Given all the comments about scanning, I must say I found it amusing that it took over ten minutes for the 7 in box 3 to be found, and then in a very roundabout manner.
1:01:30 for me. I'm nowhere near an amateur solver, but this one gave me quite a bit of trouble. I don't understand how so many people can do these puzzles so quickly, but they have my respect. It was quite a good challenge for a change.
I personally like the human error bits because you can learn as much (and for things like sudoku, sometimes more) by seeing where logic errors snag people. I'm still not great (or at least, not speedy) at solving but learning the sort of error-traps (including just missing an easy scan) that I do make and seeing ways to work around them has helped me a lot in getting steadily better. It's one of the reason this channel is a daily watch for me.
12:05 the easier way to place the 2 in box 6 is to ask "where can the 8 go in r6". because of the 8 in box 7 and the 6-7 pair in box 5, the 8 in r6 can only go in box 6. so r4c7 cannot be an 8. it also can't be a 4, because the 4 in box 5 has to be in r4, thus making r4c7 a naked single 2.
I’m happy that you’re solving the puzzles as you go along - it’s really helpful to hear your thought strategy. I’m off to buy your classic sudoku app! Thanks for a great site.
Took me almost an hour! And to solve it, it was borderline necessary to use bifurcations! Basically at the limit of steps that you can still visualise in your mind and having to fill them in!
Great puzzle! I solved it in 12:18. There's a way around the first inference chain at 11:15 - r4c7 can't be a 4, because box 5 has a pair of 4s in r4c5 and r4c6. Additionally, the 8 in row 6 can only be in box 6, so there can't be any other 8 in box 6, which means r4c7 can't be 8. This effectively makes r4c7 a naked single 2. However, after that, my solution diverged from yours. I missed the Y-wing between boxes 3 and 6, but after finding some other digits, I still had to rely on one inference chain of my own, where an 8 in r4c6 would force an 8 into r7c6.
The interesting thing is you did not need any alternate inference chain, if you spot may other aspects. It is possible to identify cell R4C7 as a naked single. Therefore we need to lock the number 4 in box 5 within the cells R4C5/6 and to lock the number 8 in row 6 within box 6 in the cells R6C7/8. This will give us the 2 in box 7 and allows us to fill in the number 2 in box 3 (including the 7), box 1, box 3 and box 8 (in that order). This further will lock the number 2 in box 4/6 in a X-wing position in the 4 cells R5/7C1/2. After that indeed we can use they Y-wing as shown by Mark to eliminate the 1 in box 3 at min.18. This will resolve column 6 where the 4 goes into R8C6. And now we can use uniqueness. Since the 4 is locked in box 7 in cells R7C1/2 and therefore can not go in the upper part of the X-wing on 2 in the cells R5C1/2. This will resolve the 4 in box 6 and the rest is straight forward. But some might argue that you don’t want to use uniqueness as well. Well, I stay interested to see how that can be resolved...
Possibly less bifurcative logic at 18:35 : Note that in box 4, 8 is either in c1 or r5. If it's in r5, we have 9 in r5c4 immediately. If it's in c1, we have a 45 pair in box 1, which puts a 4 in r2c6 (by the 4s pencil-marked in box 2), giving us an 8 in r4c6 and therefore, again, a 9 in r5c4.
The way to solve this without bifurcation is bent triple on 148 in r3c7(18), r6c7(48), and r5c8(14), which will give you 1 in r3c7, and 8 in r6c7. NB 4 automatically remains in r5c8, then normal sudoku rules. Mind you a hidden single 2 has always been available in r4c7
I did it! And all logically, too!! It took a long time though... NY Times daily Sudoku and only watching solves here is my usual jam, but I always take on the Classic Sudokus.
I got almost to the end, noticed I had an error somewhere in my last few steps. Went to undo to go back to where I made the error. Accidentally hit the reset button instead and lost everything. I think that’s the second time I’ve done that with the new software.
Agreed. It’s too close and similar looking to the undo button. And because you are often tapping/clicking many times on undo to go back several steps, I end up mashing the reset button which instantly confirms the reset accidentally since I am intending to click this button more than once thinking it’s the undo button. Other than that the new software is great!
@@IBAndreas It is really bad UX design, the confirmation button should not be in the same space as the reset button - confirm it to the left or right, make you move the mouse to confirm. Double click is too easy! Am also liking the new software that fairly major issue aside
Took me ~170 minutes to finish the grid, I almost broke my back figuring out the crosswing on 4s and the way 3648 interact within the box 6 and with the r9c7. Glad I solved it.
I honestly don't know why Mr Goodliffe seems to always resolve to the "what if"-machine. I mean it's impressive at times, but doesn't result in a very interesting solve when your go to strategy is looking at sometimes near endless chains, before trying to find logical patterns. I enjoyed this puzzle, but found my own solve which I think was much smoother. Well, that's my two cents anyway.
