well done!! this was definitely one of my harder ones so i think you managed very well here (it was tricky to make the basics smoother than this while keeping it intact) thank you again for all the kind words, i feel like an introduction like todays is even possibly too kind. youre the best as always
Normally when I see a Simon video under 45 minutes or so, I'll give the puzzle a shot (assuming I'm not too busy). Your puzzles are an exception to that rule though, because previous experiences have taught me that your logic is so beautiful that it flies over my head every time. I respect your humility, but just know that every word of praise that Simon said in this video is justified
Usually when he says that it's to satisfy or not violate a rule of Sudoku. As this puzzle evidences, there's clearly other logic that can be done in a Sudoku puzzle that isn't strictly a direct application of the rules.
That’s just incredible. shye doesn’t miss!!! The fact that such beautiful logic can still be scraped out of such a simple ruleset is beyond. Also I love how Simon referred to the 6 and the 9 as the “funny digits” lol
@@shye229 This could be solved logically by an algorithm if someone decides to implement it. The common algorithms only check for patterns on digits, but you can also check for patterns on colours. In this case, you could give R6C1 a colour, say red, and ask where 'red' goes in rows 1, 5, and 9. This is equivalent to asking which cells in these rows can contain a 6 or a 9 and don't see R6C1. You'll end up with a swordfish in columns 4, 6, and 9, and that means that R4C9 is not red. From there on, the puzzle solves with normal Sudoku. There was a similar puzzle a while ago (was it by Jovi_al?) that could be solved with colouring as well (IIRC involving a finned X-wing). Now, checking for all of these colouring patterns would add significant computational overhead, but in theory it can be done.
yea it totally could be, and the best bet for which one most likely to implement it is YZF (he has added a few of my ideas in already) but theres also the question of "when is it too much techniques?", because overloading a solver will slow it down
I love when you said, "We did it in a strange and weird and wonderful way." That is the best summary of watching you solve that I've ever heard. Much love Simon
I somehow knew this would be a shye puzzle before even looking at it, and I'm not disappointed. This logic was so satisfying to solve, and it's immensely helpful knowing that what I think looks suspicious is actually the right place to look. This definitely makes shye the author of my top three classics.
To scan, or not to scan, that is the question: Whether 'tis better for the mind to ponder The rings and arrows on outlandish grids, Or to take a break from a sea of puzzles And by occasion pause them. To scan - to solve, Once more; and by that move to say we love The beauty and the thousand natural ways That grids are heir to: 'tis an appreciation Devoutly to be sought. To scan, to think; To mark, perchance to solve - ay, there's the rub: For in that path to solve what tricks may come, When Shye has built in some new clever bit? ‘t Must give us pause-there's the respect That means celebrity for such great art. For who can find the cells and rows that count? The setter's right, the kind man's videos, The thrills of the break-in, the pattern’s law, The novelties in classic form That patience merits for that worthy man, When he himself makes his frustration known With a quiet “Bobbins…”
@christopher b. ... which is something entirely different from "I love effing this community"... just to highlight how important it is for us to order our words carefully 🤣
After some initial help I was able to solve the puzzle well. Thank you also at this point, Simon and Mark, for your accompaniment, which I have enjoyed for about a year and a half. I learned a lot... in math and also in the English language. When I'm engrossed in your puzzles, as a German I increasingly no longer think in my native language, but in English. 🙂
A few months ago, I mentioned that I’d like to see more classic sudoku on the channel (meaning maybe one per month). If that means one Shye puzzle per month, I’m quite happy to watch you solve it.
This was beautiful and a very good demonstration of the value of the restraint in pencil marking in classics. But also a good demonstration of the beauty in a carefully constructed classic, and in a solve by someone who appreciates such beauty. Thank you, Simon.
Tough puzzle. Couldn't do it without getting help from your video. I really like that you are doing Classic Sudokus vs all the non traditional puzzles.
shye, definitely take a bow! Simon as well. It took me nearly 82 min to solve this one. Very challenging but a classic so doable no matter how long I had to stare and color and consider possibilities. What a break from everything else as well! I work the classic sudoku in my local newspaper every day and the only ones that are even somewhat challenging are the Friday/Saturday **** and Sunday *****. Sometimes those aren't even challenging, then once in awhile I really have to work at them. The ones I have to work on for awhile and try various things are the most fun!
I like how you gave us a step-by-step about how you get started in classic sudokus at the beginning. Had you not gotten distracted by the top and bottom rows, you might have found the solution even faster with your method.
I made a real mess of this one; it ended up taking me over an hour (1:06:43) as I kept focusing on which of two cells in Box 7 that "2" could go; it was only after some bifurcation that I started seeing the usefulness of the 6s and 9s; even then, I couldn't get to the beautiful logic that Simon used to deduce r4c3 first. I'm happy that I solved this, but I'm not proud of how I did. Tough classic (for me, anyway)!
Just two days ago i was redoing the shye puzzle in the App, because i was the mood for some classic. So i'm very happy to see a nice classic Sudoku again today!
I remember the 500K puzzle, the heart shaped. It took me 101+ minutes to solve it, after I watched the beginning of the video for some ideas in which direction to start. But it was worth it.
