After Simon pointed out the border of the 5 x 5 square in the middle seeming to be important, it then occured to me to throw Phistomophel's Theorem at it (Simon says it's an " axiom", but no, it's a theorem) & it fell apart quite easily. However, I don't think the theorem was known in Jan 2020; also, I felt a bit guilty using such an outre result! ;-)
Yes, Phistomophel was really the only technique I used. But it came up half a dozen times!!! Spoiler alert!!! After populating the grid, I found there were 3 9s on the ring, so I was able to place 9s in the corners; then same thing with the 6s; then a 5 (with a neat elimination); then going the other way, I was able to place a 2 on the ring by looking in the corners; same with a 7; then returning to using the ring to place the corners, I knew the last 4 digits in the corners were 1,4,5,6 and eliminating a 3 from the candidates cracked it. I've been looking for a puzzle to use this technique on and rarely find one. There were literally at least 6 instances--every time I got stuck!
For example, at 9:30 when Simon admits to "struggling", he has the Phistomofel ring completed. The two sixes in the ring demand two sixes in the corners. A quick analysis precludes r8c9 being a six because it would knock out both corners it sees. So it's r9c8 instead. You also need 2 fives in the corners which are quite restricted. And you already have both twos in the corners, which eliminates all twos as candidates in those squares in one fell swoop.
Jeremy McGuire I actually did. Never spotted the jellyfish. I instead used an empty rectangle on 6 in box 7, eliminating a 6 in box 3, forcing it into column 9 to get the 6 in box 9. Then I looked at 3’s in box 8 and found something extra hidden: if it’s in the middle, that makes r9c6 a 48, forming a 348 bent triple in box 6 because r9c3 can no longer have a 6. That led to a 5 in r9c7. If, on the other hand, the 3 in box 8 had been on bottom, obviously then r9c7 had to be a 5 still. More simple sudoku solved it from there.
I was inspired by watching these for several months to finally try my hand at the puzzle before watching. It’s my first time ever trying a puzzle of anywhere near this difficulty. It took me over an hour, but I was able to solve it! Every time I spotted a trick, I could hear Simon’s voice in my head describing how it worked. 😅
Derek Neal thank you for your kind words! I was very proud when I hit check and it didn’t find any errors. Hopefully some day I’ll be able to solve these at a pace comparable to the regulars here.
I actually had made the same mistake. I got stuck at the jellyfish pattern, and when I came back to the video and looked at the comments, I read yours and noticed my mistake lol. Maybe if I make the same mistakes as experts I'm on the right track.
In every sudoku i have watched him to solve have had similar problem. He is trying too hard to find depth, neat tricks and patterns and then he doesn't see and hear the screaming elephant in the room. Yes he is master solver no doubt.
I just did a phistomefel on the middle square (doesn't contain a 3 => corner 2x2 squares can't contain a 3, either) which is pretty much the same it turns out. On the other hand, y-wing was was new to me so I never stood a chance :/
Seeing the ring pattern appear at the 6:00 mark is even more significant now that a later video published in mid April revealed that these exact same digits most also appear in the four 2x2 corner boxes.
yes I was about to comment the same thing! I'd love for him to re-solve this puzzle now knowing what he knows and see how much faster his time would be
Its so much easier to identify the x wing, y wing, and skyscraper on your puzzles! I still have trouble seeing it when I do my own. Thanks for another great example!
The 3 can also be placed using Phistomefel's theorem. We can also place the 6 in a similar way (ie, instead of the skyscraper we use Phistomefel's theorem)
managed to do it as per the creator's solve path, but paid the price, just over 90 mins!, had some naked singles missing then found the skyscraper, but then took me so long to find the jellyfish, and then the bent triple, but still very pleased i got the intended solve
Anyone else looking at the phistomofel ring at around 13 and thinking okay what digits are in the corners. No threes. I actually used it to infer the 7 in R3C3, because Simon had one and I didn't. But I knew I needed one from the corner!
That central square outline highlights a recently found method as well, where the numbers in that outline must be the same numbers that appear in the 2x2 squares at the corners of the puzzle. There is no 3 in the central square outline, therefor there cannot be a 3 in row 9 column 2. So you can duck the jellyfish by applying an even more obscure property of the puzzle.
Interesting enough, the property is less obscure when one realises a proof: an application of none other than the jellyfish to an arbitrary positioning of the 3 within the central box - i.e wherever the 'internal' 3 is located there will remain 4 rows, drawn from rows 3-6, with columns 1,2,8,9 needing a '3' in each case. Hence, the intersection of rows 1,2,8,9 with columns 1,2,8,9 (the corner 2x2 squares) cannot contain a 3.
yes, and that gives a break into th corner 2x2's because it disabiguates the 1 and 6 in c9 (in order to have two 5's and two 6's in the corner squares to match the outline)
astonishing. Even if I'd seen the bent triple, I'd have never seen the jellyfish or skyscraper. I'd be struggling with endless chains. Genius to solve, genius to compile
This was a great one! I got the jellyfish and an empty rectangle on 6s that did the same thing as the skyscraper, but while I found the Y-wing I didn't realize the full implications of it so I had to eventually find a logic chain that proved that at least one of r3c2 and r9c7 was a 5, which eliminated the 5 from the center of the Y-wing and got me the rest of the way there.
Same here 12tone .. I found open rectangle in box 1 .. tried 6 .. and eliminated 6 from r9c3 .. which led directly to a 6 in r8c3 .. and rest of puzzle eventually gave it up without j-fish or skyscrapers .. I thought this was the big secret - especially since Simon is so fond of open rectangle technique .. surprised to see the solve .. suggests this puzzle may be even more clever than previously thought ..
I always solve the puzzle before watching the video, it took me 30 mins and 48 seconds, I only knew a few techniques like x-wing, y-wing and rectangle elemination, and i use rectangle elemination a lot in this puzzle. A lot of comments mentioned Phistomophel's theorem, I don't know what that is but I guess it's a pretty useful technique, so after this video I'm gonna search it up.
02:56:11 - Yes, another one took me nearly 3 hours. But, no help, no hints. When I got to the point where so many 8's are pencil marked, I knew some were bogus, but didn't know the right logic to eliminate the bad ones. I noticed that regardless of which pencil marked 8 i pick in the West Box, the same 8 gets eliminated from the East Box. It pointed out a problem, so I tried it thinking that cell can never be an eight, but the puzzle broke trying that. So I went back and to where I tried it just using logic only, and eventually the other 8 in the East Box was forced by normal Sudoku, I kept making mistakes at the end unwinding the pencil marks, and had to go back to resolving the 1, 5 and 6 in the Southwest Box several times, before noticing one of the pencil marked 6's can't be a six and apply all logic as it arose. Finally, that worked out. Somewhere along the line after that I recognized a deadly pattern and used uniqueness to know which way to go,
I've never heard of this Jellyfish Pattern before. However, I've eliminated the 3 from cell r9c2 early on, by using the Phistomefel ring thingy. Once I got all the numbers in the ring, I noticed it didn't include any 3s. 1Subsequently, the outer boxes couldn't include any rings either, hence, no 3 in the cell. Of course, I got stuck later because I hadn't been able to figure the either/or elimination thing you've conjured up...
Yeah, I gave this a try. Ended up taking 3 tries with a total of 4.5 hours to solve. Didn't use any of those diabolical techniques. I have a lot to learn.
This was really fun. The jellyfish and bent triple were the only exotic patterns I needed, once I saw that r3 had only one place for its 7. The rest of the puzzle just fell out for me. That jellyfish is really cool. I didn't know it had a name -- I think of it as the "N-Equations-in-N-Unknowns" rule -- 1 digit, in one of four places in a row and four places in intersecting columns. When its 2x2, we call it "X-wing". 3x3, we call it "Swordfish". Here, it's 4x4 (jellyfish). Of course, it leads me to ask if a 5x5 ever comes up and what its name is if so.
I used phistomefel ring (probably wasn't a thing when the video was uploaded) to solve it but I didn't see the bent triple so I spent a little too much time (65 mins) but I'm proud of myself! I'm new to this hard sudokus. :)
Given the name of the puzzle, I was looking for jellyfishes, and found the one you discovered, as well as one in the columns (3,4,6 and 7, although ultimately they gave the same information). I didn't spot the skyscraper, but I was able to use an empty rectangle in the top-right box and eventually solve the puzzle in 21:08; which I am very proud of. I am getting better every day thanks to this channel!
The 5-6 pair on the left was a great setup for an empty rectangle in box 3 to eliminate a 6 in box 9. I didn't expect the skyscraper after how you've gone on about how much you love empty rectangles :)
I've been watching for a while, but this was the first I've tried to solve before watching the video. I got stuck at the point of the Skyscraper. I hadn't seen that technique before. I really enjoy these videos and it has given me a hobby. This channel is a delight.
I'm watching this after you've shown the property for which the amount of each digit in the four 2x2 blocks in the corners must be the same than the amount of each digit in the square making the jellyfish, which helped me solve this puzzle since it made it easier to rule out the 3 at J2 without resourcing on the bent triplet because there was no 3 in the jellyfish square (leaving only the 45 pair) and helping spot the 6's positions without needing to resource in the sky crapper technique: having two 6 in the jellyfish square means there could be only two 6 from the four candidates in the corners' 2x2 blocks, so should the 6 were the one from the 1-6 candidates pair at the lower right corner, it'd be ruling out the other 6 in the same corner, the 6 in the upper right corner from being in the same column and the 6 from the lower left corner, leaving only one 6 among the four corners, that wouldn't match the number of 6 in the jellyfish block, hence the 1 must be in that cell instead, leaving one 6 in the same lower right corner (at J8) and the other from the 1-6 pair at the upper right corner, while ruling out the 6 from the lower left corner from being in the same row, leading to the right amount of 6 matching the amount of 6 in the jellyfish block. Hence it seems to me the property telling the amount of each digits in the jellyfish block must be the same than the amount of each digit in the four 2x2 corners seems to override the jellyfish technique altogether while helping cut corners.
Not that it made any difference in the path to finding a solution, but the threes formed a second jellyfish in columns 3,4,6 and 7. Very beautiful puzzle.
That they did, but its effect was the same as what Simon used. They removed the 3's from the four corners as well, he mostly needed that to overcome the logic
First time actually trying the puzzle before watching the video. Got it done in about 28 minutes! ^^ (albeit, I messed up 4 squares at about 25 minutes, but was able to fix them)
I solved it much like Simon did, but I don't recognize jellyfish or skyscrapers. I can see "bent triples" sometimes but I think of them as "oh, one of those". I did not notice the unusual box in the center either. My big clue was that there were a lot of places that could only be a 3, 4, or 5, so I followed that lead using Ariadne's thread and was able to eliminate possible digits and solve the thing. I thank Mr. Neal for a great puzzle!
I'm watching this a year later and I think, this would've been a much easier solve using Phistomefel's theorem/ring, because Simon got the middle square filled in quite early, so it shouldn't be an issue to fill in the corners which he didn't get resolved until the very end. But this video was recorded four months before Phistomephel's discovery, I think :)
Yes. that is my difficulty level - diabolic. thanks for the tips i.e. considering the numbers by column on its entirety...and process of elimination by column. I will apply that tip and trick!