It's not that he doesn't sometimes spot the interesting patterns. The Y-wing he spotted to get the 1 in box 3 was something I had missed, and was a better way forward than I had found at that point. But yes, I agree. His belief that if he does long chains in his head, then it doesn't count as bifurcation, is a distinction I don't agree with. His ability to remember what digits he has mentally put into the grid is impressive. But it's impressive because it's hard to do, yet he's relying on his audience to do the same along with him, if they wish to follow his solution. I would rather he put the digits into the grid, if he's following a chain, so at least we can more easily follow what he's seeing.
Great solve Mark! I'd like to see everyone with scanning advice to do live solves themselves! It's not so easy when others are scrutinizing your every move and thought. I set up a swordfish on 1's and was able to find my way through that way. I missed the y-wing in boxes 3 and 6 and in boxes 2 and 5, but looking back, they were certainly there. My time was 40:03, not a bad time for me on a very hard puzzle. Also, this is one of the few I solved before watching the video!
13:30 A 7 was placeable in box 3 from the start. Place 14 in R5C8, and the resulting bent triple eliminates 1s from the middle column of box 3. 20:00 I think that I solved it in a similar way. I was surprised to spot the bent triple on my own fairly early. Much later, I encountered the clash from placing an 8. 22:10 I didn't have an alternate-inference chain at the beginning. I fixed a 2 in box 6 as a single (naked? hidden?) instead.
SO this has taken me ages to solve but I finally did it. I wasn't a fan of the reasoning that the R4C5 cell couldn't be 8 via bifurcation. I also didn't really want to use the 'Unique Rectangle solution' on a set of 24s in the first 2 columns as that feels a bit impure as well... So i stared for ages and simplifying where I could, I kept finding Y wings that didn't help me until i found one that did: R4C1, R4C3 & R5C4 formed a 589 Y-Wing. This removed a 9 from R4C5 and showed that the 9 could ONLY be in R5C4. I'm glad I found that without the other methods, it wasn't as quick for sure but It was there to be found and I found it!
At 10:27, R5C9 = 2 means R4C7 = 8 (because of the pointing pair on 4s in box 5), so R5C4 is 8 and nowhere to put the 8 in box 3. That helps move things along.
If he had pencilmarked where 4s could go in box 5 (your pointing pair) and pencilmarked where 8s could go in row 6 (only in box 6), then r4c7 becomes a fairly straightforward naked single. It boils down to the same logic, but still simpler to see, in my view.
Starting at 9:00 there is a nice way to crack this puzzle open. The 4 can only go in 2 positions in C3, … R3 or R6. One of the positions, RC63 is restricted to 49. If we try putting the 4 in R3, it resolves the 4 in the top middle block and the 4 and 9 in the middle block. That 9 forces a 9 into R6 in the right middle block, so we now have both a 4 and 9 aiming at RC63 --.which means no number can be put in that square. Conclusion, the 4 in C3 must go in R6. That forcing test is enough to crack the puzzle open.
You weren’t the only one. It would have made life easier. Maybe Mark should re-record this with showing what happens if you do put the 7 in earlier as a comparison video.
I've been watching a few of these.. and it does seem characteristic of habitual 'deep' solvers that they don't come up for air, and clear the simpler deductions that a previous deep step has made possible.. (Which would often decrease the number of subsequently necessary deep dives)
I had a 2.4 pair in row eight which was mirrored by a 2,4 pair in row 5 that one of cells contained an eight also as a candidate! I fought against using uniqueness and found the solution in the end.
What I like about the new set-up is that I can turn off the time because I am so slow but I missed doing so this time and solved in 419:28 whatever that means. But even I go the 7 in box 3 before Mark. Love this CTC
Nice. I would have started that in the same way -- simple scan, mark pairs and triples, and mark cells with only two candidates. But no matter how hard I try not to, I always end up following that by marking up all candidates, which completely ruins the fun.
Despite some of the easy things missed, such as the 7 in Box 3 that would have given the 7 in Box 9. It's still was a difficult puzzle and there were a few techniques used that were very good examples. The use of the Y-Wing is a very powerful solving technique. I think Mark explained very well with the example given. With the variances in other puzzles, sometimes we don't get to see these techniques because they get resolved in other ways. Maybe an idea would be to create on a separate channel (for teaching) some of these techniques in detail using different examples without actually solving the whole puzzle. I'm sure many have heard the names of these different techniques but don't know them or how to spot them.
Over an hour including interruptions. I had the 4 and 6 in box 2 as pairs of candidates, and bifurcated them till I found the correct order, then my notation allowed the remainder of the puzzle to be filled in.
This took me forever. I managed to solve it bifurcating at the end, but was unsatisfied with that and backed up to where I was stuck. I eventually found an xy-chain on 4's (r6c3 and r5c8 pointing at r6c8) that eliminated the 4 from a 4, 5 set. I don't think I've ever used an xy-chain before so was quite proud of myself for finding it :)
I made it to the same length as when Mark put in the 9 in R5C4 (19:15), which IMO was bifurcating camouflaged as "chain". Does anyone have a more logical approach on how to continue from here?
The triple in column 3 gives you a 2 in C3R1, which also gives a 2 in C5R9, which sets up an X wing on 9s in column 3 and 5. That gave me some eliminations and progress.
