It's a shame there's this perception that people get "cross" with you for not fully coloring. In my opinion, it would keep things cleaner and easier to scan if you just removing the coloring once it's not needed. So in this instance, if you're coloring to represent where a specific digit goes, once you determine the exact value of the digit, the coloring is redundant and can be removed (ceremoniously, if you'd like). I do this on my stream and people don't mind at all.
+1 to this. I personally don't like to keep coloring digits that are already resolved. Keep the information "clean" so that it's easier to spot easy sudoku deduction
After smashing Simon's time on yesterday's puzzle this one took me nearly an hour and a half, double his time. A good day for more humility! I missed so many things on the way and cocked up twice. Getting the quad and quin on the diagonal felt so easy and then it all went wrong. Excellent puzzle with a grid that looks almost sweet and innocent. Thank you as always to setter and testers and solver.
This was lovely, I'm especially impressed how the diagonal communicated some info to opposite ends of the grid, so at times you had to apply knight's move logic 3-4 boxes away from where it really originated, I'm not sure I've ever seen that kind of misdirection. Simon, at 46:10 you pencil marked the 6s in box 9 then immediately forgot all about them and pencil marked 68 as candidates in r7c7 when it could only be an 8, and at 47:26 the 69 pair could have given you the 1 in row 2, but by that point you were basically just tidying up anyway. Really enjoyed watching this one, thank you sir!
I always thought the main diagonal constraint was kind of meh, but the last few puzzles you've showcased really show their potential. The way numbers can see each other from across the puzzle makes them so powerful and interesting, especially when combined with disjointed sets like in Lizzy's puzzle or the knights move in this puzzle. It's really something special.
I'm sure the title came from the software used to make it, but it also made me think of what I say most of the time I look at these sudoku, "F-ing puzzles!"
Simon figured it out, but for anyone still curious... To select all the cells of the same color you need to be in color mode and double click on a cell that is the color you want. If you want to select all the cells of the same number you need to be in one of the number modes (digit, corner, center), then double click a cell with the number you want.
If you are in colour-mode and you click on a red cell then it selects all cells that are red _or include red_ ... and if you select a cell that is multi-coloured red, yellow and grey then it selects _all_ cells that include _any_ of red, yellow or grey. (Or at least, I think that is what happens, but I'm happy to be proved wrong!). This can make it dangerous to highlight cells by colour without carefully checking that you've only got the cells you wanted!
I wonder if uncoloring squares once you have found all 9 instance of a number in the grid would aid with scanning. That way you only need to scan over the colored squares. This might backfire spectacularly if you forget to color some squares then proceed to overlook them for the rest of the puzzle.
Simon, I know you felt you had to fade out the intro music after a certain time, but it was absolutely beautiful. I'd love to hear the whole thing. Maybe you could just put it on the end of tomorrow's video?
Every time Simon says "I've run out of legible colors" when he is leaving in multiple colors that have already been identified with values just for nostalgia's sake...
4 Colors can tile any "map" or surface without touching each other... Technically youll only need 4 colors... Ill go with my nephews saying More colors better! I wonder if they can use a hatching or a shade to show different patters... we use alot of hatches and patterns in Autocad.
@@shawnward6607 4-color theorem isn't really applicable in this setting, where the colors aren't being used to delineating adjacent zones on a map but rather to representing specific unknown values with each color...
"I'm running out of colours here" ... time to ♻ recycle the reds, you've got all 9 4️⃣s in the grid now so you don't need 🟥 to mark those matching cells any more.
I started the same way as Simon, but got stuck where he was at the 17 minute mark. Never really thought about how the knight's move interacts with the diagonal, which came in useful down the line. His logic allowed me to finish the puzzle in over an hour (had the timer paused for a bit). A missing pencil mark in box 7 had me go down a wrong path for a bit, too.
This one pulverized me. After getting the 6789 onto the diagonal I decided to be ridiculous and color everything--the 24s, 135s, and the 6789 group. Then solved the rest in multi-color space with only occasional reference to numbers (in the killer cages). I didn't know if it would be possible, but it worked, even if it did look like psychedelic modern art.
I think there's a faster route to getting the first digit in place. 30 in seven cells is either 1234569 or 1234578. In the top right box, that means the uncaged cells must add up to 15, and the minimum value of the uncaged cells in the 30-cage boxes is 39, which gives you enough to place the 9 in the center box.
Sorry, I don't quite follow. Can you please explain why the uncaged cells in boxes 2, 3 and 5 have to sum to a minimum value of 39? As far as I can see, you have four given cage totals in those boxes summing to 3*30+10=100. Two pairs of uncaged cells in boxes 2 and 3 sum to 15 each, so 2*15=30. The three boxes must sum to 3*45=135. The difference between the cages + known 15 pairs and the three boxes are the cells r5c5 (in the boxes) and r4c9 (in the cages). So we end up with r5c5 - r4c9 = boxes - cages - known 15 pairs = 135 - 100 - 30 = 5. So the centre cell (r5c5) has to be 5 more than r4c9. But I don't see why they have to be 9 and 4?
@@RichSmith77 yeah, i think knights move is funny. Ive started skipping vids with rulesets (or skipping the dead parts until i see another digit once he starts looking) if he does the disjointed subset puzzles. Those make me wanna pull my hair out because I become hyper vigilant about it.
Took me days this one. I found it crazy hard. Enjoyed it tho :) r7c4 seems to the lynchpin to the entire puzzle... Simon identified it much, much sooner than I did, I spent ages looking down huge chains of logic mostly involving all of row 8, and although I did eventually fill in the 6 cage without discovering that cell, it was still important in finishing my solve. It seems like such an unremarkable cell and the fact that the puzzle hinges around it is kinda astonishing to me :)
I really like that the f cages are really only important for the break-in with the red squares being the same and the new 30 cages, to then never to be touched again in the rest of the puzzle xD
I can't believe I missed that weak number... I had the inital set up correct, but then I just didn't see that "weak number"... after that solved it in 120 mins... (yeah i looked at it for 30 mins before getting a Simon hint) very clever puzzle. I hope the guys release an app with diagonal rules set... that seems would be pretty good
ever thought of changing the colour pallet on the software? Make the purple really purple and add pink. or swap the yellow for brown. the yellow is useless.
Damn your solution is much cleaner then mine. Same break in, but I went full on multicolor and whittled down the puzzle piece by piece... And colored the entire grid before entering barely any digits. Thanks for the elegant solve.
21:25 The explanation was not given but was implied, but why can't the black square in question be the other blue, or green or purple? I think it's because it sees all three of those via knight's move or sudoku from box 5, right?
I think the way Simon viewed it was as follows. The cell in question (r7c4) was ruled out of c4 in box 5 by Sudoku and two other cells by knight's move. It also couldn't be red because we know red is elsewhere in box 8. That left three possible places where the digit in r7c4 could appear in box 5: r4c5, r6c5 and r4c6 (blue). Simon then showed why it couldn't be the blue in r4c6, since there would be no place it could then go in box 2. So that left two possible positions in box 5: r4c5 and r6c5. Now it's true those two positions do see pink, green and the other blue, but it's also true that those two remaining cells in box 5 have already been pencilmarked as 12345, so are known to be low digits. I believe it's this pencilmarking of the available cells in box 5 that tells Simon r7c4 must also be a low digit. It boils down to the same thing, since those cells were pencilmarked 12345 because they see all the high digits, either in the column or the box.
first.. look where green and purple can go in the center box.. either in column 4 which sees that box or r6c6 which sees that box by knights move.. it can't be the center blue r5c5 by knights move.. it can't be the blue in r4c6 since the number in that box is forced into column 4 in box 2 whice sees the box in question. That took me a while to figue out what simon meant.
Well at 17:37 , when you look at the 6 cage on the bottom, you can tell that 5 isn’t on the right cell because it’s a red one ( which has only the 2,3,4 possibilities) this indicates that cell isn’t 1 too. This leaves only the 2/4 possibility in 6 cage. Also the twin cell ( the one on the most top right cell of entire sudoku, that solves a lot actually). Good job
Congratulations on the new PC. I have to say I'm pretty shocked at the price, but can't argue with the 12 core CPU and latest GPU. I hope it's reliable!
I laughed hard on the 3.000 Pound Sudoku Machine. ...but i think for Video-editing it makes sense. I don't understand why he still has a HDD (why save on that, especially when you buy in this price range), but thats none of my business. Great PC nevertheless.
Holy moly, it took me a full month of waking up early to spam refresh on AMD's website just to get a Ryzen 5900X. Just goes to show: it pays to be Sudoku famous! Nice rig Simon!
