dailysudoku.com

[nataraj]

a good one for xy-wings ...

first in boxes 1 and 3 ( 68-67-78 ),
then boxes 5 and 6 ( 59-56-69 )
and finally, boxes 4 and 6 ( 57-67-56 )

if I remember correctly, the last one was in a BUG+1 situation, but if possible I try to avoid the uniqueness argument and try to find xy-wings or xy-chains instead, which should always be possible with so many bi-value/bi-location links available.

BTW, does anybody know of a BUG+1 situation that cannot be solved using xy-chain or coloring techniques?

[sdq_pete]

By way of confirmation, I came across the same 3 XY's.

Peter

[Marty R.]

I used those XY-Wings, plus one more that was extended with coloring. Before I found that last puzzle-busting Wing, I also found a couple of W-Wings. A fun puzzle.

[Earl]

I used an four step xy chain from R4C4 to R5C8 to eliminate <5> in R4C8 which opened the puzzle. I don't particularly like chains, but I saw it and it worked.

Earl

[cgordon]

I brilliantly used a combination of an Empty Rectangle on 8s (C2, T hinge in Box 5 and target C4R8); an x-wing on 7s (C68); and a skyscraper on 7s in R59 (target C9R8).

[eddieg]

"I brilliantly used a combination of an Empty Rectangle on 8s (C2, T hinge in Box 5 and target C4R8); an x-wing on 7s (C68); and a skyscraper on 7s in R59 (target C9R8)."

I got a chuckle out of that. You are brilliant on my view.

[tlanglet]

I initially found two swordfish, one on 2 and another on 6, but they did not seem to help very much. After I found the third XY-Wing, it finally fell in place.

[duffy]

I find myself never using swordfish these days: other techniques seem to make them unnecessary. I do, however, spot xy-wings more easily than ever as a result of experience, I suppose. It turns out I used exactly the three initially reported by nataraj. I would say, however, that it is not just a question of an abundance of bi-value cells, but also a kind of "diagonal adjacency" (for lack of a better phrase) of lots of bi-value cells that don't in themselves make hidden pairs. (To a degree, hidden or locked pairs are not good candidates for xy-wings; the pair can never be used directly in the wing.) Perhaps this is what nataraj is alluding to when he says there are "...many bi-value/bi-location links available."

Don D.

[crunched]

From time to time, I go back and do an archived puzzle. I am not good at finding wings, but I am getting better at it. I notice that all you "pros" used wings to solve this. I never found a wing here. But I did solve this puzzle. I found a UR consisting of 3 68s in boxes 7 & 8, along with a 678. This eliminated the 68 in that one box, leaving a 7. This solved the puzzle for me. Actually it was a UP (instead of UR), where P stands for parallelogram.

[Asellus]

crunched,

I assume you are referring to one of the archive puzzles, not the VH of 18-Apr-2008 discussed above since <6> was a given in box 8.

A parallelogram (non-rectangle) arrangement of 4 matching bivalues does not form a deadly pattern, even when limited to two boxes. So, I'm afraid you might have gotten lucky on this one.

With just a little practice, XY Wings and XYZ Wings are not difficult to spot. There is a systematic way to hunt for them in a grid. Take a look at nataraj's post here. (Earlier in that same thread is a post about hunting for X-Wings. I find it efficient to hunt for X-Wings and Skyscrapers at the same time.)

[crunched]

[Asellus]