I think he had chain logic on his mind because he kept focusing on it. For this one the 67 pair jumped out at me in row 6 c4&5. That left a 49 in r6 c3 because of the Snider (sp?) notation. He later erased those 5 pencil marks for some reason. I loved the chain logic solve though. I've never been able to do those.
8:45 a much easier way to solve this is to look at that square (R4C7) and note only 2 or 8 can go there (4 is locked in R4C5, R4C6 in box 5). This is fair until you look to find where 8 goes in row 6. With the 67 pair in box 5, and C3 being already taken by the 8, 8 has to go in R6C7, R6C8, leaving only a 2. Rest of the puzzle unties itself from there.
@@mjschryver R3C3 and R6C3 is a 4,5 pair, putting a 9 in R4C3. Method I used when solving was examining how putting a 4 in box 5 worked in chaining a whole bunch of numbers into place (as this can directly fix the 4 in box 2, and if you set it up so a 4 is in row 3, a lot more things start to fall into places). Basically, I had a chain of 5-10 iffs.
At 14:25, with the 14568 quint on row 3, there is still the possibility that the 1 for that row is the candidate in box 1 and therefore the candidate 1 for box3 3 can still be in row 1. Can I ask how you were able to dismiss row 1 column 8 as that's what stopped me from identifying your Y-wing later on - because that r1c8 was still a candidate for 1 ?
@13:30 it seems to me that it was improper to mark 4 as a possibility in r4c1 and r4c3 because the 4 had already been designated as being required in row 4 in box 5. I'm not sure if that would have helped.
At least he has the partial excuse that he hadn't bothered to corner pencilmark the 4s in box 5, so it wasn't as clear as it could have been. But at 15:43 he includes 5 as a candidate in r6c3 despite 5 already being corner pencilmarked elsewhere, both in the box and in the row. 😂
Um does it count as bifurcating if you guessed it right and completed the puzzle without having to backtrack? Still took me nearly an hour but I think that's the first time I've ever successfully finished a sudoku having guessed one of the digits on a 50/50 chance.
Yes, anytime you go 50/50 is already bifurcating, whether it turns out right or wrong. Still, I think it's more satisfying to bifurcate and finish a puzzle rather than be unable to finish it. :) I usually take a screenshot at my bifurcation point, then watch the video later to find the logical move for it.
Ah that's a great idea, although the path I took on this one was completely different from Mark's, heh. If I thought I was better at following the logical path I'd definitely do that though, even just as a reference to backtrack to if the bifurcation breaks down.
This is why I scan systematically. Numbers, rows, columns, boxes and repeat. That way I don't miss that much that is really obvious. - Still took me a very long time and a repeat just to understand what I missed in the first run.
I periodically check my work along the way by hitting the check button on the new web version. It doesn't seem to validate incorrect entries all the time. I expect it to flag entries that are incorrect even before the puzzle is fully solved. If that assumption is correct, then it's broken. Let me give you an example: I'm 80% done and I've been checking along the way. Happily solving down a complete column and I go to plug in the last number BUT... what "must" go there is already somewhere in the ROW... and it's a static number there from the puzzle source. Clearly there is an error elsewhere in the column but it is never flagged when I click "check". So am I misunderstanding the "check" or is it broken?
The check button only checks you don't currently have a repeated digit in a row, column or box. When they enter the puzzle into the software, they don't have the solution, so they cannot enter a solution for the software to check against. It's only a rudimentary checking function that checks if your entered digits violates any standard sudoku rules. Which is why it's sometimes possible to find an "alternative" solution that the check button says "looks good to me" on a variant sudoku, because it's not checking the variant rules.
I think all the comments about things you guy miss are just people getting a bit excited when they can see something that you didn't I know it gets me a bit excited even though many many of these puzzles are out of my reach otherwise haha Love the solves The logic at 11:30 or so was great the possible 2 in row 5 creating the 49 pair in box 4 but there is a cell r6c3 I think it was that can only be 4 or 9 very very hard to see and I couldn't see it I couldn't finish this one at all
I fared better on today's 'Simon' puzzle, rather hilariously. I did solve this one, but it took me 36:12 and I resorted to a late bifurcation in Box 6 to finish the job. (Before that, I *did* find a Y-wing, which was nice.) Tough classic! EDIT: After watching the video, I don't think what I did was much different from what Mark did; just in a slightly different starting location. I don't feel so bad about my solve now.
I found a clear path to the 2 in box 6 (Mark's chain was bordering on bifurcation; the 2 is a naked single once you get the 8 pencil marked in box 6 and the 4 pencil marked in box 5) and I found the Y wing, but I wasn't quite able to find the last chain that cracked the puzzle. I wonder if there was a simpler way than that last step, but I sure couldn't find it.
I hate to bifurcate but despite finding a hard-sought naked single in column 7 and a Y wing that didn't really even help me out very much, I'm _still_ stuck...
Well Mark's started to struggle with it long before I ran out of gas entirely, so I'm feeling somewhat justified in finding it so difficult. Although it's still only a 23 minute video... there must be a breakthrough that I just can't find...
Annoyingly but not surprisingly he's bifurcating instead of finding the logical path. A long deduction that's much simpler and more elegant if you consider the 8 in box 6 must be in row 6, which rules it out of r4c7, making it a naked single 2 - and proving that r5c9 can't be 2.
I just know it's going to be something ridiculously simple after all this. I've reached the glazed point where my brain isn't working and I'm probably just missing something that'll be a face-slapper when Mark spots it.
So I got stuck, watched the video, saw what he did with r5c4, and went to my favorite sudoku solver website (Sudoku Solver by Andrew Stuart). This site walks through dozens of different potential logic paths, including x-wings, y-wings, swordfish, etc. (even some complicated and obscure ones that I can't remember at the moment). Sadly, the site was down. So I went searching for other solvers that would give solution paths. None were anywhere near the quality of the Andrew Stuart one. But they all stated that they used brute force to solve. *SIGH*
took me the lenght of video, but it was 2nd try. The first was about 13 min but wrong, so i had to start again. PS. Please, more classic sudoku like this.
solved it in 10 minutes and at check i realized i had to 8s in the same column. couldnt find the mistake so had to redo it. This time the sudoku was much harder and took me 22 minutes.
Hey, Mark! Do you feel like avoiding "bifurcation" and uniqueness in these videos has slowed you down in competitions? Practice makes habit, after all.
Random suggestion, but I wish there was a simple "episode app" along with the miracle, chess, etc. that had all the puzzles, or possibly worked with the hyperlink in the description to play the puzzle with some record or added UI functionality. Specifically because I'm lazy and on mobile so the website is a little less comfortable to use than the apps ;). But it would be a fun addition. Love the series of course! We'll see 500k subs soon I bet
46 minutes even. Why is the sudoku in the thumbnail different than the one in the video? I started doing that one in my head before hitting play and was surprised that it was all wrong.
They wouldn't use uniqueness in their solves. Their aim is to logically find a solution and prove it's the unique solution. The puzzle creator couldn't assume it had a unique solution. If they were able to prove it had a unique solution, then it should be possible for me too. No shortcuts. 🙂
Robert Longo's Karate series is really quite impressive. There's Black Belt, Beyond Black Belt, and 'WAY BEYOND' Black Belt at the extreme end of that series. I can't recommend those highly enough, they are quite the workout.
![Classic Sudoku: Bonus Puzzle 4: [This is STUNNING]](http://i.ytimg.com/vi/O0uASyvv554/mqdefault.jpg)
Given all the comments about scanning, I must say I found it amusing that it took over ten minutes for the 7 in box 3 to be found, and then in a very roundabout manner.
Hahaha 😂🤣
That was the first number I put in the grid. :-D
R4C7 is a naked single due to 135679 in the rows and column, 4 being confined to R4 in box 5 and 8 being confined to C7-8 in R6
Mark must be kicking himself for missing that, as well as the 7 in box 3. Missing these two cells really affected his solve.
@@ThatGuy-dj3qr Yeah, he really wasn't bringing his a-game to this solve.
Came to the solution the same way. I was typing this in a responce and then i saw you already had posted the naked single in r4c7.
"There's the bad scanning I was planning" gave me a chuckle.
Sevens
Missed 2 row 1, col 3 for a long time!
1:01:30 for me. I'm nowhere near an amateur solver, but this one gave me quite a bit of trouble. I don't understand how so many people can do these puzzles so quickly, but they have my respect. It was quite a good challenge for a change.
I just wanted to thank both of you. Lockdown has been terrible and, although I’m terrible at puzzles, your videos have helped me pass the time.
I personally like the human error bits because you can learn as much (and for things like sudoku, sometimes more) by seeing where logic errors snag people. I'm still not great (or at least, not speedy) at solving but learning the sort of error-traps (including just missing an easy scan) that I do make and seeing ways to work around them has helped me a lot in getting steadily better. It's one of the reason this channel is a daily watch for me.
12:05 the easier way to place the 2 in box 6 is to ask "where can the 8 go in r6". because of the 8 in box 7 and the 6-7 pair in box 5, the 8 in r6 can only go in box 6. so r4c7 cannot be an 8. it also can't be a 4, because the 4 in box 5 has to be in r4, thus making r4c7 a naked single 2.
Yes! Mark always resorts to bifurcating too quickly.
I’m happy that you’re solving the puzzles as you go along - it’s really helpful to hear your thought strategy. I’m off to buy your classic sudoku app! Thanks for a great site.
Took me almost an hour! And to solve it, it was borderline necessary to use bifurcations! Basically at the limit of steps that you can still visualise in your mind and having to fill them in!
Great puzzle! I solved it in 12:18.
There's a way around the first inference chain at 11:15 - r4c7 can't be a 4, because box 5 has a pair of 4s in r4c5 and r4c6. Additionally, the 8 in row 6 can only be in box 6, so there can't be any other 8 in box 6, which means r4c7 can't be 8. This effectively makes r4c7 a naked single 2.
However, after that, my solution diverged from yours. I missed the Y-wing between boxes 3 and 6, but after finding some other digits, I still had to rely on one inference chain of my own, where an 8 in r4c6 would force an 8 into r7c6.