At 11:12, went back to my puzzle and gave it a thought, and realized both of the columns and rows making up that square were missing 3, and the corners were at corner blocks. This means each corner has to have a 3 on either a row or column making up the square, and that's how I eliminated the 3 from r9c2
Instead of skyscrapper we could also go with the fin x wing on row 2- 8 col3-9 and the fin on row 8 col 1 which eliminates 6 from row 9 col 3 and the 6pair on row 8 col 1-3 as a result, eliminating 6 on row8 col9, therefore giving us 6 on row 9 col 8
You should be, this was a headache. I screwed it up but did the "undo" back to where I knew I screwed it up and then was able to solve it easily but dang this one had me wrapped around for a while.
Very happy to see a brilliant, challenging classic Sudoku on this channel. (not that I don't like the variants!) Managed to solve in 11:59, but I have to admit I probably wouldn't have noticed the jellyfish without the thumbnail hint and the 'stinger' reference... personally was a bit of a spoiler, but nevertheless really enjoyed solving this puzzle. Thank you once more!
You missed a hidden pencil marked 3s in column 7. After you knew the 3 in row 7, the jellyfish removed the 3 from all other squares in box 9 except the 2 in column 7, which reduced the 3 in box 3 to row 3 a little earlier. But I don't know how much that would or wouldn't have helped you, though. You still did it better than I probably could have. Good job on the solve!
Woof, what a puzzle! Took me 56:38, and I had to get some video help to get past my first ever jellyfish and to point out the skyscraper, but after that was smooth sailing.
On row 2 6 can be only on c3 or c9 and because on column 9 6 can only be in r2 or r8 I used empty rectangle in box 7 on sixes to get rid of the 6 on r8c9. So for me it was jellyfish, y-wing and empty rectangle. Quite a repertuare of thecniques.
I was able to spot the A3 6 you spotted at the 21m mark, at your ~10m position, and that broke the whole puzzle open, and I'd finished it a minute later with just basic Sudoku pretty much. Except I used the positions of the 6's in rows 2 & 8 instead.
Hello. I just found your channel and really enjoy this. I've learned many cool strategies watching you. At 11:44 into this video, you used a JellyFish pattern to eliminate the 3 from row 9, col 2. I eliminated the 3 from that position using a different strategy you explained in another video. In the other video you said that the ring from 3,3 to 3,7 to 7,7 to 7,3 back to 3,3 has the same numbers as the four 2x2 boxes in the corners. Because there are no 3's in the ring, I knew there couldn't be a 3 in any of the four 2x2 boxes in the corners, so I was able to place the 4 in that spot.
4:54 - In R5, C8 you had a 7 screaming at you for a long time, because of the 37 restriction in C1 and the 7 in R9 C9. Solving that would have given you two 1's also.
It's incredible how easy they look when he starts to solve eveything, it's like "oh, yes, of course, that makes sense". The thing is, how to spot the patterns
Interesting that for once I could do this reasonably quickly, once you mentioned the square. I used Phistomafel's square and corners to place the outer corner digits, and it all went swingingly after that!
I don't think the Phistomefel theorem was a thing at this time but going back and watching this you can actually deduce at 4:40 that the empty cells in the Phistomefel ring must both be 7s.
I'm not very good at these (slow), but here is how I solved it: R4&R6-C1 are restricted to a 37 pair. R7C1 is restricted to a 13 pair and R7C2 is restricted to a 34 pair. Therefore, R7C1 cannot be a 3 (because either R4 or R6 in C1 must be a three) and is a 1, which makes R7C2 a 3. Now R3C8 is also a three because of the x-wing consisting of R3 columns 2 & 8, and R7 columns 2 & 8, and this cracks it open. [In fact, you don't even need to use the x-wing once you realize that column 9, rows 2 & 8 are restricted to a 16 pair--then row 3, column 8 becomes a 3.]
It's actually not the center 3 but just the 16 cells around the middle box5 that rule out 3 from the 16 corner cells of the grid. So this must have been pre Mephistomefels rule being known to you, Simon. Once you have established all 16 numbers in the ring you unfortunately can't just sort them to corner cells, that varies, but there's no 3 in them, therefore no three in the corners.
@@kimberlypreston6204 I created a puzzle two months ago that uses that technique. It was featured last week, and in the video Simon gives a great explanation of what it is and why it works, starts around the 21:30 mark: ua-cam.com/video/VfTFtnCBtOQ/v-deo.html
@@PuzzleQodec Awesome. Will look at it tonight! Just started doing puzzles. Didn't realize they are so relaxing and there is a whole community doing them. :)
I know this predates when the pattern was discovered, but since there is no 3 in the outer ring, then there could have been no 3 in r9c2, leading into the bent triple. But given that technique wasn't known yet, this puzzle was made a lot harder.
Amazing puzzle! Got around the Jellyfish and all by the pattern you discussed in a previous video. where the 16 cells around box 5 and the 4 2x2 squares in the edges have to match. i was able to elimante a lot of marks by this. well done Derek Neal! edit: found it: /watch?v=ZLcey7qiXv8
Didn't notice any of those things. Just had to keep making chains. Somehow managed with that. And yeah, I don't call them skyscrapers but eliminated the 6 the same way.
10:00 -- It took me forever to figure out what to do with 35 at the bottom of C7 -- Its effect on B7 and B9 (blocks). If R9C7 is a 5, C2 has a 34 pair, and C3 has a 68 pair. Otherwise, if R9C7 is a 3, It forces R7 to have a 24 pair, forcing a 3 into C2. Either case, C3 of B7 can't have a 3. B7's 3 is in C2. Once you do that, you can go around the circle ruling out 3s and finally fixing a 3. Then another application of R9C7's 35 removes a 6 from B7, and this might crack the puzzle. (End of puzzle) It does. 14:00 -- Perhaps it might have occurred to me, now that you mention something similar, to rule out 3s from the corner 4x4 blocks using whats-his-name's theorem. Then if I ever recognize the pattern (or after you mention it), I could use the "bent triple" to rule out the 3s from B9's 234 pair. 19:00 -- B9 has its 3 in C7. B3's 3 must be in R3.
very proud i could find the jellyfish on my own. i had to sleep on the puzzle though haha, and I used the other 4 columns for the jellyfish instead of the 4 rows you used edit: i also used a skyscraper on fives in columns 2 and 8 instead of yours on sixes
6:00 you now know what it is, dont you :) 1h15m37s im slow but hey its nice to be able to solve sudoku without mastering none of those techniques. This is the 3rd hard sudoku that had a "crown" pattern. It helped me solve those without more advanced techniques. Tho I had to use proof by contradicton to eliminate the 6 from 56 in r7c1.