6:30 "It will not be your friend." I've found that triple pencilmarking corner cells helps me catch things that I would find a lot harder otherwise -- triples, X-wings, swordfish, etc. 17:28 I paused here to see what I could do with the pattern I had of the bottom row 1269 quad and 69 cells in R6C1, R4C9, and R7C8. I didn't notice any advanced pattern specifically, and no chain of logic worked with me. Therefore, I quadfurcated on these options for the bottom row: :: 69 in cells 1 & 9, 12 in cells 4 & 6 (two 69 pairs) :: 12 in cells 1 & 9, 69 in cells 4 & 6 (no 69 pairs) :: 2 in cell 1, 69 in cell 9 (69 pair in column 9) :: 1 in cell 9, 69 in cell 1 (69 pair in column 1) The first three broke the puzzle. The last solved it. That was probably karma for my choice to quadfurcate. 17:58 That was nasty of me, wasn't it? I clicked to pause the video and the flaming error message flashed in. 18:40 Have you noticed that R2C2 is a given 6? 21:10 6s and 9s are the right thing to look for. Just expand your field of search. Also, centermark the bottom row. 22:20 "By Sudoku"? Would anything be otherwise in classic Sudoku? 23:10 Yes! You found the 69 cells. 23:50 I see it now. The conditional X-wings force the two 69 cells on the edges to be opposite. (Later) This time, I managed to solve the puzzle without -furcation. 6s and 9s were restricted in rows 4 and 6. Several pairs, triples, etc. later, the puzzle finally collapsed. 38:00 The first solve, I had the 69s and the bottom row. I didn't think of the top row interacting, and I doubt I would have discovered the conditional X-wing in any case. For me, the conditional X-wing did establish that the 69 cells were opposite. Nice video!
I managed to solve a Shye classic! Well, almost. With the smallest of suggested hints from watching Simon - I just needed him to start pondering on the possibility of a 6 x-wing in rows 1 and 9, even though there wasn't one. I had found that r4c9 and r6c1 were restricted to 6 or 9 only beforehand. I did give up, though, and had started watching Simon up to 11:15. Then paused the video and thought harder about 6s and 9s. Very satisfying when I then worked out r4c9 and r6c1 had to be different, and that r4c3 had only one candidate. I'm not proud. I'm counting it as a solve. 🙂
I finished in 61 minutes. I think I had a good sense on what this puzzle was from the title. However, it still took me a while to appreciate all the parts. I saw that rows 1 and 9 were the most interesting, especially when I was able to a 69 in both r4c9 and r6c1. I also noticed that many pencil marks overlapped, so if I could rule one digit in the middle of that, a chain could form. It took me a while, but I did finally notice the break-in involving those same 69's and whether they could be the same number. If they were the same, the digits in rows 1 and 9 break. That allowed me to get a digit and place 8 into r4c3. From there, it was pretty straightforward. I am so glad to have spotted that. It is some great construction. Great Puzzle!
I'm just amazed at your ability to land on the right track that fast. So many ideas that lead to nothing yet here you are exploring the right one right of the bat.
This one was really fascinating. I loaded it into Good Sudoku which insisted that it wasn't solvable, and while I was stuck on things for a little bit, Simon's observation about the 6/9 "conditional x-wing" made the whole thing fall down *immediately* thanks to Good Sudoku's solving tools - which then refused to believe that I'd *actually* solved it, so it won't actually save my solution or time (or give me a score).
Brilliant construction and well worked out by Simon. It's interesting that if you put the first derived digit C3R4 as an extra given, the puzzle drops from super-extreme to medium in difficulty. So it's really the 69 logic that eliminates two out of the three options for C3R4 that is the magic key.
I basically had to pencil mark the entire grid to notice a few cells were restricted to 6 and 9, and that either way they would be filled in, would force an x wing for the other, revealing an extra digit, and the whole thing unraveled.
But what are we coloring 6/9? ha! He started with blue and orange, then changed to purple and red, no resolving them. I guess the red/purple were to show they were different, but it was confusing show double blue in C6. Oh, the mysteries of coloring!
You've gone from doing the "Back to the Future" puzzle, which is about as polar opposite from a classic Sudoku as you can get ... to this ... showing an amazing array of skills along the way!
I love how you say there's absolutely no point in trying to scan the grid for digits, and then proceeds to do so for the first five minutes, proving yourself absolutely correct. 😅 Anyway, I'm going to have a go before I get any further into the video and the way to play is to click the link under the video. Now I get to play, so let's get cracking!
Brilliant puzzle. I got the break-in in a fairly reasonable (for me) amount of time, of which I'm rather proud, but it took me ages to sort everything out from then on. When I'm stuck I always start spraying too many pencil marks around the grid, and then can't see the forest for the trees.
Fun fact: the B-side of They're Coming to Take Me Away Ha-Haaa! was the same "song" played in reverse, and the label was a mirror image of the A-side. This was interesting. I didn't quite get there the same way, but I thought there was something hincky about 6 & 9, especially once I spotted the R6C1/R4C9 pair. Once I realised they had to be different, it was just plain sudoku. I'd be surprised if a computer solver couldn't spot this, it didn't seem too difficult to work out that those two cells had to be different. I love the way that Shye is still finding new ways to push the boundaries of classic sudoku. Whilst solving this, I had an idea for something that I've never seen in a sudoku. I wondered whether this was perhaps what Shye was doing, but it wasn't. Unlike many of my bright ideas, I wrote this one down while it was clear in my head, so that I won't forget it. I'm going to try to set my first sudoku using it, to see if it's possible. If I get into trouble setting it, I may pass the idea onto Shye.
