dailysudoku.com

[keith]

I have not yet looked at this one ...

[David Bryant]

Hi, Keith!

Just for grins, I ignored the "UR" in r18c56 ... I found an interesting "Nishio" sort of move on the "6"s instead. dcb

[keith]

Dvaid,

If you do not use the UR, there is an XYZ-wing in the lower left corner that solves it.

Keith

[Marty R.]

I thought this one wasn't up to the usual standards of his six-star ratings, as it was solvable with nothing more than the basic techniques of pairs, triples, etc. and locked candidates.

[David Bryant]

[keith]

The XYZ- wing is on <168> in R7C3. It takes out <1> in R7C1.

Let me try to explain. Both XY-wings and XYZ-wings are naked triples spread over more than one box or line. The XYZ-wing is much weaker, and much less common.

For example:

Here is an XY-wing. It eliminates Z from the cells labeled #. (The common buddies of the "wing" cells, the ones containing <Z>.)
[9/3: Edited to fix an error in the following diagram. Thanks, Tracy!]

[TKiel]

Keith,

I know you know this and it was probably just an oversight, but wouldn't an XY-wing also take out the Z candidate in the (#) cell?

[Marty R.]

Keith/David,

Obviously you guys enjoy looking around for more advanced techniques to use than are required to simply complete the puzzle. So the question I pose is this: at what point in doing a puzzle do you stop eliminating candidates with basic techniques and start looking for more advanced stuff?

[TKiel]

[ravel]

Tracy, you would convince me to use a program, but Marty ??

Marty,
to your question: it depends on, what candidates i already have filled in. Sometimes i also can see a short chain with few candidates, before i am able to spot an x-wing, xy-wing or coloring. Also xyz-wings are sometimes easy to spot, it is like looking for triples "round the corner".

[David Bryant]

[Marty R.]

Tracy/Ravel,

I haven't gotten into using the various solvers since I'm not very tech-savvy.

David,

Maybe I misinterpreted something. For me, this puzzle required only very basic techniques, but with all the talk of URs, Nishio, XY, XYZ and chains, I didn't realize that you went with basic as far as you could.

[TKiel]

[David Bryant]

[Marty R.]

[David Bryant]

Marty --

The cool thing about the "UR" techniaue -- especially the Type 1 UR -- is that it's so easy to spot and so simple to use that one can easily start thinking of it as a very basic technique. I've even used it that way on some simple puzzles in the newspaper -- not because I couldn't see some other way to go, but because I was in a hurry and the inference "have to put a 2 right here" seemed so simple. Heck -- I've even gotten into the habit of looking, when I spot a pair that lies in one 3x3 box, to see if there might possibly be another pair somewhere that lines up with this one and gives me a "UR." It's a great device for visualizing the structure of a puzzle. dcb

[Marty R.]

David, that's true, a Type 1 is easy and could be a basic technique, perhaps even easier than some techniques widely considered to be basic. But I tend to stick with what I learned when I was just starting out, and URs were considered advanced. So when I refer to basic techniques, it's pretty much limited to pairs, triples, etc., and locked candidates.

[AZ Matt]

My two cents on when you start looking for more complicated solves:

I have done very difficult puzzles that were boring and very easy puzzles that were fascinating. As this forum demonstrates, I think we are all interested in puzzles that have multiple solutions involving varying degrees of difficulty. I define boring as puzzle that plods forward on one path, and fascinating as puzzles that show opportunities for multple paths and new or rare solving techniqes.

It is true, however, that the very best puzzles are the ones that can ONLY be solved by using a new or rare technique, whether they plod to that point or offer multiple ways of getting to that point.

So my answer to the question is, "It depends on the puzzle..."

[keith]

[Marty R.]