SudokuWiki.org
Strategies for Popular Number Puzzles
| Page: | 2025 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | : | 2024 | 2023 | 2022 | 2021 | 2020 | 2019 | 2018 | 2017 |
| 2015 | 2014 | 2013 | 2012 | 2011 | 2010 | 2009 | 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 andrew@str8ts.com.Post a Comment or Question here...
Sunday 14-Dec-2025
... by: Lu, Shanghai,China
Load Sudoku: CLICK TO LOADIt can be seen that steps 3 and 4 of the solver use two completely identical AICs with Groups, while step 5 uses another slightly shorter but harder-to-find AIC with ALS. This is actually unnecessary, as the same result could be achieved using the identical AIC from the previous two steps.
This is not the first time I have observed this phenomenon, which indicates it is not extremely rare among all AICs. Therefore, I hope you can make adjustments to this behavior. Otherwise, the solver will assign multiple technique scores to the same AIC, which may affect the overall difficulty rating of the puzzle (although I haven’t noticed this happening yet).
Specifically, the solver could remove all candidates from the same group in one go when using AIC with Groups. For human solvers, this is obvious and does not require stacking additional technique scores.
For example, the output could be written as:
Saturday 13-Dec-2025
... by: Donald Zrudsky, Retired Engineer
Talking about having an archive of the last 30 days? I'm thinking that as well
Thursday 11-Dec-2025
... by: Bellmac, Bellingham, WA, USA
1. The save function does not seem to be working correctly for me. I can save a puzzle, but if I close the browser and reopen the solver webpage, the saved puzzle is not there. However, if I then do a reload, the last saved puzzle appears. I have refreshed the script, so do not know what is going on.
2. You asked for comments on the new scoring system. First, it seems to represent the comparative difficulty of puzzles fairly well, and certainly is better than the predecessor. However, I continue to believe that using a Log5 algorithm does not work very well. Most non-technical folks like to think in linear terms rather than geometric. (for example, people generally have trouble understanding the Richter earthquake scale). The difference in difficulty between a 7.0 and 7.5 score is fairly substantial using a log5 scale. I would still recommend going back to a linear scale and getting rid of the decimal point, which is kind of "pointless" in this application: (ha-ha).
Good feedback on the new scoring, thanks for that. I felt I needed a wider spectrum which means increasing the scores for hard strategies in addition to the new weights. But that meant going from a score than maxed at around 1800 to one that could get as high as 13000. I felt that the spread was good but the large numbers introduced “drama digits”. I wanted to discourage thoughts that a puzzle of 4001 points was easier than one of 4002. Now I could normalise to say 1 to 100 but as there is a long tail and no natural upper bound that’s difficult to do. I think a log suits as a way to shrink the values and remove the drama digits. If I removed the decimal place we’d have 10 or 12 grades and that is too close my six broad ones. Three decimal places is enough of a differentiator. I get that log scales are not naturally intuitive for many people. But I don’t think they have to worry about it being a log of a larger number, I’m hoping puzzles in certain score range like between 6 and 7 have a certain feel. It’s a tough problem but I also wanted to do something better than “score = hardest strategy used”.
Tuesday 9-Dec-2025
... by: Manfred, Vienna, Austria, Europe
Your last update to the Sudoku Solver does the trick: Weekly Unsolvable #675 is no longer unsolvable :)
And: I can confirm Jacob's report (see below).
Please load Daily Level 3 from today (12.9.) and step through.
After setting last candidates 3 in D3 and E4, candidate 4 in D6 gets incorrectly highlighted. Candidate 3 for removal in D6 gets not highlighted.
Thanks again for your great work.
Okay so I made a mistake making this Weekly. I was searching for a suitable unsolvable by running the generator for a long time but it was a version of the program where I'd accidentally disabled the Forcing Chains. I did wonder at the time that I was finding unsolvables a bit too quickly but didn't connect. Anyway, still a crazy tough one.
Tuesday 9-Dec-2025
... by: numpl_npm, Japan
If 5R2C9 then
Hidden Singles off & Pointing Pairs on -> 2R45C7 ( ... 2R4C7 )
( -2R4C8 -2R6C8 -2R5C9 -2R6C9 1R4C8
5R6C8 -5R6C6 -48R9C6 5R9C6 -5R9C4
[48]R9C9=[
4R9C9 8R9C4 4R7C6 8R6C9 |
8R9C9 4R9C4 8R7C6 4R6C9 ]
=2R6C6 -2R4C6 2R4C7 )
Hidden Singles on -> 4R4C7
Monday 8-Dec-2025
... by: Jacob, Netherlands
After the update (2.49 --> 2.50) the solver sometimes incorrectly highlights in yellow the wrong numbers, that have been eliminated by the newly found hidden singles.
Btw, having the strategies "Pointing Pairs", "Box/Line Reduction" and "Triples Naked + Hidden" as optional is a nice addition! What stopped you from also making "Hidden Singles" and "Pairs Naked + Hidden" optional? :)
Are you sure? I’ve stepped through the puzzle very carefully and all the new highlights are from Naked Singles. Where Hidden Singles are found they are shown in green, then the next step is to turn them into solutions which triggers the highlight of those candidates to be cleared off in the step after. Ie they are now Naked Singles. Send me a picture if I’m wrong.Update: Seen and fixed! Thanks.
I know from testing that certain mid and complex strategies require basic strategies to be cleared off, or may produce unexpected results. I think Pairs are fundamental and should be required first but relaxing the other ones. We’ll see if we get any reports of funny stuff.
Update: Added a checkbox for Pairs
Friday 5-Dec-2025
... by: 39_Chevy, USA
Friday 5-Dec-2025
... by: John, Texas
For the past 2 nights, Dec. 4 & 5, 2025, the Jigsaw solver shows the given clues in the correct positions row & column, but the pattern in the Solver is different from the pattern in the Puzzle. If I solve a cell & check with the Solver, same problem, wrong pattern from the Puzzle yielding 0 Solution Count
Tuesday 2-Dec-2025
... by: Matt, DC, USA
Would you consider changing the 17 clue requirement from an error to a warning, and still let the existing logic do what it can?
Tuesday 2-Dec-2025
... by: Leo, Germany
Could be an easy fix, to disable text selection for the buttons via CSS:
.numberpad {
(...)
-webkit-user-select: none;
user-select: none;
}
Thank you for providing lots of free puzzles!
Greetings
Leo