Andrew Stuart's sudoku solver does actually solve this one, but through some crazy forcing chains. It finds that the special cells have to be 6 and 9 instantly, and I'm guessing that it understands that in either case, the other red cell cannot be same digit as the first, but I guess that it only understands that within the context of exploring individual logic steps, and so it doesn't combine that with other logic.
I could swear I’d heard the term “marlin” in sudoku before but I don’t know where. When he asked why it was called that I thought, of course, and was fully expecting a sort of conditional/overlapping swordfish pattern. Which is a pretty apt description of what happened here. Strange, that. Though I’d admit I doubt I would have been able to solve this puzzle, myself.
I came _very close_ to figuring this one out on my own. Close, but no cigar. I could tell that it hinged on that central 69 pair being different, but I could not quite figure out how to prove it. It took help from the video to direct me to the significance of the top row, after which I was able to sort out the rest. My final time was about 64 minutes.
Though I sometimes have my doubts, I think CTC has definitely improved my skills as I used to be hopeless at these classics but happy to find my way through this one using the same logic as Simon.
29:20 for me. Finding a certain pair of numbers that were key to the solve was *nice*. And then the break in cell being my lucky number was nicer. took me a while to spot that though.
I'm kinda curious if Cracking the Cryptic would do this, but I'd be gneuinely interested to see if a sudoku puzzle I've seen in an anime is not just possible, but actually interesting to beat. The anime is Phi Brain Kami no puzzle, and the sudoku in question is season 1 episode 18. A character called Gammon/Galileo created a sudoku that meshes into other sudoku squares in the shape of a plus with the very center square being absent. If you want the exact timestamp for that episode I believe it's around 13:00 that the sudoku becomes visible.
OK who else spent ages trying to play the same trick that Simon did with the 69 pair but on box 6 and box 9 before finding the other 69 pair to play with in box 4? Nice work as ever from Shye.
Wow! I solved this myself, admittedly taking about 50 minutes. I'd noticed the distribution of 6s & 9s in rows 1 & 9 quite quickly so was looking for a jellyfish on 9s. The discovery of opposing 69 (remote pair?) in c1 & 9 allowed everything to solve relatively easily from then on. I was gratified to note that my logic largely followed Simon's brain. Maybe I've been remotely infected, 'cos I for one would probably enjoy Simon's company at a party. Not sure the reverse would be true though. Nice one Shye.
24:36 snapshot : I think there is no immediate need to find that 69 pair un the rows 4 and 6. The point is that, in row 1 and 9, therecan be only one 6 and one nine in columns 1 and 9, so there are two missing (there must be two sixes ant two nines in these rows). That gives an virtual X-wing in sixes between column (1 or 9) and 6, and another virtual X-wing in columns (1 or 9) and 4.
Is there a name for that breaking? If not it should be called the shye double finned x wing! That is absolutely amazing! Lovely puzzle! Hats off Shye! Well done Simon!
I can’t believe I solved this on my own. Shye’s ordinary sudokus kill me. I used set though. Highlighted C1,4,6,9 and R1,4,6,9 and removed the overlaps. If you had the 69 in R6C1 and R4C9 then it became really clear the 6 and 9 had a big constraint in boxes 4 and 6. Not sure if that was the intended solve but it was easier than what Simon did.
When solving this I, for whatever reason, ended up going down a path of logic (which was pretty much 1 fork of a bifurcation) where I had gone "If I can prove x, then lots and lots of stuff happens" (I've forgotten what x was exactly unfortunately, something to do with 6s and 9s). I ended up getting so much stuff from that assumption that I ended up thinking that this is surely the correct path and decided that, rather than going back to prove x and losing my progress, I would just finish the puzzle and then later go back to the start to try and prove x. The entire puzzle solved except for 2 digits, one of which was r4c9 which had no possible values. Which meant that my assumption, despite solving 98% of the puzzle, was actually false. This is utterly bonkers. 1) I have no doubt that the computer struggled really hard if there are almost perfectly valid (but not quite) configurations consistent with the clues given 2) This was the thing that made me look at r4c9 in the starting grid. Since I knew that ANY solve would have to look at it so as to eliminate the almost-perfectly-valid solution I had found.
A computer is only as 'smart' as the person who programmed it. In fact this could be solved logically by an algorithm if someone decides to implement it. The common algorithms only check for patterns on digits, but you can also check for patterns on colours. In this case, you could give R6C1 a colour, say red, and ask where 'red' goes in rows 1, 5, and 9. This is equivalent to asking which cells in these rows can contain a 6 or a 9 and don't see R6C1. You'll end up with a swordfish in columns 4, 6, and 9, and that means that R4C9 is not red. From there on, the puzzle solves with normal Sudoku. There was a similar puzzle a while ago (was it by Jovi_al?) that could be solved with colouring as well (IIRC involving a finned X-wing). Now, checking for all of these colouring patterns would add significant computational overhead, but in theory it can be done.
Simon always seems to ignore his corner pencil marks. Gets 69 in the corner.... ignores the 2 pencil mark that will give him both the 2 and the 4 in box 7.
A marlin is closely related to a swordfish. It's not QUITE a swordfish, but it looks like one. And the pattern with the 6 and 9s is kind of an offset swordfish. If the 69 pair in boxes 4 and 6 were in the same row, it would be a true swordfish.
Seems like this was a pre-record - no birthday shout outs nor Patreon monthly puzzle solver names read out. Maybe when he's back doing videos on the day of upload.
