This Fog Sudoku Is Genius.

[@martysears]

This is probably my favourite ever fog puzzle. Absolutely amazing… the concept is so far up my street it's practically in my living room. Very much the sort of thing I wish I’d have thought of myself! I do love to see counting circles logic used creatively and innovatively in this kind of way, and this is such a fine example. And the break-in digit is one of the most unexpected and satisfying break-in digits I have ever seen on the channel. Puzzle of the year for me, I think. PS: The fact the grey dot didn't end up hiding any yellow line was so much better than if it had 🤣

[@shoto5892]

The grey dot was still relevant though, I dont think simon appreciated it properly because if it hadnt been there you could have early on just fogured out that it had to move up to the upper leftmost cell in box 6, skipping a lot of logic (that admittedly was more in simons solve because he didnt pick up on some clear problems with the pink line visiting box 5) and giving you the pink line literally for free

[@andrewvincent77]

Great summary! I agree

[@sirgeremiah]

Icyfruit should be very happy to receive that high praise, Marty. I love your puzzles (though they sometimes are beyond me). I agree this one is amazing. It took me a while to wrap my head around the constraints inherent in the lines, but after that, it was amazingly satisfying. That break-in digit surprised me so much, I was fully expecting that entry to not clear the fog.

[@gi0nbecell]

Who else was perfectly amused by Simon elaborately proving that the pink line _could_ do something complicated and completely missed the point that in that case it would _have_ to pick up a 3 in box 5 which already had to appear on the segment of the yellow line, as it is either length 7 or 8, thus immediately disproving the convoluted pink line?

[@RomanQrr]

My thoughts exactly.

[@avishevin1976]

I was thinking that it was disproved because the only place the 3 could be placed on the pink line in box 5 would have prevented the yellow line from exiting.

[@puritan7473]

100%

[@iisaklussenden1513]

Yeah. I got it by realizing the only available digits in boxes 2 and 5 are 1 or 9, since everything else has to be on the yellow line.

[@zmitter4844]

tbf I did the exact same thing during my solve

[@NinjarioPicmin]

that's my favorite fog of war first digit i've put in ever i think, that was so crazy and out of the blue i absolutely loved this revelation

[@RichSmith77]

I agree. It took such a lot of detailed thinking to conclude the yellow line had to have a single cell line segment somewhere and that was the only possible place for it, and further the yellow line has to visit eight boxes so it must be an 8. So much was being confirmed by entering that single digit. If the fog didn't clear, I think I'd have given up there and then. Will the fog clear or not...phew! 😂

[@EiQ200]

I needed about 50min to come to that 8 that was such a good feeling. Really nice puzzle

[@timotab]

56:44 and onwards: Once boxes 2 and 5 are between them 7 and 8 length yellow segments, both of them will have 3 on the yellow line. the pink link cannot go into either box, because the 3 is already on another line. Pink is thus a 2-segment line

[@MaryVidal-pt8bx]

I love the joyful sounds of the children echoing Simon’s joyful reactions to uncovering the clever bits of this puzzle.

[@BozoTheBear]

Holy moly, what a puzzle!! I literally stared at this for about half an hour with no clue what to do (apart from figuring out that the yellow line had to visit eight boxes, and the order it did so), then gave up and had to do other things. But later in the evening while washing the dishes, suddenly realised there had to be a box with only one yellow-line-cell, and knew it had to be Box 7, and that there was an 8 in the corner of that box! I couldn't have run faster back to the laptop to put it in, and YESSSSSS!!!!! Then managed to finish in 51:55. This might be the trickiest break-in I've managed to figure out. Well done, IcyFruit, I loved your puzzle!

[@pick6and129]

I haven't had the time to do a sudoku with Simon in a while. It is so nice to come back and learn a new concept with/from Simon. He is so eloquent and such a great teacher. The way he is able to break down complicated issues and communicate his thoughts is amazing. I aspire to do that. I just know I will eventually get it as he has taught me so much about Sudoku and communication.

[@gusmichel7035]

I paused about 35:00. I noticed this from the get-go, but in the previous minute Simon contradicted my understanding. There does NOT have to be a two cell segment on an arbitrary line. A simple example: a length two diagonal segment which crosses a cell boundary can be 2 2 [without diagonal it would violate Sudoku rules]. Similarly segments divided by commas across boxes and avoiding Sudoku could be [3, 3 1, 3], or [3, 3 2, 3 2], [4, 3 4, 3 4, 3 4] or many larger sets. [3,3,3] could even be a purely diagonal path in three different boxes. However, if you consider the set of numbers on the longest [may be tied for longest] segment, the line must pass through a number of boxes equal to the largest such number, and through a number of cells equal to the sum of the set of numbers in that box. So you can skip small numbers under at least some conditions. If there are multiple segments of the same length, they must contain the same set of digits. I don't know whether I hope that the idea Simon has that each line which passes through n cells must have distinct segment lengths is of negative consequence or not. ***** I will add any later related comments below this line. *****

[@chipsounder4633]

At 59 mins in.. row 3 column 7 can never be a 3 because the yellow lines in box 2 and 5 are 7 and 8 cells long both requiring 3s. What a gorgeous puzzle.

