The New Sudoku Trick That Almost Nobody Knows

[@nomore6167]

At 7:32, Simon is "stuck" and asks "So, what should we do now?" Unfortunately for him, he can't use the trick that has worked for me 100% of the time -- stop trying to solve it and watch the video.

[@shawon265]

Exactly what I did! 😂

[@blanda8821]

Worked for me haha

[@matthewyoung1973]

Haha me too!

[@cluso9985]

Me too. I only needed one “little” clue. I am getting better by watching/solving but I am terribly slow 😂

[@Finsirith]

@@cluso9985 Yes, as soon as he got the big AHA (that "4" at C4R7), I stopped the video and was able to go over and finish the puzzle on my own.

[@laytonspuzzle]

nothing has striken more fear into my heart on this channel than seeing a puzzle with the only rules being "Normal sudoku rules apply"

[@jestfullgremblim8002]

True

[@tombeal364]

I downloaded the CtC app to see what I could achieve... All went well (but slow) until I came to a "normal" sudoku. I'm far from a genius but I couldn't see how to crack it and had to use the 1st hint to see how to look at solving it.

[@daniellehallihan6015]

I think it's 'struck'. Nothing has struck more fear into my heart. Yeah. That sounds right I think. Sorry, I normally try not to be a douchey grammar policewoman 😂 but I just had to drop a note here. Happy puzzling 🫡😊

[@laytonspuzzle]

@@daniellehallihan6015 Thank you, your kind words have stucken my heart

[@azzteke]

stricken

[@napvs.goldenreaper9715]

I've come across your videos within the last week and can't seem to stop consuming them. As a novice sudoku player, your ability to rationalize your way through these is just remarkable. Would you ever consider providing a sudoku video series on how to progressively improve to higher levels? Many thanks for hours of entertainment! Keep it up

[@shye229]

wonderful solve! especially given the time pressure 😅 you explained the trick exactly the way id envisioned it, always so happy to see that ^^ and happy birthday to mr anthony!

[@marshallgarey2913]

Amazing puzzle, Shye! Thanks for sharing it with all of us!

[@RustyBrakes]

Great puzzle! I feel like the box of "maybe a bent triple" combined with the "crank handle" makes a shape that looks like a cell being divided - is that where the name comes from?

[@RGS61]

Wonderful puzzle Shye! Thank you! .. I would be curious your thoughts re applying the '"slot machine" method to this one, and whether it appears to disprove the "elbow digit" theory?

[@firstlast4506]

Amazing puzzle, well done!

[@shashankjodge974]

Thank you for this gem of a puzzle

[@danielcavallarigoncalves7309]

I can't deal with the fact that Simon just says "these digits are SUS" now

[@EyeDewDie]

I saw the four vent!!!

[@nomanatcore]

Came here to say the same thing. So good.

[@luizzeroxis]

Like, why did he say that? He's clearly not in the joke, didn't seem the type to meme around, he's pretty serious. Maybe he heard other people use it and just sort of accidently internalized it? That's really weird.

[@EyeDewDie]

@@luizzeroxis i could see him playing some among us on his down time. If i was a killer in his lobby I'd kill him first. His puzzle solving skills are too good. (edited due to typo)

[@hnglbanana]

@@luizzeroxis 'sus' is also just a contraction of 'suspicious'. it's been around in english for quite a while. among us didn't invent 'sus'

[@LornaM2304]

Thanks to my fitbit I noticed that when I watch CtC videos my heart rate is at its lowest 😅 Cracking the Cryptic scientifically proven to be relaxing! Hope you enjoyed your lunch Simon 😊

[@bobh6728]

And at it’s highest when I try to solve one on my own!!!!!

[@jamessweatt985]

Are you even calm when a "cell entirely lets itself down"? ; ^ )

[@cgsbdag3235]

Do

[@ostensibly531]

Quite the opposite for me

[@lisas8614]

Makes me want to wear mine again

[@johnseddon7793]

I use dots to mark the possibilities in every square - thus 1 is in the top left corner and two to the right of it and so down to 9 at bottom right - this creates a number pattern in the form of double dots, and triple dots, etc - and the pattern of dots/numbers provides the key - or at worst identifies the way forward 🙏

[@brants.3367]

Binary fission is a method of asexual cell reproduction where it basically splits into two identical copies of itself. Thanks for the tremendous content!

[@davidc1633]

that's a good hint. If you pencil mark all the possible positions for 1 in the grid and color code them properly you are left with at most 2 options for every box and r7c4 gets resolved that way, because it's not a possible position for a 1.

