When I ran across, in www.sudowiki.org, Andrew Stuart’s treatment of a pattern he says is “poached straight from Sudopedia”, I thought there must be a serious disagreement in the Sudoku community over the use of the term “Franken fish”. Andrew’s example appears to be a specialized form of Franken fish, but from his article, entitled “Franken Sword-Fish”, one could take it to be the Franken fish. To avoid this impression, let’s label it a Sudo Franken fish, and reserve judgement on the “fish”.
By Stuart’s description, the Sudo Franken base units are three rows or three columns, two of which run through the same three boxes. The third line does not cross these boxes, and contains a slink of two X-candidates, aligned with candidates in the other base lines to form grouped slinks. In the third box intersecting the base lines, the Franken box, each base line contains a group of one to three X-candidates, shown as groups A and B.
In this pattern, the third line slink permits just one of the aligned slink partner groups to be true. One candidate in this group and one of the Franken box groups must be true. Any combination of true single slink partner, true grouped slink partner, and true Franken box group removes candidates in any of the red cells.
Spotting this pattern of three lines among a cloud of candidates gives many removals. The entire third line adjacent to the first two base lines is cleared, as well as the two crossing lines outside of the grouped slinks. Andrew shows several variations in the Sudo-Franken pattern, with their varying elimination consequences.
I wouldn’t label this pattern a Franken Swordfish, or any other fish. It’s not the configuration of candidates in units, as much as it is the specialized groups within the units, that defines this pattern. It does not involve complete base or cover units. Maybe you can relate it to something and name it appropriately. How about SudoPod? Applying it is no problem once you recognize it.
To follow up on the previous post:
Virginia reports that Sudoku Snake’s de-finned mutant swordfish has no kraken victims.
One victim is skewered by the fin. But if there is a mutant finned fish, there is also a kraken analysis. Did you do one? Here’s mine:
All nice loop victims in r4 also follow the loop in both directions to imply the fin. And finally, r6c2 and r6c5 follow the slinks of c3 and c4 in opposite directions to imply the fin.
Well, no luck this time, but that’s what sysudokies do with finned fish, regular or mutant.
Patient reader, this concludes my opening say about fish. With X-chains and fish, we have been on the X-panel for quite a long time. We will consider three more advanced techniques that are aided by the X-panel, namely coloring, alternate inference chains, and pattern overlay.
Coloring is next , and is reason enough to go for office support. I set up firstname.lastname@example.org in order to send blog readers the two small files containing the puzzle template .pptx and the bv tables, X-panels multiple and single, mutant fish finder and pattern overlay forms in a .docx. These forms are what you see throughout the blog, and are good for Sudoku solving, and for archiving and transmitting step by step solutions. Soon I am adding a page describing the sysudoku office tools, and how to use them. I don’t plan to upgrade my ©WordPress account to offer downloading these files from the blog, unless reader’s requests by email or blog comment become too much for me to handle. I do want to be in touch with solvers using these tools.
Want to work through another tough cookie with me? This one will be the second coloring example coming up. It’s another Megastar Maestro Niveau 9 from Fall 2009, number 50. I’ll have a boxmarking and line marking checkpoints in the next post.