I managed to solve it without spotting the intended logic. I was able to eliminate a 2 in r9c1 which gave me a few digits reducing r4c3 to 68 and then that whether it was 6 or 8 r9c9 was always 1. If it was 6 r9c9 was 6 and r4c9 was 9 and if it was 8 r7c7 was 8 and r7c7/8 were a 69 pair. It unravelled fairly easily from there.
I wrote my own Sudoku-cheating program, so I know exactly the problem it had with this puzzle: it is only ever considering one digit at a time, and it has no sense of curiosity. Unless you know The Trick, there is effectively a deadly pattern on sixes and nines. You only need to play about with digits for a moment to realise that R6C1 and R4C9 have to be different, but you have to think of both digits together: if you fill in either one of those cells, you quickly run out of steam trying to chase it around the grid; but writing the same digit in both cells converts just enough potential energy to kinetic energy for the two to collide! And it's making my brain hurt trying even to begin thinking about how to get a computer to deal with that sort of situation .....
well done!! this was definitely one of my harder ones so i think you managed very well here (it was tricky to make the basics smoother than this while keeping it intact)
thank you again for all the kind words, i feel like an introduction like todays is even possibly too kind. youre the best as always
Normally when I see a Simon video under 45 minutes or so, I'll give the puzzle a shot (assuming I'm not too busy). Your puzzles are an exception to that rule though, because previous experiences have taught me that your logic is so beautiful that it flies over my head every time. I respect your humility, but just know that every word of praise that Simon said in this video is justified
Does there have to be a 6 and 9 in the colored cells in boxes 2 and 8? I feel like if there isn't it breaks r4c9 and r6c1
@@BryanLu0 thats true! and you can catch both of those deductions at once with a rank0 diagram (...but the important elims are the ones simon got)
Absolutely brilliant puzzle! you're too modest :)
What a beautiful puzzle, lady S!
Ah, the joys of hearing Simon justify inferences as "by sudoku" in a classic where sudoku is the only thing to do
Usually when he says that it's to satisfy or not violate a rule of Sudoku. As this puzzle evidences, there's clearly other logic that can be done in a Sudoku puzzle that isn't strictly a direct application of the rules.
you don't have to apologize for silence. People who find this channel aren't the type that need to be stimulated every second it's cool.
I always appreciate when Simon simply ignores the given 6 in box 1 for about 10 minutes.
I love watching Simon solve these brilliant shye constructions. Way over my head but a joy to watch.
"We did it in a strange and wonderful way" - best sum up of why I followed this channel ♥
That’s just incredible. shye doesn’t miss!!! The fact that such beautiful logic can still be scraped out of such a simple ruleset is beyond. Also I love how Simon referred to the 6 and the 9 as the “funny digits” lol
I’m pretty sure computers can solve this better than me.
Same!
you might be surprised! ive thrown this into various solvers and they all resort to some pretty nasty things
I spent an hour making pencil marks, then gave up.
@@shye229 This could be solved logically by an algorithm if someone decides to implement it. The common algorithms only check for patterns on digits, but you can also check for patterns on colours. In this case, you could give R6C1 a colour, say red, and ask where 'red' goes in rows 1, 5, and 9. This is equivalent to asking which cells in these rows can contain a 6 or a 9 and don't see R6C1. You'll end up with a swordfish in columns 4, 6, and 9, and that means that R4C9 is not red. From there on, the puzzle solves with normal Sudoku.
There was a similar puzzle a while ago (was it by Jovi_al?) that could be solved with colouring as well (IIRC involving a finned X-wing). Now, checking for all of these colouring patterns would add significant computational overhead, but in theory it can be done.
yea it totally could be, and the best bet for which one most likely to implement it is YZF (he has added a few of my ideas in already) but theres also the question of "when is it too much techniques?", because overloading a solver will slow it down
I love when you said, "We did it in a strange and weird and wonderful way." That is the best summary of watching you solve that I've ever heard. Much love Simon
I somehow knew this would be a shye puzzle before even looking at it, and I'm not disappointed. This logic was so satisfying to solve, and it's immensely helpful knowing that what I think looks suspicious is actually the right place to look. This definitely makes shye the author of my top three classics.
so kind! thank you 🥺🥰
To scan, or not to scan, that is the question:
Whether 'tis better for the mind to ponder
The rings and arrows on outlandish grids,
Or to take a break from a sea of puzzles
And by occasion pause them. To scan - to solve,
Once more; and by that move to say we love
The beauty and the thousand natural ways
That grids are heir to: 'tis an appreciation
Devoutly to be sought. To scan, to think;
To mark, perchance to solve - ay, there's the rub:
For in that path to solve what tricks may come,
When Shye has built in some new clever bit?
‘t Must give us pause-there's the respect
That means celebrity for such great art.
For who can find the cells and rows that count?
The setter's right, the kind man's videos,
The thrills of the break-in, the pattern’s law,
The novelties in classic form
That patience merits for that worthy man,
When he himself makes his frustration known
With a quiet “Bobbins…”
Wonderful from you Amos!!
@christopher b. ... which is something entirely different from "I love effing this community"... just to highlight how important it is for us to order our words carefully 🤣
Brilliant and beautiful!
I love that due to the makeup of this channel, the captions tend to capitalize "Mark" in "pencil Mark."
Nice to see a purely classic sudoku tonight.