I know the thumbnail spoils that there is a jellyfish, but I'm still proud to have found it, since I learned it recently and had never applied it until now.
he must have done a video about this later than thia one, then. i mean the 4 squares in the corners have the same digits (not sum! exact digits!) as the circle he drew here around the centre box.
Great solving - unfortunately I did not spot jelly fish and XY Wing pattern! I would have had to go with the pairs ( 56 or the 37 ) and see how far it takes me using ARIADNES THREAD. Does anybody know about this ‘technique’?!
12 mins 30 or thereabouts. How to eliminate the 3 from R7C2. Phisto's Ring: the numbers in the square that surrounds the middle 3x3 are the very numbers that are going to occupy the for corner 2x2s. Guess what? There is no 3 in those 8 numbers. That being the case, R7C2 cannot be a 3. Hence you have your Y Wing!
I like how the ring theorem has made some interesting changes. Since there isn't a 3 in the ring we can rule out a 3 from r9c2 :) EDIT: I like how this is a puzzle that can really take advantage of the ring, haven't seen that in regular sudokus before. 21:30 we know the corners of the boxes needs to have 2 sixes and since both cannot go in c9 we can place them
I'm SO proud of myself that I saw Phistomefel's trick and could deduce a 3 in R7C2 on my own before Simon did! :) PS: Of course I only know about that trick BECAUSE of Simon - but let's be quiet about that. :P
Quick question! @8:06 can't you deduce that a 2 can not go into row 6 column 9 because of the unique rectangle rule? Thus putting a 2 in row 7 column 9?
The fact that there are no 3s in the ring mean you can’t have a 3 in any of the corner 4x4 boxes. Same conclusion as the jellyfish but doesn’t need the central 3...
I also just noticed that from a different point of view. Once you have that big jellyfish box, the only way to place the threes is on swastika arms. Try anything else and you see there's no way to complete the four arms. And that knocks out the threes from your corner 4x4 boxes, as you point out. Simon's a great teacher, would have never spotted anything like that if he hadn't drawn our attention.
Whenever he says, "Ah. Ahh, wow. Okay," you know something amazing is about to happen.
It's amazing how much easier this puzzle is in hindsight, with the knowledge of Phistomophel's Theorem in mind...
I was thinking that as well!
After Simon pointed out the border of the 5 x 5 square in the middle seeming to be important, it then occured to me to throw Phistomophel's Theorem at it (Simon says it's an " axiom", but no, it's a theorem) & it fell apart quite easily. However, I don't think the theorem was known in Jan 2020; also, I felt a bit guilty using such an outre result! ;-)
Yes, one of my fastest solves (13:35)
Yes, Phistomophel was really the only technique I used. But it came up half a dozen times!!!
Spoiler alert!!!
After populating the grid, I found there were 3 9s on the ring, so I was able to place 9s in the corners;
then same thing with the 6s;
then a 5 (with a neat elimination);
then going the other way, I was able to place a 2 on the ring by looking in the corners;
same with a 7;
then returning to using the ring to place the corners, I knew the last 4 digits in the corners were 1,4,5,6 and eliminating a 3 from the candidates cracked it. I've been looking for a puzzle to use this technique on and rarely find one. There were literally at least 6 instances--every time I got stuck!
For example, at 9:30 when Simon admits to "struggling", he has the Phistomofel ring completed. The two sixes in the ring demand two sixes in the corners. A quick analysis precludes r8c9 being a six because it would knock out both corners it sees. So it's r9c8 instead.
You also need 2 fives in the corners which are quite restricted.
And you already have both twos in the corners, which eliminates all twos as candidates in those squares in one fell swoop.
Glad you enjoyed it :)
Awesome puzzle my dude
I don't think I could have solved this puzzle without the hint in the title.
Jeremy McGuire I actually did. Never spotted the jellyfish. I instead used an empty rectangle on 6 in box 7, eliminating a 6 in box 3, forcing it into column 9 to get the 6 in box 9. Then I looked at 3’s in box 8 and found something extra hidden: if it’s in the middle, that makes r9c6 a 48, forming a 348 bent triple in box 6 because r9c3 can no longer have a 6. That led to a 5 in r9c7. If, on the other hand, the 3 in box 8 had been on bottom, obviously then r9c7 had to be a 5 still. More simple sudoku solved it from there.
You almost discovered Phistomefel's Theorem.
I was inspired by watching these for several months to finally try my hand at the puzzle before watching. It’s my first time ever trying a puzzle of anywhere near this difficulty. It took me over an hour, but I was able to solve it! Every time I spotted a trick, I could hear Simon’s voice in my head describing how it worked. 😅
Congratulations! This is a very tough puzzle to pick as your first one to try before watching the video. Bravo for solving it!
Derek Neal thank you for your kind words! I was very proud when I hit check and it didn’t find any errors. Hopefully some day I’ll be able to solve these at a pace comparable to the regulars here.
He solved in 20 minutes. I took 90, using Phistomophel's Ring... I still have soooo much to learn and practice 😅
Brutal indeed, jellyfish into y-wing into skyscraper? *Oof*, I wasn't even aware of 2 of these patterns, and can't recognize the other on my own yet!