[@lukasr5867]

By that point it has been kind of proven by the fog being lifted, but technically it's still an assumption that Simon made from the beginning. In theory the line could have multiple segments of the same length and skip a digit (e.g. Box 5 could be another 5-cell segment). The only thing Simon proved in the beginning was that the line couldn't skip a segment-length, because that would cause duplicate digits. But the other way around that's no problem ...

[@mats1365]

I solved this puzzle myself a few days back. For the first 40 minutes I literally didn’t write I single thing into the grid. But I’m so glad I stuck with it, cause the euphoria I experienced when I got the first digit was nothing short of magical

[@IgorPellinen]

Feeling jealous!

[@chloratum]

Now that was indeed an absolutely genius and ridiculous first digit. What a genius puzzle.

[@jonasf94114]

What makes Simon assumption correct (all lines are different lengths) is column 4: at this point, you know that all digits except 9 are on the lines thus there can’t be 2 lines with the same length

[@ADrunkenGoat]

Ah! Was looking for this explanation. I kept wondering why there coulnd't be, for example, 2 lenght-3 segments. Simon simply assumed this and turned out to be correct because of column 4!

[@TheSwitzer1234]

Yeah, but some of the logic before revealing column 4 was trickier without knowing the full structure. Even with a single 1-cell segment, the "next shortest" segment can repeat along the list of lengths (e.g. 1-2-3-4-5-5-6-7 was a possibility that tortured me a lot)

[@jonasf94114]

@@TheSwitzer1234 you are right. I did not rewind to check how soon Simon used the assumption (IIRC, up until the 4 in box 9, we don’t need it)

[@HansMaurer.]

​@jonasf94114 At 46:19, he fills in 567 in row 7. The 5 is only correct if he can be sure that there's only one segment of length 4. However, at that point, column 4 is visible enough.

[@cfitz92]

​@jonasf94114 I was looking for an explanation for that. I was thinking it could be a length 6 line and R7C9 could be 3 or 4. But because of column 4 every segment must have a different length and we already have a length 6 segment in box 4

[@HunterJE]

Late solve spoilers relating to the purpose of the grey dot: ... What Simon missed in his metagaming certainty about what the grey dot would mean is that _negative_ information can be just as helpful as positive (and so just as important to control in a controlled-reveal variant like fog). Specifically in this puzzle, without the dot hiding the fact that the line _doesn't_ go in to r6c7 it would have been possible at 55:06 when Simon placed the 8 pencil marks to instead just place the 8, which would have totally bypassed the interesting logic around the pink line.

[@Timlagor]

He could have bypassed most of that by noticing that the 3 had to be on the yellow line in the remaining boxes

[@Tagswell]

Yes, the purpose of the grey dot is to prevent the metagaming of ruling out certain paths lines might take simply by seeing the edge of a line on the corner of a cell. Since the grey dot covering where the line might go makes the presence of the line in that cell uncertain. Which unfortunately means, in the example with the pink line later, Simon *could* actually have ruled out the pink line going into box 2 based on the fact that he could not see the line in the corner of the relevant cell. But thankfully, he was able to find a proof he found more satisfying. And the presence of the grey dots in relevant corners encourages and requires that you find such alternate proofs. Perhaps this puzzle could have used more grey dots in some places.

[@artursruseckis4242]

@@Tagswell yet he missed the most satisfying (and presumably - the intended) path - pink could not have 3 nor 4 on it because those numbers were already taken by yellow line in middle boxes.

[@SarahLivne]

And yet I do think they should have alo added a grey dot on the pink line to prevent us from seeing that it doesn't bend diagonally upwards (and force us to reason that it doesn't go into box 2 for other reasons).

[@artursruseckis4242]

@@SarahLivne agree

[@IgorPellinen]

Technically two and three one-cell line segments are possible. Simon missed that. It would be diagonal line of two twos or three threes. Otherwise great solve.

[@yummymemez]

Okay, but what makes you claim he missed that possibility though

[@Sidnv]

@@yummymemez All of the analysis done at the start regarding the structure of lines ignores this possibility. It's clear he missed it.

[@artursruseckis4242]

@@Sidnv I am pretty sure IcyFruit missed that as well. The intention was clear and had to be defined in more or less not obvious way, but the definition turned out to be a bit wider than intended.

[@HansMaurer.]

​@artursruseckis4242 Probably not, see other comments regarding column 4.

[@chris5619]

@@artursruseckis4242 What makes you think IcyFruit missed that? The puzzle is of course logically solvable with only one solution, so it seems like a weird thing to assume he missed.

[@GrantDexter]

"I'm trying to make sure I'm agreeing with my brain." Lol. I had the opposite problem. ;)

[@titusadduxas]

47:34 - I spent a lot of the night thinking through the logic as at first blush I didn’t have a clue how to start. Once I’d got my head round it though everything fell into place and I’d got my first digit (totally not where I’d expected it) The FOW reveals were truly amazing and I think this is the first Icy Fruit one I’ve managed to solve! I particularly liked the grey dot!!!

[@loppemaster]

I think Simon makes a wrong assumption about the line segment lengths. I think it's true that you need a segment length of 1, and if you increase you need a length 2 and so on. But what's to stop the yellow segment lengths of going 1-2-3-4-5-6-7-7 and just not putting a 1 on the line or 1-1-2-3-4-5-6-7 and not putting a 7 on the line, that would still prevent the issue of forcing two of the same digit. Obviously once you know there is length 8 segment you know it must go 1-2-3-4-5-6-7-8 but not until then

[@RoderickEtheria]

Segments of the same size contain the same digits.

[@loppemaster]

@@RoderickEtheria I know, but what's stopping a line visiting 3 boxes from having three segments of length one, all containing the digit 3? Or from having segment lengths 1-1-2 containing digits 3 - 3 - 3, 1? Or segment lengths 1-2-2 with digits 3 - 3, 2 - 3, 2. All those are easily doable. Like I said, you need a length 1 segment and you can't skip a length but you can repeat lengths as long as you then miss out the appropriate digit like I showed above

[@NinjarioPicmin]

yeah my first thought experiment before even watching simon's thoughts was a line with 5 segments which contained a [5], [45], [345], [345], [345] and at least while testing this myself i don't see any issue with something like that existing

[@timjensen4320]

I had the same thoughts. Once you get column 4 exposed to see all but row 9 is on the line though, it forces Simon's logic to be right since every digit but 9 makes an appearance on the line.

[@Killrbladez]

Simon's logic wasn't wrong for requiring a 1 cell segment and a 2 cell segment. You always require a segment with one digit, so that confirms 8 segments exist. Then there's no option to have a second 1 cell segment, and there's nowhere else than box 1 that can be a 2 cell segment. And then placing the 7 in r2c3 tells you that you have 8 digits on the yellow line in column 4, so that tells you that the segments have to be 1-2-3-4-5-6-7-8.

[@roccov3614]

29:20 When Simon says that he hasn't got the one cell segment yet, I could see where it had to go. That is now my first digit.

[@olivier2553]

The explanation why the green line in the demonstration around 22:00 did not work is simple: the line in box 5 has a segment of length 4, so it must comprise at least the digits 1, 2, 3 and 4. The digit 4 must be present in 4 segments, but the line has only 3 segments.

[@DanielHShead]

I giggled with joy and closed the puzzle after the first digit. It was satisfying enough by itself. I came back later and watched the video.

[@Yttria]

Loved it. Finished in just under 54 minutes. It's always fun when a fog puzzle has you placing digits far away from anything shown.

[@MarkBennet10001]

Nice solve of nice puzzle. Note that at about 1 hour the short line can't have a 3 on because the long yellow lines in boxes 2 and 5 need 3, so there are not enough 3s to be found.

[@misak481]

This opening move is the most beautiful thing I have seen in a puzzle

[@stevieinselby]

I had done a lot of good thinking but hadn't articulated to myself that the yellow line needed a one-cell segment, that was the only hint I needed. The key observation for the pink line is that if it's a 3 cell line then it must take a 3 in box 2 or 5, but we know that in both of those boxes the 3 must be on the yellow line.

[@brotherhood9398]

Towards the end when Simon is wondering the nature of the pink line, is it not easier to say it can't enter box 5 because the 3 must be on the yellow line? Since we know the line takes 7 or 8 cells in that box? That was what I used to solve the puzzle but I don't think I've ever used logic Simon hasn't seen before so wondering if I've missed something.