EDIT. that was a fantastic classic sudoku. Very clever.
After some initial help I was able to solve the puzzle well. Thank you also at this point, Simon and Mark, for your accompaniment, which I have enjoyed for about a year and a half. I learned a lot... in math and also in the English language. When I'm engrossed in your puzzles, as a German I increasingly no longer think in my native language, but in English. 🙂
A few months ago, I mentioned that I’d like to see more classic sudoku on the channel (meaning maybe one per month). If that means one Shye puzzle per month, I’m quite happy to watch you solve it.
Rules: 03:15
Let's Get Cracking: 03:50
Simon's time: 33m24s
Puzzle Solved: 37:14
What about this video's Top Tier Simarkisms?!
Bobbins: 2x (12:29, 29:10)
And how about this video's Simarkisms?!
Pencil Mark/mark: 20x (04:05, 04:17, 04:36, 04:41, 04:44, 04:49, 04:59, 05:12, 05:15, 05:42, 05:54, 06:24, 06:38, 07:01, 10:02, 22:05, 25:48, 26:47, 33:16, 34:12)
Ah: 17x (04:36, 06:13, 07:53, 08:15, 08:27, 11:27, 14:47, 21:20, 21:45, 21:49, 21:57, 24:09, 24:09, 26:39, 28:37, 32:46, 34:05)
By Sudoku: 8x (14:17, 14:23, 22:15, 23:06, 30:35, 30:39, 35:14, 35:35)
Brilliant: 6x (01:26, 02:37, 22:18, 28:09, 28:12, 37:18)
Sorry: 4x (12:41, 17:30, 20:25, 30:05)
Clever: 4x (01:06, 13:49, 20:39, 26:25)
Beautiful: 4x (26:57, 28:33, 37:33, 37:36)
Good Grief: 3x (26:57, 32:46, 36:02)
The Answer is: 3x (05:06, 24:26, 24:29)
Facetious: 3x (18:31, 18:35, 18:38)
Hang On: 3x (16:42, 17:57, 26:39)
Wow: 3x (28:09, 34:08, 34:36)
What on Earth: 2x (13:39, 32:24)
Goodness: 2x (22:55, 29:22)
Apologies: 2x (04:21, 30:06)
Naked Single: 2x (27:36, 29:56)
I Have no Clue: 2x (03:25, 05:08)
Extraordinary: 2x (00:42, 01:38)
Obviously: 2x (13:15, 21:16)
Whoopsie: 2x (05:51, 29:37)
Plonk: 2x (06:38, 35:45)
Symmetry: 2x (09:36, 22:58)
Stuck: 1x (38:09)
Of All Things: 1x (35:43)
I've Got It!: 1x (36:08)
Doesn't get a Song: 1x (35:30)
Wake Up: 1x (20:57)
Almost Interesting: 1x (11:30)
What Does This Mean?: 1x (25:04)
Most popular digit and colour this video:
Six (92 mentions)
Red (6 mentions)
Antithesis Battles:
High (2) - Low (0)
Even (4) - Odd (0)
White (4) - Black (0)
Row (33) - Column (32)
FAQ:
Q1: You missed something!
A1: That could very well be the case! Human speech can be hard to understand for computers like me! Point out the ones that I missed and maybe I'll learn!
Q2: Can you do this for another channel?
A2: I've been thinking about that and wrote some code to make that possible. Let me know which channel you think would be a good fit!
Cheers!
I think you should count "words ending in AGE but not valid in Scrabble " e.g. "white marlinage" 😀
@@BigAsciiHappyStar Not sure how the AI would parse that.
Good Bot
This was beautiful and a very good demonstration of the value of the restraint in pencil marking in classics. But also a good demonstration of the beauty in a carefully constructed classic, and in a solve by someone who appreciates such beauty. Thank you, Simon.
Tough puzzle. Couldn't do it without getting help from your video. I really like that you are doing Classic Sudokus vs all the non traditional puzzles.
shye, definitely take a bow! Simon as well. It took me nearly 82 min to solve this one. Very challenging but a classic so doable no matter how long I had to stare and color and consider possibilities. What a break from everything else as well!
I work the classic sudoku in my local newspaper every day and the only ones that are even somewhat challenging are the Friday/Saturday **** and Sunday *****. Sometimes those aren't even challenging, then once in awhile I really have to work at them. The ones I have to work on for awhile and try various things are the most fun!
A classic by shye! I'm soooo ready to see something mind-blowing with the most minimal rules, I'm very very excited :3
I'm at 20:25 and the anticipation continues to grow for that aha moment where Simon realizes the full beauty of Shye's puzzle.
I like how you gave us a step-by-step about how you get started in classic sudokus at the beginning. Had you not gotten distracted by the top and bottom rows, you might have found the solution even faster with your method.
I made a real mess of this one; it ended up taking me over an hour (1:06:43) as I kept focusing on which of two cells in Box 7 that "2" could go; it was only after some bifurcation that I started seeing the usefulness of the 6s and 9s; even then, I couldn't get to the beautiful logic that Simon used to deduce r4c3 first.
I'm happy that I solved this, but I'm not proud of how I did.
Tough classic (for me, anyway)!
This is the best classic Sudoku I've seen in a long time
this means a lot to me, thank you!
A marlin is a close relative of the swordfish
Fish facts did help my solve
Just two days ago i was redoing the shye puzzle in the App, because i was the mood for some classic. So i'm very happy to see a nice classic Sudoku again today!