[@jdyerjdyer]

@@davidc1633 That was my conclusion on the title. :)

[@gr328]

Also nuclear chain reactions. It's a hint that it's a convoluted chain I think.

[@PierceArner]

I love how even before the explanation, you can see Simon's sense out where a bit of logic is hiding itself, before he unearths it for himself. It helps get a handle on what he's working at as he comes up with a way how to gradually extract it from the puzzle in a way that makes sense verbally. Brilliant stuff as always, and wishes of a most pleasant celebration of your dad's annual solar rotation!

[@Goryus]

There is an entire class of techniques like this called "Krakens". In each case, the logic is: either this chain is true, or there is a technique here (y-wing, swordfish, it can be any technique). Things like a finned swordfish-kraken can start getting truly monstrous.

[@notnotandrew]

Sounds like gussied-up bifurcation to me :)

[@killroy42]

@@notnotandrew shhhh! It's not bifurcation if you can come up with a cool name for it!

[@fredhassett8390]

@@killroy42 ,

[@RGS61]

Hi Derek - I tried your "slot machine" method here, on 1's .. unfortunately it was not able to yield much, but more interestingly the puzzle seemed to defy the rule of "elbow" digits .. Or perhaps I misunderstood Simon's explanation of the rule, that I took to mean "in a slot machine situation, where there are three candidates in box, the elbow/middle digit , can always be eliminated" ... ??

[@MrBlackHawk888]

I'm not sure that i understand what you are talking about, but i will like this comment just in case...

[@kimberlypreston6204]

Simon looks especially handsome today! Happy Birthday to Simon's dad.

[@TheNewAccount2008]

A long, long time ago, on a Sudoku channel called "Cracking the Cryptic" I learned about a technique called "Slot Machine"... When applied to 1s here, the single 4 is revealed quite quickly. ;-)

[@GrantMcLean]

When Simon noted all the diagonally offset Snyder pencil marks, my first thought was "I wish I could remember how that slot machine technique worked" 😃

[@RGS61]

I would be curious at what point in the puzzle, and what candidate "1" cell you started the slot machine from, in order to eliminate the '4'? .. Starting the slot machine, before inserting 'any' candidate cells, since we know R7C7 eventually proved to the '1' .. starting the slot machine from the three candidate 1 cells that point at R7C7 (R4C7, R7C4, R7C5) all produced possible (non-contradictory) results .. Also interesting, this puzzle seemed to contradict Derek Neal's hypothesis concerning the ability to eliminate "elbow" digits in a slot machine (or at least my understanding of it) .. In this puzzle, the only cells I was able to eliminate "1s" from using the slot machine method were R4C1 and R6C9

[@PowasaurusRex]

@@RGS61 This was exactly the same conclusion I came to (don't know the slot machine, but by solving for 1's). It wasn't until Simon said that only one of the inner corners could be a 1 that the 4 clicked and the rest fell into place.

[@charliel2967]

I'm new to this but my method is to use dots in each square. 1-top left,2-top middle and so on. It becomes easy the more you do it and i love exploring different ways to solve these. Brilliant puzzle

[@Thebrooky12]

my heart skipped a beat when simon said "sus" LOL

[@nomanatcore]

I hope he says it every video from now on.

[@tonymason6637]

"That's the world's most disappointing three". I feel bad for that three now. However, I can't help but laugh lol

[@marnak24]

I actually laughed out loud at that "world's most disappointing 3" comment. 😀

[@patricksmith3135]

My goodness! Your brain was in hyperdrive mode with this one. It was 20 minutes before you placed the first number but it all undid itself quickly afterwards. Congratulations. I enjoyed that immensely .

[@xiuyeyin]

I appreciate that Simon always mentioned his pencilmarking purpose. It reminds me to stick in this method and today I actually spotted the 1s!😌

[@inspiringsand123]