[@ninjamunk]

Think this is a very typical Simon-thing (and why we all love him). He will overcomplicate things when there are obvious solutions right in front of him. Normally it is a long line of logical conclusions in stead of simple sudoku. Here it was what you described. Either way, that's just Simon :)

[@Tagswell]

Very interesting puzzle. The line rules seem to necessitate that the lines themselves follow a lot of other rules as a result, which actually give you quite a lot of information about the lines and their digits. I'm only partway through the video now, so I'm not sure how much of it Simon will utilize or how much is necessary to utilize for the solve. But it seems like there is actually a lot of variation in this ruleset that would allow for some fun solves. After some musing to myself about it with the video paused, I'd like to compile a list of rules that lines must follow as a result of the base rules below, if anyone has any they can add or any exceptions to them that would be great. Base rules: -Lines can travel orthogonally and diagonally. -Lines cannot cross themselves or each other. -Lines are broken into segments for each box they enter. All lines have at least two segments. -Digits appearing on a given segment will appear on each segment of equal or greater length. -A Digits appears on a whole line as many times as it's value. *(Potential Spoilers Ahead)* Okay, now for the inferred rules: -Each line must contain at least one segment of length 1. (As a result of Sudoku rules.) -The largest digit that appears on a line is also the amount of segments that line contains. -The smallest digit that appears on a line is also the amount of segments of length equal to the longest length that appears on the whole line. -The shortest length segment(s) must contain the largest digit that appears on that whole line. -The longest length segment(s) must contain the smallest digit that appears on that whole line. -The length of the longest segment(s) must be equal to the number of different digits that appear on that line. -Each line must have at least one segment of each possible length between the shortest and longest length. -Each line must have at least one digit of each value between it's largest and smallest digits. Also, I'm not sure there is anything stopping a line from entering a box twice, though I am not sure whether it reentering the same box would count as a separate segment. I think by Sudoku rules it would have to count as a single segment in order to not break, but I am not certain.

[@stevieinselby]

The rules say that box boundaries separate the line into segments, therefore if a line were to exit and then re-enter a box, it would count as a new segment and so would need to repeat X within the box.

[@Tagswell]

​@@stevieinselby At least going by the rules as worded in the video, you could define a "segment" as the part(s) of a line in a single given 3x3 box. It's not intuitive, but if you do, then two separate parts of a line in the same box would be allowed within sudoku rules, as long as you count it as one "segment" within the line rules. Admittedly, it's not particularly relevant to this puzzle specifically, and sort of betrays the standard definition of a line segment, but not technically against any rules as stated here I don't think.

[@RichSmith77]

​@@TagswellBut box borders are what divide the line into segments. Since you cross the box border each time you enter a box, I don't see how it couldn't be considered starting a new line segment each time you entered a box?

[@saschabaer3327]

There’s some inferred rules (which simon also assumed, see many other comments) that are actually wrong: you can have (with some very carefully arranged diagonals) more than one one-digit segment, which allows for a line like 6-6-6-36-236-1236, which violates your last rule (since 4 and 5 are skipped over)

[@Tagswell]

@@saschabaer3327 awesome, i thought there might be a chance there was exceptions to that one! But I decided to put it anyway because of Cunningham's Law lol

[@sirgeremiah]

Simon always underestimates how well he has done. He exclaims he spent 35 minutes thinking - but he started solving 13 minutes in. On top of that, he spent a good bit of those remaninig 22 minutes explaining his thought process in great detail. I'd bet he would have gotte to that point in under 10 minutes if he were solving on his own.

[@chocolateboy300]

I finished in 168 minutes. I needed Simon's help at first to understand the implication of the rules. As soon as Simon said that a single cell must be used and that lines are limited into which boxes it can enter, I completely understood what this meant. I was so happy when I figured out where the single cell went and which one it was. I laughed when I put a digit into the bottom left corner, only to reveal two givens. This puzzle is a lot to think about. This puzzle was difficult, even fully understanding the rules. Great Puzzle!

[@HunterJE]

It doesn't get particularly used in the solve to come, but I think Simon misses a possible case when exploring minimal line lengths at the start, what's to stop a two cell line crossing a box border diagonally, in which case it could be double 2 without breaking any rules?

[@jamesmoriarty8415]

I was wondering the same thing when he started looking at the options of line lengths.

[@Tahgtahv]

Well, he only used the logic on lines that he could see. And clearly those lines contained more than one cell in a box.

[@HunterJE]

@@Tahgtahv Correct, I was merely commenting on his (unused, as I noted) assertion that the minimum line was a three cell line with 2 and 12 on it.

