What’s missed here is that the enigmas were different for a very practical reason. That being that army units sent multiple messages a day however, naval ships at most sent very few messages a day because they had to be quiet because every message you allows for direction finding apparatus technologies to locate your ship, submarine etc. so individual fleet units sent very few messages a day. The Navy operators had more time for a more cumbersome system because they weren’t having continually do this for two or three messages an hour or a daily situation report they could in fact say “we are in this sector this is what we’re seeing” and that would be it. They were very short messages in general.
I was imagining that at least part of the reason the navy had additional equipment like the three extra wheels to choose from, the big codebooks etc, was that weight isn't as much of a limiting factor on a ship. Whereas the army potentially had to have a method that could be used in lightweight mobile command posts that could be easily set up or moved at a moment's notice, perhaps even by foot if need be. But your idea makes sense.
This was not available until later in the war. One of the latest fat electrician vids covers it a bit, th-cam.com/video/-BFVAZYGQCU/w-d-xo.htmlsi=_ZWq9-geNT2mKLGO
One time codebooks are the only really secure method, even today. That's entirely unbreakable, and it can ALSO be combined with further security, so even if the codebook is captured, it's still likely to be useless.
@@fuzzywzhe There are codes and there are ciphers. The Onetimepad was a code while Enigma was a cipher, the mechanism gave predictable results, The difficulty was to find the starting setup. That starting setup could be considered as a code. But it remaind set for a day. The internal wiring of each wheel was simply a mechanical cipher and it didn't change.
So if I'm ever drafted into a war I don't care for, I'll be writing "nothing to report" to my command every day and the enemy will leave me alone. Brilliant survival strategy! 😄
My favourite story is the guy was asked to send a fake message, and decided to just press L like 200 times and the code breakers immediately saw the message and were like 'there is no L anywhere in this message, maybe the plaintext is all Ls' and solved the settings basically straight away
10:13: I love how both of them picked "random" letters and inadvertently spelled out words (in the case of wam, more of an onomatopoeia word), proof that humans are not good at randomness. Hence why for cryptographic procedures, we use things like background radiation, videos of lava lamps (cloudflare famously uses this), etc.
I mean, that could also just mean that a significant frcaction of 3-letter grams are also words, thus making it likely you'd pick words. (Not arguing we aren't trash at randomness tho, we totally are.)
I'm not really sure that is a good proof of why humans are bad at randomness. True randomness will often contain coherent messages. If you tasked humans to randomly choose a string of letters, you will get letters that specifically do not make an actual word far, far more often than a truly random process would. Because picking b, then o, then m, then b, it just doesn't feel random to a human.
Alan Turing was of course a genius and played a huge part in Bletchley but he wasn’t the only one. His story takes more prominence probably because of his situation and suicide after the war but there were others, particularly Gordon Welchman who played just as crucial a role in deciphering but his story is relatively unknown.
As you say there were others who played a part. I would like mention Tommy Flowers while not a code breaker per sey built the Colossus and subsequent machines.
Welchman's biggest contribution was probably (radio) traffic analysis, which aided the prediction of enemy intentions without needing to decipher any messages at all. He also introduced an important modification to the bombe machine that speeded up encryption significantly.
Professor Brinley Newton-John was there, being MI5 and fluent in Deutch, he was working on the Deutch language side of the decoding, as well as interrogating the captured Rudolf Hess, and his daughters would go on to be famous too, eg Olivia Newton John of movies Grease and Zanadu, and their songs and other songs too.
@@RWBHereUnfortunately, the information you provide is incorrect. John von Neumann was not directly involved in breaking the enigma at any point. He had previously done theoretical work that provided part of the framework upon which Turing was working. But Turing was amazing in that he had a phenomenal ability to pick out patterns in all of these text messages. And his ability to pick out and recognize and then match patterns was a significant aspect of Turing's work.
I read the memoires of one of the Polish codebreakers. It was so "out there" that I initially found it hard to believe (but I do now). The codebreakers used a sort of early "social engineering". It wasn't "heil hi***er", because Enigma predates the infamous Bavarian, but "wer, was, wo", German for "who, what, where". They assumed that Germans, for a test message, would use a question. Why? I don't know, I wasn't educated in a German high school, unlike the codebreakers.
If you noted this, I must've missed it. The Ringstellung isn't the initial setting of the wheels, it's the relative position of the letter to the wiring. You can turn the ring so that the letters signify a completely different wiring combination, as well as the notch for incrementing the next wheel being at a different location.
@@NuclearCraftMod It's been a long time, but I seem to recall the sender selected three letters and sent them plain text. There was a step or two after that to determine the start positions for the message itself.
The ring settings (Ringstellung) are part of the daily code. The wheel setting is initalized to the first three letters of the message (send in clear). Then the next six letters are decoded, the should be on the form ABCABC is everything is correct. Then you set your rings to ABC and decode the rest of your message. (The "long time ago" is around 6:10 in the video by the way.) edit: As they said in the video, repetition of the three letters for the start setting was dropped in 1940.
@57thorns one 4000 character message (might have been in lorenz cypher) was lost in transmission, the operator reset his machine to the same settings and resent with some errors and some shortening, British got both messages and were able to work out the entire operation of the system
The Navy Enigma was the same as the Army one (M3) up until October 1941. At that point the Navy added one more rotor to the machine for a total of 4, calling it M4. The first 3 rotors were still selected from the same pool of 8 as on the M3, and the 4th extra rotor was chosen from two special rotors.
As far as I know, this 4th rotor was "reverser". In Army reverser was fixed, in Navy it rotates. In one position it provided the same connection as Army reverser, so they have compatibility when needed.
@@johnfox2483 Not quite. The _reflector_ from the M3s was made smaller (thin reflector) and another rotor position was added, with a choice of two (Beta or Gamma). Unlike the first three rotors, the fourth rotor wasn't incrementing. To be honest, I don't know if the M4 was compatible with the M3 or not.
And - if the message was short only the first rotor was employed, as the rotor clicked round with every character, clicking the 2nd rotor after one revolution...and so on for the other rotors. Only very long (and hence very few) messages were long enough to activate the 4th rotor. So in the main (on the main😊?) the 4th rotor rarely had any significance
@@Pippins666 Not true. The longer the message, the higher the chance it would increment the 4th rotor, but any message of any length can increment the 4th rotor depending on the initial settings.
@@Pippins666 i have read somewhere, that each rotor had few special notches, which caused next rotor to rotate one position. So second rotor rotates few times during one revolution of first rotor.
I remember seeing James Grime's original enigma-related vide on numberphile and thinking that Hitler's first name was actually "Heil" (I was very young).
A friend's grandfather (ww2 veteran) had a bulldog called Hitler when I was a kid. We'd go round there for lunch on occasion and I never got the joke of him saying, "Heel Hitler!" to the dog until I was much older.
Frankly... I find the most irritating thing about the film 'The Imitation Game', was the fact it ignored Turing's family's objections to the verdict of the inquest into his death. They strenuously denied the verdict of 'suicide'.
@@LMB222 Since the nonce was to setup the machine and could be sent in plaintext, they could have used D20 filled with letters and use 6 letter initialization code to have 20^6 different nonces for every day to use in messages. Of course, due birthday paradox there would still be 50% change for at least collision after around 8000 messages so maybe make the nonce 8 random characters (no repetition of the nonce in encrypted form but feel free to send it twice in plaintext) and it would take 160000 messages per day to have a random collision. Without a single nonce collision, breaking the encryption would have been really really hard.
I can recommend _The Hut 6 Story_ by Gordon Welchman, it includes a lot of material about the people other than Turing. It did cause a bit of a stink when it was first published and really cost Welchman his security clearance after he moved over the pond.
But it was neither the endphrase, nor that guy on a lonely island. Mostly it was weatherreports from various known stations that helped. The British could observe the weather themselves and guess the text.
IIRC, weather reports were also used because they all began with "Wetterbericht" (a pretty long word, good for stringmatching) and followed a strict syntax
That's what I remember from my reading on the subject too. One of the reasons that weather reports were so useful is that the same report was sent to each branch of the German military. From what I recall, the Luftwaffe had the most lax encryption protocols and were the easiest to crack. Once that was done, it gave them a route into breaking the Wehrmacht and Kriegsmarine codes even though their encryption protocols were much more stringent. So, once again, Goering screwed over Doenitz.
Been a while since I read 'The Codebreakers' but another I recall would be they'd get the RAF to attack some penny ante weather boat or remote station not normally high up a target list who would then report that they had been attacked.
The guy wasn't on an island, he was stationed in North Africa on the edge of the Great Sand Sea, he was there just in case the Brits pushed a force through the soft desert sand, which was seen as basically impossible already. As I understand it they knew where he was and sent their LRDG patrols around him specifically to keep up the illusion that nobody could come from the Great Sand Sea
It's the typical human problem of a layman not trusting the encryption algorithm created by knowledgeable experts and adding additional rules, procedures or conventions on top of it to "improve" it, but often ending up just weakening it significally. They did that mistake 80 years ago, and many people are still doing it today. Trust the experts, they know more than you. Don't start adding your own rules to the algorithm.
I think the most delightful thing about this era of cryptography is that they had the math to do incredible security, but not the portable computing devices to implement it. So every code system is a balance between being robust to codebreaking but also being robust to human nature and large numbers of non expert operators.
21:35 But the Poles already had a bomba, designed by Marian Rejewski, on which the bombe was directly based. And the version of the bombe actually built was after an important refinement by Gordon Welchman. So it wasn't just Turing's idea; it was a group effort, like you assumed. The only reason the Poles couldn't crack the beefed up code on their own is that they did not have enough bombas. So when the invasion began, they sent all their work to the British. Later, the British sent their designs to the Americans, who built a ton of these things in Dayton and did most of the codebreaking work of the war. It was more a matter of resources than technique. Americans certainly didn't come up with the idea, and they were only able to break more codes because they had more money to build machines. But the same is true of the British to some extent. Granted, the cryptographic principle on which the bombe worked was more different from the bomba than the principle on which the American machines worked differed from the bombe, but it's a matter of degree. Polish machines could do a version of a known-plaintext attack, and their method worked on its own, with enough machines. I really think the Poles should get most of the credit here.
This just isn’t true. The Polish Bomba only worked on the basis of the key being repeated (as James describes at the beginning) - and was limited to there only being three rotors available. Turing built the British Bombe on a much more flexible basis to deal with expected text snippets being used. The Polish deserve plenty of credit for the early steps they took. But their system wasn’t even capable of dealing with the basic German enigma army protocols at the beginning of the war. They were having to work, to the extent they could, by manual methods. And that’s before you even start on Turing’s work in reverse engineering the German Navy procedures and various other innovations during the war. The Americans obviously had more resources to throw at things, but as ever with World War II, they didn’t get involved until 1942, so missed a huge chunk of it. And they built their Bombes entirely based off the Bombes Turing and his team has built - Turing literally went to Washington to contribute to and review the American efforts. Ultimately the simple fact is that the Polish efforts never intercepted wartime German messages (or even in the direct run up to war). While the British (and later on American) did, and adapted to the four rotor and other later developments. Trying to diminish the massive achievements of Turing and others at Bletchley Park is just unnecessary historical revisionism.
I have one precision to add. The transfer of the Poles' work to the British was not that simple. The British were wary towards the Poles, believing they might be infiltrated by the Germans and trying to confuse them with fake data. The Poles sent their work also to French spies who rapidly recognized its value. But French spies also knew France was close to being invaded. So its the French spies who convinced the British to take it seriously and organized an encounter to transfer a first set of documents, with the French analysis, for the British to check. And transfer the rest once the British accepted it as valid.
@@Zveebospecifically, as I recall, the Polish efforts just brute forced rotor order by having a separate bombe for each possible rotor order and running a message through all six to see what came out making sense. That worked when there were 6 possible orders, but when the possible rotor orders became 60 when the two spare rotors were added in 1938 or 336 when the Navy went up to 8 rotors to choose from in 1939 it became entirely impractical. They absolutely laid a foundation, but they were nowhere near the finished line when the war started and their efforts were slightly derailed.
This just isn’t true, American Bombs (I think Navy and Army having their own design) completely outperformed the British because they didn't use their designs.
You should do a thing on Arne Beurling. The Swedish intelligence service cracked the German code (Siemens & Halske T52 ) machines too and from the material he uncovered Ericsson retro-engineered and manufactured machines to decode the messages.
9:03 The navy Enigma M4 was different from the army and air force Enigma. It had a 4th rotor, with two to choose from, as well as two different reflectors to choose from.
6:21 What we do know is that the start setting is not starting with J or T, not having Q or S in the middle and is not having P or T last. If the receiver is a three wheeler of course.
I had to take a break to go shop groceries, and on the way to the store I realized a muich better way to pick those six letters than the German multi stage codebook. Use a die and three 3x3 letter tables. The tables are calls 12, 34, and 45. Pick one using your first roll. Then roll for the row (12 for the first, 34 for the second, 56 for the third) The roll for the column (again 12 for the first, 34 for the second, 56 for the third) If you get the 27th combination (all three dice rolls are five or six) you reroll from the start. Repeat six times to get all your letters. Speed it up using three colured dice. The tables can be just alphabetically. If you dice (or die) is good, you will get a perfectly random result, and all you need is a decent die, the tables you can write up on the spot.
Once correction: The bigram substitution cipher, applied to random letters, means that there is no longer a three character code sent in the clear. This makes it harder to decode other messages once one has been broken, since we do not know the setting used. The break gives us all the internal setting (ring setting, steckerboard in a later pass after the bomb reduces the cipher to a simple pairwise substitution cipher) and we can decode the message including those six random digits encoded with the bigram cipher. Not sure how that part was broken really. This means you need to apply the bomb machines to every message, instead of just setting up a douzen enigma copies and start decoding.
If anybody still knows anyone who worked at Bletchley Park and has not spoken to Bletchley Park museum about what they were doing at Bletchley Park they are desperate to get the information before these people's knowledge is taken to the grave. Even with the government saying you are free to talk about what went on people who worked there are still not talking or communicating what they did. There are still big gaps and understanding of what happened at Bletchley Park and esp some huts the museum has no information or understanding of what went on there. The family only found out what my nan did in WW2 only after she died becuse they found a letter of thanks and a medal. She was invited to Bletchley Park becuse she won a crossword competition and she ended up in the Japanese code-breaking side of the park.
@@SpencerTwiddy True but the triplet isn't made up by the operator which is the thing the receiver needs to decrypt -- I thought that was made up but it's taken from a codebook. My bad!
Navy vet here. The scene that REALLY got my goat was the one where they worked out where the enemy was - and is right where the allied convoy was about to be - and one of the guys knew his brother was on a destroyer in that convoy, and wanted to warn the convoy. All that HAS to be utter bo££ox! There is no way that any of those data would have been available to code breakers or anybody outside of a select few. He might have known he was on HMS xyz, but not where XYZ was, let alone which convoy he was guarding or where that convoy was. (My aunt knew my father,, her brother, was on HMS Eagle, but not that he left the ship for a medical emergency before she was sunk. The sinking was announced on the news)
If I remember correctly the regular weather reports by the U-Boats were also a massive clue. Since they could triangulate the rough position and thus deduce what they would report from their own reports of the same region. Also weather reports had a clear format.
I don’t understand the relevance of the 3 chosen letters. Was the intention for these to be used to change a setting in enigma? Did they replace the Ground word? What am I missing here?
@@silverghini2629 Thanks for watching! Watch the section again and watch the captions closely. The three chosen letters were the setting for that specific message only. The ground setting was a daily set of letters used to decrypt the message setting. They didn't want to use the ground setting for every message as that would make all the messages for that day easier to crack. So the procedure is: use the ground setting to decrypt the first six letters of the message. Say it becomes RSKRSK. Then you set your machine to RSK, and use that to decrypt the contents of the message. Hope that helps!
I assume one flaw into the naval enigma would be they had to talk to the army or if force at times so there would be some messages that went to both navy and air force units.
@@DanielHarveyDyer I don't think it's praising, or even accidentally praising, said Austrian. But yeah, YT might disagree and I get it from content creator perspective. Charlie Chaplin used the phrase 'HEIL HINKEL' in his movie Dictator. I'd prefer that, but maybe the ever-so-wise algorithm can see through that one too.
The repetition of the key phrases being the weakness that led to breaking the code just underlines a point that another WW2 history made: The Nazi's were in many aspects comically inept in carrying out the most basic of things.
Military communications do follow specific protocol of format. It is just like regular e-mail has recipient and sender mentioned so same applies to military messages even further. Same weakness was present in Allied communications as well so it was not exclusive to Germans. In purely mathematical sense Enigma was practically unbreakable. In practice it was broken through human error and convenience. Here is an example how receiver in three separate messages could be marked. Receiver is captain Klaus Schwulmeister in command of 2. Dinner Company of 3. Food Battalion. Military style: To: Commander of 2. Dinner Company To: Commander of 2. Dinner Company To: Commander of 2. Dinner Company Informal style: This goes to 2. Company commander To idiot who made whole division sick This belongs to soon-former captain Schwulmeister First is way easier to decode as it follows specific style.
OP, your observation is based. For unknown reason, probably The History (Hitler) Channel has played too many fluff pieces on German ingenuity. Fully 1/3 of the American populace believes the N@Z were technologically superior to the Allies. Because of some airplane, or rocket. Von Brauns V2 was a truly great accomplishment. As was the Saturn V which is known to designed under his leadership. Which seems to lend credence to the belief. But at almost every juncture from the Battle of Britton onward, German incompetence is overwhelmingly obvious. They failed to develop a coherent plan of action from before the war. For one: They didn't secure enough oil production or refining capability to wage a war on the rest of the world. In spite of the fact they clearly understood "blitzkrieg" tactics relied heavily on energy.
Fascist regimes tend to be inept like that. The lack of flexibility and adherence to the leader and tradition leads to the culling of actual minds that think.
there was some post-war use of the Enigma, presumably they had some different procedures for using it, but Bletchley was kept secret for a long time, so maybe not...
I wonder if it might be plausible that that one guy watching over a place never attacked might be a veteran of the great war, read and resonated with the book All Quiet on the Western Front, and everyday his messages would be the book title in german "im westen nichts neues"
Did the British ever try cracking those Enigma messages later on, that they hadn't been able to crack originally early on or during the dark periods? Some of them could still have been potentially important during the war or historically interesting afterwards. And how about now with our far more powerful computers and better understanding of the Enigma machines, and the ways they were used?
There are a few places online you can actually play with engima machines, and a collection of messages which were actual messages, so I know they have played with it in the recent history.
Enigma encoding is not considered secure by modern cryptographic standards. It can be decoded in a reasonable amount of time with our computers. However, there are several Engima (Navy 4-rotor) messages that have not been decrypted and there is ongoing work to do so.
@shura0107 Without the starting point, Short messages are really hard to crack. Even with a modern heavy computer, If it's too short it's just not got enough space to guess the pattern.
@@leechowning2712 Not only is there not enough data in the message to reconstruct the original plaintext, it may well be that there are multiple valid plaintexts that could have encoded to the same crypttext with different IVs. That is to say, collisions. At an extreme, some initial settings might encrypt "ILOVEYOURDOG" into "XDKLDWKVQWRK", while another set of initial settings would also encrypt "SENDFEETPICS" into "XDKLDWKVQWRK". When any of the initial settings are valid and they are all equally likely, deciding which one was the original plaintext and which ones are the collisions is half guesswork and half wishcasting.
I was a USAF Cryptographic Account Manager and KL7 operator. If you sat an Enigma operator down in front of my KL7 he would know instantly what he was looking at, a bit more complicated but he would know.
Something that was unclear to me when I first heard of the enigma machine quite a few years ago, is that the machine itself was not a secret, other countries had them also.
8:00 If Germany had distributed D20 dice to every operator and told that operator must throw dice 6 times to get the initial 6 letters (sent without repeation), Allied forces would have never broken the encryption during the WW2.
I've only ever really associated Enigma with a Naval requirement. Anyone who has watched documentaries and movies based on the Enigma codes and machines, will generally only ever seen Naval personnel associated with it so why the pronouncement at 00:28 ?
While Turing is credited for cracking the Enigma, the Swedish mathematician Arne Beurling cracked a more complicated version of the German crypto machines (Siemens & Halske T52), using nothing but pen and paper.
Just an observation that despite the enigma machine doing all that clever encoding and decoding, it still relies on the sender and receiver having an agreed secret set of code/decode words ie keys, for each message. The convenient thing is that the key can be very short. But today, if we can agree on the secret keys then due to technology, length of key isn't really an issue. I can just have a file that for this message maps the first key press to one list of randomly generated numbers, the next keypress to another set of randomly generated numbers and so on up to however many keypresses is more than enough. This is utterly undecipherable ever, without the key. But what we can't do today is rely on the key remaining secret. But if we could, we wouldn't need the clever subsequent encoding and decoding algorithm anyway.
I’ve always wondered how much more secure the army and Air Force enigma protocol would’ve been if they’d handed out 20 sided lettered dies to the operators. Surely this would have eliminated lot of the human error in choosing their three letter codes? Perhaps even more effective than the navy protocol?
Along with the nothing to report guy there was also a guy who always sent a weather report and he always started it Weather Report. They also could guess what the weather might be and use that.
It's written that the German Navy personnel were much more disciplined in following their procedures and training and made fewer mistakes versus their army or Luftwaffe code clerks, where a Bletchley Park person thought of how a code clerk who was hung-over, tired, or bored and not paying attention could make mistakes and learned to search for those mistakes and then decode them (the mistake or subsequent corrective action) made it possible to break the coding via manual methods.
was this the same system the high ranking officials used? I thought I heard they had an entirely more advanced system, that could also transmit automatically. And They built Collossus to break that?
Why the hell would they repeat the secret setting? I know you said, to be sure. But that seems like an enormous flaw. Like, let's assume you received a code book and an enigma machine and were intercepting the messages. You didn't know about the secret setting. But every single message you receive starts with the first three letters repeated, then a bunch of nonsense. I don't think it would have taken any serious problem solving to just try those three letters as the machine setting. Really, it's the typing it out twice that's the problem. It's just a massive hint that the first three letters of every message are important. Why not do something like type a random sentence and use the first, third, and fifth word's first letters as the setting? Since we would use recognizable words there doesn't seem to be a need to type the sentence twice to ensure accuracy. And this would significantly obfuscate the obvious three character setting.
"But every single message you receive starts with the first three letters repeated, then a bunch of nonsense." Possible misunderstanding here -- the message didn't start with three repeated letters followed by nonsense. Those three repeated letters were *encrypted*. So a message might look like: GJAOENFEQOC... GJAOEN decrypts to ABCABC using the ground setting. FEQOC... decrypts to the message content using the ABC setting. So it's not as if the message setting ABC was plain to see at the start of a transmission. Also, it's worth nothing a couple of things: The Nazis were also kind of arrogant and believed the Enigma to be uncrackable, as a code machine in and of itself. As James mentioned, the Nazi forces stopped using this protocol early in the war, as the Polish cracked Enigma by recognising that the first six characters have repititions (i.e. exploiting a flaw in the protocol rather than cracking Enigma itself). After that, they just sent the message setting once. Hope that helps!
@AnotherRoof No I understand that you need the base setting to get the secret code. My argument is that once you have that book, the "secret" setting is so absurdly obvious as to be completely pointless. There are so many simple ways to obfuscate that second setting.
@TheTonyMcD Then I'll refer you to the latter two points and what James said about how capturing codebooks was very rare (and only lasted a month at most). Don't know what else to say -- the Nazis were (thankfully) bad at security it seems
@@TheTonyMcD The starting letters were used to synchronise the sending and receiving machines. At midnight, all machines were set to the same day settings: choice of rotors, ring settings, order of the rotors in the machine and plugboard settings. As time passed, machines would get "out of sync" (i.e. the rotors would move as messages were typed.) Enigma decrypts by typing in the encrypted text on a machine set exactly as the sending machine, i.e. the rotor positions when used to encrypt (because the actual rotor choice, order, ring settings etc were set for everyone and only changed at midnight - except in 1942 for submarines when they changed twice a day) So the receiver has to know what the senders rotor position was when the message was encoded. This is how they did it: choose three letters, e.g. GAF. Set the Enigma rotors to show GAF in the windows and then select three more letters, e.g. TVE, and type them in to get, say, MKP. The sender would set their machine back to TVE and type in the message. At the start of the message the sender would send GAFMKP followed by the encrypted message. The receiver would then set their machine to show GAF and type in MKP to reveal TVE. They would then set their machine to show TVE and type in the rest of the message revealing the clear text. So actually it doesn't really matter per-se about the choice of the initial 6 letters and their randomness, the Bombe would be able to break it anyway. By choosing something less random for your 6 letters, e.g. CILLIE (your girlfriend's name), and using it repeatedly just makes it somewhat easier to break - it becomes easier for the cribsters to decide on a crib to use in breaking the message due to past history of that sender. I say somewhat because although you might recognise the first 6 letters, CILRDF to really be CILLIE (i.e. you know the cleartext of RDF) you still need to know the rotors used, the ring settings, the rotor order and plugboard settings to actually break that message and any others sent on the same network. The Germans cryptologists, who were not stupid, would know that including anything non-random would increase the likelihood of a code being broken, because that's just how it is, so went to lengths for the navy to reduce the randomness. The problem of course is you have a miltary consisting of butchers, bakers and candlestick makers so the process has to be simple to work, the drawback being humans are rubbish at random. Note that the Allies were no better off in this regard and the Germans broke plenty of their codes as well. The reality is in breaking Enigma the Allies didn't actually need non-randomness to do so, it just gave better clues. Of course, if you had the sheet of daily settings and the Germans didn't know, then you could read all messages sent on that network for as long as the sheet lasted - no cryptology work was needed at all.
The Imitation Game - It is a 2014 American period biographical thriller film directed by Morten Tyldum and written by Graham Moore, based on the 1983 biography Alan Turing: The Enigma by Andrew Hodges.
Was't the plans for the bomb presented to British inteligence in July of 1939 by Polish code brakers? They build bomb with 6 stages , however Polish cod brakers handed plans to build bomb with 60 stages. What the diffrence between Polish bomb decoding maschine and Turing bomb decoding maschine? The is a lot of confusion.
Mike from Numberphile did some demos of bruteforcing it th-cam.com/video/RzWB5jL5RX0/w-d-xo.html with some statistical analysis. Colossus seems to have started that type of computer assisted analysis , albeit on the Lorentz cipher
The Poles originally created the Bombe, Turing and Welchman, and a host of others improved on it until we had the monster that is in the NSA Museum in DC.
Why does everyone only ever talk about breaking Enigma? I don’t think I’ve ever heard anyone even mention the inventor of Enigma… even less make a movie about them.
That’s why they screwed up without a pre-code. They would become redundant and start repeating the same code over and over and over again, which made them vulnerable.
If the British didn't have the navy pairs table, would they have to decrypt each message individually? Since they couldn't work out the general day code?
Not having the pairs table means they would have to repeat their work next day. But being able to decrypt one message means they already had their settings of the day.
Did the German engineers who designed and built the Enigma system ever get any credit for their brilliant creation? It was human error/habit which eventually defeated it, not the concept or the performance. I'm British, but I still hugely respect the superb engineering which went into Enigma.
@@strummergr I agree and I'm fascinated by the device itself. However there was one flaw the Allies could exploit, which is that a letter never gets mapped to itself when encrypting, and this turned out to be a massive weakness!
Actually, it wasn't human error that broke it, Enigma was fundamentally flawed. The 26 combinations that couldn't exist out of a possible 159+ quintillion was all that was needed. Human error gave clues to help with obtaining daily settings but the allies were still breaking these even when those errors didn't exist. I suppose in fairness you could argue that human error in putting rotors into the Polish postal system helped a great deal because without knowledge of rotor wirings you are stuck. But day to day operational errors didn't really contribute.
If I ever have stupid lottery money, I'm going to have an Enigma Machine made for me from scratch. (while an actual original would be cool, I'd rather an identical replica because authentic ones belong in museums) Maybe even hire someone to fix the flaws of the device (like a letter can never be itself) and make additional code wheels beyond the 8 that the Navy had.
that would mean capturing a navy operator alive (unlikely), identifying that person as an operator and then getting him to talk. managing one of those? unlikely but possible, doing all three? i doubt that.
But what was the likelihood that by the time the other side has decoded a message, it was already too late and the others had already carried out that plan, so it was too late to intervene or whatever?
At the start of the war this was the case - essentially, the work being done was useful in identifying in how it could be better exploited by breaking messages even weeks old. However, once the Bombes came on stream in early 1940 and the codebreaking became more streamlined, Enigma networks were more rapidly broken. It wasn't long before most Enigma networks were broken every single day and in many cases in the first 8 hour shift starting at midnight!
@TheLucnicLord Essentially, the "secret" was sent encrypted with the message along with the clue to unencrypt it. I explain in a comment above how it works.
@@kevingeier3385 There are three rotors in the machine but five options for which to use. Since we're talking about the start of the war we didn't mention the fourth rotor added later. Hope that helps!
How and why are different words and sometimes they are inappropriately used. The hosts are discussing HOW they differ, not WHY. Why they were made differently? Only those who chose to do so can answer that and most of them are now dead.
The fact is that Turing really acted on a thought that many knew. Codes and other information needs to be processed and the resulting information discovered very quickly. Although many of the enigma codes were being broken and other information about the codes was known, the solution of the codes came after the textual information they contained was uselessly old. German businesses were using the Enigma device before the war to discourage corporate espionage. Polish mathematicians were decoding some of the messages back then. But it would have been the same result finding out after the information was useful. Turing had designed a device he thought would work. But it didn't. Someone at Bletchley Park knew the other piece of the puzzle. He or she knew about a British Telecom technician named Fellows (I believe). Once that person knew what Turing was trying to achieve he went home and returned a week later with a schematic for the circuitry that did what Turning said was needed and at an blinding speed. So it wasn't a single person who solved the problem, it was a single person, Turing, that put them on the right track. But, even more importantly, Turning set humanity on a whole new road where today we rely on electronic circuitry to solve many of the question that can be asked. George Bernard Shaw ( in "Back to Methuselah") had it right when he had the snake in the garden of Eden say, "Some men look at things that are and wonder why? I think of things that never were and say, "Why not." (or words to that effect)(also, I believe, an expression used frequently by Robert F. Kennedy, who also likely got it from the same place, Shaw being an Irish author and all.)
If you believe Peter Wright (autobiography: spy catcher) there was more to it than that. He implied Turing talked too much... and to the wrong people, although he never clearly stated exactly whatever happened
@@andyj2106 We know. James is talking about *three substitutions* at around 20:10. A -> 1 -> 2 -> 3. 1 is not A, 2 is not 1, and 3 is not 2. James is saying that there's a 1/26 chance that 3 is A.
@@AnotherRoof I guess your comment isn't very clear to me. I think the OP here is saying that in any particular position, there is a 1:25 chance of a particular letter being lit when a key is pressed. Maths isn't my strongpoint, but I think the 1:26 combination comes from the fact that James is talking about loops and that in any particular position, there is a 1:25 chance of a letter appearing on the lampboard PLUS the key pressed giving 1:26. In other words the probability is derived because it is calculated across the cleartext and the enciphered text? In fact at *any* position in the message there is a 1:26 chance of a letter appearing when including cleartext and enciphered text? Can I surmise that when you say that "a letter can become itself (and did) after three substitutions" that was not a definitive statement and you were referring to the example James used? This is the bit of your comment I'm not clear on, or indeed A->1->2->3. At any given starting position, a loop could occur from position 2 onwards (e.g. using HEIL as the cleartext, H -> E; E -> H would be a possibility) and that at any position after the first key press there is a 1:26 chance of a loop arising?
@@andyj2106 The only time a probability of 1 in 26 or 1 in 25 comes up in the video is in the aforementioned section, so I assume that's what OP is referring to. I think your understanding is correct. If we did find a loop after two keystrokes, then we can use the methodology James discusses. I used three keystrokes because that's the example in the video, but the loops could be any length greater than 1.
@@richardgrego We didn't. James mentioned this when we discuss the army protocol and how the Polish method relied on that. Turing and the others built on their work but cracked it in a way that didn't depend on that protocol and thus could be applied to naval messages etc. Hope this helps!
What’s missed here is that the enigmas were different for a very practical reason. That being that army units sent multiple messages a day however, naval ships at most sent very few messages a day because they had to be quiet because every message you allows for direction finding apparatus technologies to locate your ship, submarine etc. so individual fleet units sent very few messages a day. The Navy operators had more time for a more cumbersome system because they weren’t having continually do this for two or three messages an hour or a daily situation report they could in fact say “we are in this sector this is what we’re seeing” and that would be it. They were very short messages in general.
I was imagining that at least part of the reason the navy had additional equipment like the three extra wheels to choose from, the big codebooks etc, was that weight isn't as much of a limiting factor on a ship. Whereas the army potentially had to have a method that could be used in lightweight mobile command posts that could be easily set up or moved at a moment's notice, perhaps even by foot if need be. But your idea makes sense.
The contents might be,
Ship A position A observed X, leaving out the words "ship position observed". That is padding you don't need.
This was not available until later in the war. One of the latest fat electrician vids covers it a bit,
th-cam.com/video/-BFVAZYGQCU/w-d-xo.htmlsi=_ZWq9-geNT2mKLGO
One time codebooks are the only really secure method, even today. That's entirely unbreakable, and it can ALSO be combined with further security, so even if the codebook is captured, it's still likely to be useless.
@@fuzzywzhe There are codes and there are ciphers.
The Onetimepad was a code while Enigma was a cipher, the mechanism gave predictable results,
The difficulty was to find the starting setup. That starting setup could be considered as a code. But it remaind set for a day.
The internal wiring of each wheel was simply a mechanical cipher and it didn't change.
So if I'm ever drafted into a war I don't care for, I'll be writing "nothing to report" to my command every day and the enemy will leave me alone. Brilliant survival strategy! 😄
Cute, but modern encryption is totally different.
My favourite story is the guy was asked to send a fake message, and decided to just press L like 200 times and the code breakers immediately saw the message and were like 'there is no L anywhere in this message, maybe the plaintext is all Ls' and solved the settings basically straight away
WHOA! That's some high-level applied game-theory 🧐
@@DanielHarveyDyerIn general the inability of a letter to be encoded as itself was a big weakness of the Enigma
@OP …and from the process of sending that same message every day, the enemy will also crack the cipher…? 😅
James Grime is a very engaging fellow and has a very pleasant voice
10:13: I love how both of them picked "random" letters and inadvertently spelled out words (in the case of wam, more of an onomatopoeia word), proof that humans are not good at randomness. Hence why for cryptographic procedures, we use things like background radiation, videos of lava lamps (cloudflare famously uses this), etc.
I mean, that could also just mean that a significant frcaction of 3-letter grams are also words, thus making it likely you'd pick words. (Not arguing we aren't trash at randomness tho, we totally are.)
These were sent in groups of 5 letters, were they not.
I'm not really sure that is a good proof of why humans are bad at randomness. True randomness will often contain coherent messages. If you tasked humans to randomly choose a string of letters, you will get letters that specifically do not make an actual word far, far more often than a truly random process would. Because picking b, then o, then m, then b, it just doesn't feel random to a human.
Alan Turing was of course a genius and played a huge part in Bletchley but he wasn’t the only one. His story takes more prominence probably because of his situation and suicide after the war but there were others, particularly Gordon Welchman who played just as crucial a role in deciphering but his story is relatively unknown.
As you say there were others who played a part. I would like mention Tommy Flowers while not a code breaker per sey built the Colossus and subsequent machines.
Van Neumann was another important character.
Welchman's biggest contribution was probably (radio) traffic analysis, which aided the prediction of enemy intentions without needing to decipher any messages at all. He also introduced an important modification to the bombe machine that speeded up encryption significantly.
Professor Brinley Newton-John was there, being MI5 and fluent in Deutch, he was working on the Deutch language side of the decoding, as well as interrogating the captured Rudolf Hess, and his daughters would go on to be famous too, eg Olivia Newton John of movies Grease and Zanadu, and their songs and other songs too.
@@RWBHereUnfortunately, the information you provide is incorrect. John von Neumann was not directly involved in breaking the enigma at any point. He had previously done theoretical work that provided part of the framework upon which Turing was working. But Turing was amazing in that he had a phenomenal ability to pick out patterns in all of these text messages. And his ability to pick out and recognize and then match patterns was a significant aspect of Turing's work.
I read the memoires of one of the Polish codebreakers. It was so "out there" that I initially found it hard to believe (but I do now).
The codebreakers used a sort of early "social engineering". It wasn't "heil hi***er", because Enigma predates the infamous Bavarian, but "wer, was, wo", German for "who, what, where".
They assumed that Germans, for a test message, would use a question.
Why? I don't know, I wasn't educated in a German high school, unlike the codebreakers.
If you noted this, I must've missed it. The Ringstellung isn't the initial setting of the wheels, it's the relative position of the letter to the wiring. You can turn the ring so that the letters signify a completely different wiring combination, as well as the notch for incrementing the next wheel being at a different location.
Yeah bro that shit is too complicated...
Why were they not just using Telegram or Signal like everybody else?
What were the wheel settings initialised as before the specific message setting was sent?
@@NuclearCraftMod It's been a long time, but I seem to recall the sender selected three letters and sent them plain text. There was a step or two after that to determine the start positions for the message itself.
The ring settings (Ringstellung) are part of the daily code.
The wheel setting is initalized to the first three letters of the message (send in clear).
Then the next six letters are decoded, the should be on the form ABCABC is everything is correct.
Then you set your rings to ABC and decode the rest of your message.
(The "long time ago" is around 6:10 in the video by the way.)
edit: As they said in the video, repetition of the three letters for the start setting was dropped in 1940.
@57thorns one 4000 character message (might have been in lorenz cypher) was lost in transmission, the operator reset his machine to the same settings and resent with some errors and some shortening, British got both messages and were able to work out the entire operation of the system
"Are we the baddies?" is a great message to send and decode. :D
I'm guessing a reference to the Mitchell & Webb skit 😂
Long live the Bumbling Badger of Mediocrity.
The Navy Enigma was the same as the Army one (M3) up until October 1941. At that point the Navy added one more rotor to the machine for a total of 4, calling it M4. The first 3 rotors were still selected from the same pool of 8 as on the M3, and the 4th extra rotor was chosen from two special rotors.
As far as I know, this 4th rotor was "reverser". In Army reverser was fixed, in Navy it rotates. In one position it provided the same connection as Army reverser, so they have compatibility when needed.
@@johnfox2483 Not quite. The _reflector_ from the M3s was made smaller (thin reflector) and another rotor position was added, with a choice of two (Beta or Gamma). Unlike the first three rotors, the fourth rotor wasn't incrementing. To be honest, I don't know if the M4 was compatible with the M3 or not.
And - if the message was short only the first rotor was employed, as the rotor clicked round with every character, clicking the 2nd rotor after one revolution...and so on for the other rotors. Only very long (and hence very few) messages were long enough to activate the 4th rotor. So in the main (on the main😊?) the 4th rotor rarely had any significance
@@Pippins666 Not true. The longer the message, the higher the chance it would increment the 4th rotor, but any message of any length can increment the 4th rotor depending on the initial settings.
@@Pippins666 i have read somewhere, that each rotor had few special notches, which caused next rotor to rotate one position.
So second rotor rotates few times during one revolution of first rotor.
Legendary collab!
Enigmatic you could say.
I love it when James Grime (and other Numberphilers) do crossovers on other TH-cam math(s)-themed content-creators' channels!!
i did not expect to hear james grime say "heil hitler" like 20 times in a row.
I remember seeing James Grime's original enigma-related vide on numberphile and thinking that Hitler's first name was actually "Heil" (I was very young).
A friend's grandfather (ww2 veteran) had a bulldog called Hitler when I was a kid. We'd go round there for lunch on occasion and I never got the joke of him saying, "Heel Hitler!" to the dog until I was much older.
Frankly... I find the most irritating thing about the film 'The Imitation Game', was the fact it ignored Turing's family's objections to the verdict of the inquest into his death. They strenuously denied the verdict of 'suicide'.
The Navy were basically using a password manager prefilled with pseudo-randomly generated passwords, handed down from their higher-ups
There wasn't anything much better back in the 1920's (yes, enigma predates the Nazis )
@@LMB222 Since the nonce was to setup the machine and could be sent in plaintext, they could have used D20 filled with letters and use 6 letter initialization code to have 20^6 different nonces for every day to use in messages. Of course, due birthday paradox there would still be 50% change for at least collision after around 8000 messages so maybe make the nonce 8 random characters (no repetition of the nonce in encrypted form but feel free to send it twice in plaintext) and it would take 160000 messages per day to have a random collision. Without a single nonce collision, breaking the encryption would have been really really hard.
I can recommend _The Hut 6 Story_ by Gordon Welchman, it includes a lot of material about the people other than Turing. It did cause a bit of a stink when it was first published and really cost Welchman his security clearance after he moved over the pond.
Thanks, I learned new things. Knew about the weakness but didn't know about the protocol.
But it was neither the endphrase, nor that guy on a lonely island. Mostly it was weatherreports from various known stations that helped. The British could observe the weather themselves and guess the text.
It wasn't mostly weather reports, it was everything they could use, and then they brute forced it when computers got fast enough.
IIRC, weather reports were also used because they all began with "Wetterbericht" (a pretty long word, good for stringmatching) and followed a strict syntax
That's what I remember from my reading on the subject too. One of the reasons that weather reports were so useful is that the same report was sent to each branch of the German military. From what I recall, the Luftwaffe had the most lax encryption protocols and were the easiest to crack. Once that was done, it gave them a route into breaking the Wehrmacht and Kriegsmarine codes even though their encryption protocols were much more stringent.
So, once again, Goering screwed over Doenitz.
Been a while since I read 'The Codebreakers' but another I recall would be they'd get the RAF to attack some penny ante weather boat or remote station not normally high up a target list who would then report that they had been attacked.
The guy wasn't on an island, he was stationed in North Africa on the edge of the Great Sand Sea, he was there just in case the Brits pushed a force through the soft desert sand, which was seen as basically impossible already. As I understand it they knew where he was and sent their LRDG patrols around him specifically to keep up the illusion that nobody could come from the Great Sand Sea
It's the typical human problem of a layman not trusting the encryption algorithm created by knowledgeable experts and adding additional rules, procedures or conventions on top of it to "improve" it, but often ending up just weakening it significally.
They did that mistake 80 years ago, and many people are still doing it today.
Trust the experts, they know more than you. Don't start adding your own rules to the algorithm.
I think the most delightful thing about this era of cryptography is that they had the math to do incredible security, but not the portable computing devices to implement it. So every code system is a balance between being robust to codebreaking but also being robust to human nature and large numbers of non expert operators.
@@DanielHarveyDyer Hagelin machines were compact with included printer and manual grid systems gave easy non-sequential encoding
I wonder how secure message is if it first "coded" with some easy cipher like pendulum or reverse and then ciphered with actual machine and sent?
Brilliant video. I'm quite good at maths, post grad in engineering, but don't have a clue about this sort of stuff. Alan was a genius.
Math is pure logic.
Code is pure logic.
@@Triple_J.1 there's also an art to it, the best proofs are also elegant, as is the best code.
21:35 But the Poles already had a bomba, designed by Marian Rejewski, on which the bombe was directly based. And the version of the bombe actually built was after an important refinement by Gordon Welchman. So it wasn't just Turing's idea; it was a group effort, like you assumed. The only reason the Poles couldn't crack the beefed up code on their own is that they did not have enough bombas. So when the invasion began, they sent all their work to the British. Later, the British sent their designs to the Americans, who built a ton of these things in Dayton and did most of the codebreaking work of the war. It was more a matter of resources than technique. Americans certainly didn't come up with the idea, and they were only able to break more codes because they had more money to build machines. But the same is true of the British to some extent.
Granted, the cryptographic principle on which the bombe worked was more different from the bomba than the principle on which the American machines worked differed from the bombe, but it's a matter of degree. Polish machines could do a version of a known-plaintext attack, and their method worked on its own, with enough machines. I really think the Poles should get most of the credit here.
This just isn’t true. The Polish Bomba only worked on the basis of the key being repeated (as James describes at the beginning) - and was limited to there only being three rotors available. Turing built the British Bombe on a much more flexible basis to deal with expected text snippets being used.
The Polish deserve plenty of credit for the early steps they took. But their system wasn’t even capable of dealing with the basic German enigma army protocols at the beginning of the war. They were having to work, to the extent they could, by manual methods.
And that’s before you even start on Turing’s work in reverse engineering the German Navy procedures and various other innovations during the war.
The Americans obviously had more resources to throw at things, but as ever with World War II, they didn’t get involved until 1942, so missed a huge chunk of it. And they built their Bombes entirely based off the Bombes Turing and his team has built - Turing literally went to Washington to contribute to and review the American efforts.
Ultimately the simple fact is that the Polish efforts never intercepted wartime German messages (or even in the direct run up to war). While the British (and later on American) did, and adapted to the four rotor and other later developments. Trying to diminish the massive achievements of Turing and others at Bletchley Park is just unnecessary historical revisionism.
It's all just robotic safe cracking
On today's lock picking lawyer, we'll be using this 40's vacuum tube computer to decrypt launch codes...
I have one precision to add. The transfer of the Poles' work to the British was not that simple. The British were wary towards the Poles, believing they might be infiltrated by the Germans and trying to confuse them with fake data. The Poles sent their work also to French spies who rapidly recognized its value. But French spies also knew France was close to being invaded. So its the French spies who convinced the British to take it seriously and organized an encounter to transfer a first set of documents, with the French analysis, for the British to check. And transfer the rest once the British accepted it as valid.
@@Zveebospecifically, as I recall, the Polish efforts just brute forced rotor order by having a separate bombe for each possible rotor order and running a message through all six to see what came out making sense.
That worked when there were 6 possible orders, but when the possible rotor orders became 60 when the two spare rotors were added in 1938 or 336 when the Navy went up to 8 rotors to choose from in 1939 it became entirely impractical.
They absolutely laid a foundation, but they were nowhere near the finished line when the war started and their efforts were slightly derailed.
This just isn’t true, American Bombs (I think Navy and Army having their own design) completely outperformed the British because they didn't use their designs.
You should do a thing on Arne Beurling. The Swedish intelligence service cracked the German code (Siemens & Halske T52 ) machines too and from the material he uncovered Ericsson retro-engineered and manufactured machines to decode the messages.
Super cool to have James join you. I'm sure he appreciates your potential. Either way, good shit. Thank you for making such quality content ❤️
Great video, you hear a lot about the different procedures but i haven't actually heard them described, and the challenges they caused. Thanks!
What British had to do? They had to do second meeting with Poles in 1940 in Paris. Btw. First code broke by the Poles was Kriegsmarine code.
0:24 - somewhere Grace Hopper is feeling very proud of the Navy
9:03 The navy Enigma M4 was different from the army and air force Enigma. It had a 4th rotor, with two to choose from, as well as two different reflectors to choose from.
Yep. This guy conveniently omits a bunch of things.
6:21 What we do know is that the start setting is not starting with J or T, not having Q or S in the middle and is not having P or T last. If the receiver is a three wheeler of course.
Very nice to see Dr. Grime again. I met him in Odense, Denmark, a couple of years ago. Very nice bloke, indeed.
20:43
Couldn't several shifting "compensate" each others?
How can you tell the chance decrease when you add more shiftings?
I had to take a break to go shop groceries, and on the way to the store I realized a muich better way to pick those six letters than the German multi stage codebook.
Use a die and three 3x3 letter tables.
The tables are calls 12, 34, and 45. Pick one using your first roll.
Then roll for the row (12 for the first, 34 for the second, 56 for the third)
The roll for the column (again 12 for the first, 34 for the second, 56 for the third)
If you get the 27th combination (all three dice rolls are five or six) you reroll from the start.
Repeat six times to get all your letters.
Speed it up using three colured dice.
The tables can be just alphabetically.
If you dice (or die) is good, you will get a perfectly random result, and all you need is a decent die, the tables you can write up on the spot.
Once correction:
The bigram substitution cipher, applied to random letters, means that there is no longer a three character code sent in the clear.
This makes it harder to decode other messages once one has been broken, since we do not know the setting used. The break gives us all the internal setting (ring setting, steckerboard in a later pass after the bomb reduces the cipher to a simple pairwise substitution cipher) and we can decode the message including those six random digits encoded with the bigram cipher. Not sure how that part was broken really.
This means you need to apply the bomb machines to every message, instead of just setting up a douzen enigma copies and start decoding.
If anybody still knows anyone who worked at Bletchley Park and has not spoken to Bletchley Park museum about what they were doing at Bletchley Park they are desperate to get the information before these people's knowledge is taken to the grave.
