This is a sysudoku grading review of the Super Tough puzzles available on the KrazyDad website.
KrazyDad is the creation of the very creative Jim Bumgardner. It offers for download a huge number of free Sudoku puzzles at levels ranging from Easy to Insane, along with a mind boggling array of other kinds of puzzles. That’s www.krazydad.com. The Super Toughs are almost Insane, being next to the highest level. Let me just say I’m glad I started there. If Jim likes what I’ve done here, and after I’ve posted a bit more Sudoku theory I have in mind, I just might go Insane.
My grading survey sample of 10 puzzles were the #5 puzzles from volume 5, books 1-10. There are eight puzzles in a book, and 100 books in volume 5, so my review sample is a systematically scattered smattering. The conventions of the review table below are explained on the review page.
The KrazyDad survey Super Toughs graded widely across the advanced methods, from bv scan to advanced coloring. The notable exception was the book 10 #5, which ran dry inbox marking at number 5, with only four numbers unassigned.
Super Tough starting clues are generous, and line marking tends to be easier than the Sheldon and Longo puzzles just reviewed. A large bv field usually supports interesting bv scan activity.
The review sample brought some remarkable coincidences of XYZ wings, XY-chains and finned fish occurring in pairs. There were also multiple techniques removing the same candidate and removing more than one candidate from the same cell. I’ll show these in my snapshots next post, but right now, let’s look into those alternative solutions of Super Tough v.5 b.1 #5.
Did you find the UR? I missed it at first and went on to coloring to encounter a very interesting color wrap. I’m reporting that first, then the somewhat unlikely UR.
My blue/green cluster traps an indecisive 4-candidate, leaving a lot of uncovered bv toward the southwest. The red/orange clashes with it in r9c9, and in the S box. The bridging logic is
r9c9 => !(green & red) => blue | orange
No candidates see the toxic sets, but
Orange =>(W5,W6=>C6m) and extends blue/green, forcing two blue 8’s in c7.
The green and orange candidates win. Now repeat after me:
I believe in coloring .
I believe in coloring.
. . .
A 3 or 4 in r3c4 forces the deadly rectangle, but neither does it alone, as the removal requires. The branched forcing chain shows that 4 in r9c4 is capable of that feat.
The highlighted candidates are an equivalent form of marking I have used earlier in the blog. I prefer the more explicit forcing chain. This is also an example of Denis Berthier’s xynt-chain.
The UR removal is not enough. We reach an impasse after
S5=>(SE5=>SW1=>SW4=>SW3, C5=>W6, W5, Sr8np23=>Sr7np48).
But the UR has enriched the bv network, and there is good reason to return to the updated bv map. I think you should see both the XYZ map and the bv map at this point.
In the XYZ map, the crossed out hinges mark the many “almost” XYZ-wings that were possible, but unfruitful. My snapshots of the next post show many successful ones across the 10 survey puzzles. On the bv map, the XY net is built on two main curves, red and black, plus purple and green extensions and connections. The curves limit and organize the search, making long XY-chains easy to find. An extreme solving method becomes merely an advanced one. No big deal.
The resulting naked pair in the N box triggers a raging epidemic through the bv:
After the upcoming Super Tough snapshots, we’ll conclude this KrazyDad review with a second puzzle, volume 5, book 8, number 5, the second from the review to reach the coloring level. It has coloring logic I had no examples for, in my coloring posts. If you would like to work through it with me, here is the starting grid. I’ll have the box marking trace next post.
I’m looking forward to reviewing KrazyDad’s Insane collection. I think you will too, after you see my Super Tough snapshots. Don’t wait that long to download these collections.