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 andrew@str8ts.com.

Post a Comment or Question here...

Wednesday 9-Mar-2011

... by: Sandy H, Chicago, IL

I am a daily user of this wonderful website, but am having trouble after upgrading my operating system from Windows Vista Home Premium to Windows 7 Home Premium. For some reason, the 2 print buttons and the e-mail button do not work when I am using Internet Explorer 8 or 9. However, everything works fine when I use the Mozilla Firefox browser. If there is a setting in Windows 7 or IE that I can change to solve the problem, or if you have any other ideas on what might be wrong, I hope you will let me know. Thank you for building and maintaining such a helpful website.

Andrew Stuart writes:

Do popups work at all in IE? You may have to enable them, somewhere in "options". Also, the popup may not be popups at all, but opening a new "tab" on the browser. Can you check that? I don't have that precise OS/browser to test myself, but nothing has changed this end...

Tuesday 8-Mar-2011

... by: Alan, South West England

What a fantastic site. It's made me realise I've only really done fairly simple Sudokus in the past, the solver took me to exactly the place I was stuck on doing the Times Super fiendish Sudoku and introduced me to pointing pairs (which seem obvious now you've shown me). Having dealt with them, the rest fell into place, so even the Super fiendish didn't go beyond level 6 pointing pairs. I can't see me ever catching Finned Sword-Fish! Definitely one to share.

Andrew Stuart writes:

Appreciate the kind words and glad it's working well for you : )

Solver gives no solution and no grading to this puzzle but rotate 90 degree anti-clockwise do give results....

Andrew Stuart writes:

The "Solution Count" for this puzzle is zero - which means there is a mistake in the puzzle or your entry of it.

Rotating a puzzle manually can also introduce errors.

I will today or tomorrow introduce a new version of the solver with a couple of left and right rotate 90 degrees buttons. Theoretically it should lead to the same solution and strategies used but the practice will be interesting. Often the search for an elimination on the tougher strategies will return the first instance of a result, but it searches top left to button right, so it may cause branching in a different direction

Sunday 6-Mar-2011

... by: JULIO, SPAIN

The solution count is a index of difficulty of sudoku? If not, can you implemented it. thank`s and regards. Julio

Andrew Stuart writes:

No, "Solution Count" is merely a quick brute force method of determining the number of solutions - any puzzle with more than one solution is broken and cannot be graded. I provide this so that people who get a contradiction or the solver cannot complete can know that the problem is with the puzzle, not the solver. To determine the grade (using the grading methods I employ), hit the "Grader" button

Friday 4-Mar-2011

... by: CynJ1171, Connecticut-USA

Just a thought about something that makes me crazy sometimes (lol like today!) - on the Killer Solver- if you accidentally hit the link to the "Strategy Guide" under the puzzle it opens in the same window and if you hit "back" to return- all your work in the solver is gone. Is it possible to make that link open in a *new* tab or window? It's really frustrating to lose all your work on a puzzle with miss-click.

On a side note- I adore Str8ts and Killer (although these take me forever to finish!)!

~CynJ1171

Andrew Stuart writes:

All the links on the solver pages have "_new" on them so they open in a new tab or window. However, where this fails is if the solver page is itself in a tab or window called "_new". This can happen if you navigate to the strategy pages and then go to the solver.

However, I am aware of this problem and it's not something I should leave to the users awareness. I will be implementing a auto-save cookie soon so that the current game is always re-loaded. This might break if you try and have two puzzles in two windows, but we'll see

In the first step everything is okay: SINGLE: D5 set to 1, unique in Row SINGLE: E6 set to 2, unique in Row

But there is bug in the second step: SINGLE: C5 set to 2, unique in Column SINGLE: E4 set to 3, unique in Row SINGLE: E7 set to 7, unique in Row (Here!) SINGLE: F2 set to 1, unique in Row SINGLE: G8 set to 1, unique in Row SINGLE: H4 set to 1, unique in Column

E7 can be set to 7 only after setting E4 to 3 (it means, E7 can be solved only in the next step, not this).

Andrew Stuart writes:

Generally and for all non-basic strategies the eliminations take place on the current state of the board - no 'look-a-heads' allowed. But with singles there can be so many at once it would be tedious to show them in series. So in this case - I am setting a single as I go through the search for singles - which may in some cases produce another single. It doesn't happen often - and it follows what a human would do with pen and paper. It's a shortcut but it doesn't break any rule - those new found singles would be got in the next step any way. Hope that clarifies.

