The New Sudoku Players' Forum

[nd]

It occurs to me to ask another question about Killer Sudoku puzzles: is it possible to design puzzles that omit sums for some portions of the diagram, or have cages that overlap--& for the puzzles still to be soluble & unique? It occurs to me that if the standard setup (requiring every cell to belong to exactly one cage, no more no less) were varied then this might make for an interesting, elegantly sparser variation, which would also be in some ways closer to the "original" sudoku (where many cells are left blank with no initial hints at all). It would also make the 45-sum rule trickier to apply, which might give some extra zest/difficulty to puzzles....

Anyway, I know there are some puzzles where you could leave out a few of the caged sums & they'd still be soluble (I say this because I've occasionally worked from partially illegible puzzles--there were a few days where the Times' puzzle online was really hard to decipher), but I wonder how many could be omitted before it became impossible to solve.

[tso]

Good idea. No reason at all overlapping enclosures wouldn't work. This is, in fact, what the Killer is to begin with! Think of the 9 rows, 9 columns, 9 boxes as special cases of the numbered enclosures for which the sum of 45 is assumed. No duplication is allowed within ANY enclosure. Still, overlapping smaller enclosures could be interesting.

Standard Cross Sums is nothing but overlapping sum-enclosures.

[Pi]

Have you tried Kakuru, they have those in the guardian and work on a similar principle to killer sudoku

[dalek]

Have you tried Kakuru, they have those in the guardian and work on a similar principle to killer sudoku

Yes. I have found them to be much harder, as they do not follow sudoku rules - or, not exactly. Also, on at least one occasion the printed puzzle has contained one or more mistakes (as corrected on the following day in the corrections column) so it doesn't fill you with confidence..........it's a shame because theguardian (as we must now call it) is my paper of choice. I am perservering with the kakuro puzzles but definitely prefer the killers.

[domt]

The independant also has an interesting variation with 4x4 boxes including 0-9 and A-F. Same sudoku rules apply but takes a good bit longer. Handy if you have a couple of hours to kill!

Dom

[Pi]

The Independent should forget numbers and just go with the letter A-P

[andylong]

t occurs to me to ask another question about Killer Sudoku puzzles: is it possible to design puzzles that omit sums for some portions of the diagram

Yes, there's one at the bottom of the page at
http://www7a.biglobe.ne.jp/~sumnumberplace/22326860/

[nd]

Thanks for the pointer--I look forward to digging into that one!

[udosuk]

That particular Killer Sudoku/Sum Num Place puzzle looks more difficult than it really is... surely took me less than 29 minutes to finish it once I figured out how to start. And it's got that artificial taste inside... you could feel the "challenge" laid by the puzzle creator (Ms Misawa?) as the cells are uncovered step by step. Don't get me wrong, it's an outstanding puzzle and I recommend everybody to try it out... but once you've solved it once all the secrets are known and there is no point to tackle it again, unlike some others where you could look for a better/shorter/more logical solution. I think "linear" is the word one Japanese puzzler described their puzzle design principle... not only the solution is unique, but the way to it is too (sort of).

In this page by Ms. Misawa about "kakuro/cross sums", I've managed to solve the bottom one (11x11) with much more strife. If that killer puzzle worths 4 stars of difficulty, then this kakuro surely worths 10! Even the last 10-20 cells could cost quarters of hours to figure out! I guess there is a reason why kakuro is more popular than sudoku in Japan...