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. Please feel free to drop me a note on the side of the page or try the Facebook
comment box. Or you can email me directly at andrew@str8ts.com.
Post a Comment or Question here...
... by: Craig, USA
Load Sudoku: CLICK TO LOAD
++++  29 7 6  59 25 1  3 48 48   13 8 4  36 7 369  19 2 5   123 35 2359  23 4 8  7 6 19  ++++  5 13 38  7 6 2  89 14 49   6 19 7  59 8 4  2 15 3   4 2 89  1 3 59  6 58 7  ++++  28 6 35  4 9 35  18 7 128   2378 4 23  236 1 367  5 9 268   279 59 1  8 25 567  4 3 26  ++++ It doesn't find the xwing in columns 3 & 4 on 2's?
Andrew Stuart writes (11Oct2009):
I stepped through to the position below. I don't see how you eliminated the 2 in G3 to make the xwing. My next step would have removed
SINGLES CHAIN (Type 1): Removing 2 from H1
Making it beyond your board above
... by: Bob Kukla, Dallas Texas
I use your website everyday. It would be nice to have your program on my laptop for use without using the Internet. Do you sell it in that form?
Thanks.
Andrew Stuart writes (11Oct2009):
I did make a windows version but I couldn't finish it. Not to the standard I'd have liked. It is something I do want to finish when I have time
... by: PLH, New York
Load Sudoku: CLICK TO LOAD
Hello,
Thanks for the great service you are providing us!! I tried your site for the first time a couple of days. Today being frustrated not seeing the XWing in this puzzle I entered the information (twice) and twice got an error after clicking on "next step". I cannot figure out what is creating this other than a bug.
Could you please check it out. Thank you. Philippe
Andrew Stuart writes (11Oct2009):
Someone was hammering the server and it made that part stop. Xwings is the first strategy that calls this server side service. Its working for me at the moment
... by: donged, usa
Just want you to know how much I appreciate the incredible service you provide. Every time I reach a point of frustration, and am sure there is no solution to the puzzle, I go to this site and PRESTO  there it is. I would have stopped attempting the tough puzzles long ago, if not for you. THANK YOU.
Andrew Stuart writes (4Oct2009):
Thx!
... by: Florence, UK
Load Sudoku: CLICK TO LOAD
I was just using your solver with the sudoku above and it suggested the next move was linked to "XCYCLE on 5 (Discontinuous Alternating Nice Loop, length 10): 5[D1]5[D2]=5[G2]5[G4]=5[F4]5[E5]=5[E7]5[J7]=5[J1]5[D1]  Discontinuity is two weak links joined at D1, 5 can be removed from that cell" I am just learning about XCylcle's and wanted to check 2 things: a) do the links have to alternate between weak and strong and b) why isn't F4E5 a strong link? Thanks.
Andrew Stuart writes (19Sep2009):
Have a read of this page: Weak_and_Strong_LinksString links can be weak links to allow the alternations  its just weak links which can't be strong
... by: JPAV, USA
Load Sudoku: CLICK TO LOAD
Unless I'm missing something, this is a puzzle where the 'diabolical' XYZ Wing strategy can't be applied correctly, but the 'tough' Swordfish strategy can (which is the one your solver applies first). The XYZ strategy seems the easier to identify for me, so I'm used to applying it before Swordfish, but here it doesn't seem to work. I get to the wing represented by E2F3H2, which should allow me to remove 4 from D2. That creates 2 more wings for D1D2H1 and D1D2H2, the latter allowing me to remove 5 from E2. That leaves 5 as the solution for E8, which is wrong. Thus, it appears the Swordfish strategy must take precedence over the XYZ Wing strategy in this case. Am I doing something incorrectly?
Andrew Stuart writes (19Sep2009):
D1D2H1 and D1D2H2, the latter allowing me to remove 5 from E2
D1D2H1 is a valid XYZwing but you cant eliminate on E2 because E2 cannot see all three cells.
D1D2H2 is not a valid XYZwing because the hinge cell only contains 2 numbers.