The interesting thing is you did not need any alternate inference chain, if you spot may other aspects. It is possible to identify cell R4C7 as a naked single. Therefore we need to lock the number 4 in box 5 within the cells R4C5/6 and to lock the number 8 in row 6 within box 6 in the cells R6C7/8. This will give us the 2 in box 7 and allows us to fill in the number 2 in box 3 (including the 7), box 1, box 3 and box 8 (in that order). This further will lock the number 2 in box 4/6 in a X-wing position in the 4 cells R5/7C1/2. After that indeed we can use they Y-wing as shown by Mark to eliminate the 1 in box 3 at min.18. This will resolve column 6 where the 4 goes into R8C6. And now we can use uniqueness. Since the 4 is locked in box 7 in cells R7C1/2 and therefore can not go in the upper part of the X-wing on 2 in the cells R5C1/2. This will resolve the 4 in box 6 and the rest is straight forward. But some might argue that you don’t want to use uniqueness as well. Well, I stay interested to see how that can be resolved...
Possibly less bifurcative logic at 18:35 : Note that in box 4, 8 is either in c1 or r5. If it's in r5, we have 9 in r5c4 immediately. If it's in c1, we have a 45 pair in box 1, which puts a 4 in r2c6 (by the 4s pencil-marked in box 2), giving us an 8 in r4c6 and therefore, again, a 9 in r5c4.
I did in 20 minutes and 48 seconds! can say I'm quite proud of this time
The way to solve this without bifurcation is bent triple on 148 in r3c7(18), r6c7(48), and r5c8(14), which will give you 1 in r3c7, and 8 in r6c7. NB 4 automatically remains in r5c8, then normal sudoku rules. Mind you a hidden single 2 has always been available in r4c7
I did it! And all logically, too!! It took a long time though... NY Times daily Sudoku and only watching solves here is my usual jam, but I always take on the Classic Sudokus.
I got almost to the end, noticed I had an error somewhere in my last few steps. Went to undo to go back to where I made the error. Accidentally hit the reset button instead and lost everything. I think that’s the second time I’ve done that with the new software.
Agreed. It’s too close and similar looking to the undo button. And because you are often tapping/clicking many times on undo to go back several steps, I end up mashing the reset button which instantly confirms the reset accidentally since I am intending to click this button more than once thinking it’s the undo button.
Other than that the new software is great!
@@IBAndreas It is really bad UX design, the confirmation button should not be in the same space as the reset button - confirm it to the left or right, make you move the mouse to confirm. Double click is too easy! Am also liking the new software that fairly major issue aside
Or, use the key binding for undo (Cmd-Z on Mac, presumably Ctrl-Z on Windows or Linux)
Took me ~170 minutes to finish the grid, I almost broke my back figuring out the crosswing on 4s and the way 3648 interact within the box 6 and with the r9c7.
Glad I solved it.
I honestly don't know why Mr Goodliffe seems to always resolve to the "what if"-machine. I mean it's impressive at times, but doesn't result in a very interesting solve when your go to strategy is looking at sometimes near endless chains, before trying to find logical patterns. I enjoyed this puzzle, but found my own solve which I think was much smoother. Well, that's my two cents anyway.
It's not that he doesn't sometimes spot the interesting patterns. The Y-wing he spotted to get the 1 in box 3 was something I had missed, and was a better way forward than I had found at that point.
But yes, I agree. His belief that if he does long chains in his head, then it doesn't count as bifurcation, is a distinction I don't agree with. His ability to remember what digits he has mentally put into the grid is impressive. But it's impressive because it's hard to do, yet he's relying on his audience to do the same along with him, if they wish to follow his solution. I would rather he put the digits into the grid, if he's following a chain, so at least we can more easily follow what he's seeing.
Great solve Mark! I'd like to see everyone with scanning advice to do live solves themselves! It's not so easy when others are scrutinizing your every move and thought. I set up a swordfish on 1's and was able to find my way through that way. I missed the y-wing in boxes 3 and 6 and in boxes 2 and 5, but looking back, they were certainly there. My time was 40:03, not a bad time for me on a very hard puzzle. Also, this is one of the few I solved before watching the video!
13:30 A 7 was placeable in box 3 from the start. Place 14 in R5C8, and the resulting bent triple eliminates 1s from the middle column of box 3.
20:00 I think that I solved it in a similar way. I was surprised to spot the bent triple on my own fairly early. Much later, I encountered the clash from placing an 8.
22:10 I didn't have an alternate-inference chain at the beginning. I fixed a 2 in box 6 as a single (naked? hidden?) instead.
Oh dear, Mark is clearly suffering from sudoku fatigue...
SO this has taken me ages to solve but I finally did it.
I wasn't a fan of the reasoning that the R4C5 cell couldn't be 8 via bifurcation. I also didn't really want to use the 'Unique Rectangle solution' on a set of 24s in the first 2 columns as that feels a bit impure as well... So i stared for ages and simplifying where I could, I kept finding Y wings that didn't help me until i found one that did:
R4C1, R4C3 & R5C4 formed a 589 Y-Wing. This removed a 9 from R4C5 and showed that the 9 could ONLY be in R5C4. I'm glad I found that without the other methods, it wasn't as quick for sure but It was there to be found and I found it!