At 6:00 "It feels like there should be something I should be appr3ciating about this". We all know now that's Phistomofel!!
It’s crazy how he saw this grand pattern hidden in plain sight but didn’t notice a simple 8 in column 9
For didactic reasons. Otherwise he would probably not have needed the diabolics.
Seemed to be so tunnel-visioned on finding the three he couldn't see that 8 screaming at him
I actually had made the same mistake. I got stuck at the jellyfish pattern, and when I came back to the video and looked at the comments, I read yours and noticed my mistake lol. Maybe if I make the same mistakes as experts I'm on the right track.
In every sudoku i have watched him to solve have had similar problem. He is trying too hard to find depth, neat tricks and patterns and then he doesn't see and hear the screaming elephant in the room. Yes he is master solver no doubt.
I just did a phistomefel on the middle square (doesn't contain a 3 => corner 2x2 squares can't contain a 3, either) which is pretty much the same it turns out. On the other hand, y-wing was was new to me so I never stood a chance :/
The intro music was... Entertaining.
Darn it I was gonna make this joke.
Scott Joplin wrote the theme "The Entertainer". Was also used in the movie THE STING.
Thanks. Was searching for the hustler
One of my favorite "classical" songs.
@@leojf3828 yes he and we all know that hence he made that joke
Seeing the ring pattern appear at the 6:00 mark is even more significant now that a later video published in mid April revealed that these exact same digits most also appear in the four 2x2 corner boxes.
yes I was about to comment the same thing! I'd love for him to re-solve this puzzle now knowing what he knows and see how much faster his time would be
I saw that - and it's the first time I've ever used that trick to help eliminate numbers. Came in handy!
Very nice, and maybe easier to spot than the jelly fish. No 3s in the ring means no 3s in the 2x2 corner boxes.
Its so much easier to identify the x wing, y wing, and skyscraper on your puzzles! I still have trouble seeing it when I do my own. Thanks for another great example!
The 3 can also be placed using Phistomefel's theorem. We can also place the 6 in a similar way (ie, instead of the skyscraper we use Phistomefel's theorem)
I don't think Phistomefel had developed his theorem at this point, though. Otherwise, Simon would've mentioned it
@@billionai4871 yep, it appeared 3 months later ua-cam.com/video/ZLcey7qiXv8/v-deo.html
And while I'm thinking about it - thank you for creating this channel, Mark and Simon.
After a few months of watching I’ve finally done one of these hard classics in a half hour finally
would like to see you redo this puzzle using Phistomefel Ring.
managed to do it as per the creator's solve path, but paid the price, just over 90 mins!, had some naked singles missing then found the skyscraper, but then took me so long to find the jellyfish, and then the bent triple, but still very pleased i got the intended solve
Anyone else looking at the phistomofel ring at around 13 and thinking okay what digits are in the corners. No threes. I actually used it to infer the 7 in R3C3, because Simon had one and I didn't. But I knew I needed one from the corner!
That central square outline highlights a recently found method as well, where the numbers in that outline must be the same numbers that appear in the 2x2 squares at the corners of the puzzle. There is no 3 in the central square outline, therefor there cannot be a 3 in row 9 column 2. So you can duck the jellyfish by applying an even more obscure property of the puzzle.
Interesting enough, the property is less obscure when one realises a proof: an application of none other than the jellyfish to an arbitrary positioning of the 3 within the central box - i.e wherever the 'internal' 3 is located there will remain 4 rows, drawn from rows 3-6, with columns 1,2,8,9 needing a '3' in each case. Hence, the intersection of rows 1,2,8,9 with columns 1,2,8,9 (the corner 2x2 squares) cannot contain a 3.
yes, and that gives a break into th corner 2x2's because it disabiguates the 1 and 6 in c9 (in order to have two 5's and two 6's in the corner squares to match the outline)
YES! And I finally found a puzzle that I could apply this to!!!
astonishing. Even if I'd seen the bent triple, I'd have never seen the jellyfish or skyscraper. I'd be struggling with endless chains. Genius to solve, genius to compile
This was a great one! I got the jellyfish and an empty rectangle on 6s that did the same thing as the skyscraper, but while I found the Y-wing I didn't realize the full implications of it so I had to eventually find a logic chain that proved that at least one of r3c2 and r9c7 was a 5, which eliminated the 5 from the center of the Y-wing and got me the rest of the way there.
I used trial-and-error to eliminate a 6 from r9c3.
Same here 12tone .. I found open rectangle in box 1 .. tried 6 .. and eliminated 6 from r9c3 .. which led directly to a 6 in r8c3 .. and rest of puzzle eventually gave it up without j-fish or skyscrapers .. I thought this was the big secret - especially since Simon is so fond of open rectangle technique .. surprised to see the solve .. suggests this puzzle may be even more clever than previously thought ..
I always solve the puzzle before watching the video, it took me 30 mins and 48 seconds, I only knew a few techniques like x-wing, y-wing and rectangle elemination, and i use rectangle elemination a lot in this puzzle. A lot of comments mentioned Phistomophel's theorem, I don't know what that is but I guess it's a pretty useful technique, so after this video I'm gonna search it up.
You could have used the technique that uses the 16 numbers around the center are the same 16 numbers in the 4 corners. Phistomophel's Theorem
That technique was not known when this puzzle came out.
Should the Phistomefel ring be called the Neal ring?