I remember the 500K puzzle, the heart shaped. It took me 101+ minutes to solve it, after I watched the beginning of the video for some ideas in which direction to start. But it was worth it.
6:30 "It will not be your friend." I've found that triple pencilmarking corner cells helps me catch things that I would find a lot harder otherwise -- triples, X-wings, swordfish, etc.
17:28 I paused here to see what I could do with the pattern I had of the bottom row 1269 quad and 69 cells in R6C1, R4C9, and R7C8. I didn't notice any advanced pattern specifically, and no chain of logic worked with me. Therefore, I quadfurcated on these options for the bottom row:
:: 69 in cells 1 & 9, 12 in cells 4 & 6 (two 69 pairs)
:: 12 in cells 1 & 9, 69 in cells 4 & 6 (no 69 pairs)
:: 2 in cell 1, 69 in cell 9 (69 pair in column 9)
:: 1 in cell 9, 69 in cell 1 (69 pair in column 1)
The first three broke the puzzle. The last solved it. That was probably karma for my choice to quadfurcate.
17:58 That was nasty of me, wasn't it? I clicked to pause the video and the flaming error message flashed in.
18:40 Have you noticed that R2C2 is a given 6?
21:10 6s and 9s are the right thing to look for. Just expand your field of search. Also, centermark the bottom row.
22:20 "By Sudoku"? Would anything be otherwise in classic Sudoku?
23:10 Yes! You found the 69 cells.
23:50 I see it now. The conditional X-wings force the two 69 cells on the edges to be opposite. (Later) This time, I managed to solve the puzzle without -furcation. 6s and 9s were restricted in rows 4 and 6. Several pairs, triples, etc. later, the puzzle finally collapsed.
38:00 The first solve, I had the 69s and the bottom row. I didn't think of the top row interacting, and I doubt I would have discovered the conditional X-wing in any case. For me, the conditional X-wing did establish that the 69 cells were opposite.
Nice video!
I managed to solve a Shye classic!
Well, almost. With the smallest of suggested hints from watching Simon - I just needed him to start pondering on the possibility of a 6 x-wing in rows 1 and 9, even though there wasn't one. I had found that r4c9 and r6c1 were restricted to 6 or 9 only beforehand. I did give up, though, and had started watching Simon up to 11:15. Then paused the video and thought harder about 6s and 9s. Very satisfying when I then worked out r4c9 and r6c1 had to be different, and that r4c3 had only one candidate.
I'm not proud. I'm counting it as a solve. 🙂
I finished in 61 minutes. I think I had a good sense on what this puzzle was from the title. However, it still took me a while to appreciate all the parts. I saw that rows 1 and 9 were the most interesting, especially when I was able to a 69 in both r4c9 and r6c1. I also noticed that many pencil marks overlapped, so if I could rule one digit in the middle of that, a chain could form. It took me a while, but I did finally notice the break-in involving those same 69's and whether they could be the same number. If they were the same, the digits in rows 1 and 9 break. That allowed me to get a digit and place 8 into r4c3. From there, it was pretty straightforward. I am so glad to have spotted that. It is some great construction. Great Puzzle!
I'm just amazed at your ability to land on the right track that fast. So many ideas that lead to nothing yet here you are exploring the right one right of the bat.
This one was really fascinating. I loaded it into Good Sudoku which insisted that it wasn't solvable, and while I was stuck on things for a little bit, Simon's observation about the 6/9 "conditional x-wing" made the whole thing fall down *immediately* thanks to Good Sudoku's solving tools - which then refused to believe that I'd *actually* solved it, so it won't actually save my solution or time (or give me a score).
The use of the 6/9 pattern was nice.
some of these I wouldn't think to solve in a 100 years.... I wouldn't have gotten that 6-9 issue at all. Brilliant is all I can say....
I took one glance at the title and the thumbnail and said out loud "Surely this is a Shye puzzle".
One of the first time I can solve a Shye classic!
Pretty happy about that!
I spotted the same logic, but used a sword fish with the 69 pairs
I saw 'Marlin' and went, "[Expletive form of 'oh dear'], thats a kind of swordfish, I'm terrible at those"
That was a nice solve. Very clever puzzle.
Brilliant construction and well worked out by Simon. It's interesting that if you put the first derived digit C3R4 as an extra given, the puzzle drops from super-extreme to medium in difficulty. So it's really the 69 logic that eliminates two out of the three options for C3R4 that is the magic key.
What a unique and new rules.
The classic Tomb Raider games are some of my all time favorites! Hearing a Tomb Raider reference put a gigantic smile on my face!
18:00 for me. Lovely classic sudoku!
I basically had to pencil mark the entire grid to notice a few cells were restricted to 6 and 9, and that either way they would be filled in, would force an x wing for the other, revealing an extra digit, and the whole thing unraveled.
But what are we coloring 6/9? ha! He started with blue and orange, then changed to purple and red, no resolving them. I guess the red/purple were to show they were different, but it was confusing show double blue in C6. Oh, the mysteries of coloring!
You've gone from doing the "Back to the Future" puzzle, which is about as polar opposite from a classic Sudoku as you can get ... to this ... showing an amazing array of skills along the way!
I love how you say there's absolutely no point in trying to scan the grid for digits, and then proceeds to do so for the first five minutes, proving yourself absolutely correct. 😅
Anyway, I'm going to have a go before I get any further into the video and the way to play is to click the link under the video. Now I get to play, so let's get cracking!