[@minamagdy4126]

Fun fact, the same is also true for three 3's in a diagonal. In fact, what can be proven is that there can be no two equal-length segments that aren't one long. That doesn't preclude the possibilities stated here, and even including other digits (it's easy to construct examples with 4 and either 1 or 2, maybe even 5 with either, although I haven't come upwith one for 1 and 5)

[@willemm9356]

@@minamagdy4126 I disagree. You could for example have a line that goes through three boxes with one 1-length segment and two 2-length segments. (3 | 3-2 | 3-2) What can be proven is that you can't have a jump in line-segment lengths. That is, each length from 1 to the longest length has to appear at least once.

[@shagiephoto]

Consider a hypothetical line: (r3c3) to (r2c4 to r3c5) to (r4c6) -- this has 1:2:1 length and could be filled with the numbers 3 : 3, 1 : 3

[@Tepalus]

22:14 for me. That has to be the fastest I have EVER had for a puzzle that took Simon close to an hour! What a beautiful one!🎉

[@jamesrahner8920]

probably my favorite break-in of all time. wow. stunning.

[@sainio]

At 40:52, why the yellow line cannot go row 8, col 5 and then row 7, col 6?

[@ilsekleibscheidel7219]

Because at that point those cells are already revealed as blue, aren't they?

[@jeremyblack1929]

Truly an Epic first digit. It's one of those that at a glance you would never assume is the puzzles first digit. This is one of those puzzles that really shows how fog of war can make otherwise completely absurd rules actually approachable. Love it.

[@steve470]

54:29 for me, including a solid 15 minutes of figuring out what properties any such line must have. Nifty mental workout! Sad that Simon missed some major logic. Around the 21:34 mark, what he proved that each line must have *at least* one 1-cell segment, *at least* one 2-cell segment, on up to the longest segment on the line. He did not prove---and it is not generally true---that there must be *exactly* one segment of any particular length, but proceeded as though it must be true. (Counterexamples: you could have a three-segment line with the digits 3/13/3 - or even 3/3/3. You could have a six-segment line with 6/6/6/36/236/236. Those don't violate any of the given rules.) However, once you see enough of column 4 to know that there are eight different digits on the line (and that there's a box the line can't enter), you now know that this particular line doesn't skip any digits, so must have each possible length (because a duplicate length will result in a low digit being skipped).

[@maoman4855]

Yes, this is true if you imagine every possible configuration of what the line could do in a completely blank grid. However, the uncovered spots we were provided very carefully made any such line impossible, at least as far as I can tell. Go ahead and try it yourself--reset the puzzle and try to draw in a yellow line that connects the given revealed sections and yet allows something _other_ than a 1-2-3-4-5-6-7-8 segment line without breaking any rule. I certainly couldn't find one. That's part of the brilliance of this puzzle, I think. The tiny morsels of hints we're given at the start are _just enough_ to figure out all we need to know about how the yellow line works. It's genius.

[@ahiskali1]

3/13/3 is impossible, because it couldn't get to another box to place the last 3. Same logic with the other line

[@steve470]

@@ahiskali1 Maybe not with the line layouts in this particular grid, but you absolutely could if you had a different line. Example: 3 in r7c2 (box 7), 1 in r6c3 (box 4), 3 in r5c3 (box 4), 3 in r4c4 (box 5)

[@mirthemw]

38:43 that worried face then happy face is so on point haha

[@mcstruthers]

This may be one of the first of the CTC puzzles I’ve done that actually seemed pretty straightforward to me. Everything just seemed to follow through with clear logic!

[@henriparenteau6496]

I love the grey dot covering part of the line, it always bothers me just a little when you can see just a tiny little bit of a line that's going diagonally under two bits of fog because it feels like I shouldn't be seeing it! Love it, awesome puzzle and great line constraint I hadn't seen before!

[@chubaloegames5426]

Thank you for this explanation, finished the puzzle without understanding this part from the ruleset. And thought this dot made no sense at all. Now I understand!

[@Green9Monkey]

35:18 for me. Desceptively simple rules, but did not take long to see how they worked together to tightly limit the possibilities.

[@LadyEmilyNyx]

I sppreciate you not killing spiders on sight, like so many unfortunately do.

[@audreyclark7938]

This is one of a very few puzzles I was able to solve at all. Usually, after 30 minutes or so, I give up. This one, I kept at it and was able to solve. For a beginner like myself, it was satisfying to know that it is possible for me to solve something once I kept at it.