02:56:11
- Yes, another one took me nearly 3 hours. But, no help, no hints. When I got to the point where so many 8's are pencil marked, I knew some were bogus, but didn't know the right logic to eliminate the bad ones. I noticed that regardless of which pencil marked 8 i pick in the West Box, the same 8 gets eliminated from the East Box. It pointed out a problem, so I tried it thinking that cell can never be an eight, but the puzzle broke trying that. So I went back and to where I tried it just using logic only, and eventually the other 8 in the East Box was forced by normal Sudoku, I kept making mistakes at the end unwinding the pencil marks, and had to go back to resolving the 1, 5 and 6 in the Southwest Box several times, before noticing one of the pencil marked 6's can't be a six and apply all logic as it arose. Finally, that worked out. Somewhere along the line after that I recognized a deadly pattern and used uniqueness to know which way to go,
I've never heard of this Jellyfish Pattern before. However, I've eliminated the 3 from cell r9c2 early on, by using the Phistomefel ring thingy. Once I got all the numbers in the ring, I noticed it didn't include any 3s. 1Subsequently, the outer boxes couldn't include any rings either, hence, no 3 in the cell.
Of course, I got stuck later because I hadn't been able to figure the either/or elimination thing you've conjured up...
Yeah, I gave this a try. Ended up taking 3 tries with a total of 4.5 hours to solve. Didn't use any of those diabolical techniques. I have a lot to learn.
This was really fun. The jellyfish and bent triple were the only exotic patterns I needed, once I saw that r3 had only one place for its 7. The rest of the puzzle just fell out for me. That jellyfish is really cool. I didn't know it had a name -- I think of it as the "N-Equations-in-N-Unknowns" rule -- 1 digit, in one of four places in a row and four places in intersecting columns. When its 2x2, we call it "X-wing". 3x3, we call it "Swordfish". Here, it's 4x4 (jellyfish). Of course, it leads me to ask if a 5x5 ever comes up and what its name is if so.
I used phistomefel ring (probably wasn't a thing when the video was uploaded) to solve it but I didn't see the bent triple so I spent a little too much time (65 mins) but I'm proud of myself! I'm new to this hard sudokus. :)
Given the name of the puzzle, I was looking for jellyfishes, and found the one you discovered, as well as one in the columns (3,4,6 and 7, although ultimately they gave the same information). I didn't spot the skyscraper, but I was able to use an empty rectangle in the top-right box and eventually solve the puzzle in 21:08; which I am very proud of. I am getting better every day thanks to this channel!
The 5-6 pair on the left was a great setup for an empty rectangle in box 3 to eliminate a 6 in box 9. I didn't expect the skyscraper after how you've gone on about how much you love empty rectangles :)
I've been watching for a while, but this was the first I've tried to solve before watching the video. I got stuck at the point of the Skyscraper. I hadn't seen that technique before. I really enjoy these videos and it has given me a hobby. This channel is a delight.
27:09 for me. Solved it very similar to you, getting the 6 in row 3 out of the skyscraper logic and unwinding the rest of the puzzle from there.
I'm watching this after you've shown the property for which the amount of each digit in the four 2x2 blocks in the corners must be the same than the amount of each digit in the square making the jellyfish, which helped me solve this puzzle since it made it easier to rule out the 3 at J2 without resourcing on the bent triplet because there was no 3 in the jellyfish square (leaving only the 45 pair) and helping spot the 6's positions without needing to resource in the sky crapper technique: having two 6 in the jellyfish square means there could be only two 6 from the four candidates in the corners' 2x2 blocks, so should the 6 were the one from the 1-6 candidates pair at the lower right corner, it'd be ruling out the other 6 in the same corner, the 6 in the upper right corner from being in the same column and the 6 from the lower left corner, leaving only one 6 among the four corners, that wouldn't match the number of 6 in the jellyfish block, hence the 1 must be in that cell instead, leaving one 6 in the same lower right corner (at J8) and the other from the 1-6 pair at the upper right corner, while ruling out the 6 from the lower left corner from being in the same row, leading to the right amount of 6 matching the amount of 6 in the jellyfish block. Hence it seems to me the property telling the amount of each digits in the jellyfish block must be the same than the amount of each digit in the four 2x2 corners seems to override the jellyfish technique altogether while helping cut corners.
Not that it made any difference in the path to finding a solution, but the threes formed a second jellyfish in columns 3,4,6 and 7. Very beautiful puzzle.
That they did, but its effect was the same as what Simon used. They removed the 3's from the four corners as well, he mostly needed that to overcome the logic
BTW: I've always wondered what a skyscraper pattern told me. Now I know. Thanks :D
First time actually trying the puzzle before watching the video. Got it done in about 28 minutes! ^^ (albeit, I messed up 4 squares at about 25 minutes, but was able to fix them)
I solved it much like Simon did, but I don't recognize jellyfish or skyscrapers. I can see "bent triples" sometimes but I think of them as "oh, one of those". I did not notice the unusual box in the center either. My big clue was that there were a lot of places that could only be a 3, 4, or 5, so I followed that lead using Ariadne's thread and was able to eliminate possible digits and solve the thing. I thank Mr. Neal for a great puzzle!
I'm watching this a year later and I think, this would've been a much easier solve using Phistomefel's theorem/ring, because Simon got the middle square filled in quite early, so it shouldn't be an issue to fill in the corners which he didn't get resolved until the very end. But this video was recorded four months before Phistomephel's discovery, I think :)
14:59 by myself and I LOVED the reference to The Sting. My favorite Paul Newman movie.
Yes. that is my difficulty level - diabolic. thanks for the tips i.e. considering the numbers by column on its entirety...and process of elimination by column. I will apply that tip and trick!
Simon, could you solve this puzzle again using the phistomofel ring?
At 11:12, went back to my puzzle and gave it a thought, and realized both of the columns and rows making up that square were missing 3, and the corners were at corner blocks. This means each corner has to have a 3 on either a row or column making up the square, and that's how I eliminated the 3 from r9c2
Instead of skyscrapper we could also go with the fin x wing on row 2- 8 col3-9 and the fin on row 8 col 1 which eliminates 6 from row 9 col 3 and the 6pair on row 8 col 1-3 as a result, eliminating 6 on row8 col9, therefore giving us 6 on row 9 col 8
chasing that 56 around the board was intense. very proud of myself for solving this one.
