In this post a SASdC trial plays like a regular hard puzzle from other collections.
The trial amounts to adding two clues, and seeing what happens. But it isn’t guessing, and instead of making a small gain, the trial solves the puzzle. Having given that away, here is the last ultrahardcore, UHC 399, for your homework.
Except it’s not the last. I slipped up on the spacing of the ultrahardcore review puzzles and did not cover the last 100, so I’ll add right page ultrahardcore 443 and 487. After that, I may just halt collection reviews and turn to refinement and publication of the Sysudoku Guide. I can be tempted to do another review if you can come up with another tough collection I can use as an excuse to keep drawing grids. I’m only 83.
On UHC 355, Beeby gets the nod to go to work on the Single Alternate trial after the follow up and reload of the previous post. As in all trials, new clues are marked as centered pencil marks. The first result is ALS_74.
Human ALS_XZ spotting depends on limiting attention to ALS with two singles as value groups, or a single plus group in a box for Z or aligned outside for X.
Next is an AIC ANL with two AIC nodes. It shows why the almost nice loop is more powerful than the logically equivalent 1-way or boomerang from 4r8c9. The latter two don’t remove 4r7c8.
An XY ANL wraps green! That’s remarkable in an ultrahardcore.
Then a hidden unique rectangle of 25 at r19c56,which depends on the corner slinks in one UR value and the opposite side slink in the other. The removal gets us to
Beeby’s “simple” AIC ANL. The 1-way starter is 7r8c4, where you’re looking at three 7 candidates to finish a chain. For a boomer, the ANL is a happy accident, because you’re looking for a closing confirming ANL.
I’m noting this as prep for the Guide, which is for human solving, not solver interpretation.
Another AIC ANL and an rasily spotted Type 1 UR.
The previous remarks apply to the AIC.
The blue ALS is eligible for spotting the overlapped ALS_42, because the 2 group aligns with an outside 2-candidate.
Somehow the human solver has to bring these two ALS together from her collection of eligibles. When you can do that, you’re wasting your valuable time solving Sudoku puzzles.
Just before the end, Beeby does the wildly inhuman. It’s a boomer from 6r7c2. When it reaches 4r7c1, that candidate is recognized somehow as the single value group in the weirdly constructed ALS 346259 in r23567c1 slinked to the 9 value group seeing 9 in the starter cell r7c2,
Then an ironic trigger for the collapse, a short AIC ANL.
One post late, the 10th Heine review puzzle, ultrahardcore 399, starts next week.
Systematic Sudoku 