15:26 "Don't see how we prove that though." Well for one thing, the 12345 in R8C2, R7C3 and R6C4 all see that square by sudoku and knights move. So at least three digits are going to be eliminated from R8C5. The question is, at what point would R8C5 no longer be able to be from the digits 234 at which point it's no longer possible for it to be red. By the way, R9C4 also sees those same cells along the diagonal as well as R9C1! So whichever way the puzzle goes, those four cells cannot all be 1245! One of them has to be a 3! Ergo, there's no 3 in R1C9!
The thumbnail LOL! Note to self: don't open CTC pop-ups while you are on a conference call and not on mute. 🤣 Explaining to a co-worker why you are laughing about a sudoku is not easy.
40:30 R2C5 is identified as a 3 or 5, and Simon says he doesn't know if it is orange or purple, but it must be purple. Orange is locked in the same column in box 5
I do love watching him struggle with things while ignoring the blindingly obvious for several minutes...
3 года назад
Congratulations on the new computer! What a huge leap, if I think back 2 years, where your mouse kept skipping, the software was creepy, etc. I think this machine should give video editing a decent boost too ...
Thanks to Simon’s training from many previous killer sudoku videos, I knew the way to start this was to do the maths on the 2/3/5 cages. Combined with the knowledge that the uncaged digits added up to 15, that meant that r4c9 (a.k.a. the “red” squares) must be a 4 and r5c5 must be a 9). Imagine my surprise when I watched the video, only to find that Simon never even considered adding up the values in those cages and in fact dismissed the 30 cages fairly early as having no further value. It was interesting to see a totally different (albeit more roundabout) method to get those digits, but also somewhat disconcerting to see Simon completely miss out on that logic.
Not sure I see why that forces r4c9 to be 4 and r5c5 to be 9? Doesn't it just fix that r5c5 has to be 5 more than r4c9? (Cage totals= 3 * 30 + 10 = 100 Known uncaged totals = 2 * 15 = 30 Boxes 2,3,5 = 3 * 45 = 135 Boxes = cages + known uncaged + r5c5 - r4c9 So 135 = 100 + 30 + r5c5 - r4c9 Therefore r5c5= r4c9 + 5 I seem to be missing a step.)
@@RichSmith77 I believe you are correct. I appear to have gotten lucky at that stage in the logic when I incorrectly concluded that r5c5 must be 9 and r4c9 a 4. At that point in the logic, making r5c5 an 8 or 7 with r4c9 being 3 or 2 would have been valid. I’m still surprised that Simon didn’t attempt the maths on those cages (given that this is usually his first instinct with this type of puzzle), but as it turns out, the only benefit would have been a bit down the road, allowing you to get the value of both cells once one of them had been worked out.
36:00 Simon missed a little piece of logic concerning 135 : The 135 in R6C4 sees all of the three cells R4C345, therefore they can't be all 135. Thus R4C5 is a 2 and R6C5 is 135.
He should do a video where the guitar playing stays in the background. That would be so relaxing.
3 года назад
It is very difficult to read the pencil marks in black fields. Perhaps the software can be improved so that dark colored cells have light pencil markings... A further improvement could be that the selected cell does not completely take on any mixed color, but only has a colored frame (e.g. yellow-black dashed)
When and where can we order these t-shirts ? "Let's get cracking" "The answer to the Ultimate Question of Life, the Universe, and Everything is 45" "Nori-Nori" "I may have to use some coloring" "Everything is wrogn"
At 15:08 you colluded that at row 9, column 5 red could only be 2 or 4. That would mean that the sum to 6 could also only be 2 or 4. And that would mean that r9, c1 could not be red else you would have a triplet of 2 & 4. So, either r9,c1 or r9c5 is not red. Only at 22:31 you found out that r9c1 could not be red. A minute further and you would have found that red = 4 by looking at the 8 & 10 sums!
I was going to make a joke yesterday about doing a Sherlock off the Reichenbach Falls rather than attempt the solve but I decided against it. ...Only to see a _Deer Hunter_ reference today. Beautiful intro btw Added: I did actually attempt this one, got about as far as the 8:10 mark myself, I just can't see what to do next
Black is my favourite to use in sudoku’s that need colouring, but Simon always avoids it. But liked the 2 second appearance in this video. Edit: 35.10 ; I object!
35:16 "I'm running out of colors" In terms of sheer information, and in respect of some color-challenged viewers, you might try simply _un_-coloring some cells. For example, the fact that at this point 8 is still pale green is a bit redundant. In fact I'd be quite aggressive about making cells pale again once their value is known and/or pencilmarked. Just a suggestion. Love this puzzle and this presentation so thank you!!
Well I was shouting at my screen! ;-) (16 min and earlier). Please please look at the 10 cage. You can remove 9 from one of the blues. /// It seems to me the puzzling community has created a new thing -- call it "Meta-Sudoku" or "Ultra-Sudoku." With all of the variants, and the adding of a layer of solving technique involving the use of color (which is not bifurcation but rather is allowing the solver to deduce the location of unknown digits - something new to the art of solving sudoku)..... I know that's an incomplete sentence, and my thoughts aren't fully formed, but I've been simmering on this for a while, and this video made it clear -- what Simon is doing is not "sudoku." It's something new. Meta-Sudoku or Ultra-Sudoku is born! ////// And - as the new solving techniques (mainly coloring) become more well known, the setters create puzzles with less and less information in them, requiring the use of the new techniques.
I let the power of mutli-colouring go to my head, and ended up with a psychedelic camouflage pattern, as if Zigby was trying to hide the magic round-about.
![Math for Game Devs [2022, part 1] • Numbers, Vectors & Dot Product](http://i.ytimg.com/vi/fjOdtSu4Lm4/mqdefault.jpg)
![Math for Game Devs [2022, part 1] • Numbers, Vectors & Dot Product](/img/tr.png)
It's a shame there's this perception that people get "cross" with you for not fully coloring. In my opinion, it would keep things cleaner and easier to scan if you just removing the coloring once it's not needed. So in this instance, if you're coloring to represent where a specific digit goes, once you determine the exact value of the digit, the coloring is redundant and can be removed (ceremoniously, if you'd like). I do this on my stream and people don't mind at all.
Actually he started getting blinded by all the colors in the end. He wasn't seeing lots of obvious sudoku deductions.
Exactly! And if you really want to you can always colour them back in once the puzzle is completed :P
Rangsk, where do you stream?
Honestly I feel like him keeping the coloring going after the use is over really makes the puzzle look messy.
+1 to this. I personally don't like to keep coloring digits that are already resolved. Keep the information "clean" so that it's easier to spot easy sudoku deduction
Any day we hear Simon playing the guitar is a good day.
He has improved a load over this series. Love the deer hunter theme
Relaxing intro. Simon is a genius at puzzle solving, and a wonderful guitar player as well.
For those interested, song in the intro is Stanley Myers "Cavatina", which was a part of the score for The Deer Hunter(1978)
Thanks for the knowledge bomb! :D
Thank you!
I just knew it as the music from The Gallery.
Thank you! Was scouring the comments for the answer to my question!
Those two blues that are different numbers are turning on my anxiety
They are the same number, just rotated... if that helps ;)
Simon: "red is in the eleven cage"
me: "ok, red definitely is less than eleven" 🤔
After smashing Simon's time on yesterday's puzzle this one took me nearly an hour and a half, double his time. A good day for more humility! I missed so many things on the way and cocked up twice. Getting the quad and quin on the diagonal felt so easy and then it all went wrong.
Excellent puzzle with a grid that looks almost sweet and innocent. Thank you as always to setter and testers and solver.
The thumbnail game has become next level. I actually, genuinely laughed out loud.
Hitting the like button already after the intro.
Same! That was lovely!
Hearing Simon saying he plays Overwatch has smiling so much. Online Gaming is for everyone!
This was lovely, I'm especially impressed how the diagonal communicated some info to opposite ends of the grid, so at times you had to apply knight's move logic 3-4 boxes away from where it really originated, I'm not sure I've ever seen that kind of misdirection. Simon, at 46:10 you pencil marked the 6s in box 9 then immediately forgot all about them and pencil marked 68 as candidates in r7c7 when it could only be an 8, and at 47:26 the 69 pair could have given you the 1 in row 2, but by that point you were basically just tidying up anyway. Really enjoyed watching this one, thank you sir!
I always thought the main diagonal constraint was kind of meh, but the last few puzzles you've showcased really show their potential. The way numbers can see each other from across the puzzle makes them so powerful and interesting, especially when combined with disjointed sets like in Lizzy's puzzle or the knights move in this puzzle. It's really something special.
Every video that starts with Simon playing his guitar gets x100 times better.
I absolutely love the smiles that Simon has as he plays the guitar in the intro! I love that he's enjoying playing.
Smart, English accent, and can play the guitar, this guy’s a total dreamboat.