[@jinkela7295]

1:00:51 Because you need a 3 and a 4 on the yellow line segment in boxes 2 and 5, r3c8=1

[@mstmar]

Challenge accepted. Spoilers for the bonus puzzle ahead. this is a proof that you need an odd number of cells of one color on the border (at least for a 9x9). even x even have different rules which we'll also see If you take a ying yang puzzle and draw a border around one of the colors, then delete the colors. focus on any 2x2 square, what will you see in the 4 inner border segments? you can't see an odd number of segments, because that would mean that you'd have an odd number of color changes going around, so the first square would be different colored than itself. you can't see 0 segments, because that would mean all colors are the same aka a 2x2 which is not allowed. you can't see 4 segments, because that would create a checkerboard (long time viewers of the channel would know this secret). that leaves exactly 2 border segments connecting to EVERY center of a 2x2. if we zoom back out what can we say about the border. well it can't be 2 or more loops, as that would isolate sections. this means means we have one continuous path that goes through every vertex (the center of the 2x2s) in the grid before going to the edge on both ends (as long as there's 1 color change). Now if we color the vertices in a checkerboard pattern, the path would alternate between the 2 checkerboard colors. ignoring the edge for now, we have 64 vertices, and so the path starts and ends on a different color before dipping out of to the edge. that means that the first and last vertex on the path have to be different colors. How do we achieve that? By having an odd number of cells of the same color (not counting the corners) This works for any odd x odd sized grid as there will be an even x even number of center vertices. for even x even, we can show the opposite actually. there are 1 more cell of one color than the other, and so our inner path will need to start and end on the same color (the one that has more). this means that we need an even number of cells of one color between the color change (not counting the corners). not only that, but it's also restricted in where the color change can happen, as it has to happen on the correct bishop color.

[@mumushanshi]

Stunning. This comment needs to be pinned.

[@In_42_Space]

That was a complicated start, but love the way the line logic worked. Genius indeed. Lovely solve.

[@EdithKFrost]

@38:05 You have to be more careful. Although the length of segment can't jump, it could repeat. It's not necessarily true to have length 1 through 8.

[@thnderleg]

might be the most enjoyable fog sudoku! The beginning was very intimidating but once I quickly realised the logic of what numbers go on the yellow line I found the first 8 and then it just snowballed into awesome fun!

[@LaytonBehelit]

5:15 I tried it out and I think it only is true for odd size grids, otherwise you can just make yin and yang into two "forks" which don't obey this rule.

[@mstmar]

ya for even sized girds it's the opposite. it needs to have even runs of each color instead and odd runs are broken

[@JakeRoeder]

I had to borrow some of Simon's insights at the beginning, but from my placing the first digit, it was all my own logic. Such a fun puzzle and more proof for people who still, for some reason, don't like fog of war puzzles that they can be far more than a guided tour.

[@Gonzalo_Garcia_]

18:21 for me. What a great puzzle, loved it!!

[@nightwishlover8913]

How did you manage that without knowing what the grey dot means?

[@Gonzalo_Garcia_]

@@nightwishlover8913 The rules do mention it: "A grey dot may hide part of a line". I believe it's there only to create some uncertainty on where the line may go in box 6.

[@nightwishlover8913]

@@Gonzalo_Garcia_ Thanks - I already got that.

[@artursruseckis4242]

Simon missed a very important thing when thinking about the pink line. Of course he found a very complicated way of how to eliminate 1 from 19 pair below, that's exactly what we would expect from Simon, especially when the intended path is trivial. Here it goes: If pink line visits more than two boxes, it has to visit box2 or box5 (or both). Either way it will require a number 3 (or 4) on a pink line in those boxes. But we know, that boxes 2 and 5 contain yellow lines of length 7 and 8, i.e. both numbers 3 and 4 are already taken by yellow line and are not available to be used by the pink one! The only free numbers in those boxes are 1 and 9. There is no pink in boxes #2 nor #5 and thus pink is limited to two boxes and that immediately disambiguates it because of 2 in column seven.

[@Zwangsdemokrat]

The grey dot is a fantastic touch, loved it.

[@cjbarthvdm]

@39:05 I don´t understand why Simon connects the yellow lines in box 4. I don''t see why it isn´t it possible that the lines connect also using r4c1. Then there are 2 positions for 8: r4/5c1.

[@bobson_dugnutt]

r4c1 is not in fog. We know it's a blank cell.

[@MisterFizzer]

I was smiling like a maniac while solving this puzzle, such a genuinely fun experience. When the 7 in box 7 revealed new given digits it felt unwrapping a present. 🥰

[@Coyotek4]

Took me about 40 minutes over two sessions ... challenging fog! Nice puzzle!

[@stevenape377]

Yes! My son and I did it! Definitely the hardest puzzle we solved together. But we loved it so much. We couldn't believe when we finally found the break-in.

[@Antinomiste]

Spectacular. Was hard to get my head around, but I was never really stuck, loved it, longest Simon solve I attempt and succeed it I think

[@chesterfield9682]

"Cardboard box? You were lucky!"

[@Escviitash]

1:02:22 No, there doesn't have to be a 2-cell segment on the line. R3C4 + R4C8 + R5C6/R6C6 = 3, and R5C7 = 1 will create a 3-segment line with three 3s and one 1, which will obey all the rules. If you ignore the given lines you would even be able create a 5-segment line with five 5s and one 1 e.g R3C4 -> R4C3 and straight down to R8C7 which will have four 1-cell segments and a 2-cell segment, or a 3-segment line with just three 3s e,g R3C3 -> R6C4 -> R7C5 which will have three 1-cell segment,. And you can even combine these example lines into one line containing five 5s, three 3s and one 1 which will have two 1-cell segments, two 2-cell segments and a 3-cell segment. The conclusion that there have to a segment of each length from 1 to X on an X-segment line is utterly wrong. The reason that the pink/purple line can't be a 3-segment line or longer is that the 3s in box 2 and 5 have to be on the yellow line, thereby making it impossible to get a third 3 for the pink/purple line

[@fruit3193]

The last thing I expected was to uncover multiple given digits. That was phenomenal.

[@virtuous-sloth]

I loved the fog reveals in this one, but there were still some thinking to be done as you went along

[@olivier2553]

At 59:10 a simpler explanation could be that every digits 2 to 8 are used by the yellow line in boxes 2 and 5 so the pink line could not get any 3 or 4 in those boxes, so the pink line is only 2 boxes long.

[@PauxloE]

→ 16:51 "So we have to have lines of at least length 3" - Couldn't a diagonal line of length 2 also work (with two 2s on it)? → 45:49 "So we basically know that [r6c7]'s going to be where the line goes but we're not allowed to assume it" - Or actually it could be a red herring to make it harder for us.

[@bopcity5785]

I think it must be said that Simon's scanning was pretty good here

[@Scott2148]

Simon you are brilliant, and you share your thoughts out loud so we can learn. But I had an amusing thought - a Simon self-parody - could you stack the dish washer with this same level of reflection, logic and fore thought. A parody tutorial - if may require lashings of chocolate cake first. Thank you and good wishes to you and yours.

[@houligan5254]

I love these fog puzzles where you can get a digit in the middle of the fog somewhere. The slalom puzzle was like that too.

[@diederikvandedijk]

I think Taron refers to the roller coaster, not the actor.

[@joergstoye695]

Yes. Taron (I'll not drop links, but easy to find) is a famous (for what it is) coaster that meanders all over one theme area in a German theme park. It's hard to not read the final grid as homage.

[@olivier2553]

Moving your mouse cursor around the spider, would it spook it? Like gecko would follow laser pointer light on a wall and try to catch it

[@leefisher6366]

I immediately thought of Taron (actually Taren) Capel from Dr Who lore, who is himself a reference to Karel Capek, the first person to use the word 'Robot' in his play R.U.R. My mind is weird too, Simon - You Are Not Alone.

[@werefrogofassyria6609]

39 minute solve. The trick was to recognize the line had a different segment in each box it visits, and it goes fast there. It was great that the first digit was nowhere near the initial clues.

[@johnh2052]

Looking at the pink line, it can't visit 3 boxes; neither box 2 nor box 5 can contribute a 3 as they're already needed for the yellow lines.

[@henriquecaridade]

Simon at the start proved that the every line must have a 1 cell segment but and after that you have to increase by one every time but i don't think that's right. I believe that Simon only proved that the line can't "jump" number or in other words you couldn't have segments with lengths 1 2 4 5 for example but simon didn't prove that they have to be diferent lengths. You could have lengths such as 1 2 2 3 3 4 5 or am i missing something? Example line segments: length 1: 4 length 2: 4 3 length 2: 4 3 length 3: 4 3 1 Edit: I'm asking because he at 46:33 says that the box 4 segment has all the digits 3 to 8 which is only true if you can't have repeated lenghts. At 52:00 he discards the length 6 option from the box 9 segment because he already has a length 6 segment in box 4. At 55:30 he assumes the lengths of boxes 2 and 5 are the 7 and 8 length segments.

[@markp7262]

37:31 finish. A very fun idea, and well executed!

[@AWanderingSwordsman]

52:53 for me but I did cheat a little by trying numbers to see if I was right about the paths. Fun puzzle. I love fog.

[@frankjiang1857]

Finished in 38:05. The trickiest part is the break-in, but once you figure that out with the wealth of information coming from out of the fog, it becomes rather straightforward. The hint I will give for the break-in is think of the geometry of the yellow line and what that implies for how many digits are in each box. Fun puzzle!