At 10:27, R5C9 = 2 means R4C7 = 8 (because of the pointing pair on 4s in box 5), so R5C4 is 8 and nowhere to put the 8 in box 3. That helps move things along.
And of course he spots it (by a slightly different route) within the minute, nice!
If he had pencilmarked where 4s could go in box 5 (your pointing pair) and pencilmarked where 8s could go in row 6 (only in box 6), then r4c7 becomes a fairly straightforward naked single. It boils down to the same logic, but still simpler to see, in my view.
oh my goodness, I was screaming at the start of the video for the 7s, they were so obvious but regardless amazing puzzle!!!
Thanks for the approachable sudokus!!
Thumbnail: ‘Approachable Sudoku’, title of the video: ‘Tough Sudoku’! 🤯
Starting at 9:00 there is a nice way to crack this puzzle open.
The 4 can only go in 2 positions in C3, … R3 or R6.
One of the positions, RC63 is restricted to 49.
If we try putting the 4 in R3, it resolves the 4 in the top middle block and the 4 and 9 in the middle block.
That 9 forces a 9 into R6 in the right middle block, so we now have both a 4 and 9 aiming at RC63 --.which means no number can be put in that square.
Conclusion, the 4 in C3 must go in R6.
That forcing test is enough to crack the puzzle open.
Screaming "PLACE THE 7 IN BOX 3" for 10 minutes my god 😂
You weren’t the only one. It would have made life easier. Maybe Mark should re-record this with showing what happens if you do put the 7 in earlier as a comparison video.
I've been watching a few of these.. and it does seem characteristic of habitual 'deep' solvers that they don't come up for air, and clear the simpler deductions that a previous deep step has made possible.. (Which would often decrease the number of subsequently necessary deep dives)
It took me over an hour before I gave up to check your reasoning for why R5C4 couldn't be an 8. I knew I was missing a chain, but couldn't spot it.
I had a 2.4 pair in row eight which was mirrored by a 2,4 pair in row 5 that one of cells contained an eight also as a candidate! I fought against using uniqueness and found the solution in the end.
Plugged this into the app I usually play on (regular sudoku only) and got it in 8:57
19:59... Nice to have an occasional classic puzzle.
I'm still a beginner but I was fortunate to spot the 2 in box 6 pretty early on and it was a walk to the end from there. Sub 20!
How did you disambiguate the 8/9 in box 5?
Human error a-go-go. That's why I like you guys.
I come for the sudoku, I stay for the human error a-go-go.
12:38 for me! I’m pumped! And NO bifurcation.
What I like about the new set-up is that I can turn off the time because I am so slow but I missed doing so this time and solved in 419:28 whatever that means. But even I go the 7 in box 3 before Mark. Love this CTC
Nice. I would have started that in the same way -- simple scan, mark pairs and triples, and mark cells with only two candidates. But no matter how hard I try not to, I always end up following that by marking up all candidates, which completely ruins the fun.
That sounds eerily familiar. I know I shouldn't, but I have to feel like I'm making progress. 😂
Following this channel for the human live solves. Much more interesting! 😊
Despite some of the easy things missed, such as the 7 in Box 3 that would have given the 7 in Box 9. It's still was a difficult puzzle and there were a few techniques used that were very good examples. The use of the Y-Wing is a very powerful solving technique. I think Mark explained very well with the example given. With the variances in other puzzles, sometimes we don't get to see these techniques because they get resolved in other ways. Maybe an idea would be to create on a separate channel (for teaching) some of these techniques in detail using different examples without actually solving the whole puzzle. I'm sure many have heard the names of these different techniques but don't know them or how to spot them.
Over an hour including interruptions. I had the 4 and 6 in box 2 as pairs of candidates, and bifurcated them till I found the correct order, then my notation allowed the remainder of the puzzle to be filled in.
10:06 for me. As you can imagine I'm mighty pleased with that time!
This took me forever. I managed to solve it bifurcating at the end, but was unsatisfied with that and backed up to where I was stuck. I eventually found an xy-chain on 4's (r6c3 and r5c8 pointing at r6c8) that eliminated the 4 from a 4, 5 set. I don't think I've ever used an xy-chain before so was quite proud of myself for finding it :)
I made it to the same length as when Mark put in the 9 in R5C4 (19:15), which IMO was bifurcating camouflaged as "chain". Does anyone have a more logical approach on how to continue from here?
The triple in column 3 gives you a 2 in C3R1, which also gives a 2 in C5R9, which sets up an X wing on 9s in column 3 and 5. That gave me some eliminations and progress.
And that 2 in row 1 also sets up a weird 19 pair in row 1
>No arcane, complicated rule
Brain : *visible confusion*
waiting for him to find that duck of a 7 in box 3 is excruciating.
54:27, the first time I use the ring to solve it, then when I got stuck I traced to see any pencil marks that could be errased.
Please more killers/greather than
I think he had chain logic on his mind because he kept focusing on it. For this one the 67 pair jumped out at me in row 6 c4&5. That left a 49 in r6 c3 because of the Snider (sp?) notation. He later erased those 5 pencil marks for some reason. I loved the chain logic solve though. I've never been able to do those.
