SET
(Eingestellt am 16. März 2025, 06:00 Uhr von ICHTUES)
Rules:
- Normal Sudoku and variant rules apply.
- Diagonals: Digits along marked diagonals cannot repeat.
- Arrow: Digits along an arrow sum to the digit of the attached circle.
- Disjoint Groups: Corresponding cells in each box must contain the digits 1-9 exactly once.
- Let a digit's entropy describe if it is low (123), medium (456), or high (789). Let its modularity be its remainder when divided by 3.
Let a SET be a set of three digits such that their entropy and modularity are either all the same or all different. Some examples of SETs are: 789 (same entropy, different modularity), 258 (different entropy, same modularity), 168 and 348 (different entropy and different modularity).
In this puzzle, the three rows and three columns of each box all form SETs.
Additionally:
Solve online:
Sudokupad
Lösungscode: row 2
Kommentare
am 8. April 2025, 21:50 Uhr von tesseralis
Very nice. Even knowing the 'trick', there were some bits that I had to think about. Thank you for spreading the joy and madness of SET :D
am 8. April 2025, 10:41 Uhr von TheNineElements
Solve Time: 50:08
I rated this 4/5 difficulty.
This was a really interesting idea for a puzzle! The "break-in" was definitely the hardest part. This is somewhat expected/obvious since you (the solver) are required to figure out which SETs types and arrangements could exist in the puzzle - all before being able to make progress with placing digits/candidates.
I think a 4* overall difficulty rating is fair for this puzzle. Within the first few minutes of my attempt, I very quickly realized how the chosen selection of SET type(s) worked geometrically. Additionally, I understood which sorts of SETs could coexist (and that one of the SET types must be in the puzzle somewhere) but I found it really quite hard to "prove" that point. This "proof" was definitely the most time consuming part of the solve for me. This part of the solve was probably 4-5* difficulty, leaning towards 4*.
Once I had thoroughly convinced myself that my logic was sound, the remainder of the solve was much easier and probably 2-3* difficulty IMO. Even once I knew which SETs "belonged where", there was still a nice puzzle of actually figuring out the actual digit placements - which had a few tricky steps.
That is what led me to the 4/5 overall rating, on the slightly easier side of 4*.
The actual geometry of the puzzle and the arrow placements was very clever and creative, and it really was surprising that so few given clues managed to uniquely resolve this sudoku. The geometry of the rows/columns that arose from the SET rule was very interesting and surprising (at first). As an unusual side note: this ruleset actually results in a nearly identical geometric/digital pattern that arises from the combined anti-king+knight move ruleset. It obviously isn't truly identical (don't want to provide too much detail to avoid spoilers here), but I did find that interesting.
Once I had fully grasped how the rules worked, it was very satisfying to have such a big payoff once I determined the first SET type & "placement". My favorite aspect of the puzzle, though, had to be the logical interfacing of the arrows with the diagonals' geometries. It was really cool how the arrows meshed together with the rest of the ruleset and followed the same "theme" - all while disambiguating both the possible SET type(s) AND the actual digit values in the puzzle. This was a really creative idea & the construction/design was well-executed.
This was a very cool and interesting puzzle, thanks for sharing!
am 22. März 2025, 00:58 Uhr von adalg
Very enjoyable. After using your tesseralis link for training I found the path surprisingly straightforward and rewarding. A very catching methodology.
am 19. März 2025, 06:11 Uhr von waffles_the_dog
I was going to comment that if you're not aware of the game SET you might think SET is involved in the solve, but I see you've added a comment to that effect :D
I found the overarching logical structure of entropy-vs-modulus interesting and fun (many overlapping colors).
I found the later parts of the solve still suffered from the standard "disjoint groups" task of re-checking the disjoint group constraint.