Hope that helps
... by: TMH, USA
Load Sudoku: CLICK TO LOAD
Two things: the sudoku I've entered above ("Diabolical" because it uses the "almost locked sets" strategy), and that you can add iCab to your "Tested on:" banner at the top. That is what I use and it works perfectly.
Andrew Stuart writes (19Sep2009):
Thanks TMH, I'll add that next update, good to know
... by: Dave Scott, US
Load Sudoku: CLICK TO LOAD
This was a medium level puzzle generated at Road Runner, I believe listing you as the creator. It seems that recent puzzles have more frequently completely blocked solving by applying the three rules (no repeats in row, column, or square). And those weren't identified as particularly hard or fiendish. This forces more complicated strategies, such as the one in this puzzle that when the possibles are listed the nine toward the bottom of the puzzle is the only possible entry for that cell. The Solver goes on to display multiple changes at once and when rows and columns are simultaneously affected, sometimes as many as 67 cells solved at one time, it becomes very confusing to try to follow everything going on at once. Suggest perhaps changing the single step (or make it an option) so that only one cell is changed for any given step/cycle. An alternate would be to suggest changing only rows or columns at the same time, but not both, etc. Thanks, Dave
Andrew Stuart writes (10Sep2009):
Yep, I provide these puzzles, although someone else designed the player. Regards the solver, its is generally designed for diffcult puzzles. The multiple solutions you are experiencing in one step are easy singles which don't impact on each other. They are all in the same round because they can all be found at the same time. If I made this part step by step it would lengthen an already long process of stepping through. If I made it an option there's still the the quandrary of which ones to show first (the order would be top left to bottom right) but the human solver would be finding these singles all over the place. What I've done is a compromise over a lot of conditions :)
... by: mimi, Vienna, Austria
Load Sudoku: CLICK TO LOAD
Hi Andrew, why does the TrialandError method not solve this Sudoku sent in by Geoffrey Howell, see Thursday 30Jul2009? I tested the Sudoku. A twostep trial is enough to solve it, so rather an easy one, well, relatively speaking. br, mimi
Andrew Stuart writes (10Sep2009):
I don't have many trial and error strategies (well 2) so I can't solve this yet. I would *like* to solve it logically without guessing. Its useful to have these however, thanks for sending in
... by: EHK, USA
Solver no longer works beyond simple techniques. No longer free? Where / how do we sign up and pay?
Andrew Stuart writes (10Sep2009):
If its not working the the service on the server has got gumbed up and I need to kick it. No plans for a paid access to the solver. Thanks for the offer though :)
I'll give it a restart
... by: Chris Armstrong, Ladner, British Columbia, Canada
I have been doing the difficulty 4, 5, 6 Sudoku puzzles since the Vancouver Sun started publishing them in spring 2005, I believe. I NEVER LOOK AT AN ANSWER, since I keep at them until they complete successfully. I also have done the most difficult ones in 2 books. I recently purchased Will Shortz "ferocious sudoku" and have done several of the "Beware, Very Challenging" ones. I enlarge ALL puzzles to about 5.5 inches square on 8.5x11 paper, for easy pencilling. On several, I have got as far as I can go with my own limited logic, and I obviously have another "trick" to learn (which will NOT happen on my own, I am afraid !). I can complete ANY of these by making 2 copies, picking ONE NUMBER of a pair, and seeing if the puzzle completes or deadends (then carry on making double copies, as required). Obviously, if the 1st number is wrong, the other MUST be right, etc., etc. I have never read any "rules", but I would like to be able to complete WITHOUT making ANY copies!! MAY I EMAIL YOU A PUZZLE in which I can go no further, and you can show me ONE additional move (and the logic behind it !). Maybe that is all I need to complete ALL with NO COPIES !? If not, I will be back for another bit of logic ! Will Shortz put me on to this web site ! THANKS
Andrew Stuart writes (10Sep2009):
It is best if you load the puzzle into the solver and step through to the position you are stuck at. The next step will be the kint you are looking for, I hope. Of course the solver might take a different route from you as it should take the leastpath. Give it a go.