Friday 25-Feb-2011

... by: kmor, Ireland

Certain suduko puzzles are unsolvable unless another solution methodology is used. What I am talking about is 'blind substitution'. When one gets to the level where no other methods are available, chose the first unsolved square, and blindly use one of the available candidates in that square. We dont totally disregard the other available candidates, but instead, save the puzzle as is and then repeat the list.. until either we are sure that the candidate we chose MAY be valid, OR an error is caused by duplicate numbers in row/col/square. If an error is created, we can return to the state where we substituted that candidate, and are sure that that number we entered cannot be a correct value. We try this for all candidates in this square eliminating invalid numbers.

Andrew Stuart writes:

There's lots about 'trial and error' strategies such as yours - which generally we want to replace with logical pattern based strategies.

"The whole idea of logical strategies is that we want to find a pattern or reason for making an elimination and/or solution at every point in the puzzle. I, and many others, are against the idea of guessing, or trialling a candidate and seeing if it produces an error. Mr. Crook has made an good attempt at showing how paper and pencil methods are perfectly viable and his paper does bring formality to the simpler aspects of the puzzle, but this is not the Holy Grail.

In a nutshell, Mr. Crook confuses a priori knowledge with a posteriori knowledge. His solution is gained through experience of the puzzle – the exploring of the paths and finding out which numbers don’t violate the rules. This is an a posteriori method. Logical strategists such as myself are looking for rules which give us surety before confirmation – an a priori method where we deduce the correct answer because we have identified a pattern that equates to a rule. Crook's theorems are such rules and are good solid a priori stuff - but he fails to show how to identify preemptive sets in ALL situations and resorts to guessing."

Tuesday 22-Feb-2011

... by: JD, Maryland, USA

Your site is wonderful - helps me tremendously incorporate more sophisticated strategies into my solving repertoire. However, once I get in the diabolical range, it seems just as quick and easy to simply choose a square with only 2 options left, try one, and follow the chain until either the puzzle is solved or a contradiction is reached. In the latter case, I then choose the other beginning option and follow it through to solve the puzzle. So my question is - how does one see the more complicated chains you list as strategies quick enough to make them worthwhile to use instead of the version of trial and error that always works? Are there keys to noticing when a diabolical or more advanced strategy is going to be likely to work?

Andrew Stuart writes:

Interesting question. I guess it's a question of logical purity. A good number of solvers, myself included, prefer to know why an elimination is correct rather than having a solution founded on trial and error, or guessing. So it may be more efficient to bust a bottleneck with an inspired guess but it doesn't advance our understanding. So it's a question of degree - in how much you want to know. There's quite a lot of debate about this - and about the extent to which some strategies seem to merge from purely logical into trial and error.

Saturday 5-Feb-2011

... by: Bill Harrison, Virginia, USA

Hello Andrew,

I want to express my appreciation for your outstanding website that makes it possible for me to analyze and modify my Sudoku solutions. I attacked (and solved) my first Sudoku puzzle in the Richmond, Virginia, newspaper at Christmas time in 2009; so you see, I am a quite recent entrant into this particular puzzle genre.

Having gotten hooked, I solved the daily Sudokus until around the middle of last year. As a programmer I began to think that creating the puzzles themselves would prove more challenging than simply solving someone elses'.

My computer printed out my first successful 81-digit Sudoku last July. However, creating a 30-or-so digits puzzle by randomly knocking out 50-or-so digits from my Sudokus has turned out to be rather frustrating. It was too easy to generate a redundant (i.e. non-unique) puzzle. But your Sudoku Solver has helped me immeasurably!

Entering my trial puzzles into your solver, I can judge their levels of difficulty by how many of your 35 strategies are needed to solve them. Also, it is relatively easy to see the puzzle combinations that cause trouble (and redundancy) by stepping through your strategies and then tweaking the puzzle distributions slightly to "fix" any difficulties.

Again, thanks so very much for your splendid site!

