recently learned). I found this puzzle particularly difficult as the {28} UR, forcing a 1 at r8c3, a defective technique I will not use as it fails to prove the uniqueness of the solution, was so obvious.

This was Sam being really nasty. More please, Sam.

Glassman