Methodology in Solving Sudokus

Advanced methods and approaches for solving Sudoku puzzles

Methodology in Solving Sudokus

Postby McBainCo » Sun Aug 14, 2005 3:25 pm

To truly be a logical puzzle, Suduku should only be solved by logical steps ( i do accept Nishio as a logical step, in that it is testing for a hypothese).

The difficulty which I have is that among the range of technics available, it is sometimes difficult to select the appropriate one.

I have come across numerous puzzles which require sophisticated solving technics.

The problem is that the selection of the technic is not a logical step.

When I am stuck, I usually check for:

1. X-wing
2. Swordfish
3. Colour
4. Nishio

This sequence has led me in numerous difficulties ( in various puzzles) because although all conditions were right for using sworsdfish (i.e all the candidates were identified), I still was unable to solve the puzzle.

This suggests, that not only the solution of a sudoku has to be unique, but also the sequence of technics to solve it.

I am not comfortable with that conclusion
Posts: 1
Joined: 14 August 2005

Postby lunababy_moonchild » Sun Aug 14, 2005 9:22 pm

It is my humble opinion, and only my opinion, that the use of various techniques is overrated, at least somewhat. And that the use of said techniques over-complicates the puzzle.

It seems to me that the rules of the game do not imply the need for solving techniques per se, just logic : insofaras 'that can go there because.....'.

Obviously people are free to use such techniques as they desire, should they desire to do so and perhaps my opinion is coloured - excuse the pun - by the fact that I really can't be bothered to learn said advanced solving techniques and that I haven't gotten any further than Fiendish level.

Posts: 659
Joined: 23 March 2005

Re: Methodology in Solving Sudokus

Postby Moschopulus » Mon Aug 15, 2005 1:21 pm

The puzzle is in choosing the shortest route to the answer. Even computer programs have to solve puzzles by doing certain things in a certain order. If one program does these things in a different order to another program, it may finish quicker or slower.

McBainCo wrote:To truly be a logical puzzle, Suduku should only be solved by logical steps ( i do accept Nishio as a logical step, in that it is testing for a hypothese).

If you allow testing for a hypothesis as a logical step then here is a logical method which will solve any sudoku.

The total number of possible grids is known to be 6,670,903,752,021,072,936,960. Make a list of all these grids. You go through these grids one by one until you find one that fits your puzzle. At each stage you are hypothesising that "this grid is the solution".
This is completely logical and will always work. (Of course nobody would actually try this but it illustrates the point)

Now, depending on how you ordered the grids, you might finish in 1 second or in millions of years.
Posts: 256
Joined: 16 July 2005

Return to Advanced solving techniques