Rules: 02:42 Let's get cracking: 03:02 And how about this video's Simarkisms?! Stuck: 4x (04:49, 08:41, 16:38, 16:41) Beautiful: 4x (01:40, 17:57, 23:16, 27:22) Useless: 3x (09:55, 16:26, 16:29) Sorry: 3x (06:37, 11:17, 12:17) Hang On: 3x (15:14, 18:14, 25:28) The Answer is: 2x (21:13, 23:26) Nonsense: 2x (18:14, 25:20) Gorgeous: 2x (02:34, 24:58) Bother: 1x (12:39) Clever: 1x (27:01) Recalcitrant: 1x (25:08) Incredible: 1x (01:24) Take a Bow: 1x (27:24) By sudoku: 1x (21:04) Approachable: 1x (01:03) Famous Last Words: 1x (03:18) FAQ: Q1: What is a Simarkism? A1: A Simarkism is something that Simon and Mark typically or frequently say. Q2: How do you do this so fast? A2: I'm not made of flesh and blood, but of sand ... Q3: Why don't you include 'XX' and 'YY'? A3: Probably it's already on the list ('Scooby-Doo' for example), but not mentioned in this video. But if you think it's not, tell me what you'd like me to include and there's a good chance I'll add it! Q4: You missed 'XX' at 'YY:ZZ'! A4: That could very well be the case! Human speech is hard to understand for computers like me, especially British sometimes! Point out the ones that I missed and maybe I'll learn! Q5: Could you turn these statistics into videos? A5: I've been playing around with the idea and I'm open to input as to what people would like to see. Let me know if you are interested in this and/or have suggestions.

[@wariolandgoldpiramid]

good bot

[@TheMaxymon]

Reddit bots on yt? Love it

[@multiarray2320]

@@TheMaxymon its a yt bot

[@cypher686]

No one cares

[@deltaangelfire]

how to find the trick: look for the first "beautiful" (that's after the intro :P )

[@chuckdecker3765]

Simon abandoned marking 1s at 11:30 too soon! If he continued around, he'd have noticed you get only 2 possible places for 1 in every box. One color for one pattern of ones, one for another, and the 14 in r7c4 is instantly resolved and the puzzle collapses after that into a simple elimination. He thought he was misleading us, but, it was the direct path to the solution.

[@Paolo_De_Leva]

At that time there was not enough information to get 2 possible places for 1 in every box.

[@onestone9871]

I still don´t understand and hope someone will help me out. Why has there to ´be a 1 in one of those squares of row seven`. Why can´t there be a one in square 4 of row seven, where is the reason for that. Help, help, help😀

[@fl4792]

@@onestone9871 you need to look at cell 2 and 8. 1 can only exists in r1c4 and r3c5 for cell 2. and 1 can only exists in r7c5 and r6c9 for cell 8. with this restriction, r7c4 cannot be 1. this is the method this comment uses. not the one Simon use in his solution. Hope this help

[@Kleyguerth]

@@onestone9871 He explains it quite well in 18:51. You either have a 1 in r3c3, which forces the 1 in r7c5, or you get a bent triple which forces the 1 in r7c7. There's no way to put a 1 in r7c4 without breaking column 5. A reversed way of seeing it: if r7c4 is a 1, r7c7 is a 9, which due to the "bent triple" forces r3c3 to be a 1. Now you have nowhere to put 1 in column 5.

[@lisaoburke1]

Life lesson from Sudoku: The answer is there! Persist. Trust. Don't give up and you will find it! Thanks for the videos. Love it.

[@daniellehallihan6015]

Not only was it beautiful watching this come together, with your offset 1s and hypothetical bent triple, but your pure joy and enthusiasm are contagious. I love the respect and awe that you show toward Shye (whoever she is, I'm new to this channel) and the puzzle and its solution. You can just tell you genuinely love what you're doing. And that really speaks to me, as a lover of puzzles myself. The satisfaction I get from solving a tough logic puzzle is so wonderful. And I'm really becoming obsessed with sudoku and learning techniques from you guys. Great channel, great content, and the english accents are the icing on the puzzle cake 😄🧩🎂

[@benl8962]

Never thought a sudoku video could be so captivating! To watch you come up with the solution is one of the most intriguing things ive seen in a long time! This has motivated me to pick up sudoku's myself again

[@brianmcadam443]

20:47 solve here, which I am extremely happy with for a shye classic. I've decided that my actual "technique" for any of these classics where there's probably some strange technique is to pencil mark literally the entire grid and then look for cells with not many candidates and how they might interact, because it's usually impossible to spot these things with the usual Snyder notation, etc. I spotted the almost-bent-triple on R3C3 through R7C7 and exactly one them would be a 1. Then added in C5 to get something kinda like an X-wing pointing at R7C4, and that was enough to break it.

[@sarahsorlien]

yes solved in 20 min with all pencil marks, hidden doubles (46 & 79) and chains but I really like watching the logic here

[@BrBill]

Exactly how I did it, but I took a few minutes longer.

[@angelavonhalle5144]

I am a fan of this system depending on the set up, does it take longer?sometimes. If time isn't a problem, why not solve sudoku this way, sorry about the simple tactic. It is still a good way to train concentration.

