This post finishes the snapshots of significant solving events from Extreme review puzzles found in Paul Stephens’ books Mastering Sudoku Week by Week, and Sudoku Addict Workbook.
In the last post, we left Addicts 138 with removals you could follow up for a 3-chain removal in r7c7. The next snapshot is the coloring after that.
It’s truly remarkable to have a unique rectangle generating a color bridging conflict. Blue and red leave a deadly rectangle in r59c79, therefore either green or orange is true. The unfortunate 3-candidate sees both.
With this removal the red/orange cluster expands to paint the 3-candidate in r5c9 orange, removing the one in r5c7, and thereby merging orange and blue. Traps of the enlarged blue/green cluster confirm blue.
The next happening I have for you occurs in Addicts 146. It did provide a hidden pair in line marking, but then an ordinary XYZ and XY change was putting me to sleep, when I decided to return to an implausible UR before coloring.
I was trying forcing chains from the 1-candidate in r2c1 to see if it could possibly see the 2 in r9c5, when they joined up, I don’t remember where, to form an almost nice loop eliminating my 1-candidate! The ANL is almost an XY chain, but actually combines XY chains with one, optionally two slinks. So it is an AIC ANL.
Fortunately, the removal also removed the threat of a deadly rectangle. At right is a trace to the coloring of Addicts 146. The 9-wing and 17-wing smile at you from the next snapshot. The 9-wing removes two 8-candidates; the 17-wing, two 8-candidates.
Then below, the finned 8-wing removes two more 8-candidates from r6. When the two-cluster coloring is completed, blue and red candidates in the same cells announce that either green or orange is true. With of the removals from r6, when 8 in r6c8 is banished for seeing green and orange, the removal confirms orange.
Our snapshot of Addicts 150, the final puzzle and our last review puzzle, is a bit of a bummer. Paul’s introduction anticipates a swordfish to start, with many advanced techniques then required. Paul even hints of a WXYZ-wing, which I have probably passed over many times, but have never found.
The 7-candidate starting the chain manages to “see” both the 7-candidate and the 4-candidate that prevent the deadly rectangle, so it must go. But then the bv riddled puzzle collapses in a heap.
It’s a case of a puzzle being extremely tough under Stephen’s basic solving, but offering up too many bv under slink marking. Its not a fault, really. We want the possibility of a collapse anywhere along the way.
Next up in the blog is a review of Will Shortz’ toughest puzzles from Ferocious Sudoku. Perhaps you’d like to see how far you can go with the box marking of Beware 160: