A Miracle Thermometer Sudoku

** TODAY'S PUZZLE **
Some recommendations are so fulsome that we obey them instantly! Marty Sears told us this sudoku by Michael Lefkowitz ("juggler") was a stone cold classic and so we had a go. You'd be correct in thinking that the grid seems rather empty and the rules somewhat open-ended but there is a unique solution and it is humanly-findable!!
(During editing we noticed a small logical error very late in the video (58mins) right as Simon says "That's the droid we're looking for" (it turned out it wasn't the droid after all!!). The puzzle is virtually solved at this point but, in case anyone is interested, the easy way to continue without the error is to notice that r1c7 could not be a 5 at this point as it could not attach to a 6. No droids were harmed as a result of this error but Simon apologises nevertheless.)
Play the puzzle at the following link:
sudokupad.app/bnq909vwgm
Rules:
Fill the grid with the numbers 1-7 such that no number repeats in a row or column. The grid is completely covered by thermometer lines. You cannot see them because they are invisible. Nevertheless, numbers increase along thermometers starting from their bulb end. Thermometer lines move orthogonally and diagonally, and do not cross themselves or other thermometers, except at the white diamonds, where they must cross themselves or another thermometer. Thermometers do not pass through Xs. Thermometer bulbs do not touch each other, even diagonally. All thermometers have the same length.
You can find Delia's recipe for Chocolate Guinness Cake here (but remember: no walnuts and double the icing):
www.deliaonline.com/recipes/o...
** ROLLER COASTER NURIKABE by the_cogito **
Simon's 2.5 hour solve of Roller Coaster Nurikabe by the_cogito is now available on Patreon. Check the puzzle out at the link below:
logic-masters.de/Raetselporta...
Also there's our January Patreon competition: All The Fun Of The Fair featuring sudokus themed around fairground rides!
Join us on Patreon (for as little as $2):
/ crackingthecryptic
** KICKSTARTER - DELIVERED 24 December 2023 **
Our Kickstarter has finished and we've created the finished novella/puzzle collection for you. It was delivered on Christmas Eve so check your inboxes!
www.kickstarter.com/projects/...
** ORDER OUR NEW BOOK AT THIS LINK **
If you would like to get hold of a physical copy of our new book, Cracking The Cryptic's Greatest Hits - Volume 2, there is still a chance. Visit the link below:
bean-games.pledgemanager.com/...
▶ SUDOKU PAD - Use Our Software For Your Puzzles ◀
You can input classic sudoku puzzles into our software and help support Sven, the programmer responsible for the wonderful user interface we all use to play these puzzles everyday. The app also comes with 12 handmade puzzles from us:
iOS:
apps.apple.com/us/app/svens-s...
Steam: store.steampowered.com/app/17...
Android:
play.google.com/store/apps/de...
ALSO on Amazon: Search for “SudokuPad”
▶ Contents Of This Video ◀
0:00 Theme music & Puzzle intro
1:45 Magic Square Scrabblegram
3:26 Watercoloured Simon
3:41 The World's Best Chocolate Cake?
4:34 Broken Secrets by TotallyNormalCat
5:09 Happy Birthdays
7:35 Rules
10:57 Start of Solve: Let's Get Cracking
▶ Contact Us ◀
Twitter: @Cracking The Cryptic
email: [e-mail]
Our PO Box address:
Simon Anthony & Mark Goodliffe
Box 102
56 Gloucester Road
London
SW7 4UB
(Please note to use our real names rather than 'Cracking The Cryptic'.)
**************************************************************
▶ CTC FAN DISCORD SERVER◀
/ discord
▶ OUR BACK CATALOGUE - ALL CATEGORISED WITH LINKS!◀
tinyurl.com/CTCCatalogue
▶ NEW CRACKING THE CRYPTIC MERCHANDISE◀
teespring.com/en-GB/stores/cr...
▶SEND US PUZZLES TO SOLVE/CONTACT US◀
[e-mail]
▶FOLLOW US◀
Twitter: @crypticcracking
Instagram (for how to solve daily clues from The Times): crackingthe...
We also post the Wordle In A Minute videos on TikTok.

[@thejuggler42]

I'm so glad you enjoyed my puzzle! It's certainly an honor to be in a CTC video, I really am chuffed to bits. I have more in the works that take this concept even further, so stay tuned. :)

[@matiasrustagte]

I solved this problem on my own and got a different solution. It doesn't break the rules, as far as I can see. 6715432 5326741 4172365 2431576 3254617 1647253 7563124 i.imgur.com/B1z7ObC.png

[@martysears]

@@matiasrustagte do you have a 5-6 and a 1-2 that cross eachother illegally on the border between rows 5 and 6 / columns 5 and 6?

[@matiasrustagte]

@@martysearsNow I see :)

[@MaxHaydenChiz]

Very cool puzzle concept. Looking forward to whatever you make next.

[@Paolo_De_Leva]

I solved your *magnificent* puzzle quite easily by using the concept of *broken thermometers,* as defined by KH Schneider, to inject maximum power into one of my favourite solving techniques: *placeholder digits.* I explained my approach in a separate comment. See also my playlist for examples of puzzles solved with placeholders by Simon and Mark. I am curious to know whether you used placeholders to test your puzzle while you were building it. I know that some constructors do use them, but I think this technique is *widely underestimated* and some might not appreciate its elegance and power. Some may even be irrationally biased against it. For instance, one of them does not like it because, in his opinion, it "does not preserve ambiguity." On the contrary, they preserve it because they are just symbols, each representing two or four different digits. They are equivalent to letters or colours, except that for most people (including Simon and Mark) they are easier to scan and interpret. They are not a "guess"❗ Even Simon and Mark rarely use placeholders, although they know how to use them properly (see *The first flurry of snow in a sudoku,* for instance). Dozens of times in the past they got heavily bogged down due to their intricate notation, that I was able to simplify by means of placeholders.

[@theodosiskostouros2974]

Loved the puzzle! Absolutely brilliant! However i think the reasoning that allowed you to complete the final stretch of the puzzle at 58:19 doesn't account for the fact that the 6 in R4C6 could get a 5 from R4C5. I don't really know what the correct next step would be in that part of the puzzle but I'm confident you would have figured it out if you gave it a bit more thought. Very enjoyable watch as always!

[@chrissolnordal9421]

The correct step at that point would have been to notice that r1c7 could not be a 5, because it would need to access both a 4 and a 6 if it was, and that was not possible. Hence r1c7 is a 2, and r3c7 is a 5 (as Simon serendipitously deduced).

[@bobfish7699]

@58:20 - Lucky guess - There was another 5 to the left..

[@rainerzufall42]

But where does the 5 on the right get her 6 from? Isn't possible!

[@The_ElegyOfSins]

​@@rainerzufall42 that is correct, Simon did phrase it incorrectly though, as he argued the 6 needs a 5 as he drew the line. It confused me too, but may only have been a brain-mouth miscommunication

[@FleckerMan]

Good catch; From the description: "(During editing we noticed a small logical error very late in the video (58mins) right as Simon says "That's the droid we're looking for" (it turned out it wasn't the droid after all!!). The puzzle is virtually solved at this point but, in case anyone is interested, the easy way to continue without the error is to notice that r1c8 could not be a 5 at this point as it could not attach to a 6. No droids were harmed as a result of this error but Simon apologises nevertheless.)"

[@rainerzufall42]

@@FleckerMan Correct. r1c8 is the next step...

[@vanciervancier]

bless@@FleckerMan

[@monkeyunit4533]

Love it when the puzzle takes a certain direction and I'm like: Simon will pull out his lovely box of crayons and start colouring any minute now

[@freya28733]

I loved this puzzle, I made the colloring less confusing by just having 2 collors, 1 for low(123) and one for high(567). It also helped make some deductions easier because i knew there had to be 3 of each collor in every row and column

[@rainerzufall42]

A faster argument for not having 49 thermos, is, that the 49 bulbs have to be separated, which is not possible!

[@Censeo]

Also the grid is completely covered by termos according to the rules. Well then all numbers in the grid are in thermos 😅

[@rainerzufall42]

@@Censeo But that would still be possible without that rule. 7 thermos with 1, 7 with 2, etc. Rules, I should say, bulb rule and crossing rule!

[@Censeo]

@@rainerzufall42 yeah, im dense this evening, lol

[@dNIGHTCROWLERb]

Another deduction could be that you can only get 3 bulbs in the middle 5x5 so you need 4 more, therefore corners can't be bulbs, then you can see blue cannot be 1

[@HunterJE]

Also at a metagaming/uniqueness level, given that all the other rules besides the latin square constraint just tell how the thermos can move between cells a grid full of one-cell thermos surely would not result in a uniquely solveable puzzle, not enough to count on on its own but definitely a good first sign that something is going to prevent it...

[@martysears]

So pleased that this got featured, these kind of geometry puzzles are my absolute favourite... great solve from Simon too - he did it pretty much exactly the same way I did, but he made fewer mistakes than me! Well done Michael, look forward to more of your puzzles

[@marklindsey1995]

Please don't beat Simon up about the 65 @ 58:18. Simon caught it during edit and put the logical solve path in the video description. Spoiler......R1C7 is limited.

[@rainerzufall42]

Yes, it's also missing a 6 in the environment...

[@andrewbiliter6693]

True, although, funnily enough, the correction in the video description refers to a cell called "r1c8", which is not an existing coordinate in a 7x7 grid.

[@ChavDog-wo2qu]

Simon does have a face that reminds me of football iwant too kick it,

[@patrickdematosribeiro1845]

Thanks for making the comment I was just going to ask about that point.

[@justincrane8825]

That still doesn’t allow you to immediately deduce which 5 the 6 in r6c4 should take though, does it?

[@laincoubert7236]

what i absolutely loved about this was how every single time i expected a line to go a certain way because of other information the line went in a totally different direction (upon other deductions). it's easier to just do sudoku at the very end and then be surprised how all of the lines connect

[@KevFrost]

There's a previous Simon technique that I used: Given that each thermometer is always 1-7, each end must be either 1 or 7 and these are interchangeable if we ignore the bulb rule. So lets assume r1c2 is a 1 as a placeholder, then we can change all 1s to 7s, all 2s to 6s etc when we have the final disposition of thermos when we can see whether the bulb rule makes the "1" end or the "7" end the bulb

[@Paolo_De_Leva]

That's the same technique I used. Very powerful in this case. For the first time in my life, I solved faster than Simon. See my separate comment for details.

[@NettoTakashi]

Where Simon started using colors, I switched to placeholders-- letters A through G, where A and G were 1 and 7, B and F were 2 and 6, etc. This helps keep the relationship between different cells clear-- A is consecutive with B is consecutive with C, etc.-- while keeping track of the fact we don't yet know which ends of the thermos are the bulbs and which are the tips.

[@RoselynTate]

I think my colouring choices made this puzzle a lot easier for me -- I decided to leave 4s white, like Simon, and then I used (slightly simplified for lack of indigo/violet distinction on the colour palette) colours of the rainbow, and then I put them in order -- red, orange yellow for 17, 26, 35 on one side of 4, and then green, blue, purple for 35, 26, 17 on the other side. It turned out quite pretty, and the fact that I happened to make 1s red was a bonus, lol, but having the colours have an order naturally in my brain made this puzzle a lot less head-wrecking -- watching Simon's solve is breaking my brain a little bit simply because I'm struggling more to keep track of which colours go into which. Final time 28:22 for me!

[@inspiringsand123]

Rules: 01:42 Let's Get Cracking: 10:58 Simon's time: 50m17s Puzzle Solved: 1:01:15 What about this video's Top Tier Simarkisms?! Bobbins: 2x (23:35, 46:40) The Secret: 2x (03:22, 04:43) Knowledge Bomb: 1x (20:48) And how about this video's Simarkisms?! Ah: 12x (10:06, 17:49, 21:58, 23:26, 24:09, 28:52, 32:15, 35:18, 45:49, 46:52, 46:52, 57:39) Hang On: 10x (18:40, 18:40, 18:40, 27:47, 28:52, 31:03, 38:00, 52:21) By Sudoku: 8x (21:43, 25:32, 27:34, 47:20, 51:47, 53:31, 53:40) Obviously: 8x (02:56, 10:26, 13:10, 16:34, 28:47, 29:29, 38:40, 39:26) Brilliant: 6x (01:49, 06:19, 41:36, 44:49, 44:51, 1:01:23) Beautiful: 5x (24:38, 24:41, 49:48, 1:02:09, 1:02:23) What Does This Mean?: 5x (09:06, 20:59, 44:22, 44:27, 53:50) Pencil Mark/mark: 5x (16:31, 54:05, 54:30, 54:40, 56:54) Cake!: 5x (03:53, 03:56, 04:13, 04:18, 06:56) Sorry: 4x (06:31, 06:33, 52:27, 55:17) Astonishing: 4x (10:44, 10:47, 10:50, 10:53) Nonsense: 3x (31:03, 31:08, 38:57) Shouting: 3x (01:46, 03:29, 03:39) I've Got It!: 3x (43:17, 43:17) Whoopsie: 3x (43:00, 43:00, 43:00) Bother: 2x (25:34, 57:39) Stuck: 2x (36:08, 36:08) Fascinating: 2x (1:01:28, 1:01:32) Incredible: 2x (02:08, 1:01:34) In Fact: 2x (15:18, 44:08) Unique: 2x (10:45, 1:01:37) Weird: 2x (24:20, 38:47) Apologies: 1x (32:26) Lovely: 1x (07:30) Gorgeous: 1x (47:17) Hypothecate: 1x (38:12) Take a Bow: 1x (1:02:29) Come on Simon: 1x (32:58) Bizarre: 1x (59:22) Unbelievable: 1x (04:00) Most popular number(>9), digit and colour this video: Twenty Six (23 mentions) One (93 mentions) Green (53 mentions) Antithesis Battles: Even (7) - Odd (1) White (3) - Black (1) Column (22) - Row (17) FAQ: Q1: You missed something! A1: That could very well be the case! Human speech can be hard to understand for computers like me! Point out the ones that I missed and maybe I'll learn! Q2: Can you do this for another channel? A2: I've been thinking about that and wrote some code to make that possible. Let me know which channel you think would be a good fit!

[@kalahatze]

The only good bot on KZbin

[@ChavDog-wo2qu]

Patronising cockwomble simple Simon is a bot.😊😊

[@_cheesestraws_]

Tom Lehrer released all his songs into the public domain recently, which was extremely cool of him.

[@ipudisciple]

I don't know why this comment is on this video, but I'm very pleased to find this out, so thank you.

[@_cheesestraws_]

@@ipudisciple a passing comment at the beginning of the video about copyright strikes for singing Tom Lehrer songs. Only barely relevant, but cool enough to be worth mentioning anyway :-)

[@sjm6280]

It may have been nice to use related colors each side of 4, for instance light colors one side of 4 and dark colors the other side of 4

[@RichSmith77]

Or just use two. One for one side of 4, one for the other.

[@KennethBouman]

I used identical colors for 17, 26, 35, then differentiated the sides with corner letters. 17 is either a ‘G’ or an ‘A’ in my designation, but both were red. Then when I found 2 instances of A too close, I knew G must be 1, and the rest of the puzzle fell into line 😊

[@zondebok980]

Yeah, I used 2 colors exactly like this and it was much clearer, imo. Essentially each digit was only high or low, but the ones on the same side of the 4 were all the same

[@Zalaniar]

I thought it would be clever to use red orange yellow for one side, then green blue purple for the other, meeting at a 4 that could just stay white.

[@steve470]

33:33 for me. Cool puzzle! It seems that Simon just decided to do this puzzle on hard mode, forcing himself to remember which colours connect to which. Would have been a lot easier to use placeholder digits, or even just to use two colours for low and high (not knowing initially which is which, but knowing that, say, red 17 connects to red 26 connects to red 35 connects to 4 connects to blue 35 connects to blue 26 connects to blue 17 - which seems way easier than keeping track of six colours and their sequence).

[@kathyjohnson2043]

I expected him to use low /high colors or perhaps a low/high flash

[@stangerrits6712]

He just loves colouring, doesn’t he 😂

[@Anne_Mahoney]

I was expecting him to use different colored lines for different thermos, though that turned out not to be necessary.

[@thejuggler42]

I likewise used a high-low coloring scheme when I was play testing it, and it has exactly the advantages you point out. In Simon's defense, though, his way does make sudoku inferences easier to spot, because you can more quickly see what colors are missing from a row or column.

[@steve470]

@@thejuggler42 Interesting. I find colors very difficult to scan and see what's missing. Maybe it varies from person to person. Either way, that was a really enjoyable puzzle. Thank you for creating and sharing it!

[@adagraves]

Took me over twice as long. Though impressed myself with how quickly I got that the thermos had to be size 7 and that got me a 17 pair in the bottom left almost immediately.

[@Ciiran]

Loved this one. The most difficult puzzle I've ever managed to solve without help from Simon. I went in exactly the same path as him, but much slower since I had to back up and rework several parts when I encountered conflicts.

[@markp7262]

34:17 finish. I colored my pairs as high-4-low instead of individual colors. It was a lot easier for me than remembering which color leads to which color on the thermos. And I missed the logic that Simon found @55:55, but was able to solve without using it fortunately. It clicked with me almost immediately that the puzzle had to have a pair of 7s touching, or it would be able to flip thermos. But I was surprised that it resolved itself as early as it did. Such a fun puzzle, excellent!

[@alli97253]

I did mine as a rainbow so I wouldn't have to remember, just follow the next color in the sequence. Plus that naturally splits high/low into warm/cool colors

[@giladooshlon]

It was quite difficult for me to find the way into this puzzle, but eventually I got there. Very clever setting and satisfying solve (38 minutes).

[@JakeRowland337]

I started this puzzle using colors but what's cool is you can just arbitrarily pick which ends of a thermo are bulbs and which are tips. Then, if you ever end up with two 1's touching, you know you got it backwards and can just swap them all around. Much easier to keep track of for me than colors.

[@jonhansen9622]

What a fun puzzle! I really do enjoy the different size grids that some of these setters come up with, it's a neat change to watch

[@Nevir202]

11:52 also, bulbs can't be adjacent, and by definition a 1 length thermo would be nothing but bulb, so no chance of 49 of them in there.

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

since the start its obvious the number of thermos is 7 because all cells are occupied by thermos

[@Nevir202]

@@_-_-Sipita-_-_ Except that, as Simon said, it could be 49 1 cell thermos, if not for the rule that they can't be adjacent.

[@rainerzufall42]

@@Nevir202 ... or the rule with the crossings... well, my main (simple) argument was also the bulbs!

[@luketurner314]

Was about to comment the exact same thing

[@Sasha-1313]

Helen Arney has an updated version of the periodic table song which includes the elements which *have* been discovered since Lehrer wrote the original.

[@Akatsuki69387]

Wait. Wait. We discovered some since the original was written?

[@duck_is_everywhere]

Greetings from Australia. This is an appropriate puzzle as we have just had a run of days in the high 30s. And some places in the Pilbara have had a few days at about 47C.

[@Anne_Mahoney]

We should get back to the high 30s here in Boston this week -- of course, I mean 30s Fahrenheit. Which is cold. 😺

[@ryaneakins7269]

That's a perfectly cromulent mark at 34:35. None of the thermos can leave the grid.

[@holliambria9870]

This way of coloring is so much clearer than what I did!

[@chocolateboy300]

I finished in 106 minutes. It's pretty incredible how such simple rules combined with a simple grid to produce magic. I really enjoyed this one, but this was hard to think about. I quite enjoyed proving that the bottom left cell had to connect to the r3c2 line. Great Puzzle!

[@bait6652]

Nice ruleset almost faltered with the top region having 2 Es pointing to one F w/o seeing the 2nd F

[@elLooto]

36:10 In todays episode of _Simon does things the hard way_ he declares himself stuck after examining 3 of the 4 lines he has to work with. Hint: 4 is next to 35.

I tried it first the same way as Simon, putting 17, 26, 35 etc. everywhere, trying to color etc. It took like an hour and I gave up. So I reset the board, assumed which one is 1 and which one is 7 (with possibility of flipping if that went wrong), and it was a 14:20 super easy solve. I got 1-7 backwards so I had to flip it halfway, but that's fine. 7 colors are just so much harder to reason about than 7 digits, even if final assignment of those digits is only revealed later.

[@srwapo]

I started doing it your way, then went to pencil marking, eg, 17, 26, but coloring them to show they are different digits. (26 green touches 35 green touches 4 touches 35 purple touches 26 purple) I wish I stayed with just assuming a number and fixing it at the end, honestly. My method made scanning difficult.

@@srwapo It's just so difficult to see anything when a cell has 3+ colors. I'm pretty sure the right way is either assume digits, or use letter tool with ABCDEFG.

[@Paolo_De_Leva]

That is one of my favourite tricks. I used it to solve this puzzle as well, and for the first time in my life I solved faster than Simon. It is called *placeholder digits.* I have been using it successfully for years. I collected CTC videos about that in my playlist. I have an important advice for you: when you start using placeholders, and you realize in the middle of your solve that you must switch to the opposite configuration, the quickest and most enjoyable strategy is delaying the transformation as long as you can. If possible, keep solving for the "wrong" configuration up to the very end. This simplifies the transformation and makes it more enjoyable (you do not have to deal with center and corner pencilmarks). In this case, the "wrong" configuration is perfectly valid if you assume you are using *inverted thermometers,* with *7* in their bulb and *1* in their tip. In other words, *1s* can touch diagonally, while *7s* cannot. At the end, I was able to transform in about 20 seconds, and it was fun❗

[@andrewgrant6516]

Is it a bad time to mention I don't like chocolate cake? Chocolate is already perfection. Using it as an ingredient in anything merely devalues the chocolate more than it enhances the other. Eat chocolate, have a cake, but not both at the same time!

[@RichSmith77]

I like chocolate. I like cake. I see no downside to combining them. 🙂 (Although I can do without icing sugar/frosting).

[@bbgun061]

But then what flavor would your cake be, if it isn’t chocolate?

[@williammorris7279]

​@@bbgun061Real vanilla, or almond, or raspberry, or lime, or pistachio, or coffee, or yuzu, or walnut, or salted caramel, or matcha, or orange, or .......

[@estherwestbroek]

I have the same feeling about strawberry and cake vs strawberry cake... I'm fine with chocolate cake. I guess it's really personal

[@estherwestbroek]

Did it. Took extra long to have 7 different colours and 7 differently coloured lines. It lookes beautiful!

[@dustpan5356]

My method of coloring was red-orange-yellow-4-green-blue-purple. Made it easy to keep track of the order while also having pairs of colors (17 was red-purple, 26 was orange-blue, 35 was yellow-green).

[@zirco77]

I admire how Simon can keep remember what color goes next in his probably-colorblind-friendly scheme. For practical purposes, I just used a much easier "rainbow order" color scheme 😉 Thank you @thejuggler42 for this puzzle, I really enjoyed it and I'm impressed at how people can come up with a new yet simple ruleset. Kudos!

[@derrickstewart8127]

I used coloring cells to keep track of each thermo (sometimes I knew a cell was part of a thermo, but not exactly where in it). I kept track of high/low parity with A and B pencil marking. There are three A's on one side of the 4, and three B's on the other side. I had a much easier go of it than Simon. Instead of trying to keep track of what color is next to what, it's as simple as A's next to A's and B's next to B's.

[@57thorns]

The colouring at the start(?) 28 minutes in would have been much easier using just two colours...

[@longwaytotipperary]

Sorry about the bad weather- I am so ready for springtime! That Guinness cake looks delicious- let us know if you make one!

[@user-ox9oo6wp6d]

I think coloring the digit by polarity, and coloring 4 with a third color would make the puzzle clearer.

[@cmsg77]

Exactly my approach. Indeed used only two colours. No colour required for 4. Though it did make Snyder difficult.

[@AexisRai]

I just used letters mode, and declared the bottom left cell to be A and the one above it to be G, and replaced them all with the correct numbers at the end when I disambiguated which was the bulb.

[@titusadduxas]

1:57:09 - That was very interesting. I realised early on that the bulbs were what was going to disambiguate the thermos, so rather than colour individual pairs I just selected one option and when it crashed quite quickly switched to the other one. It was still very challenging though.

[@frankjiang1857]

Finished in 86:34. I went down the wrong route of trying to find where all 17s 26s 35s and 4 could be. I wish I had started off by coloring boxes and seeing what eliminations those boxes had made. Fun puzzle!

[@Gonzalo_Garcia_]

18:55 for me. What a fantastic puzzle, really loved this one!!

[@pzaan]

This certainly seemed like an opportunity for coloring and placeholders. I settled on separate coloring for each thermo line and then added orange / blue shading for digits above and below 4. The line colors made it easier to see what fragments might connect to each other. (And made the logic assumption in 58:19 stand out.) Loved the puzzle and the way you could progress from the left corners to the rest of the grid.

[@Kinada]

This was fun and only took me a bit over and hour. I expected to get stuck longer but once you get through some of the early deductions you can restrict things pretty well using colors and making sure to label everything.

[@erikpaige2165]

Love this one!❤

[@77kaczka77]

Thanks

[@CrackingTheCryptic]

Thank you so much!

[@KyleBaran90]

The way I remember magic squares is "816" for the top row, which then informs me that even numbers go in the corners, and odd numbers go on the edges. Next is "357" which is the odd numbers increasing in value, then "492". Taken together, the even numbers decrease and the odd numbers increase in *reading order*.

[@crab0traps0now]

I hope "my brain isn't telling me how to do this" does not catch on as a replacement for "I don't know how to do this." I can tolerate Simon using it (or overusing it), but if everyone starts saying it, my brain is telling me that it will drive me up the wall!

[@prahas777]

At 24 minutes I had a talk to the screen moment. Simon looked at 4 in R4C2, but not 4 in R5C1. If he had looked at the latter he would have seen that the puzzle breaks, thereby placing a 4 in the former cell. 🥰

[@anaayoung9142]

After see how you solve it, I thought: I should had tried 😅

[@Akatsuki69387]

This is a neato puzzle. I like this one.

[@MateiGall]

Little mistake at 58:18; the 6 can get the 5 also from r4c5. But the 5 in r1c7 has no 6 near it, so it forces 2 in r1c7 and a 5 in r3c7.

[@ericsjoberg8167]

Indeed, I noticed this too.

[@edwurtle]

it's much easier if you just guess one of the thermos directions, then swap the order when and if the bulb conflict occurs. In other words use the number 7 as a place holder for blue 17 cell and 1 as place holder for the red 17 cell and so forth for 2,3,4,5,6.

[@user-wc5re2ml9j]

I think it'd be less confusing to use the letters ABCDEFG instead of colors. You can leave the determination as to whether they are 1234567 or 7654321 until it's doable, and it's much easier to remember which are consecutive.

[@Rach881101]

59:42 for me. Beautiful puzzle!

[@musicuber2825]

I love how you got lucky at 58:30 with the 5. Instead of figuring out that R1C7 can't be 5, you assumed R4C6 HAD to go NE, disregarding it could possibly go straight W.

[@musicuber2825]

Either Way FANTASTIC solve on such a cool puzzle feature. Congrats on the solve!!

[@six_5000]

77:06 cuz I restarted only to realize I didn’t even need to 😮

[@srwapo]

47:08, but I looked at the video to see why I made a few mistakes. I kept messing up because I drew lines with different colors and kept thinking they couldn't touch solely because of the colors I used not matching. I also made a weird mistake where I had a thermo going from 2 or 6 to 3 or 5 back to 2 or 6 instead of 4 and had to undo like ten minutes of work.

[@jeffreyhall3944]

Ughhhh I incorrectly assumed that 2 thermos crossing at an X HAD to be 2 unique thermos. I didnt consider that it could be one thermo doing a little loop-de-loop

[@rmjarvis]

Rather than coloring, I found it useful to use A-G at the start until it became clear which of A,G was the 1.

[@SomeRandomGuyOnYouTube]

Simon hates letters, though. Only uses them as a last resort.

[@ingvarsuigin609]

Pretty sure that Tom Lehrer's songs are not only not copyrighted, but available for free on his website.

[@difoolgc5211]

Yup en.wikipedia.org/wiki/Tom_Lehrer#All_songs_in_the_public_domain

[@hockeyhacker97]

34:57... How long is it going to take him to notice where the 4 that connects to the orange in the diamond because it can't connect to the 4 that is already there because if it did the thermometer would have to split in two and snake. and so there is a given 3rd 4 that has been there for a very long time at this point.

[@zirco77]

Simon got lucky with a non-logical deduction at 58:13. At that point, the 6 could have taken the 5 right next to it as well. There was instead a rather nice deduction to make with the 7 and 6 x-wings and the crossing to disambiguate.

[@debrawilden1971]

Wonder if any had the following suggestion? And why Simon didn't use it. He started with all those colors, making a very messy grid. Then he didn't remove them once he discovered the associate #. I used the pen tool to draw the thermos in a diff color. (I did repeat colors, but ones that were far from others.) I also colored the 1/7 pairs orange and all else green, like in sandwich sudokus. That made things very easy to see. The puzzle was actually completed and ticked as correct before I entered all the thermos because I used sudoku to fill in all the #s.

[@Pablo360able]

This one took me two and a half hours. So many false starts due to misunderstanding the implications of the bulb rules... I got it eventually, but man. I used one color for all the numbers less than 4, and another for the numbers greater than 4, plus a separate color scheme for the individual lines.

[@Ffxfan197]

On the bright side in this weather... no Maverick! :D

[@cenedrariva4990]

I did most of this one on my own!!

[@Oblivion776]

Where is the thumbnail image from? Google Lens gives me a bunch of tarot and astrology images but none of them are quite right.

[@thefallenarm589]

Did i miss something or do Simon assumes thermos can't branch ?

[@bobh6728]

Good point.

[@jimi02468]

What annoys me is that Simon didn't use grey lines for the thermometers

[@karlmortenlunna2417]

Really fun puzzle. 14:40 for me today

[@Ciiran]

58:10 Isn't it an incorrect deduction that the 6 in r4c6 must go to a 5 in r3c7? He mentioned earlier that there is a 5 in r4c5.

[@barrynicholson5804]

About the Cake. I'm Kate's Dad and over the years I have been lucky enough to eat lots and lots of The Cake.😊😊😊😊

[@dindul3843]

toward the end i think you made an assumption that r4c6 can only get the 5 from r3c7 when it could have got a 5 from r4c5 maybe thats a small mistake but you still got the correct solve

[@markcoburn8269]

The mistake was spotted prior to the video being posted. There's a comment about it in the video description.

[@Manigo1743]

With 49 thermometers it is pretty hard for the bulbs not to touch each other.

[@Wakaflockabach]

He is so much better than me, yet when I see something that he doesn't I am screaming at the TV until he sees it lol

[@pairot01]

You actuqlly onoy needed 2 colors. Low side vs high side of number 4.

[@bobh6728]

The apology is inaccurate. It says the way to solve at 58:00 is to realize that r1c8 cannot be a five. That is very obvious from the beginning since r1c8 is not even in the puzzle since it is a 7x7 grid! Apology accepted anyways.

[@woif00]

Simon got lucky at 58:15 since he missed that the 6 could connect to the 5 to the left of it

[@woif00]

The way to proceed from there is to prove R3C4 can't be 2 (if it were R3C5 is 3 by thermo necessity and R3C6 would also have to be 3 by sodoku chain R2C4=6 - R2C5=4 - R2C6=7 - R3C6=3) Then there's only one path left for the 1 in R4C4

[@merkdirwas]

the colors make a huge mess- it would be enough to keep track of the lower and the higher digest ... like warm / cold which you use regulary - that would make the following much easier and the deductions more obvious - just my thaught - maybe next time :) thx a lot

[@Hagriddler]

Too bad Simon didn't consider 1 big 49 cell thermo 😂

[@thescrewfly]

Simon's cake baking channel when?

[@mstmar]

a small mistake at 55:10. simon says that the only place r4c6 can get a 5 is in r3c7, but it can also get the one in r4c5. to finish up the puzzle, you need to ask where the 1 in r4 gets its 2. it has to be in the start of row 3, and so r3c7 is not a 2 and so is a 5. that 5 can only connect to the 6 in r4c6, which proves that simon was correct all along

[@RichSmith77]

The mistake was spotted prior to the video being posted. There's a comment about it in the video description.

[@adamheywood113]

it could also be one thermometer of length 49....

[@rainerzufall42]

I think, in this case "numbers between 1 and 7" implied natural numbers. Thus no 49 length thermo here...

[@isaacthek]

That you also love Tom Lehrer only adds to my high esteem of you.

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

42:25 for me. very fun

[@SekGuy]

I'm really bugged that Simon isn't using the default colour palette

[@gianlumndrll5999]

What does it means to be hatched lol?

[@yanturgeon1384]

Happy birthday anyone it is

[@trudain]

45 mins for me. I used letters A to G until the sequence was clear.

[@pendax]

There is a bug in the pen tool. Does anyone know where it can be reported? (drawn line can change color when they shouldn't)

[@hipparchos]

Another possibility would be one thermometer 49 cells long😂

[@rainerzufall42]

... with 49 different digits between 1 and 7...

[@MrCoxmic]

not with the x restriction

[@rainerzufall42]

@@MrCoxmic That's another restriction...

[@srwapo]

@@rainerzufall42 (thinking about the recent puzzle with different NUMBERS, not digits...)

[@rainerzufall42]

@@srwapo So was I! But then I checked, that it's just 1, 2, 3, 4, 5, 6, and 7 here, not PI or E...

