JoergWausW Daily subscriber Best completion time: 14:52Time on first attempt: 33:34Used 'show wrong moves'
I had a couple interruptions. But it seems to be doable in under 4 minutes!
Posted 10th Oct 2023 at 09:26
winterer Daily subscriber Rated puzzle: ModerateBest completion time: 2:22Time on first attempt: 15:25
@Joerg: tongue firmly in cheek, I presume?
Posted 10th Oct 2023 at 11:32
Elisabeth Daily subscriber Rated puzzle: HardCompletion time: 1:48:12
No winterer, look at the charts!!! Mind boggling:)
Posted 10th Oct 2023 at 12:36 Last edited by winterer 10th Oct 2023 at 13:09
winterer Daily subscriber Rated puzzle: ModerateBest completion time: 2:22Time on first attempt: 15:25
Yeah, "if it's too good to be true..." If it was sports, I would suspect doping.
(as a proof of concept, I improved my time to 2:22 "without any aid"... thankfully, the charts only list the time for the first attempt, so my initial 15:25 is still recorded there)
Posted 13th Oct 2023 at 16:16
JoergWausW Daily subscriber Best completion time: 14:52Time on first attempt: 33:34Used 'show wrong moves'
I guess I know how to do these pretty fast 8 minutes is clearly possible for a Toroidal Sudoku if it goes really well (here is one that's not free: https://www.puzzlemix.com/rankings.php?id=94795 ). But way quicker, especially when it's rated harder, I consider 8 minutes a bare minimum for this one.
I just did it again - after a weak and a couple other puzzles and a three day's break from any puzzle you might say I don't remember anything from this specific one. Mistakes happened.... again. Ran into two sixes in a row, backtracked it twice. Running into 2 same numbers somewhere else each time... "Show wrong" -> I put a second 6 in the pink area for no reason. Fixed that and played it to the end without more mistakes. Just under 15 minutes. Without the mistake I might have made it under 14 maybe under 13. But under 4? As winterer showed: it's less than twice the time you need to just fill it out...
I know that I'm slow with anything that has "killer" in the name - and I don't like those days with at least three puzzles of that kind at once (random...?). There must be a way of doing those, that I don't know yet to complete these in half of my time, I'm used to that. But if there is another Toroidal-way to cut my assumed possible best time in half that no other "usual suspect" (who place in the top 5 mostly all the time) has figured out, then I get suspicious.
Posted 16th Oct 2023 at 08:55 Last edited by winterer 16th Oct 2023 at 15:53
winterer Daily subscriber Rated puzzle: ModerateBest completion time: 2:22Time on first attempt: 15:25
As Joerg did, I replayed it after a while, and it took again around 15 minutes. A time around 8 minutes would require a lucky break, and a time of 4 minutes would require several, IMHO. Which I think is unlikely.
Regarding my 2:22 attempt: that time was not just entering the solution, but actually solving the puzzle with the help of an external solver program. About 1.5 minutes to enter the hints in the required format (number + coordinates), 0.25 seconds to solve, the rest to enter the solution back into the player. So, just copying the solution from a pre-solved instance should be faster than 4 minutes (or take rather slow typing, or use a solver very much slower than mine).
My guess is rather play it without login (it's a free puzzle after all), remember some key positions, then do the speed-run with login. That might give the "lucky breaks" required to achieve four minutes.
Since the non-free puzzles do not allow this, I regard those times as more valid. If someone would achieve a time a quarter of the regular top players in a non-free puzzle, I'd be much more impressed. (although there are workarounds there, too, to game the results: external solver or a little knowledge of JS...)
@Joerg: Regarding the "killer" sudokus. There's no trick, just knowing the possibles combinations for a given sum by heart. And looking for "dents" or "noses", where the rows/cols/blocks don't add up to a multiple of 45. Plus the usual sudoku tricks (X-wings etc), if applicable. The rest (for me) is 20 years of practice... :-) My bane are the kakuros - somehow, I can't work out the intersections fast enough, even though it's all about those same sums that I have to handle in the killer sudokus.