[@Zardox2]

Happy Birthday to Simon's Dad! And... Great puzzle, Shye! I got stuck, and had to watch... but I was VERY close to seeing the bent pseudo-triple... just couldn't quite work it out. But, a year ago I wouldn't have even known where to BEGIN to look... so, I guess I'm learning!

[@ib9rt]

Actually, the bent pseudo-triple was a red herring. You didn't need it to solve the puzzle. It was enough to spot all the rows, columns and boxes where a 1 could only appear in one of two places, and this was enough to eliminate the 1 from the 14 cell that Simon identified. (Coloring the alternate 1's as red/blue pairs made the logic particularly easy to spot. I think this is where the binary fission from the title came in.)

[@fisharepeopletoo9653]

Me after pausing Simon at the start then finishing this puzzle and going back to watch the video: "C'mon Simon I'm already done." Also me: takes over an hour to complete

[@TheFabulousCube]

I've been working on a puzzle generator which, of course, includes a solver and a grader. My solver caught the 2-5 and 4-6 hidden pairs. That's it. The grader put it as 5 star (anything the solver can't solve) . The grader *did* tell me there is a solution, and only one solution. But it used a recursive "pick one and try to solve it" method. I love the symmetry of this puzzle! Me personally solving it, I also caught a locked pair of 1s in column 7 (rows 4 & 7). I'm not sure it helped, I spent way more than 20 minutes on it. Nice one!

[@th.nd.r]

Shye. A genius setter. Wow. Every time, it’s just so ridiculously elegant. Well done. And Simon, wonderful solve, way to spot the trick and to analyze it and break it down so well, wonderful job as always.

[@MyReligionIs2DoGood]

I solved it by using an WXYZ-Wing (1478 in B5C5G4H5), eliminating the 1 in A4 and G5, followed by the Y-Wing on 789 in C3C7G3, eliminating 9 in G7. Took about an hour. However, the way you solved it is more elegant, and gives me yet _another_ nice technique to use. Thanks a lot! :) What will this technique be called? My suggestion would be 'Bent X-Wing'.

[@TalkingInTheDark]

I did it using the WXYZ as well. If you see three in a line like ab, abc and abd then look for WXYZ wings.

[@Raven-Creations]

Very interesting setting. The bent triple wasn't required; the crank shape formed by the 1s was sufficient. If you set R3C3=1, you get a complete set of 1s. If instead you set R7C7=1, you get a different, non-overlapping set. These form two skyscrapers, which can eliminate 1s from R4 in box 4 and R6 in box 6. Crucially, it also rules out 1 from R7C4, making it 4. Following through on this placement, the puzzle quickly unravelled.

[@manfredwitzany2233]

I have finished the idea about 1s at the beginning of the solve. It was possible to color all boxes with exactly two possibilties for the 1. This directly removed 1 from r7c4 and forced it to be a 4 without the nearly bent tripple. But of course your solve was much more beautyful.

[@rabidsamfan]

That is what I did.

[@VanpyroGaming0]

I did the same thing. Follow the 1 pencil marks, and you can reduce the 1 down to 2 possibilities in box 8. R7C5, and R9C4. Eliminating it from R7C4

[@_MrL_]

Video popped up on my YT home screen … the puzzle immediately looked interesting, so I screen shotted the home screen and cleaned it up in photo editing then dropped into books. (i didn’t want ANY clues) I didn’t even open your video. Using a touch pencil only on an iPad screen it took me 12minutes 38 seconds to solve. I then went to look for your video and struggled to find it because I didn’t open it. When I eventually found it, I wanted to see how you solved it… I found myself frustrated straight away that you didn’t see the pair of 6’s in box 2 as for me the 6’s were the key to the quick solve.

[@sammorley5679]

Another masterclass from Shye, absolutely incredible! Please Shye make a setting video for this one too, I found the last one absolutely fascinating!

[@jaredwhite4934]

Hi Sam

[@owo8770]

Does shye have a channel?

[@alexanderstelmakh8906]

@@owo8770 She had an account on KZbin, no puzzles ((

[@thekingcuber]

A while ago you guys had a video of a technique called slot machine which I used to determine that there couldn't be a 1 in the cell. Very useful and I find myself continuing to gravitate towards it when doing puzzles I have a tough time starting out on

[@chris_the_scientist]

The initial number pattern on the grid looks like a cell or nucleus about to divide. Maybe that is why it is called binary fission. It also has some neat patterns in the solve.

[@perpendicularfifths7312]

love shyes puzzles, very excited!

[@JoeyNiklas]

I'm totally not sure how these vids popped into my algorithm, but I'm loving this. I applied many of these techniques when I'm doing puzzles; however, I never knew that there was specific names for them.