This is a truly hard to solve puzzle, but it's so amazingly good. I was Euphoric when I solved it.
Brilliant puzzle. I got the break-in in a fairly reasonable (for me) amount of time, of which I'm rather proud, but it took me ages to sort everything out from then on. When I'm stuck I always start spraying too many pencil marks around the grid, and then can't see the forest for the trees.
Bravo Simon!! Great job.. Loved the solve 👍
Fun fact: the B-side of They're Coming to Take Me Away Ha-Haaa! was the same "song" played in reverse, and the label was a mirror image of the A-side.
This was interesting. I didn't quite get there the same way, but I thought there was something hincky about 6 & 9, especially once I spotted the R6C1/R4C9 pair. Once I realised they had to be different, it was just plain sudoku. I'd be surprised if a computer solver couldn't spot this, it didn't seem too difficult to work out that those two cells had to be different.
I love the way that Shye is still finding new ways to push the boundaries of classic sudoku. Whilst solving this, I had an idea for something that I've never seen in a sudoku. I wondered whether this was perhaps what Shye was doing, but it wasn't. Unlike many of my bright ideas, I wrote this one down while it was clear in my head, so that I won't forget it. I'm going to try to set my first sudoku using it, to see if it's possible. If I get into trouble setting it, I may pass the idea onto Shye.
more classics please
lovely as always
So that was lovely and took me a long time to find by Schneider pencil marking was so helpful.
Andrew Stuart's sudoku solver does actually solve this one, but through some crazy forcing chains. It finds that the special cells have to be 6 and 9 instantly, and I'm guessing that it understands that in either case, the other red cell cannot be same digit as the first, but I guess that it only understands that within the context of exploring individual logic steps, and so it doesn't combine that with other logic.
Great content. The config around the centre row was nice to watch as you tackled it out.
24+ minutes in and I am yelling at Simon to fix the erroneous blue entered in R1C1.
GREAT puzzle!!
I found 6 and 9 quite early but instead of check yellow squares I was looking for some x wings for ~2h :D
Is not a Marlin a type of Swordfish ?
I could swear I’d heard the term “marlin” in sudoku before but I don’t know where. When he asked why it was called that I thought, of course, and was fully expecting a sort of conditional/overlapping swordfish pattern. Which is a pretty apt description of what happened here. Strange, that. Though I’d admit I doubt I would have been able to solve this puzzle, myself.
25:17 for me, that was absolutely elegant!!
So, I asked my computer to solve this. Yeah, it can solve it better than I can.
I came _very close_ to figuring this one out on my own. Close, but no cigar.
I could tell that it hinged on that central 69 pair being different, but I could not quite figure out how to prove it. It took help from the video to direct me to the significance of the top row, after which I was able to sort out the rest.
My final time was about 64 minutes.
Though I sometimes have my doubts, I think CTC has definitely improved my skills as I used to be hopeless at these classics but happy to find my way through this one using the same logic as Simon.
Magnificent!
shye has got so many tricks up their sleeves one wonders how many shirts they're wearing
10:49 for me the 6 9 pair was doing wonders
29:20 for me. Finding a certain pair of numbers that were key to the solve was *nice*. And then the break in cell being my lucky number was nicer. took me a while to spot that though.
32:33 this time around. A little slower, but still good.
I'm kinda curious if Cracking the Cryptic would do this, but I'd be gneuinely interested to see if a sudoku puzzle I've seen in an anime is not just possible, but actually interesting to beat. The anime is Phi Brain Kami no puzzle, and the sudoku in question is season 1 episode 18. A character called Gammon/Galileo created a sudoku that meshes into other sudoku squares in the shape of a plus with the very center square being absent. If you want the exact timestamp for that episode I believe it's around 13:00 that the sudoku becomes visible.
OK who else spent ages trying to play the same trick that Simon did with the 69 pair but on box 6 and box 9 before finding the other 69 pair to play with in box 4? Nice work as ever from Shye.
Wow! I solved this myself, admittedly taking about 50 minutes. I'd noticed the distribution of 6s & 9s in rows 1 & 9 quite quickly so was looking for a jellyfish on 9s. The discovery of opposing 69 (remote pair?) in c1 & 9 allowed everything to solve relatively easily from then on. I was gratified to note that my logic largely followed Simon's brain. Maybe I've been remotely infected, 'cos I for one would probably enjoy Simon's company at a party. Not sure the reverse would be true though.
Nice one Shye.
Once I got the 69s left and right, I just tried each out a few steps. 9 on the right was unproductive but 6 led me all the way home.
To my big surprise I managed it in 26:52. Will now take a look on Simon's solving path.
Notably, while there were no threes in corners today, two of the corners were multiples of three, and the other two add up to a multiple of three
Ah, the joys of numerology. No one making off-colour remarks about the 69 pattern?
love the numbering, and layout, of the 12 3 45 in the top line (R1)
24:36 snapshot : I think there is no immediate need to find that 69 pair un the rows 4 and 6. The point is that, in row 1 and 9, therecan be only one 6 and one nine in columns 1 and 9, so there are two missing (there must be two sixes ant two nines in these rows). That gives an virtual X-wing in sixes between column (1 or 9) and 6, and another virtual X-wing in columns (1 or 9) and 4.
14:54 for me. Awesome puzzle!!