I'm sure the title came from the software used to make it, but it also made me think of what I say most of the time I look at these sudoku, "F-ing puzzles!"
Simon figured it out, but for anyone still curious... To select all the cells of the same color you need to be in color mode and double click on a cell that is the color you want. If you want to select all the cells of the same number you need to be in one of the number modes (digit, corner, center), then double click a cell with the number you want.
If you are in colour-mode and you click on a red cell then it selects all cells that are red _or include red_ ... and if you select a cell that is multi-coloured red, yellow and grey then it selects _all_ cells that include _any_ of red, yellow or grey. (Or at least, I think that is what happens, but I'm happy to be proved wrong!). This can make it dangerous to highlight cells by colour without carefully checking that you've only got the cells you wanted!
@@stevieinselby You're right, it's dangerous to use in conjunction with multi-color mode!
35:10 "I'm running out of colors" But keeping the reds and not making real use of blue
Now I can't stop thinking about who Simon would play in Overwatch. I'm leaning towards a tank main
Probably sigma who's all about equations and hears some musics and melodies
I’d imagine more big picture healing
I wonder if uncoloring squares once you have found all 9 instance of a number in the grid would aid with scanning. That way you only need to scan over the colored squares. This might backfire spectacularly if you forget to color some squares then proceed to overlook them for the rest of the puzzle.
Simon, I know you felt you had to fade out the intro music after a certain time, but it was absolutely beautiful. I'd love to hear the whole thing. Maybe you could just put it on the end of tomorrow's video?
This has become the best asmr channel on RUclips for me
Every time Simon says "I've run out of legible colors" when he is leaving in multiple colors that have already been identified with values just for nostalgia's sake...
4 Colors can tile any "map" or surface without touching each other... Technically youll only need 4 colors...
Ill go with my nephews saying More colors better!
I wonder if they can use a hatching or a shade to show different patters... we use alot of hatches and patterns in Autocad.
@@shawnward6607 4-color theorem isn't really applicable in this setting, where the colors aren't being used to delineating adjacent zones on a map but rather to representing specific unknown values with each color...
@@HunterJE well put. And I felt kind of smart there
Yeah I would really prefer he get rid of the colors for digits that are already resolved to make the puzzle "cleaner"
@@esonsee me too - scanning woul be much easier without colours that do nothing anymore to advance the solve
We need a bonus video of just simon’s guitar! It was so beautiful
wow... hearing you play cavatina... takes me back sooo many years to when i played that in concert at school!
Downright touching musical intro there Simon, thank you
What a delight the intro music! You played with your heart, i felt it!
Beautiful logic. Thoroughly enjoyed. Thank you.
"I'm running out of colours here" ... time to ♻ recycle the reds, you've got all 9 4️⃣s in the grid now so you don't need 🟥 to mark those matching cells any more.
That PC should just about be powerful enough to run a sudoku!
Gotta hit those high frame rates for speed-solving.
Need to future proof it for the puzzles yet to come.
I cant even complete a simple 9x9 sudoku but im obsessed with watching these killer sudokus
"That one can't be 4 because its in a cage with a 4" And a column and box and row.
"Let's get the parts of speech correct when we're describing sudoku colors and their powers." Love it.
"And it can't be that one because that one, believe it or not, sees that one via the medium of madness." Loving the quotes today.
"7 is in fact purple, so those three squares can now become what they always wanted to be."
"That square's an 8. 8 goes in the middle box there. Let's color the 8s in green to make them feel at home with their friends."
I started the same way as Simon, but got stuck where he was at the 17 minute mark. Never really thought about how the knight's move interacts with the diagonal, which came in useful down the line. His logic allowed me to finish the puzzle in over an hour (had the timer paused for a bit). A missing pencil mark in box 7 had me go down a wrong path for a bit, too.
This one pulverized me. After getting the 6789 onto the diagonal I decided to be ridiculous and color everything--the 24s, 135s, and the 6789 group. Then solved the rest in multi-color space with only occasional reference to numbers (in the killer cages). I didn't know if it would be possible, but it worked, even if it did look like psychedelic modern art.
I think there's a faster route to getting the first digit in place. 30 in seven cells is either 1234569 or 1234578. In the top right box, that means the uncaged cells must add up to 15, and the minimum value of the uncaged cells in the 30-cage boxes is 39, which gives you enough to place the 9 in the center box.
