The New Sudoku Players' Forum

[Yogi]

2...8.419......865..8...72317..2.358...8.16728.2...194.8....931541.7.286923618547
This puzzle is supposed to be solvable with a Finned X Wing (Hoduku). Clearly it would exist between Boxes 1245, but I can’t find it.
Maybe there is another elimination to come before the X Wing becomes visible.

[Leren]

Code: Select all

*--------------------------------------------*
| 2   f356  567 | 357     8    3567  | 4 1 9 |
| 347  139  479 | 123479  349  23479 | 8 6 5 |
| 4-6 *1569 8   | 1459   *4569 4569  | 7 2 3 |
|---------------+--------------------+-------|
| 1    7    469 | 49      2    469   | 3 5 8 |
| 34   359  459 | 8       3459 1     | 6 7 2 |
| 8   *356  2   | 357    *356  3567  | 1 9 4 |
|---------------+--------------------+-------|
| 67   8    67  | 245     45   245   | 9 3 1 |
| 5    4    1   | 39      7    39    | 2 8 6 |
| 9    2    3   | 6       1    8     | 5 4 7 |
*--------------------------------------------*

Grouped Skyscraper/ Finned X Wing (6) c25/r36 fin r1c2 => - 6 r3c1; stte

Leren

[Yogi]

Very good, thank you. It’s very obvious once it’s been pointed out.
Could you please fill out the meaning of ‘Grouped,’ and ‘Grouped Skyscraper’ a little as applied here.

Yogi

[Leren]

That's just my own term referring to the fact that one of the Strong links in the Skyscraper is not a simple one.There is a Grouped Strong link between the 6's in r12 and r6 in Column 2.

Don't know whether any one else uses this terminology, but you'll always find that a Grouped Skyscraper is finned X Wing and vice versa.

Have a look at this page on Andrew Stuart's site that discusses Grouped Strong links.

Leren

[SpAce]

StrmCkr has made a great list of the different variants of these short X-Chain patterns here (as a reply to one of your own earlier posts), including the grouped types. You'll see that the terms Grouped Skyscraper and Grouped 2-String Kite were used in that discussion, and not just by Leren.

I think those terms are quite standard, at least on this forum. Leren might remember that I also protested them like a year ago, but I've since changed my mind. I now think they're accurate and descriptive terms, and thank Leren for setting me straight!

ERs, on the other hand, are usually grouped anyway (in the hinge box) and the non-grouped variant is the special case there. For example, in Hodoku the non-grouped variants are not even reported as ERs by default, but it's an option. ERs could have group nodes in two boxes, actually.

[Yogi]

OK. Thanx to Leren. The Andrew Stuart discussion was interesting but It seems that (for now) for this P&P solver the term ‘Grouped,’ is a mindfield (pun intended) and it is easier to stick to the well-understood names like ER, Kite, Skyscraper, X Wing etc. The whole idea in my approach is to initially identify the candidates which could be applicable in any of these short-chain eliminations. I then look for the CP building blocks that the chains can be constructed out of, and finally check to see if any of them form those patterns. The more different types like Finned and Sashimi I know about the better. If all that comes to nothing, at least I still have the CPs which I can use to try other ideas with.
Good of SpAce to remind me of that contribution to my earlier post. I‘m now in a much better position to appreciate what StrmCkr was pointing out.
That's probably about all for now.