dailysudoku.com

[nataraj]

One xy-wing (56-46-45 box 2 and 3) and a UR (45 in boxes 8,9) did it.


first the xy-wing makes r1c7=2, then (r6c7=5, r56c8={2,8} ) r7c8<>2, which leaves the UR:

one of r78c6 must be 6 or 8, together with "68" in r4c6 makes r9c6=5 and the puzzle is solved (I can never remember the UR types, I just know this is one of them)

[andras]

Actually I don't think you don't need the UR - I found that the xy-wing on its own was enough to break it.

Another very satisfying one, though.

John

[Johan]

The xy- wing mentioned by nataraj was the final step for me.

Note that the ER in Box 7 for digit <8>, also eliminates <8> in R9C6.

[nataraj]

[sdq_pete]

For the record, following the XY, further down the line I came across:
an X-wing on 6 in Rows 2 and 5 and
a W-wing on 25 in Blocks 1 and 3.

But I gather they were not really necessary.

Peter

[cgordon]

I used an X-wing R25C24 for <6> . Then an ER: R4/Box7 that removes <8> from R7C1.

[storm_norm]

[nataraj]

[storm_norm]

nataraj,

I believe if you go back to the x-wing on the 6's, there is actually a jelly fish on the 6's as well. it does nothing more than the x-wing did. but its interesting to find one.

[cgordon]

Storm Norm: You sure like them Swordfish, Jellyfish and Finned stuff. For me they are like looking for a needle in a haystack. If no other methods existed, I'd chuck it all in.

[nataraj]

this puzzle's got it all ...

amazing.

norm, I can understand your preference for fish'n'fowl, though myself I'm not opposed to more kinky chain based solutions ...

And I realize that the list of possible attacks was more a manifestation of the scientific mind than the actual desire to use them all at the same time.

Couldn't resist to quote that line from "Chicago", though

edit: for those of you don't have the dvd ready, here's the scene: http://www.weshow.com/fr/p/9707/catherine_zeta_jones_chante_cell_block_tango
(at time -3:30)

[nataraj]

on a more serious note (and seeing cgordon's post) - the methods I use really depend on what the grid looks like:

when there are still a lot of candidates in many cells, I start looking for xwing, fish etc., basically the search is on a number by number basis. That's when I find even rather long ("multi-coloring") chains.

If (like in this particular case) there are many cells with only two candidates, I try the UR, xy(z)- w- wing and xy-chain approach first

Once I've solved a puzzle I rarely go back to look for alternative ways

[storm_norm]

anyone who has had an easy sudodu book ( white belt sudoku by michael Rios for example) knows how it feels to kick out puzzle after puzzle. well for me, it was about looking for the 1's, for the 2's, the 3's and so on. x-wing strategy fits into that candidate search and until I researched online, I had no idea what the method was called.

however, coloring can come just as natural.
everyone can remember the first time they saw a sudoku they couldn't solve by scanning alone. oops, what do I do now?

ok, hmm, lets start filling in these candidates and see where I went wrong. OH NO !!! I did nothing wrong. what do I do now?

Cue the locked pair, A.K.A. the Naked Pair.

remember that puzzle?? the one where it looked like two solutions were possible in your budding sudoku career? you were shocked. Naturally, you said "impossible" if I put that number here and that number here, it doesn't work ( imagine that ). your first coloring experiment.

One locked set in a sudoku made you go berserk and now you are on the net looking up sudoku... why didn't I take the BLUE PILL ???!!!
for those of you who like the Matrix.

and down the rabbit hole you went.

welcome to the real world of sudoku. you can't go back to the way you were as an innocent bystander in the sudoku matrix. nope. you are free, free at last from ignorance and condemned all the same for knowing the truth...

once you were in a fantasy world with all the sudoku answers
now you can't skip over any of the biggest questions.

[DennyOR]

From Johan's position, the 56-64-48-85 XY chain did it for me.