[@jestfullgremblim8002]

Yeah, it was kinda weird for me as well. I was used to stuff like Minesweeper and other puzzle games and the techniques in those other puzzle games had no names. They were just patterns found with logic that could be applied in many levels, just like in Sudoku but no one named them

[@djeidot]

As soon as r7c4 is found to have 1 and 4, there's a bent quadruple in r7c4 r8c5 r3c5 and r2c5 which allows to remove the 1 in r1c4 and thus place a 1 in r3c5, making the 78 89 79 bent triple work. You can do it without resorting to Simon's technique.

[@Stephen-Fox]

Needed to use Simon's heavy prompting about the corners around 14 minutes, but was able to spot the naked single 4 in R7C4 thanks to the way the 178 in R3C3 forced 1 into either R7C5 or R7C7 thanks to the almost Y-Wing and spindle. Very entertaining bit of logic, at which point the puzzle collapses very satisfyingly.

[@samus88]

What does she mean she's "really proud of this one"? She should be proud of ALL her Sudokus!!

[@mattiascrowe2549]

Figured out an interesting thing on the phistomefel ring is that you can only place 1s in certain places (~16min) because you have to account for 1s in corner squares

[@bluerizlagirl]

We've seen bent triples, but this is more like a pushed-over X-wing! By which I mean, it's forming a parallelogram instead of a rectangle, as though the top row of an X-wing on ones in r3c5, r3c7, r7c5 and r7c7 had been pushed over to the left by two cells .....

[@Coyotek4]

I'll admit it; I'm not great at 'classic' Sudokus at this site. I was able to find a solution (time: 17:37), but only because I bifurcated and got lucky; I missed the trick in the video completely. Nice puzzle!

[@sylviacarlson3561]

I really liked that whole 1 logic. It was fascinating. Thank you Shye and Simon. LOVE Sudoku!

[@Playmaker6174]

Another classic lesson from the wonderful shye, focus mode on :)

[@BarryMerilion]

Hi Simon, Sorry to say ,I solved this puzzle in 14mins. My method is write any numbers that are obvious then starting at number One ,cross check all the ones ,then twos through nine. when that done I usually start with the squares with the most numbers in them and cross check again one through nine and using [dots] for pencil marks , in empty squares 123 top of square 456 through the middle and 789 for bottom of square. when all squares are filled with dots then cross check again and find what can not go in that square therefore only one number is left that it can be. This may seem as clear as mud to you, but I know what I mean😅. Time was actually 12 mins because I listening to you then drew the grid on a piece of paper

[@joshwalsh6194]

I loved the hail Mary 1 guess and quick delete @ 17:51 and then a total work around to figure out how to put it back in using the new knowledge without going straight for it absolute magician

[@chocolateboy300]

I finished in 39:31 minutes. Classic sudokus that are handmade are always so fun to solve. This one had so many cool parts. Seeing that 6 couldn't go into r2c1 due to the impact of 4 having no space in box 8 was fun to see. It took me way longer, but the next part on seeing that r3c3 couldn't be a 7 as this also makes r2c5 a 7 and breaks in either row 7 or row 8 depending on where the 3 went in box 8 was also amazing to see. If the 3 went in r8c6, then the 7's rule out a place for 7 in row 8. If 3 went in r9c6, then this forces 5 into r7c5, forcing 4 into r7c4, forcing 1 into r9c4, forcing 1 into r7c6. In column 3, the starting 7 forced an 8 into r7c3 and now we can see that in row 7, no place is available for 9. That is so cool to see. Great Puzzle!

[@guihermeborracha7412]

It took me about an hour to solve this, getting some of the tips in the video and checking to see if I'd done right. I do not know any techniques, but I'll try to learn what I can. This was my first puzzle, I hope the first of many.

[@BruceLarson-uv9lo]

This was difficult(!) but I eventually proved that r2c1 cannot be 6 when r7c7 is either 1 or 9, and the rest of the solve was easy. Simon, you are incredible!

[@tofmill3640]

"These three cells are a bit sus." - Simon Better sharpen up your scanning if you think at least one of them vented.

[@sudokucafe]

Love Love Cracking The Cryptic. I have learned so much watching their video. I am addicted to Sudoku. I probably do 8 to 10 puzzles a day.

[@maverickstclare3756]

My backtracking (i.e. bifurcating) solver took 17s for this one, the previous Shye ones were around 28s, so this is not so bad

[@vfatil798]

You know your a CtC fan when you recognize the setters favorite tricks and just stare until you figure it out.