Sorry, I don't quite follow. Can you please explain why the uncaged cells in boxes 2, 3 and 5 have to sum to a minimum value of 39?
As far as I can see, you have four given cage totals in those boxes summing to 3*30+10=100. Two pairs of uncaged cells in boxes 2 and 3 sum to 15 each, so 2*15=30. The three boxes must sum to 3*45=135. The difference between the cages + known 15 pairs and the three boxes are the cells r5c5 (in the boxes) and r4c9 (in the cages).
So we end up with r5c5 - r4c9 = boxes - cages - known 15 pairs = 135 - 100 - 30 = 5.
So the centre cell (r5c5) has to be 5 more than r4c9. But I don't see why they have to be 9 and 4?
Love (and put like) every time you start with the guitar.
Ok. 46:20 cracked me up. Simon talks about knights move breaking it, and then immediately pencils 68 in r7c7 after putting 6 into r8c5.
...and after corner pencilmarking where 6s can go in box 9 seconds earlier. 😂
@@RichSmith77 yeah, i think knights move is funny. Ive started skipping vids with rulesets (or skipping the dead parts until i see another digit once he starts looking) if he does the disjointed subset puzzles. Those make me wanna pull my hair out because I become hyper vigilant about it.
43:20 What stares me dummy in the face? It's 2 and 6 in R6. Simple scan, far too mundane for Simon ;-) I love that man.
Another 'hour-plus' solve for me; I finished in a time of 1:08:14. Still, I got there with no guessing and *thoroughly* enjoyed the solve.
Nice puzze!
Took me days this one. I found it crazy hard. Enjoyed it tho :)
r7c4 seems to the lynchpin to the entire puzzle... Simon identified it much, much sooner than I did, I spent ages looking down huge chains of logic mostly involving all of row 8, and although I did eventually fill in the 6 cage without discovering that cell, it was still important in finishing my solve. It seems like such an unremarkable cell and the fact that the puzzle hinges around it is kinda astonishing to me :)
I really like that the f cages are really only important for the break-in with the red squares being the same and the new 30 cages, to then never to be touched again in the rest of the puzzle xD
I can't believe I missed that weak number... I had the inital set up correct, but then I just didn't see that "weak number"... after that solved it in 120 mins... (yeah i looked at it for 30 mins before getting a Simon hint)
very clever puzzle.
I hope the guys release an app with diagonal rules set... that seems would be pretty good
ever thought of changing the colour pallet on the software? Make the purple really purple and add pink. or swap the yellow for brown. the yellow is useless.
Agreed. I also wish we had more than 9 colors.
I don't know why I find it funny to imagine Simon sitting by a high-end gaming PC solving sudoku
Damn your solution is much cleaner then mine.
Same break in, but I went full on multicolor and whittled down the puzzle piece by piece... And colored the entire grid before entering barely any digits.
Thanks for the elegant solve.
That computer is a BEAST. It could probably run two sudokus at once.
I know envy is one of the "seven deadly sins", but I am most definitely envious of Simon's mind and his deduction skills.
21:25 The explanation was not given but was implied, but why can't the black square in question be the other blue, or green or purple? I think it's because it sees all three of those via knight's move or sudoku from box 5, right?
I think the way Simon viewed it was as follows.
The cell in question (r7c4) was ruled out of c4 in box 5 by Sudoku and two other cells by knight's move. It also couldn't be red because we know red is elsewhere in box 8. That left three possible places where the digit in r7c4 could appear in box 5: r4c5, r6c5 and r4c6 (blue). Simon then showed why it couldn't be the blue in r4c6, since there would be no place it could then go in box 2. So that left two possible positions in box 5: r4c5 and r6c5.
Now it's true those two positions do see pink, green and the other blue, but it's also true that those two remaining cells in box 5 have already been pencilmarked as 12345, so are known to be low digits. I believe it's this pencilmarking of the available cells in box 5 that tells Simon r7c4 must also be a low digit.
It boils down to the same thing, since those cells were pencilmarked 12345 because they see all the high digits, either in the column or the box.
first.. look where green and purple can go in the center box.. either in column 4 which sees that box or r6c6 which sees that box by knights move.. it can't be the center blue r5c5 by knights move.. it can't be the blue in r4c6 since the number in that box is forced into column 4 in box 2 whice sees the box in question. That took me a while to figue out what simon meant.
