Feedback and Questions

Hi just want to know how to import/create the custom shape in "Killer Jigsaw Solver". I can create the cages, but cannot the shape. I notice there's custom one to create or import, but failed.

So how should I create it?

WHY DO I GET THE RESPONSE "reset_yes_no: no row called RUR1"

Hello,

I've been enjoying your solver and your many lessons on the techniques.

I took up sudoku a few months ago and I sit comfortably at the diabolical level now, doing the level 4 sudoku of the day whenever I can.

However, I've noticed that I could not solve some of the last few daily puzzles, and when I gave up and used the "solve path" option, I saw that it required X-Cycles with ALS, listed as diabolical. Is that considered diabolical considering ALS are listed as an Extreme level technique? In my view, combining any AIC (including X-Cycles) with other techniques such as ALS or UR increases the difficulty by a lot.

It's just a silly question, but I hope you can share your insight.

Rafael

Reply to TenPeter,
You don't have to always use rows to cover the S cell candidates. We may use 1 row and 1 column, or 2 columns.
In this case the 8s can be covered by 2 columns so it is a valid exocet even with the 8.

Hi Andrew,
I'm using this solver (https://www.sudokuwiki.org/KillerJigsaw.aspx) for a long time, not to "cheat", but as an easy means for clearing candidates.
Therefore I'm using only the first two steps of the solver:
* Check for solved squares
* Show Possibles
Now there are two concerns:

Number 1:
Apparently something has changed recently. We had a conversation about this topic years ago, and for a long time you could do the step "Show Possibles" without being forced by the solver to resolve any single-candidate-square into a solved square.
This is convenient with difficult/extreme puzzles, because you could solve a square and apply all implications of this step without considering the whole board over and over again (which you would have to do, when the solver sets some solved squares simultaneously and you lose track). Nowadays the solver enforces every single-candidate-square to be resolved to a solved square before proceeding with the next step.

Number 2:
This is a real bug. When the solver does the step "Show Possibles" with the result "No" (having found no candidate to clear), it proceeds with the next strategy steps even when you have changed some squares in the meantime (regardless whether the change was clearing a candidate or solving a square). In other words: after the result "No" the solver won't be reset to start by any change you make on the board.

best regards
FHessel

Hi Andrew
Been visiting your site for quite some time now. I just realized that it would be cool to have a 6x6 Mini Sudoku Solver and additional puzzles added to your site. With Mini Sudoku becoming popular thanks to LinkedIn games, I would love to see support for Mini Sudoku.

Solved the #6487 Level 3 puzzle for October 5. The new scoring system gives this a Tough Grade of 6.0 (old score 132, Tough). I am at a bit of loss how this puzzle gets a score this high, under either system. No tough strategies are needed. 23 steps are required: 15 naked singles, 4 hidden singles, 2 naked pairs, and 2 naked triples. So, this is a pretty simple puzzle that ought to be no higher than Moderate, based on my experience.

Also, I do not really understand the new numeric difficulty scale. From the discussion, it appears that total range goes to 9+ for Sudoku, which is open-ended. Most scales have a defined fixed length (e.g. 1 to 10 or 1 to 100). And I am not sure why there is a decimal point, which tends to create an impression of precision, which may not be warranted. I think it would be better to eliminate the decimal and go to a 1 to 100 scale.

Continuing the discussion with Ayumu below. A Junior Exocet pattern requires no more than 2 cover houses for a candidate Base Pair digit. The subject Exocet has the digit 8 on three different rows (F,G&J) in the cover lines thus it requires 3 cover houses to cover the digit 8. This can be seen in the screen capture that you provided. This implies that this is not a valid Exocet pattern. If your were to eliminate the digit 8 from the Base Pair, then the remaining three digits (2,3,4) would form a valid Exocet.

Solved Level 3 puzzle for today (#6481, September 29, 2025). The current scoring system release gives this a Tough Grade of 122. This seems way too high considering the puzzle solution has 19 Naked Singles, 6 Hidden Singles, and 2 Naked Pairs, none of which are Tough Strategies or above.

However, I noticed that your new beta version 4.7 scoring system gives this a Moderate score, which seems much more appropriate. So, looks like you are making good progress on revising the scoring system.

Hi,
when trying to grade this one, your program exits with D5 cell empty.
has one solution.
think its a bug.
best regards