[@karenjohnson7878]

I have never watched anyone do a sudoku before. I love them and it just popped up in my feed. I thought I would give it a go. It was so frustrating. I kept talking to the screen. Constantly going "you keep missing it"

[@asamirid]

seeing simon struggle is very pleasing, i finaly can feel good about myself, amazing attempt to take over the crucial puzzle 💚💚..

[@longwaytotipperary]

Happy birthday Simon's dad!

[@vegasjosh8249]

I feel so much better knowing Simon got stuck at the beginning. I went straight to the page and started the puzzle before watching the video (which is what I normally do). After 8 minutes, I had all of the squares penciled with the two digit combinations, but was not able to actually fill in a single digit on any box. I consider myself to be a "Medium" level player. (Beginner, Medium, Hard, Expert, Master). Simon of course is at the Master Level, so I felt much better about myself knowing that he went to "three and four digit" pencil marks before he ever filled in his first box. If I spend another 30 minutes, I am hoping to get my first box filled in. haha

[@pwfd]

Great puzzle, I did a kind of a different logic to break in though: If the 1 is on r3c3, you get a 78 pair in box 2 that forces 4 into r8c5, forcing 1 into r7c4 that will clash with the 1 in r1c4, so 1 in r3 is at c5 Don’t know if that’s considered bifurcation, but to me it seems like not a long string to follow Great solve btw, as usual

[@xinmyname]

Would love to see a setting video for this from shye! I got a bit of an "atom smashing" vibe from it. e.g. a 2 would smash into a 25 pair, solving both. Perhaps that's the fission reference? Loved it! And Happy Birthday to Simon's Dad!

[@RGS61]

After seeing the pattern of 1's, a "slot machine" reveals R7C4 as a "4", which breaks the puzzle without requiring any pencil marks .. Having said that I love the elegance of the "crank handle" and particularly enjoy that kind of elimination logic .. Well done Shye! Yet another wonderful classic!! .. More please!!

[@Finsirith]

"Slot machine"???

[@topilinkala1594]

@@Finsirith Slot machine is kinda guessing game. Start with grid with no pencilmarks and look at the digits. If you find same digit in three boxes that don't see each other you're good to go. In this example 6 is one of such givens but 9 is not. Now mark where the digit could go in the other six boxes. If there are lots of two places you've good to go. Take one of those and imagine what happens in each case. You might find out that there is a cell where the number must be or that there is a cell where the number can't be. Either way you might get some information. If one digit don't work try another one. In this puzzle there are 1, 3, 4 and 6 all possible slot machine candidates.

[@DougCube]

A more standard approach when stuck there is to see the Finned Swordfish of 1s in rows 1,4,9. The fin is r4c7 and the elimination is 1 in r6c9. Another approach is to spot the Alternating Inference Chain if you were to put a 1 in r6c9 (using r6c3, r3c3, r3c5, r7c5, r7c7, and r4c7). A third approach is the WXYZ Wing when you try to put 1 in r1c4 that would eliminate all candidates (1 and 4) from r7c4 via the 78 pair in box 2 forcing a 4 into r8c5.

[@ianpearse4480]

Wow, that was hard. Great share. Love the diagonal triple trick.

[@TheLefty5o2]

Started casually getting into sudoku lately and tried doing this one before watching the video and couldn't pin down even a single number after like 20mins. Started to watch the video after fumbling for so long and find out it is actual a pretty difficult puzzle. So I don't feel nearly as bad anymore lol

[@fvni]

A deceptively difficult puzzle. This is one of the best of your solves, kudos Simon.

[@pixllo]

21 minutes no bifurcation. Nice puzzle, it felt pretty straightforward to me. I don't really get the title so let's see what technique Simon will summon since my solve didn't need anything fancy (well there was a cell with only 14 as possibles in box 8 and 1 was broken, and a couple hidden pairs/triplets but that's it). Edit : after viewing the video it seems we had the same idea same conclusion different proof and different execution, same result.

[@Paolo_De_Leva]

How could you prove that "1 was broken" in r7c4?

[@pixllo]

@@Paolo_De_Leva well from memory on the same row it placed a 9 and then an 8 on column 3. The 9 made r3c7 a 7. Possibles for R3C3 were 1,7,8 so now 7 and 8 are both ruled out, only 1 left. This breaks box 2 you can't place 1 anymore. Or you could prove it like Simon did, it's more elegant but more convoluted as well Edit: also my solution was just focusing on weak cells just like Simon (which was the most logical next thing to try at that point) and noticing they interacted with each other on the specific case of 1 being placed. The remembering basically a 5-6 long sequence of digits with their placements, again nothing incredible compared to what some guys can do around here and it helps train my memory at the same time! I can never spot Y wings for the life of me but I can do that and with persistence a little bit of luck/intuition depending on the puzzle I manage to overcome my inabilty haha.