Well at 17:37 , when you look at the 6 cage on the bottom, you can tell that 5 isn’t on the right cell because it’s a red one ( which has only the 2,3,4 possibilities) this indicates that cell isn’t 1 too. This leaves only the 2/4 possibility in 6 cage. Also the twin cell ( the one on the most top right cell of entire sudoku, that solves a lot actually). Good job
Nice intro today!
43:30 And in box 4 the two cells R5C1 and C3 are a 1-purple pair, so R6C1 is a 6, there is a 68 pair in box 6 thus the 15 cage in box 9 is 456.
Simon, I love the intro. Keep up the practice!
96:41 for me. Stupid long time. Enjoyed every second of it. Ended up using all the colours and filled in the entire grid with colours.
Great setting and great solve!!!!
That PC is an absolute monster
Congratulations on the new PC. I have to say I'm pretty shocked at the price, but can't argue with the 12 core CPU and latest GPU. I hope it's reliable!
I laughed hard on the 3.000 Pound Sudoku Machine.
...but i think for Video-editing it makes sense. I don't understand why he still has a HDD (why save on that, especially when you buy in this price range), but thats none of my business. Great PC nevertheless.
i asked for so long and at last we can partially color, thank you :)
Holy moly, it took me a full month of waking up early to spam refresh on AMD's website just to get a Ryzen 5900X. Just goes to show: it pays to be Sudoku famous! Nice rig Simon!
My mum took me to see the Deer Hunter for my birthday, based solely on the fact I had learned to play Cavatina on the piano. I'm still traumatised.
'Oh, my goodness, I can see what it is going to do!' That was the moment when my five minute demi slumber found an end
15:26 "Don't see how we prove that though."
Well for one thing, the 12345 in R8C2, R7C3 and R6C4 all see that square by sudoku and knights move. So at least three digits are going to be eliminated from R8C5.
The question is, at what point would R8C5 no longer be able to be from the digits 234 at which point it's no longer possible for it to be red.
By the way, R9C4 also sees those same cells along the diagonal as well as R9C1!
So whichever way the puzzle goes, those four cells cannot all be 1245! One of them has to be a 3! Ergo, there's no 3 in R1C9!
Really pretty melody at the start!
The thumbnail LOL! Note to self: don't open CTC pop-ups while you are on a conference call and not on mute. 🤣 Explaining to a co-worker why you are laughing about a sudoku is not easy.
40:30 R2C5 is identified as a 3 or 5, and Simon says he doesn't know if it is orange or purple, but it must be purple. Orange is locked in the same column in box 5
What a beautiful intro today!
I do love watching him struggle with things while ignoring the blindingly obvious for several minutes...
Congratulations on the new computer! What a huge leap, if I think back 2 years, where your mouse kept skipping, the software was creepy, etc. I think this machine should give video editing a decent boost too ...
With this new PC, I can't wait for the ray-traced sudoku graphics!
Thanks to Simon’s training from many previous killer sudoku videos, I knew the way to start this was to do the maths on the 2/3/5 cages. Combined with the knowledge that the uncaged digits added up to 15, that meant that r4c9 (a.k.a. the “red” squares) must be a 4 and r5c5 must be a 9). Imagine my surprise when I watched the video, only to find that Simon never even considered adding up the values in those cages and in fact dismissed the 30 cages fairly early as having no further value. It was interesting to see a totally different (albeit more roundabout) method to get those digits, but also somewhat disconcerting to see Simon completely miss out on that logic.
Not sure I see why that forces r4c9 to be 4 and r5c5 to be 9? Doesn't it just fix that r5c5 has to be 5 more than r4c9?
(Cage totals= 3 * 30 + 10 = 100
Known uncaged totals = 2 * 15 = 30
Boxes 2,3,5 = 3 * 45 = 135
Boxes = cages + known uncaged + r5c5 - r4c9
So 135 = 100 + 30 + r5c5 - r4c9
Therefore r5c5= r4c9 + 5
I seem to be missing a step.)
@@RichSmith77 I believe you are correct. I appear to have gotten lucky at that stage in the logic when I incorrectly concluded that r5c5 must be 9 and r4c9 a 4. At that point in the logic, making r5c5 an 8 or 7 with r4c9 being 3 or 2 would have been valid. I’m still surprised that Simon didn’t attempt the maths on those cages (given that this is usually his first instinct with this type of puzzle), but as it turns out, the only benefit would have been a bit down the road, allowing you to get the value of both cells once one of them had been worked out.