8:45 a much easier way to solve this is to look at that square (R4C7) and note only 2 or 8 can go there (4 is locked in R4C5, R4C6 in box 5). This is fair until you look to find where 8 goes in row 6. With the 67 pair in box 5, and C3 being already taken by the 8, 8 has to go in R6C7, R6C8, leaving only a 2. Rest of the puzzle unties itself from there.
How did you disambiguate the 8/9 in box 5?
@@mjschryver R3C3 and R6C3 is a 4,5 pair, putting a 9 in R4C3.
Method I used when solving was examining how putting a 4 in box 5 worked in chaining a whole bunch of numbers into place (as this can directly fix the 4 in box 2, and if you set it up so a 4 is in row 3, a lot more things start to fall into places). Basically, I had a chain of 5-10 iffs.
Didn’t have to bifurcate to get the 9 in box 5. I instead came across deadly 24 pair in columns 1&2 in box 4&7 which moved the 4 in box 4 to r6c3
Thumbnail: Approachable Sudoku
Title: Tough Sudoku
Given the video time of almost 30 mins for Mark doing a classic... I know which one is lying!
At 14:25, with the 14568 quint on row 3, there is still the possibility that the 1 for that row is the candidate in box 1 and therefore the candidate 1 for box3 3 can still be in row 1. Can I ask how you were able to dismiss row 1 column 8 as that's what stopped me from identifying your Y-wing later on - because that r1c8 was still a candidate for 1 ?
@13:30 it seems to me that it was improper to mark 4 as a possibility in r4c1 and r4c3 because the 4 had already been designated as being required in row 4 in box 5. I'm not sure if that would have helped.
At least he has the partial excuse that he hadn't bothered to corner pencilmark the 4s in box 5, so it wasn't as clear as it could have been. But at 15:43 he includes 5 as a candidate in r6c3 despite 5 already being corner pencilmarked elsewhere, both in the box and in the row. 😂
Who else spent the first 14 minutes screaming at him to put a 7 in rank 2 file 8?
Nice puzzle n enjoyed a lot. Thanks
Um does it count as bifurcating if you guessed it right and completed the puzzle without having to backtrack? Still took me nearly an hour but I think that's the first time I've ever successfully finished a sudoku having guessed one of the digits on a 50/50 chance.
Yes, anytime you go 50/50 is already bifurcating, whether it turns out right or wrong. Still, I think it's more satisfying to bifurcate and finish a puzzle rather than be unable to finish it. :)
I usually take a screenshot at my bifurcation point, then watch the video later to find the logical move for it.
Ah that's a great idea, although the path I took on this one was completely different from Mark's, heh. If I thought I was better at following the logical path I'd definitely do that though, even just as a reference to backtrack to if the bifurcation breaks down.
It's good to see a classic
This is why I scan systematically. Numbers, rows, columns, boxes and repeat. That way I don't miss that much that is really obvious. - Still took me a very long time and a repeat just to understand what I missed in the first run.
Yes I do the same. Lcr and tmb then cross meet then I look for rows colums or blocks with 5 or more digits in it
I got stuck and had to watch until he hit the Y-wing, which was enough of a boost to finish it for me. Tough one indeed!
Tough? Must be lucky. New to this. Took 19 minutes. How did you do?
I periodically check my work along the way by hitting the check button on the new web version. It doesn't seem to validate incorrect entries all the time. I expect it to flag entries that are incorrect even before the puzzle is fully solved. If that assumption is correct, then it's broken. Let me give you an example: I'm 80% done and I've been checking along the way. Happily solving down a complete column and I go to plug in the last number BUT... what "must" go there is already somewhere in the ROW... and it's a static number there from the puzzle source. Clearly there is an error elsewhere in the column but it is never flagged when I click "check". So am I misunderstanding the "check" or is it broken?
The check button only checks you don't currently have a repeated digit in a row, column or box.
When they enter the puzzle into the software, they don't have the solution, so they cannot enter a solution for the software to check against.
It's only a rudimentary checking function that checks if your entered digits violates any standard sudoku rules. Which is why it's sometimes possible to find an "alternative" solution that the check button says "looks good to me" on a variant sudoku, because it's not checking the variant rules.
I also noticed the seven in box 3 on my initial pass, which immediately gave me row 7 column seven as a seven as well
I think all the comments about things you guy miss are just people getting a bit excited when they can see something that you didn't I know it gets me a bit excited even though many many of these puzzles are out of my reach otherwise haha
Love the solves
The logic at 11:30 or so was great the possible 2 in row 5 creating the 49 pair in box 4 but there is a cell r6c3 I think it was that can only be 4 or 9 very very hard to see and I couldn't see it I couldn't finish this one at all
Took me 73 minutes and ended up having to bifurcate towards the end but seeing this solve, I'm proud that I even made it at all!
I fared better on today's 'Simon' puzzle, rather hilariously. I did solve this one, but it took me 36:12 and I resorted to a late bifurcation in Box 6 to finish the job. (Before that, I *did* find a Y-wing, which was nice.)