[@Paolo_De_Leva]

@@pixllo Interesting. Thanks for explaining. Simon's convolute method was certainly smart, fascinating and elegant, but not more than yours.

[@Heistergand]

21:28 "That's the world's most disappointing three!" Made my day.

[@amoswittenbergsmusings]

I have caught Simonitis: I get physically unwell at the thought of bifurcation. Not a great time as result: 33.12 But I found the break-in on the corners of the Phistomefel ring. Another Simon-induced habit I formed: I look for patterns in the geometry. If Shye would have hidden the kraken in cells that are not in an obvious geometrical pattern, I doubt I would ever have spotted it. The way it was, it was telegraphed. Subtly, but telegraphed it was! That was great fun!

[@ronmasters751]

Bravo! Bent Triple v. Crank eliminates shared digit in a pair. Makes me dizzy just to write it!

[@angieperalta4873]

I did it by my self, i have to say that it took me time. Thanks for the puzzle

[@the_credible_hulk]

Thoroughly enjoyed working through that with you (I just used paper & pen). Not sure I would have discovered those offset pairs of 1s without you because I was still looking at row seven as having three possible places for a 1.

[@matthallyoutube]

ya an r4c7 still can have a 1

[@kapochung9990]

First time to solve Shye new technique sodoku. Proud of myself. 😁

[@SmartHobbies]

I love CTC's solving style and Shye's puzzle so much I made a video on my channel where I analyze each step of this solving video. In addition to explaining the strategies used, I add a couple of pause the video moments and some alternate solving paths, similar to how someone analyzes a chess game.

[@mattbendzinski4945]

love to watch this. I wouldn't have caught that in forever. Just amazing.....

[@martinvanwijngaarden7235]

At 6:20 there is a strong pair in D6: 1-3-8-9 in the box, 2-4-6 pencil marked, 5-7 only possibilities.

[@jonhansen9622]

Happy Birthday to Simon's Dad! Great solve today!

[@davidh.4944]

46 minutes exactly, and I finally managed to solve one of these "spot the secret" ones on my own. I'm doubt I used the exact trick as intended, however. I spent a good while playing around with the 1789 corners, then noticed how R3C5 interacted with them and the 1s in the central chute, such that one of its three candidates could be eliminated immediately, and the other two forced 1 in R7C7. After that it was just a matter of filling out the rest of the grid.

[@enoemos1566]

fist time 10 minutes. 2:nd time A little later 6 minutes.

[@ronp5615]

I think most people miss the corollary. That. Is. All. Happy solving!

[@laurabarber6697]

It was fun to think a long with you. Shows me that my thinking is on the right track.

[@spiritwaterwolf]

I know an additional little trick, when searching for possible placings, you will only find two possibilities of a particular number in either block or line. In a sense there you shouldn't have 3 identical numbers in the box or line. As you noticed two identical numbers sharing a spot in a block means you've got your duplicates taken care of

[@dwanechristensen6422]

...If r3c3 is not a 1, then r7c7 must be a 1... only had to go through that about 13 times before it all made sense! Wow! Great puzzle, and thank you for the patient clear explanation!

[@LorisLaboratory]

Birthday wishes to your Dad! Great solve, thank you!

[@tuurepaananen7729]

11:27 for me, almost immidiately found the logic after realising that there are no power in the corner pencilmarks.

[@TheTruth-si5zq]

No you didn’t

[@_-_-Sipita-_-_]

26:45 for me. I don't really knew what the technique was but i found: 1 digit pencilmark loop elimination. 46 chains combo 78 RC pencilmark elimination 45 strange looking thingy 3 pencilmark double-effect What i do is to see what happens if i do a move that affects multiple numbers.

[@Not3xactly]

Brilliant solve. You always make it seem so easy

[@wydadiyoun]

probably the most genius classic sudoku I have ever seen! I regret giving up and watch the video instead!

[@JalebJay]

28:25 time, not sure if my pattern would be called bifurcation or not, but used simple logic putting 2 values in the same square to solve.

[@randomviewer896]

What an awesome techinque! I can't imagine how difficult this would be to spot if the rows and columns were shuffled...