36:00 Simon missed a little piece of logic concerning 135 : The 135 in R6C4 sees all of the three cells R4C345, therefore they can't be all 135. Thus R4C5 is a 2 and R6C5 is 135.
He should do a video where the guitar playing stays in the background. That would be so relaxing.
It is very difficult to read the pencil marks in black fields. Perhaps the software can be improved so that dark colored cells have light pencil markings...
A further improvement could be that the selected cell does not completely take on any mixed color, but only has a colored frame (e.g. yellow-black dashed)
"The medium of madness"
When and where can we order these t-shirts ?
"Let's get cracking"
"The answer to the Ultimate Question of Life, the Universe, and Everything is 45"
"Nori-Nori"
"I may have to use some coloring"
"Everything is wrogn"
Love that guitar intro!
What's the point in colouring after you've determined the number the colour represents? Just seems like clutter
At 15:08 you colluded that at row 9, column 5 red could only be 2 or 4. That would mean that the sum to 6 could also only be 2 or 4. And that would mean that r9, c1 could not be red else you would have a triplet of 2 & 4. So, either r9,c1 or r9c5 is not red. Only at 22:31 you found out that r9c1 could not be red. A minute further and you would have found that red = 4 by looking at the 8 & 10 sums!
Oh my god, for a second I saw two Simons. He has a doppelganger
I love that music for the intro, what is it called? I want to get the tabs to try it myself.
Simon, are you using a special version of the solver app? I don't see a "Conflict Checker" option in my settings (tried in Firefox and Chrome).
Quote of the day: 'if I replace this red square with this red square I get these sort of rugby posts'
Or: 'so those three squares can become what they always wanted to be'
I was going to make a joke yesterday about doing a Sherlock off the Reichenbach Falls rather than attempt the solve but I decided against it.
...Only to see a _Deer Hunter_ reference today.
Beautiful intro btw
Added: I did actually attempt this one, got about as far as the 8:10 mark myself, I just can't see what to do next
Love the guitar intro....
The creator's name could be a reference to the signature move of the Pokemon Snorlax.
Black is my favourite to use in sudoku’s that need colouring, but Simon always avoids it. But liked the 2 second appearance in this video.
Edit: 35.10 ; I object!
Sheeeeeesh that pc is godly holy moly am I jealous of you
"Let me just highlight this square for a second" {turns the square pitch black}
35:16 "I'm running out of colors"
In terms of sheer information, and in respect of some color-challenged viewers, you might try simply _un_-coloring some cells. For example, the fact that at this point 8 is still pale green is a bit redundant. In fact I'd be quite aggressive about making cells pale again once their value is known and/or pencilmarked. Just a suggestion. Love this puzzle and this presentation so thank you!!
That thumbnail. God I hope there is a Christopher Walken impression in here! haha
46:10 marks 6 in two places in box 9
46:17 completely forgets about it and marks a 6-8 pair in another square in box 9
;) :P
talk about a beast of a PC for sudoku. So when do we get to do Sudoku in VR?
Well I was shouting at my screen! ;-) (16 min and earlier). Please please look at the 10 cage. You can remove 9 from one of the blues. /// It seems to me the puzzling community has created a new thing -- call it "Meta-Sudoku" or "Ultra-Sudoku." With all of the variants, and the adding of a layer of solving technique involving the use of color (which is not bifurcation but rather is allowing the solver to deduce the location of unknown digits - something new to the art of solving sudoku)..... I know that's an incomplete sentence, and my thoughts aren't fully formed, but I've been simmering on this for a while, and this video made it clear -- what Simon is doing is not "sudoku." It's something new. Meta-Sudoku or Ultra-Sudoku is born! ////// And - as the new solving techniques (mainly coloring) become more well known, the setters create puzzles with less and less information in them, requiring the use of the new techniques.
So now you'll be playing Overwatch in 4K at full 60FPS at max settings, heck yeah!
I too like that creators name. Mainly because Snorlax reminds me of me
Black isn't he problem - blue on black is. Just get your programmer to make the text white when the background is black.
I'm still hopeful to see the solve of "everything is wrogn" sequel 😅
Great puzzle, thanks, despite making the precise mistake in pencil marking 7s that Simon warns against.
I let the power of mutli-colouring go to my head, and ended up with a psychedelic camouflage pattern, as if Zigby was trying to hide the magic round-about.
So many colours my brain Hertz.
Your info card says "Thermo All" not "Thermo App", I think it's a typo.