Posted 16th Oct 2023 at 18:22
JoergWausW Daily subscriber Best completion time: 14:52Time on first attempt: 33:34Used 'show wrong moves'
Never thought about the JS-way, I looked at the well programmed source once, but as soon as I do it, the code is reloaded, so it would count as started twice - no use for cheating that way. But there is probably another way to get to the source of that cheat...
@killer: my times got better over time, but are still way off. I think I placed first once, and didn't know what happened there.... @kakuro: you have to check for noses there, too - sometimes you have to sum up into 70 or more to get a needed digit... I don't like to have to do it.
..and of course my "favorite trick": exclude numbers that would not lead to a unique solution. Example for a regular Sudoku: In one row you have two cells with only two digits 1 and 5 left. In another row you have only two fives in the same two columns. X-Wing already applied. Then you can also delete the 1 from those two cells... because 15 - 15 again would not be uniquely solvable. (I hope this example was clear enough...)
Posted 17th Oct 2023 at 07:27
winterer Daily subscriber Rated puzzle: ModerateBest completion time: 2:22Time on first attempt: 15:25
I don't think JS is reloaded when you open it in the browser's "programmer tools". And if it were, you could also write your own browser plugin to extract whatever info you need...
I was surprised by the amount of stuff in the player code. I had assumed that nowadays, everything is done "in the cloud", including the solution check. Maybe the player is showing its age here. On the other hand - I prefer to do as much as possible locally, as I don't trust the stability of net connections (and in general like to have control over my data). Maybe I'm showing my age here... :-) Plus, for accurate timing, you absolutely need to exclude network latency.
Regarding your favourite trick: yes, clear enough for me. I use it sometimes, too, but it's rare that it applies.
My favourite one (for regular sudokus) I found years ago. I think it's a version of "X-chain".
If, for a given digit, you have an even number of 3x3 blocks that are connected cyclically through possible positions for that digit in the same rows/cols, and you can reduce the number of possible positions in these blocks to two each, then the numbers of left- and right-slanting positions (i.e., either looking like "/" or like "\") within those blocks must be even.
E.g., if two blocks contain two possible positions forming a backward slant "\", and two other blocks have three possible positions forming an "L", I know that I can exclude the digit from the "L" corners, giving four backward slants.
Another example: if I have three backward slants already and the fourth block has four possible positions in a square, I know that I can reduce it to another backward slant by excluding the top right and bottom left position of the square, as three backward slants + one forward slant is not possible.
It's a visual help with which I can avoid checking those x-chains individually.
Sorry: You must log in (create a free user) in order to be able to post comments on this puzzle.
You can however view other players' statistics and comments in the tables above.
Post comment
Key
A yellow/light blue highlight in the time distribution charts highlights your time, where relevant.
Rating scores out of 10.0 show the average difficulty rating chosen by users, where 1.0 is "Easy" and 10.0 is "Hard".
If a puzzle is opened more than once, including by loading from a saved position, then this is potentially a significant aid so it is listed as being completed with 'multiple sessions' for the purpose of the best time/average rating displays above.
Minor aid is defined as no more than one use of 'Check solution' when incomplete and/or no more than one use of 'Check solution' when wrong; and/or using highlighting aids (show repeated digits, show broken inequalities and show valid/invalid placements [slitherlink] only). Major aid is any and all other use of the solving aids except for 'show wrong'.
Last edited by winterer 10th Oct 2023 at 13:09
If it was sports, I would suspect doping.
(as a proof of concept, I improved my time to 2:22 "without any aid"... thankfully, the charts only list the time for the first attempt, so my initial 15:25 is still recorded there)
8 minutes is clearly possible for a Toroidal Sudoku if it goes really well (here is one that's not free: https://www.puzzlemix.com/rankings.php?id=94795 ). But way quicker, especially when it's rated harder, I consider 8 minutes a bare minimum for this one.
I just did it again - after a weak and a couple other puzzles and a three day's break from any puzzle you might say I don't remember anything from this specific one. Mistakes happened.... again.