[@EZal17]

I’ve only watched up until 18:00, but that moment gave me enough itch to go and write this: What you’re saying right now is that if there is no 1 in cell C3, with 178 in pencil (counting the top left as A1, top right as I1, bottom right as I9, etc), then there has to be 1 in cell G7, with 19 in pencil. But this can also be seen simply from penciling out 1’s in every square. Indeed, 1 in G7 implies 1 in D9, implying in turn 1 in E3, and vice versa, 1 in C3 also means 1 in D1 and in E7

[@EZal17]

In fact, there’s a more fun story going on with ones here, which you can’t see until you’ve added all the pencil marks. The chain turns into some sort of a cycle, so that you can go along the squares in the following order (or backwards, of course): 4-1-2-8-9-6-4-… (This cycled chain overall looks a bit like an eight tilted 45° to the left.) The thing about this cycle is that either a 1 in A4 or a 1 in I6 lead to an impossible situation just from following down this cycle, and erasing these two pencil marks breaks up the cycle into a straight chain: 4-1-2-8-9-6. Which only leaves us with two fewer pencil marks and no new solid knowledge 😆

[@volkerysilvia]

The bent triple (commonly called XY-wing or Y-wing) wasn't needed at all for the argumentation around the 1s - the Zed-shape in combo with knowing the possible spots for 1s in the opposing corners already told us that we had a 1 at exactly one of those two inner corners (and the opposing corner box had to have the 1 in its outer corner). The Zed is basically a distorted X-wing, and the external constraints on it allow using X-wing logic to remove the 1 from the 14-cell.

[@genoh9632]

At 20:08 I looked at r7c5 (1458), which can't be a 1, because if it is it creates three 78 pairs in column 5. => r3c5 is a 1

[@winstonheppolette8734]

Coming to this puzzle recently, I found that 20 mins into the video, I had reached the same point as Simon but, rather than take his route via the 'bent triplet ', I concentrated on the cell at the intersection of col 4 and row 7. This cell contains either a 1 or a 4. If 1 is entered there, it leaves the cell at the bend of Simon's triplet with zero possibilities. Therefore, it must be a 4! Sorry to make redundant Shye's rather elegant plot.

[@piotrmorag2597]

My trick for times when i feel like i have to guess is to pick the option that gives you the least :) It works like 80% of the time. I'm currently at min X and i look at the 2 and 5 pair at the top right. When i choose 5 on the right and 2 on the left we can see that we got: 2s in the bottom row, that gives us 3 next to the 5, 2 in the box on the other side, some info about other numbers and prob. more, but when we choose the 2 on the right and 5 on the left at the start, we can see that we get almost nothing done so that would be my guess. Edit: after skiping the video to find the result, it wasn't really surprising :)

[@jimzsblue]

By continuing on with identifying where the 1' could go, I was able to put the 4 in r7c4: If there is a 1 in r7c7, for instance, the remaining 6 missing 1 can all be placed. If there is NOT a one in r7c7, then it is in r1c1, and again the remaining 6 missing 1s can all be placed. In either case, a one is not in r7c4. So at 15:10, the 4 goes there, and pretty quickly the correct set of 1s locations are identified.

[@RichardBlignault]

I have tried this puzzle about 6 times now with no success yet. Today I gave it to my wife and she solved it in less than 20 min. I watched her and she only used 3 pencil numbers twice the rest were all 2 only. I was shocked. So she did it again in about the same time. No other logic used other than her own common sense.

[@livedandletdie]

That was a fun puzzle, as soon as I found out that the r3c3 couldn't be a 1, the entire puzzle basically solved itself.

[@danbylaw]

I didn't catch the "crankshaft," so the way I ended up working it out was to ask whether R3C3 and R7C7 could both be 1. Since that would leave nowhere for a 1 in box 8, I treated R3C3 and R7C7 as a conjugate pair for simple coloring. R7C4 saw both colors, which ruled out a 1 there, and the rest of the puzzle proceeded more straightforwardly.

[@gliderguld]

Hi Simon ... late entry, but very inspired by your explanations. I've observed, that you often push a number "one off" (one to large or to small), which leads me to suspect, that you are using an alphabetic keyboard. I think a numeric keypad and a little practice would benefit you.

[@angelavonhalle5144]

I find that 1)finding an easy number then 2) doing the rows,then the verticals, then the boxes is faster than this trying all over the place. You look for the doubles, of course, then eliminate the other numbers. In my opinion, / Use the dots marking too.

[@shawon265]

That 1 really was a tough one! (pun intended) I colored one branch of the possibilites of 1 with one color (green) and the alternative branch with another color (magenta). That meant all other cells got rid of the possibility of 1 (equivalent to pencil marking all 1's)

[@tarunkumaraggarwal4706]

20 minutes before giving up for the first digit. This Classic sudoku is tough!!