[@Pritchie45]

55 minutes

[@laranjacomacerola__]

BRAZIL MENTIONED!!!

[@theredstoneengineer6934]

74:28 for me

[@Paolo_De_Leva]

For the first time in my life I solved a few minutes faster than Simon (45:33). Typically I am much slower. Guess what was my trick❗ Actually, I am impressed that Simon could solve so "quickly" without using my trick.

[@thejuggler42]

I'm curious what your trick was!

[@Paolo_De_Leva]

​@@thejuggler42 ...I recently explained it on Discord as well...

[@Paolo_De_Leva]

@@thejuggler42 Basically, I used the concept of *broken thermometers,* as defined by KH Schneider, to inject maximum power into one of my favourite solving techniques: *placeholder digits.* See also my playlist.

[@Paolo_De_Leva]

I have been learning and writing about *placehoder* digits since Mark amazed me by using them in his video *Mmm ... muffins,* three years ago. They are applyable to many different kinds of puzzles (e.g. rotationally/mirror symmetric or based on German wispers, renbans, nabners, modular lines...). They may hugely improve your scanning by simplifying your *pencilmarks.* For instance, in a cell along a *German whisper line* you can use *2-3* instead of *2-3-7-8,* or *4* instead of *4-6.* In this case, you can use single digits *1, 2, 3, 4, 5, 6, 7* as symbols to represent a sequence of bi-value cells, along a thermo *1-7, 2-6, 3-5, 4, 3-5, 2-6, 1-7* In other words, *1* and *7* represent *blue 1-7* and *red 1-7* Similarly, *2* and *6* represent the two flavors of *2-6* *3* and *5* represent the two flavours of *3-5.* *4* represents itself. You can also use letter placeholders (A, B, C, 4, E, F, G) for the same purpose, but for most people (including Mark and Simon) they are significantly harder to scan and deal with.

[@Paolo_De_Leva]

I have an important advice for you: whenever you use *placeholders,* and you realize in the middle of your solve that you must switch to the opposite configuration, the *quickest* and *most enjoyable* strategy is delaying that transformation as long as you can. If possible, keep solving for the "wrong" configuration up to the very end. This simplifies the transformation and makes it more enjoyable (you do not have to deal with center and corner pencilmarks). In this puzzle, the "wrong" configuration is perfectly valid if you assume you are using *inverted thermometers,* with *7* in their bulb and *1* in their tip. In other words, *1s* can touch diagonally, while *7s* cannot. I solved this way and at the end, I was able to transform in about 20 seconds, using: _y_ = 7 - _x_ It was fun❗

[@ellaenchanted2399]

First 10?

[@MaierFlorian]

Unfortunately, Simon makes the assumption at the end that the 6 needs a 5 and only gets it from r3c7, but it could as well get it from r4c5. I'm sure there is a way to prove the one he did not use wrong but he missed to do so :(

[@enchanter1520]

At 58:10 if anyone is wondering.

[@CrackingTheCryptic]

This is mentioned in the Video Description today. We noticed this one during editing. The easy way to continue at this point is to note that r1c7 cannot be a 5 now as there is no 6 it could attach to. And this tidies it up. Apologies.

[@rainerzufall42]

That 5 in r3c7 needs a 6, and this was the only one available!

[@MissKapanadze]

@@rainerzufall42that 5 only gets placed logically by considering r1c7, otherwise it could be 25

[@rainerzufall42]

@@MissKapanadze But r1c7 could not be a 5 (so it's a 2), because it's lacking a 6 in the "8-environment" (i.e. r1c6, r2c6, and r2c7)...

[@RoderickEtheria]

Finish coloring!

[@RichSmith77]

I'm not sure I can remember a past video where Simon didn't finish the colouring.

[@RedBarchetta2019]

I don't know - maybe the special ruleset that says place the digits 1-7 rules out 49 1-cell thermometers? Classic Simon. Loves to ignore special rulesets to prove things that don't need proven.

[@yapayzeka3183]

You can put any number in a 1-cell thermo

[@MissKapanadze]

How does that ruleset rule out 49 1-cell thermos? 😂 you could absolutely have 49 1-cell thermos with digits 1-7 in them. That’s not even a special ruleset, that’s just normal sudoku. The only way to disprove that is to consider the following: 1) in 49 1-cell thermos, all the cells would be bulbs and would touch therefore breaking the rule about the bulbs 2) the diamonds need a crossing and 1-cell thermos can’t cross Either of the above rules disproves that option, but otherwise it’s a possibility and would break the puzzle due to trivial solutions.

[@angelmendez-rivera351]

This comment belongs in r/ConfidentlyIncorrect.

[@RichSmith77]

​@@angelmendez-rivera351He did start his comment with "I don't know..."

[@bbgun061]

@@MissKapanadze a one cell thermo isn’t a thermo at all. By definition, a thermo needs two digits in order for there to be an increase.