Is there a name for that breaking? If not it should be called the shye double finned x wing! That is absolutely amazing! Lovely puzzle! Hats off Shye! Well done Simon!
I miss classic puzzles
I can’t believe I solved this on my own. Shye’s ordinary sudokus kill me. I used set though. Highlighted C1,4,6,9 and R1,4,6,9 and removed the overlaps. If you had the 69 in R6C1 and R4C9 then it became really clear the 6 and 9 had a big constraint in boxes 4 and 6. Not sure if that was the intended solve but it was easier than what Simon did.
simon: ... if you're better than me.
me: yeah, right
When solving this I, for whatever reason, ended up going down a path of logic (which was pretty much 1 fork of a bifurcation) where I had gone "If I can prove x, then lots and lots of stuff happens" (I've forgotten what x was exactly unfortunately, something to do with 6s and 9s). I ended up getting so much stuff from that assumption that I ended up thinking that this is surely the correct path and decided that, rather than going back to prove x and losing my progress, I would just finish the puzzle and then later go back to the start to try and prove x.
The entire puzzle solved except for 2 digits, one of which was r4c9 which had no possible values. Which meant that my assumption, despite solving 98% of the puzzle, was actually false. This is utterly bonkers.
1) I have no doubt that the computer struggled really hard if there are almost perfectly valid (but not quite) configurations consistent with the clues given
2) This was the thing that made me look at r4c9 in the starting grid. Since I knew that ANY solve would have to look at it so as to eliminate the almost-perfectly-valid solution I had found.
At 25:00, I expected Simon to realize that with those two 69 cells we now have virtual swordfishes on both 6’s and 9’s! So frustrating! :-D
He did say his fish facts were lacking from the outset, rather missed the clue of the name as a result.
No way would I have gotten that breakin. Brilliant.
A computer is only as 'smart' as the person who programmed it. In fact this could be solved logically by an algorithm if someone decides to implement it. The common algorithms only check for patterns on digits, but you can also check for patterns on colours. In this case, you could give R6C1 a colour, say red, and ask where 'red' goes in rows 1, 5, and 9. This is equivalent to asking which cells in these rows can contain a 6 or a 9 and don't see R6C1. You'll end up with a swordfish in columns 4, 6, and 9, and that means that R4C9 is not red. From there on, the puzzle solves with normal Sudoku.
There was a similar puzzle a while ago (was it by Jovi_al?) that could be solved with colouring as well (IIRC involving a finned X-wing). Now, checking for all of these colouring patterns would add significant computational overhead, but in theory it can be done.
26:55. Worked really hard to find the discovery then quick solve
4:13 The fuzz coming to arrest Maverick for crimes against sudoku?
wow, The Werefrog didn't find that 69 pair in boxes 4 and 6, and that was the whole key to the puzzle. Once you get that, it all falls together.
29:55 hello, just wondering why was there a 6 in r1c6? Couldn't 6 be also in the r1c9?
that cell couldnt be any other option (placing 6 in r1c9 would mean r1c6 cannot be any correct digit)
@@shye229 you're right! Btw great puzzle you created!
I got this by guessing at some 50/50’s so curious what the actual solutions supposed to be lol
"It's almost swordfishy" in a puzzle called Marlin 😉
Simon always seems to ignore his corner pencil marks. Gets 69 in the corner.... ignores the 2 pencil mark that will give him both the 2 and the 4 in box 7.
A marlin is closely related to a swordfish. It's not QUITE a swordfish, but it looks like one. And the pattern with the 6 and 9s is kind of an offset swordfish. If the 69 pair in boxes 4 and 6 were in the same row, it would be a true swordfish.
Maybe a White Marlin is something like 'swordfishy'!
I was really hoping Simon would correct his gaffe from yesterday using the erroneous 7 pencil marks and getting lucky.
Seems like this was a pre-record - no birthday shout outs nor Patreon monthly puzzle solver names read out. Maybe when he's back doing videos on the day of upload.
Wow love some classic every now n then
Simon wasn’t wearing his new spectacles. I’m guessing that added at least 10 minutes to his time. 😅
Aha, that's the reason he completely ignored the center box for so long!
I believe Marlin is a type of Swordfish?
I managed to solve it without spotting the intended logic.
I was able to eliminate a 2 in r9c1 which gave me a few digits reducing r4c3 to 68 and then that whether it was 6 or 8 r9c9 was always 1. If it was 6 r9c9 was 6 and r4c9 was 9 and if it was 8 r7c7 was 8 and r7c7/8 were a 69 pair. It unravelled fairly easily from there.
Amazing
Finish in 17:11.
A haiku
Numbers nine and six
Keys to ancient mystery
Have a Scooby Doo.
-Fnaire Otter
Isn’t a Marlin a kind of sawfish? If so, this bodes ill for me.
I wrote my own Sudoku-cheating program, so I know exactly the problem it had with this puzzle: it is only ever considering one digit at a time, and it has no sense of curiosity. Unless you know The Trick, there is effectively a deadly pattern on sixes and nines. You only need to play about with digits for a moment to realise that R6C1 and R4C9 have to be different, but you have to think of both digits together: if you fill in either one of those cells, you quickly run out of steam trying to chase it around the grid; but writing the same digit in both cells converts just enough potential energy to kinetic energy for the two to collide!
And it's making my brain hurt trying even to begin thinking about how to get a computer to deal with that sort of situation .....
Only people with poor understanding in mathematics will overestimate the power of computers.