SudokuWiki.org
Strategies for Popular Number Puzzles
 Search:
 Page: 2023 2022 2021 2020 2019 2018 2017 2015 2014 2013 2012 2011 2010 2009 : 1 2 3 4 5 6 7 : 2008 2007 2006

# Feedback and Questions

I've received a lot of interesting comments and questions from Sudoku fans over the last few years and this page is where I try to answer them. I'm also directing Str8ts feedback here. Please feel free to drop me a note on the side of the page. Or you can email me directly at .

Part of the World you're from

Email Address (optional) so I can reply directly if necessary (it will not be displayed here)

Any Sudoku you want to publish here for easy loading into the solver - 81 characters, use '.' or '0' for unknowns.

 public - comment/question will be added to feedback column private - email comment/question directly to Andrew Stuart, don't display here

Comment, question or feedback:

letters you see:

Many thanks to all the people who have helped improve the solvers and strategies with their feedback!

## ... by: Vijay, Michigan

I discovered a uniqueness strategy that is not currently in the solver, and you may want to code it. In this sudoku:

D2,E2,E3,F2 have candidates 67,36,35 whereas
D7,E7,E8,F7 have candidates 67,36,57,135

Looking at the two columns we can conclude, F7 must be 1, otherwise the sudoku will lose its uniqueness.

Once again, great site!

Andrew Stuart writes:

I'm coded your strategy and I'm afraid it fails. :(

I have a counterexample here

...9...86......5......5...13261.9.547185..6934..36..1.2....3.6..4.69..7.67..1.4..

G3 = 1/9

G7 = 1/8/9 <-- should be able to set to 8

H3 = 1/3

H7 = 1/3

And 8 is the wrong answer for that cell. The reason you have gone astray is that unique rectangle are built from four cells with potentially the same pair. Your strategy calls for two different pairs in a rectangle + one other candidate, and you are implying that that extra candidate must always be the answer. In your example it is true. And I have found several examples where it does work - but only accidentally. 1/9 + 1/9 + 1/3 + 1/3 (in my example) does not imply a violation of uniqueness. Each set of two pairs will be determined by candidates elsewhere on the board.

Good try though.

## ... by: J, US

In your naked triples section you list possible triples as the following:

123, 123, 123
123, 123, 12
123, 12, 23
12, 23, 13

Would the following also be triples or not:

123, 12, 13
123, 13, 23
123, 123, 23
123, 123, 13

If they are not, what is the reasoning behind it? Following the logic from the first four, it seems these last four would also be triples???

Thanks!

Andrew Stuart writes:

Yes, all those are valid triples. I didn't set out an exhaustive list

## ... by: APPLECORE, USA

You may have answered this directly or by inference but: Is every Suduko puzzle solvable by rules of logic, without relying on trial and error? Or are some puzzles only solvable by trying a possible solution?

Andrew Stuart writes:

Every puzzle *ought* to be solvable by logic - it is after all a confined finite puzzle. But it has great depth and I do not know how to solve every puzzle with the strategies I have. Others may have more but I don't know of a general theory or a universal solver yet, other than brute force. People also split hairs over what constitutes logical as well, so the area is murky. Currently if I produce a set of puzzles at random and then grade them, about 0.2% will be unsolvable by my methods.

## ... by: Rick Cromer, Texas, USA

Here's another contender for the easiest Sudoku puzzle

Andrew Stuart writes:

To many clues I'm afraid (39). There has to be a maximum otherwise a sudoku with one number missing could be the "easiest". So partally finished puzzles are going to grade easier than full ones. Etc

Maximum clues are 32

## ... by: walt r, usa

kenduko27 doesn't seem to get the right answer and it rejects the correct answer???

Andrew Stuart writes:

I found a javascript bug which was preventing the solver from continuing (KenDoku only). I've fixed it and it should be going OK now

## ... by: Dick, USA

Problem solver indicated Unique 5 at D9 would lead to solution. I maintain this is an unsolvable puzzle since a "1 at D9" would also lead to a solution.

Andrew Stuart writes:

Dick, please click Solution Count, which finds 5 solutions. The solver is for sudokus with a unique solution. Specifically, the Unique Rectangles strategy, used by the solver, is not valid for sudokus which have multiple solutions. When given a sudoku with multiple solutions, the solver may be unable to find any solution or it may find one of the valid solutions.

Cool site.

## ... by: Geoffrey Howell, Australia

Dear Andrew,

I found this puzzle that your systems says cannot be solved yet has one solution. I believe you're interested in such examples.

Geoffrey

PS I use a compacted form to hold my Sudoku puzzles: A is one zero, B is two zeroes etc. It takes about half the space.

Andrew Stuart writes:

The message "Run out of known strategies" is valid. Not all sudokus with unique solutions can be solved with the known strategies. As new strategies are developed and adopted by the sudoku community, I plan to implement them in the solver, as I have done in the past.

## ... by: Gene, California

While solving a puzzle, one of your steps refers to "SINGLES CHAIN (Type 1)"
I can't find a description of this logic on your web page.
Can you help?
Thanks, Gene Tissot

Andrew Stuart writes:

http://www.sudokuwiki.org/Singles_Chains