killer su dokus

Advanced methods and approaches for solving Sudoku puzzles

killer su dokus

anyone got any tips on how to complete these, especially how to get the first few numbers on the more difficult puzzles?
kuff28704

Posts: 13
Joined: 31 August 2005

Re: killer su dokus

kuff28704 wrote:anyone got any tips on how to complete these, especially how to get the first few numbers on the more difficult puzzles?

Every row, column and block must sum to 45 - you are not limited to the sums initially given.

As an example, look at the "deadly" one:

Top right block has a group adding to 11 and a group adding to 21
.'. The remaining two must add to 45-(11+21)=13.
.'. In the centre right block, the remaining two cells of the group of 4 adding to 26 must sum to 13.
.'. The bottom left cell in this block must be 45-(7+13+6+11)=8
.'. In the bottom right block, we have 6 cells that we are told sum to 25
.'. The remaining 3 cells sum to 20 - but those 3 cells are part of a group of 5 that we are told sum to 31, and we already know one of those numbers (8)
.'. The top right cell in the bottom middle block must be 31-8-20=3

There are other places in this puzzle you can take a similar approach - (note that the puzzle is symmetrical).
whohe

Posts: 32
Joined: 28 May 2005

wow! thanks so much, that really has unlocked how to do them for me!

i'll post again if i get stuck, lol
kuff28704

Posts: 13
Joined: 31 August 2005

Hi Guys,
First post, great board. Had got that far on the working out of these great puzzles. But there must be more tricks out there to find.have managed the tricky one but have yet to attempt the others. My initial thought is there is bound to be puzzles that you cannot employ this technique initially so where do you start then?
domt

Posts: 16
Joined: 01 September 2005

This puzzle is a hybrid of Sudoku and Cross Sums. Search the web for "cross sums" and "stategy" or "tactics", etc. Some (but not all) of what you find may be helpful. For example, here and here and here.

The TIMES left out one important rule of this puzzle -- no digit may appear twice in any of the numbered enclosures. Without this rule, some puzzles may have multiple solutions.
tso

Posts: 798
Joined: 22 June 2005

tso wrote:The TIMES left out one important rule of this puzzle -- no digit may appear twice in any of the numbered enclosures. Without this rule, some puzzles may have multiple solutions.

Are we absolutely SURE that is a rule? It would be logical, and is very helpful - for example, I had assumed that an L-shaped block adding to 6 could be 1,4,1 if it crossed the boundary between two 3x3 squares. This rule would mean it could only be 1,2,3.
nputtick

Posts: 3
Joined: 22 May 2005

I'm sure, though no one else seems to believe me. I've seen several dozen puzzles, with no exception. Read my comments in this thread .

But ... you don't NEED to be sure it's a rule for it to help! Feel free to solve the puzzle without applying this rule. As long as you have a logical basis for each placement, you'll either find the solution or come to a standstill that can only be bypassed by assuming no duplicates in one enclosure. At that point, you will have both solved the puzzle, and PROVED THAT THE RULE IS REQUIRED for a unique solution.
Last edited by tso on Sun Sep 04, 2005 2:34 am, edited 1 time in total.
tso

Posts: 798
Joined: 22 June 2005

i agree it is definatley a rule. the puzzles would seem sumwhat easier if it wasn't. If it wasn't multiple solutions would be avalible and therefore no competition sudokus could be set !!!!
shezzer

Posts: 1
Joined: 03 September 2005

After you showed that the "Sum Doku" that you linked to somewhere else was made unique with this rule, I was convinced...
Lardarse

Posts: 106
Joined: 01 July 2005