dailysudoku.com

[Earl]

An xy-wing solves R3C9 and the puzzle.

Earl

[Marty R.]

Basics did it for me, I just never got to a point where I needed to look for anything.

[sdq_pete]

I reached the following position:

[Wendy W]

I did it exactly the way that Peter did: xyz on 125. Puzzle was very straightforward today.

[cgordon]

Marty found that only basics were needed - so it is not a VH. I often find it hard to take the basics to their limit. In this one, I happened to see a Type 4 UR which saved the trouble.

[sdq_pete]

By the way, I noticed later that in the position I posted, there is an X-wing on 1's (rows 1 and 7). And

[cgordon]

[Clement]

UR {2,6} in r37c23. 2 in r9c5 leads to a deadly pair {2,6} in Box 1 and Box 7 leaving 9 as the only candidate in r9c5 solving the puzzle.

[Marty R.]

[crunched]

There was an x-wing on 1 in columns 6 & 9
There was an x-wing on 5 in columns 7 & 9

[tlanglet]

We seem to have a variety of code sets after basics. The codes from Earl and Peter are different, Marty solved the puzzle, and mine is similar to Peter's but with the <29> in r7c5 deleted by the <269> subset in box 8. Also, I have an additional candidate <2> in r7c2 , which is apparently like Clement has, to form a Type 4 UR on <26> in r37c23.

[Wendy W]

Marty, that has happened to me many times. It it all depends on the sequence of methods you use. I used to follow the same sequence like clockwork: after basics I'd look for x-wings, then I'd move on to xy's and xyz's. Recently I've been looking for xy's and xyz's first just because they're more fun. I used to fill in the basics using rows; sometimes now I'll do columns or boxes.
Vive la difference.

[sdq_pete]

If a cell has two candidates and you jump (unknowingly) to a conclusion about the definitive value, then you have a 50% chance of being right.

Be that as it may, perhaps Wendy's remark is the chance I was looking for to make a little remark about the X techniques. I don't know how others experience it, but I find the ease of spotting X patterns to be - generally speaking - in inverse proportion to the complexity. Thus I generally find XYZ's the easiest to spot; then XY, then X-wings.
I guess this has to do with the degree of constraint: the XYZ is a highly constrained pattern - for an XYZ to be possible, there first has to be a triple and a subset bivalue within one block; only then need you look for a third bivalue outside that block.
X-wings I generally find difficult to spot, despite the logic of the pattern being simple. XY's ought to be easy to spot, but sometimes there are just so many bivalues hanging around, it's like looking for a needle in a haystack.
And of course, in a randomly generated puzzle, the more complex the pattern the less likely it is to occur, presumably.

Just wanted to get that off my chest.

Peter

[cgordon]

[Marty R.]

[DennyOR]

X-wings are first to try on my official list of advanced techniques, probably because that was the first advanced technique I learned, but also because they're pretty useful. In practice I usually look for other things first because X-wings are tedious, especially if you think you ought to be looking for finned X-wings and swordfish at the same time.

I think the most tedious part of a sudoku is getting to the stage where you can begin to use advanced techniques, so, after having to do all that work, I prefer a puzzle where I need to find two or more advanced techniques to solve it. I'd also like to see the very hards get gradually harder over time (which they may be).

I've often thought that the ideal sudoku puzzles would be ones that are pre-completed up to the point where advanced techniques are required.

[Wendy W]

Denny, that's an interesting thought, but for me doing the basics gets me "acquainted" with the day's puzzle in a way that just starting with advanced techniques wouldn't. It's part of the ritual. By doing the basics, you get a sense without consciously realizing it what numbers are predominating, which boxes are going to be a problem, that sort of thing.
x-wing was the 1st advanced technique I learned, too, but I've recently been learning the many adjunctive benefits of xy's and xyz's -- for instance, spotting naked triples.
Sudoku is not about the end result; it's about the process.

[Kdelle]

Denny,

I agree...and therein lies the beauty of the "sweep" function. At least the most tedious work is done!

I like "auto pencilmarks" at BB...but even their Super Hards are not as difficult for me as some of the Very Hards on this site.

I do think I'm getting spoiled, though. Like Wendy said, if you do it all right from the beginning, you get a sense of the puzzle.

I'm curious as to how many of us use the "Sweep" and "Auto pencilmarks" to get to the meat of the game.

Kathy

[wapati]

[DennyOR]

I print out the sudokus and do them on paper. That's probably very old-fashioned of me.