Best regards, Bill Harrison f8bill@gmail.com

Andrew Stuart writes:

Many thanks for writing in, appreciate your comments and compliments. Glad you're having fun making puzzles, its a very interesting area

## Wednesday 9-Mar-2011

## ... by: Sandy H, Chicago, IL

## Tuesday 8-Mar-2011

## ... by: Alan, South West England

## Monday 7-Mar-2011

## ... by: Lewis, Montreal

## Monday 7-Mar-2011

## ... by: Philip Khew, Malaysia

Load Sudoku: CLICK TO LOADRotating a puzzle manually can also introduce errors.

I will today or tomorrow introduce a new version of the solver with a couple of left and right rotate 90 degrees buttons. Theoretically it should lead to the same solution and strategies used but the practice will be interesting. Often the search for an elimination on the tougher strategies will return the first instance of a result, but it searches top left to button right, so it may cause branching in a different direction

## Sunday 6-Mar-2011

## ... by: JULIO, SPAIN

If not, can you implemented it.

thank`s and regards.

Julio

## Friday 4-Mar-2011

## ... by: CynJ1171, Connecticut-USA

On a side note- I adore Str8ts and Killer (although these take me forever to finish!)!

~CynJ1171

However, I am aware of this problem and it's not something I should leave to the users awareness. I will be implementing a auto-save cookie soon so that the current game is always re-loaded. This might break if you try and have two puzzles in two windows, but we'll see

## Friday 25-Feb-2011

## ... by: Ilmārs Cīrulis, Latvia

Load Sudoku: CLICK TO LOADSINGLE: D5 set to 1, unique in Row

SINGLE: E6 set to 2, unique in Row

But there is bug in the second step:

SINGLE: C5 set to 2, unique in Column

SINGLE: E4 set to 3, unique in Row

SINGLE: E7 set to 7, unique in Row (Here!)

SINGLE: F2 set to 1, unique in Row

SINGLE: G8 set to 1, unique in Row

SINGLE: H4 set to 1, unique in Column

E7 can be set to 7 only after setting E4 to 3 (it means, E7 can be solved only in the next step, not this).

## Friday 25-Feb-2011

## ... by: kmor, Ireland

Have a read of this:

http://www.sudokuwiki.org/Crooks_Algorithm

and I quote myself here:

"The whole idea of logical strategies is that we want to find a pattern or reason for making an elimination and/or solution at every point in the puzzle. I, and many others, are against the idea of guessing, or trialling a candidate and seeing if it produces an error. Mr. Crook has made an good attempt at showing how paper and pencil methods are perfectly viable and his paper does bring formality to the simpler aspects of the puzzle, but this is not the Holy Grail.

In a nutshell, Mr. Crook confuses a priori knowledge with a posteriori knowledge. His solution is gained through experience of the puzzle – the exploring of the paths and finding out which numbers don’t violate the rules. This is an a posteriori method. Logical strategists such as myself are looking for rules which give us surety before confirmation – an a priori method where we deduce the correct answer because we have identified a pattern that equates to a rule. Crook's theorems are such rules and are good solid a priori stuff - but he fails to show how to identify preemptive sets in ALL situations and resorts to guessing."

## Tuesday 22-Feb-2011

## ... by: JD, Maryland, USA

## Saturday 5-Feb-2011

## ... by: Bill Harrison, Virginia, USA

I want to express my appreciation for your outstanding website that makes it possible for me to analyze and modify my Sudoku solutions. I attacked (and solved) my first Sudoku puzzle in the Richmond, Virginia, newspaper at Christmas time in 2009; so you see, I am a quite recent entrant into this particular puzzle genre.

Having gotten hooked, I solved the daily Sudokus until around the middle of last year. As a programmer I began to think that creating the puzzles themselves would prove more challenging than simply solving someone elses'.

My computer printed out my first successful 81-digit Sudoku last July. However, creating a 30-or-so digits puzzle by randomly knocking out 50-or-so digits from my Sudokus has turned out to be rather frustrating. It was too easy to generate a redundant (i.e. non-unique) puzzle. But your Sudoku Solver has helped me immeasurably!

Entering my trial puzzles into your solver, I can judge their levels of difficulty by how many of your 35 strategies are needed to solve them. Also, it is relatively easy to see the puzzle combinations that cause trouble (and redundancy) by stepping through your strategies and then tweaking the puzzle distributions slightly to "fix" any difficulties.

Again, thanks so very much for your splendid site!

Best regards,

Bill Harrison

f8bill@gmail.com

Best of luck