Tough classic!
EDIT: After watching the video, I don't think what I did was much different from what Mark did; just in a slightly different starting location. I don't feel so bad about my solve now.
I found a clear path to the 2 in box 6 (Mark's chain was bordering on bifurcation; the 2 is a naked single once you get the 8 pencil marked in box 6 and the 4 pencil marked in box 5) and I found the Y wing, but I wasn't quite able to find the last chain that cracked the puzzle. I wonder if there was a simpler way than that last step, but I sure couldn't find it.
Classics are the best. :) Nice puzzle.
yeah bring back some more classic puzzles -Does solving variant sudokus have adverse effect on solving classics?
It took me 13 mins to get to the position 18 mins for Mark. Great. However, I had to watch the video to see the Y-wing. Bobbins!
That was tough. I have a horrible time finding bent triples and y wings. Took me an hour and a half, but I didn't quite bifurcate.
R6C7 when reduced to either 4 or 8, it cannot be a 4 because then R5C8 would be 1, R3C7 would be 1 and there would be no place in C7 for an 8.
...normally classic puzzles don't give me too much trouble, but this one's a toughie...
I hate to bifurcate but despite finding a hard-sought naked single in column 7 and a Y wing that didn't really even help me out very much, I'm _still_ stuck...
Well Mark's started to struggle with it long before I ran out of gas entirely, so I'm feeling somewhat justified in finding it so difficult. Although it's still only a 23 minute video... there must be a breakthrough that I just can't find...
Annoyingly but not surprisingly he's bifurcating instead of finding the logical path. A long deduction that's much simpler and more elegant if you consider the 8 in box 6 must be in row 6, which rules it out of r4c7, making it a naked single 2 - and proving that r5c9 can't be 2.
I just know it's going to be something ridiculously simple after all this. I've reached the glazed point where my brain isn't working and I'm probably just missing something that'll be a face-slapper when Mark spots it.
Yep, 18:58 is what I was missing.
Thank you for not apologizing.
So I got stuck, watched the video, saw what he did with r5c4, and went to my favorite sudoku solver website (Sudoku Solver by Andrew Stuart). This site walks through dozens of different potential logic paths, including x-wings, y-wings, swordfish, etc. (even some complicated and obscure ones that I can't remember at the moment). Sadly, the site was down. So I went searching for other solvers that would give solution paths. None were anywhere near the quality of the Andrew Stuart one. But they all stated that they used brute force to solve. *SIGH*
Mr. Magoo -
You done it again!
took me the lenght of video, but it was 2nd try. The first was about 13 min but wrong, so i had to start again.
PS. Please, more classic sudoku like this.
23 minutes, the longest in Mark's history
Did he say "human error a gogo"?! LOL!
solved it in 10 minutes and at check i realized i had to 8s in the same column. couldnt find the mistake so had to redo it. This time the sudoku was much harder and took me 22 minutes.
Thank you.
Man, I don't think I've had to look for a Y-wing on this channel in almost a year.
63 minutes of horribly hard work for me.
at least you finished, I couldn't
Took me a couple hours. Ended up using phistomefel to finally finish.
I had a different convoluted bit of logic on the 8s at a later point in the puzzle.
Solved it with one hint from the video, because I tried it without chaining first.
17 minutes 45 seconds here. Pretty pleased with the time, even though it won't be winning any prizes. Decent puzzle, but not particularly memorable.
Hey, Mark! Do you feel like avoiding "bifurcation" and uniqueness in these videos has slowed you down in competitions? Practice makes habit, after all.
I think Mark did a bit of bifurcation
Random suggestion, but I wish there was a simple "episode app" along with the miracle, chess, etc. that had all the puzzles, or possibly worked with the hyperlink in the description to play the puzzle with some record or added UI functionality.
Specifically because I'm lazy and on mobile so the website is a little less comfortable to use than the apps ;). But it would be a fun addition.
Love the series of course! We'll see 500k subs soon I bet
Ludi sudokon en tornado... kiel laŭta!
48 mins and chuffed!😄
Strewth! That was tough.
Oh! So many given numbers! But tough...
46 minutes even.
Why is the sudoku in the thumbnail different than the one in the video? I started doing that one in my head before hitting play and was surprised that it was all wrong.
18.36, probably my fastest ever. Thought Mark over thought it too much and missed the obvious a few times.
Have you ever considered playing renju?
“There’s the bad scanning I was planning” lol.
I found Y wing and then uniqueness to solve. Mark not in the right mood it seems.
They wouldn't use uniqueness in their solves. Their aim is to logically find a solution and prove it's the unique solution.
The puzzle creator couldn't assume it had a unique solution. If they were able to prove it had a unique solution, then it should be possible for me too. No shortcuts. 🙂
Took a while but no bifurcation used there ! Had to note many things and use few different techniques
7 in box 3 was evident from the start and would made it easier, as it would give a 7 right in box 9 as well.
Nice.
Robert Longo's Karate series is really quite impressive. There's Black Belt, Beyond Black Belt, and 'WAY BEYOND' Black Belt at the extreme end of that series. I can't recommend those highly enough, they are quite the workout.