Ran into two sixes in a row, backtracked it twice. Running into 2 same numbers somewhere else each time...
"Show wrong" -> I put a second 6 in the pink area for no reason. Fixed that and played it to the end without more mistakes. Just under 15 minutes. Without the mistake I might have made it under 14 maybe under 13. But under 4?
As winterer showed: it's less than twice the time you need to just fill it out...
I know that I'm slow with anything that has "killer" in the name - and I don't like those days with at least three puzzles of that kind at once (random...?). There must be a way of doing those, that I don't know yet to complete these in half of my time, I'm used to that. But if there is another Toroidal-way to cut my assumed possible best time in half that no other "usual suspect" (who place in the top 5 mostly all the time) has figured out, then I get suspicious.
Last edited by winterer 16th Oct 2023 at 15:53
A time around 8 minutes would require a lucky break, and a time of 4 minutes would require several, IMHO. Which I think is unlikely.
Regarding my 2:22 attempt: that time was not just entering the solution, but actually solving the puzzle with the help of an external solver program. About 1.5 minutes to enter the hints in the required format (number + coordinates), 0.25 seconds to solve, the rest to enter the solution back into the player. So, just copying the solution from a pre-solved instance should be faster than 4 minutes (or take rather slow typing, or use a solver very much slower than mine).
My guess is rather play it without login (it's a free puzzle after all), remember some key positions, then do the speed-run with login. That might give the "lucky breaks" required to achieve four minutes.
Since the non-free puzzles do not allow this, I regard those times as more valid. If someone would achieve a time a quarter of the regular top players in a non-free puzzle, I'd be much more impressed. (although there are workarounds there, too, to game the results: external solver or a little knowledge of JS...)
@Joerg: Regarding the "killer" sudokus. There's no trick, just knowing the possibles combinations for a given sum by heart. And looking for "dents" or "noses", where the rows/cols/blocks don't add up to a multiple of 45. Plus the usual sudoku tricks (X-wings etc), if applicable. The rest (for me) is 20 years of practice... :-)
My bane are the kakuros - somehow, I can't work out the intersections fast enough, even though it's all about those same sums that I have to handle in the killer sudokus.
@killer: my times got better over time, but are still way off. I think I placed first once, and didn't know what happened there....
@kakuro: you have to check for noses there, too - sometimes you have to sum up into 70 or more to get a needed digit... I don't like to have to do it.
..and of course my "favorite trick": exclude numbers that would not lead to a unique solution.
Example for a regular Sudoku: In one row you have two cells with only two digits 1 and 5 left.
In another row you have only two fives in the same two columns. X-Wing already applied. Then you can also delete the 1 from those two cells... because 15 - 15 again would not be uniquely solvable.
(I hope this example was clear enough...)
I was surprised by the amount of stuff in the player code. I had assumed that nowadays, everything is done "in the cloud", including the solution check. Maybe the player is showing its age here. On the other hand - I prefer to do as much as possible locally, as I don't trust the stability of net connections (and in general like to have control over my data). Maybe I'm showing my age here... :-) Plus, for accurate timing, you absolutely need to exclude network latency.
Regarding your favourite trick: yes, clear enough for me. I use it sometimes, too, but it's rare that it applies.
My favourite one (for regular sudokus) I found years ago. I think it's a version of "X-chain".
If, for a given digit, you have an even number of 3x3 blocks that are connected cyclically through possible positions for that digit in the same rows/cols, and you can reduce the number of possible positions in these blocks to two each, then the numbers of left- and right-slanting positions (i.e., either looking like "/" or like "\") within those blocks must be even.
E.g., if two blocks contain two possible positions forming a backward slant "\", and two other blocks have three possible positions forming an "L", I know that I can exclude the digit from the "L" corners, giving four backward slants.
Another example: if I have three backward slants already and the fourth block has four possible positions in a square, I know that I can reduce it to another backward slant by excluding the top right and bottom left position of the square, as three backward slants + one forward slant is not possible.
It's a visual help with which I can avoid checking those x-chains individually.
You can however view other players' statistics and comments in the tables above.