You should be, this was a headache. I screwed it up but did the "undo" back to where I knew I screwed it up and then was able to solve it easily but dang this one had me wrapped around for a while.
Great solve. Much appreciated.
12:48 Simple Colouring Chain is possible now (D9 is 8 as hidden single).
...... although I have to admit that spotting the jellyfish and sky scraper is more elegant .....
Very happy to see a brilliant, challenging classic Sudoku on this channel. (not that I don't like the variants!)
Managed to solve in 11:59, but I have to admit I probably wouldn't have noticed the jellyfish without the thumbnail hint and the 'stinger' reference... personally was a bit of a spoiler, but nevertheless really enjoyed solving this puzzle. Thank you once more!
You missed a hidden pencil marked 3s in column 7. After you knew the 3 in row 7, the jellyfish removed the 3 from all other squares in box 9 except the 2 in column 7, which reduced the 3 in box 3 to row 3 a little earlier. But I don't know how much that would or wouldn't have helped you, though. You still did it better than I probably could have.
Good job on the solve!
At 20:30, no matter which of the 2 positions the 6 is put in C1, the 6 in C9 ends up in RC29 through that forcing test.
Woof, what a puzzle! Took me 56:38, and I had to get some video help to get past my first ever jellyfish and to point out the skyscraper, but after that was smooth sailing.
Pre-Phistomefel sudoku is so primitive, lol
On row 2 6 can be only on c3 or c9 and because on column 9 6 can only be in r2 or r8 I used empty rectangle in box 7 on sixes to get rid of the 6 on r8c9. So for me it was jellyfish, y-wing and empty rectangle. Quite a repertuare of thecniques.
The *sting* of a jellyfish is only *entertaining* when it happens to someone else. This puzzle teaches you to be careful where you *place* your *3's.*
I was able to spot the A3 6 you spotted at the 21m mark, at your ~10m position, and that broke the whole puzzle open, and I'd finished it a minute later with just basic Sudoku pretty much. Except I used the positions of the 6's in rows 2 & 8 instead.
Hello. I just found your channel and really enjoy this. I've learned many cool strategies watching you. At 11:44 into this video, you used a JellyFish pattern to eliminate the 3 from row 9, col 2. I eliminated the 3 from that position using a different strategy you explained in another video. In the other video you said that the ring from 3,3 to 3,7 to 7,7 to 7,3 back to 3,3 has the same numbers as the four 2x2 boxes in the corners. Because there are no 3's in the ring, I knew there couldn't be a 3 in any of the four 2x2 boxes in the corners, so I was able to place the 4 in that spot.
Clever find on the 2's. I would not have noticed that. Then again I'm a beginner!
LOVE the new intro music lol
The Entertainer - Scott Joplin
4:54 - In R5, C8 you had a 7 screaming at you for a long time, because of the 37 restriction in C1 and the 7 in R9 C9. Solving that would have given you two 1's also.
oh wow, i had no chance to spot that Jellyfish! i got upto there no problem, but then had to turn to the video. very good solve Simon!
It's incredible how easy they look when he starts to solve eveything, it's like "oh, yes, of course, that makes sense". The thing is, how to spot the patterns
Interesting that for once I could do this reasonably quickly, once you mentioned the square. I used Phistomafel's square and corners to place the outer corner digits, and it all went swingingly after that!
44m 36sec - jelly, some xwings a y wing and an empty rectangle on 6s at the end
I don't think the Phistomefel theorem was a thing at this time but going back and watching this you can actually deduce at 4:40 that the empty cells in the Phistomefel ring must both be 7s.
After filling in the 1 and 9 in box 1, an empty recangle consideration for sixes gives the 6 in r9c8, which might be a bit easier than the skyscraper.
I once got stung by a jellyfish when I was careless in placing a number 1
I'm not very good at these (slow), but here is how I solved it: R4&R6-C1 are restricted to a 37 pair. R7C1 is restricted to a 13 pair and R7C2 is restricted to a 34 pair. Therefore, R7C1 cannot be a 3 (because either R4 or R6 in C1 must be a three) and is a 1, which makes R7C2 a 3. Now R3C8 is also a three because of the x-wing consisting of R3 columns 2 & 8, and R7 columns 2 & 8, and this cracks it open. [In fact, you don't even need to use the x-wing once you realize that column 9, rows 2 & 8 are restricted to a 16 pair--then row 3, column 8 becomes a 3.]
It's actually not the center 3 but just the 16 cells around the middle box5 that rule out 3 from the 16 corner cells of the grid. So this must have been pre Mephistomefels rule being known to you, Simon. Once you have established all 16 numbers in the ring you unfortunately can't just sort them to corner cells, that varies, but there's no 3 in them, therefore no three in the corners.
The same jellyfish was available on columns 3,4,6,7, which would have given you the same result - getting rid of that 3 from c2,r9, among others.
Watching this after the whole Phistomofel's ring discovery is quite fun.
Solved it using Phistomefel's theorem. I now wonder if this particular video inspired it.
What does Phistomefel's theorem mean. I'm a noob and google didn't even know.
@@kimberlypreston6204 I created a puzzle two months ago that uses that technique. It was featured last week, and in the video Simon gives a great explanation of what it is and why it works, starts around the 21:30 mark: ua-cam.com/video/VfTFtnCBtOQ/v-deo.html
@@PuzzleQodec Awesome. Will look at it tonight! Just started doing puzzles. Didn't realize they are so relaxing and there is a whole community doing them. :)
@@PuzzleQodec Just viewed it! Nicely explained. I actually understood it. Thanks
I tried this app for the first time. I love how the pencil marks work. 21:40.