[@angec9908]

I feel so smart (after Simon worked out all the complicated stuff 😂). I figured out the the pink line couldn’t have a 3 on it because the 3 in boxes 2 and 5 had to be on the yellow line since one was 8 and one was 7 length. Both include the 3 (8-3 or 8-2).

[@k.c.8658]

Bees, wasps, spiders, flies…Simon lives in a virtue jungle.

[@bevobexley4087]

You might say IcyFruit is "Taron-ifying"

[@stevesebzda570]

64 mins Thanks for the start, Simon.

[@lukamiler5824]

This was such a beautiful puzzle. The whole time it felt like the puzzle was cheering me on while I was solving it. As soon as the first digit is placed everything goes so beautifully smoothly.

[@eddieharwood7788]

That was utterly brilliant. I stared at it then went to bed. In the night I woke up clear how the lines worked, and thus it was immediately obvious that 8 boxes were visited and in which order. The realisation that there was only one box that could have a single cell was such a joy to reveal the 8. The grey dot was a brilliant spoiler on which way the line went in box 6. After looing at it for some considerable time last night without getting anywhere my total time of 76 minutes including that part so once you realise how it works it is not such a difficult puzzle. The other thing that made me laugh was revealing given digits - I had been wondering how on earth I was going to get the middling digits.

[@GargantuanMonster]

Took me a while to do this puzzle, but I had to go back and warch Simon's solve just to see his reaction to the sudden given digits

[@CharlesGregory]

The spider was obviously drawn to all the number 8s!

[@Afterthoughtbtw]

30 minutes for me. Loved that. I genuinely, really, terrifically struggle with IcyFruit puzzles. Maybe the FoW mechanic gave me some inoculation or something, because I actually felt like I understood what I was doing(!)

[@Ziboo30]

34:10 I was block here because I though the line was allowed to go from R7C4 to R6C3 in diagonal. nowhere in the rules it says that you cannot do that. Right ?

[@BaSYaVSkI]

Very interesting puzzle, but when you start getting the 8-s, it simplifies the beauty which felt so difficult just before getting it. I did not use the grey dot.

[@ronjohnson6916]

Simon confirmed as arachnophile.

[@TheMrtummnus]

I love that the grey circle was in the puzzle just as a troll - making everyone assume it was hiding a pesky yellow line!

[@biaberg3448]

I love logical puzzles like this one. Far better than the puzzles with a lot of math and “If - then” thinking.

[@HolyChez]

27 minutes for me. Once I figured out how the line constraints limited the options, it flowed nicely.

[@mlahut]

Once box 9 has been broken into, the pink line can be resolved from that point. If the pink line were to contain a 3, then it would need to use a 3 in either box 2 or box 5, and in both cases those 3s are booked to be on the yellow line already.

[@bothieGMX]

This video was a premiere for me: After some years now, first Sudoku on this channel I solved faster than Simon (me: 48:43, Simon: ~55:47).

[@ericpraline1302]

There was some intriguing logic in this puzzle, part of which I got but I was too stupid to spot the corollary for ages. Of course Simon clocked it immediately!

[@jurgnobs1308]

haha how did I know Simon likes spiders? or at least doesnt want to hurt them

[@mcleanmachut]

Taron is animal genus for those conical sea snails, a reference to the repeating/reducing nature of the lines.

[@Muhh1982]

59:14 you can't put a 3 in the pink line because both 3s in boxes 2 and 5 belong to the yellow line.

[@trudain]

I watched to 30' to when I finally understood the consequences of the rules. Then finished in 21 mins. Thanks!

[@srwapo]

42:40, I got that the lines had to be one longer for every box they went into and that they could only go into each box once, but I somehow didn't use that knowledge to figure out that the yellow line had to end in box 1, so I peaked at the video. I also made some horrible mistakes, using bad logic to figure out how many cells the line had to hit in several boxes, but somehow my pencil marks didn't lead to bad deductions and when I saw how the lines actually went, I was able to fix the mistakes without actually typing in any digits incorrectly. I'm shocked by that.

[@adipy8912]

35:35 Knowledgebomb: 35 minutes is not the same as 22 minutes

[@steve470]

He obviously thinks very hard about birthdays, livestream plugs, etc. 😉

[@SasukeCrafter]

25m 6s solve. The implications of this ruleset are such a novel idea I loved seeing this puzzle unfold. Im a little sad that this ruleset might be limited in how many good puzzles can be constructed from it without feeling repetitive, but brilliant fun nonetheless!

[@LarkyLuna]

The 7 and 8 length yellow lines take from 8 up to 2 at least, so the 3s and 4s on their boxes are taken before the pink like can do anything, so its biggest digit is 2