Oh wow, i didn't know about this jellyfish thing, thanks for giving me this new technique ! After having this one the rest isn't too bad
I know this predates when the pattern was discovered, but since there is no 3 in the outer ring, then there could have been no 3 in r9c2, leading into the bent triple. But given that technique wasn't known yet, this puzzle was made a lot harder.
Amazing puzzle! Got around the Jellyfish and all by the pattern you discussed in a previous video. where the 16 cells around box 5 and the 4 2x2 squares in the edges have to match. i was able to elimante a lot of marks by this. well done Derek Neal!
edit: found it: /watch?v=ZLcey7qiXv8
The new intro music caught me by surprise!
Why did you change it? :D
There was an old movie called The Sting which used ragtime music.
@@rabidsamfan Oh, that's so clever! Thanks for letting me know!
Demerion The String is one of the best movies (in my view) of all time. Watch it but don't ever, ever let anybody tell you the ending in advance.
@@malcolmbacchus421 Will do that :)
@@rabidsamfan And here I thought it was celebrating the 20's.
Didn't notice any of those things. Just had to keep making chains. Somehow managed with that. And yeah, I don't call them skyscrapers but eliminated the 6 the same way.
I didn't need the skycraper,Y wing or the jelly fish. After the 56 pair i found swordfish on 6. And that was enough to solve this puzzle.
Watching this after having watched the video about the 4 corners square are the same digit as the square around the centre. Would make it a lot easier
10:00 -- It took me forever to figure out what to do with 35 at the bottom of C7 -- Its effect on B7 and B9 (blocks). If R9C7 is a 5, C2 has a 34 pair, and C3 has a 68 pair. Otherwise, if R9C7 is a 3, It forces R7 to have a 24 pair, forcing a 3 into C2. Either case, C3 of B7 can't have a 3. B7's 3 is in C2.
Once you do that, you can go around the circle ruling out 3s and finally fixing a 3.
Then another application of R9C7's 35 removes a 6 from B7, and this might crack the puzzle. (End of puzzle) It does.
14:00 -- Perhaps it might have occurred to me, now that you mention something similar, to rule out 3s from the corner 4x4 blocks using whats-his-name's theorem. Then if I ever recognize the pattern (or after you mention it), I could use the "bent triple" to rule out the 3s from B9's 234 pair.
19:00 -- B9 has its 3 in C7. B3's 3 must be in R3.
I did try this alas to no avail! But I love the way you solve difficult puzzles.
beautiful how the 3s unwind using geometry
very proud i could find the jellyfish on my own. i had to sleep on the puzzle though haha, and I used the other 4 columns for the jellyfish instead of the 4 rows you used
edit: i also used a skyscraper on fives in columns 2 and 8 instead of yours on sixes
6:00 you now know what it is, dont you :) 1h15m37s im slow but hey its nice to be able to solve sudoku without mastering none of those techniques. This is the 3rd hard sudoku that had a "crown" pattern. It helped me solve those without more advanced techniques. Tho I had to use proof by contradicton to eliminate the 6 from 56 in r7c1.
I know the thumbnail spoils that there is a jellyfish, but I'm still proud to have found it, since I learned it recently and had never applied it until now.
So clever, I’d never have seen that.
at 19:17 the puzzle seems broken if you put a 6 in the 56 pair at r8 c1 (skyscraper situation) - so it must be a 5
Great puzzle!
Solved this in 38 minutes with that ring/corner method as little help and maybe a little dissociating
he must have done a video about this later than thia one, then.
i mean the 4 squares in the corners have the same digits (not sum! exact digits!) as the circle he drew here around the centre box.
Great solving - unfortunately I did not spot jelly fish and XY Wing pattern! I would have had to go with the pairs ( 56 or the 37 ) and see how far it takes me using ARIADNES THREAD. Does anybody know about this ‘technique’?!
This would have taken me 2 years, 6 months, 12 days, 4 hours, 12 minutes and 49 seconds to solve.
12 mins 30 or thereabouts. How to eliminate the 3 from R7C2. Phisto's Ring: the numbers in the square that surrounds the middle 3x3 are the very numbers that are going to occupy the for corner 2x2s. Guess what? There is no 3 in those 8 numbers. That being the case, R7C2 cannot be a 3. Hence you have your Y Wing!
I like how the ring theorem has made some interesting changes. Since there isn't a 3 in the ring we can rule out a 3 from r9c2 :)
EDIT: I like how this is a puzzle that can really take advantage of the ring, haven't seen that in regular sudokus before.
21:30 we know the corners of the boxes needs to have 2 sixes and since both cannot go in c9 we can place them
Whats wild is this puzzle predates phistomefel's ring theorem. So either the setter knew about it or unintentionally showcased it.
15:09 bent triple, ywing, also phistomefel pre phistomefel
I'm SO proud of myself that I saw Phistomefel's trick and could deduce a 3 in R7C2 on my own before Simon did! :)
PS: Of course I only know about that trick BECAUSE of Simon - but let's be quiet about that. :P
that was stunning
41:05 - I only noticed the jellyfish as "diabolical technique" - but at one point I used trial-and-error to eliminate one possibility.
Quick question! @8:06 can't you deduce that a 2 can not go into row 6 column 9 because of the unique rectangle rule? Thus putting a 2 in row 7 column 9?
The fact that there are no 3s in the ring mean you can’t have a 3 in any of the corner 4x4 boxes. Same conclusion as the jellyfish but doesn’t need the central 3...
I also just noticed that from a different point of view. Once you have that big jellyfish box, the only way to place the threes is on swastika arms. Try anything else and you see there's no way to complete the four arms. And that knocks out the threes from your corner 4x4 boxes, as you point out. Simon's a great teacher, would have never spotted anything like that if he hadn't drawn our attention.