Even with the government saying you are free to talk about what went on people who worked there are still not talking or communicating what they did.
There are still big gaps and understanding of what happened at Bletchley Park and esp some huts the museum has no information or understanding of what went on there.
The family only found out what my nan did in WW2 only after she died becuse they found a letter of thanks and a medal.
She was invited to Bletchley Park becuse she won a crossword competition and she ended up in the Japanese code-breaking side of the park.
12:45 - actually, you were correct! “A” and “Z” are stand-ins for what someone would’ve made up in this case.
@@SpencerTwiddy True but the triplet isn't made up by the operator which is the thing the receiver needs to decrypt -- I thought that was made up but it's taken from a codebook. My bad!
Navy vet here. The scene that REALLY got my goat was the one where they worked out where the enemy was - and is right where the allied convoy was about to be - and one of the guys knew his brother was on a destroyer in that convoy, and wanted to warn the convoy. All that HAS to be utter bo££ox! There is no way that any of those data would have been available to code breakers or anybody outside of a select few. He might have known he was on HMS xyz, but not where XYZ was, let alone which convoy he was guarding or where that convoy was. (My aunt knew my father,, her brother, was on HMS Eagle, but not that he left the ship for a medical emergency before she was sunk. The sinking was announced on the news)
If I remember correctly the regular weather reports by the U-Boats were also a massive clue. Since they could triangulate the rough position and thus deduce what they would report from their own reports of the same region. Also weather reports had a clear format.
When the 4th rotor was added, how did they continue to crack it, as the bomb machine was build with 3 rotors in consideration?
I don’t understand the relevance of the 3 chosen letters. Was the intention for these to be used to change a setting in enigma? Did they replace the Ground word? What am I missing here?
@@silverghini2629 Thanks for watching! Watch the section again and watch the captions closely. The three chosen letters were the setting for that specific message only. The ground setting was a daily set of letters used to decrypt the message setting. They didn't want to use the ground setting for every message as that would make all the messages for that day easier to crack.
So the procedure is: use the ground setting to decrypt the first six letters of the message. Say it becomes RSKRSK. Then you set your machine to RSK, and use that to decrypt the contents of the message.
Hope that helps!
I assume one flaw into the naval enigma would be they had to talk to the army or if force at times so there would be some messages that went to both navy and air force units.
Could you do a video on the differences between the Lorenz Cipher and the Enigma? ... and maybe something about what Beurling did?
Proposal: start using HELLO WORLD for example text when talking about breaking German codes in the war, js
Why?
To avoid getting demonitized by accidentally praising the 20th century least liked Austrian.
@@DanielHarveyDyer
I don't think it's praising, or even accidentally praising, said Austrian. But yeah, YT might disagree and I get it from content creator perspective.
Charlie Chaplin used the phrase 'HEIL HINKEL' in his movie Dictator. I'd prefer that, but maybe the ever-so-wise algorithm can see through that one too.
Maybe it should be Gut Morgen
@@charlesloeffler333Sorry to be that German: It’s “Guten Morgen” because “Morgen” is masculine and “gut” is in the accusative case here.
Love the Mitchell and Webb shoutout :3
The repetition of the key phrases being the weakness that led to breaking the code just underlines a point that another WW2 history made: The Nazi's were in many aspects comically inept in carrying out the most basic of things.
Military communications do follow specific protocol of format. It is just like regular e-mail has recipient and sender mentioned so same applies to military messages even further. Same weakness was present in Allied communications as well so it was not exclusive to Germans.
In purely mathematical sense Enigma was practically unbreakable. In practice it was broken through human error and convenience. Here is an example how receiver in three separate messages could be marked. Receiver is captain Klaus Schwulmeister in command of 2. Dinner Company of 3. Food Battalion.
Military style:
To: Commander of 2. Dinner Company
To: Commander of 2. Dinner Company
To: Commander of 2. Dinner Company
Informal style:
This goes to 2. Company commander
To idiot who made whole division sick
This belongs to soon-former captain Schwulmeister
First is way easier to decode as it follows specific style.
OP, your observation is based.
For unknown reason, probably The History (Hitler) Channel has played too many fluff pieces on German ingenuity.
Fully 1/3 of the American populace believes the N@Z were technologically superior to the Allies. Because of some airplane, or rocket.
Von Brauns V2 was a truly great accomplishment. As was the Saturn V which is known to designed under his leadership. Which seems to lend credence to the belief.
But at almost every juncture from the Battle of Britton onward, German incompetence is overwhelmingly obvious.
They failed to develop a coherent plan of action from before the war.
For one: They didn't secure enough oil production or refining capability to wage a war on the rest of the world. In spite of the fact they clearly understood "blitzkrieg" tactics relied heavily on energy.
Fascist regimes tend to be inept like that. The lack of flexibility and adherence to the leader and tradition leads to the culling of actual minds that think.
there was some post-war use of the Enigma, presumably they had some different procedures for using it,
but Bletchley was kept secret for a long time, so maybe not...
I wonder if it might be plausible that that one guy watching over a place never attacked might be a veteran of the great war, read and resonated with the book All Quiet on the Western Front, and everyday his messages would be the book title in german "im westen nichts neues"
9:21 From mid-1942 the German navy used a 4-rotor Enigma known as the M4.
Did the British ever try cracking those Enigma messages later on, that they hadn't been able to crack originally early on or during the dark periods?
Some of them could still have been potentially important during the war or historically interesting afterwards.
And how about now with our far more powerful computers and better understanding of the Enigma machines, and the ways they were used?
There are a few places online you can actually play with engima machines, and a collection of messages which were actual messages, so I know they have played with it in the recent history.
Enigma encoding is not considered secure by modern cryptographic standards. It can be decoded in a reasonable amount of time with our computers. However, there are several Engima (Navy 4-rotor) messages that have not been decrypted and there is ongoing work to do so.
@shura0107 Without the starting point, Short messages are really hard to crack. Even with a modern heavy computer, If it's too short it's just not got enough space to guess the pattern.
@@leechowning2712 Not only is there not enough data in the message to reconstruct the original plaintext, it may well be that there are multiple valid plaintexts that could have encoded to the same crypttext with different IVs. That is to say, collisions.
At an extreme, some initial settings might encrypt "ILOVEYOURDOG" into "XDKLDWKVQWRK", while another set of initial settings would also encrypt "SENDFEETPICS" into "XDKLDWKVQWRK".
When any of the initial settings are valid and they are all equally likely, deciding which one was the original plaintext and which ones are the collisions is half guesswork and half wishcasting.
I was a USAF Cryptographic Account Manager and KL7 operator. If you sat an Enigma operator down in front of my KL7 he would know instantly what he was looking at, a bit more complicated but he would know.
Something that was unclear to me when I first heard of the enigma machine quite a few years ago, is that the machine itself was not a secret, other countries had them also.
They were available in a two rotor form commercially before the war for sensitive commercial communications.
8:00 If Germany had distributed D20 dice to every operator and told that operator must throw dice 6 times to get the initial 6 letters (sent without repeation), Allied forces would have never broken the encryption during the WW2.
Make a vid about what Arne Beurling did?
I've only ever really associated Enigma with a Naval requirement. Anyone who has watched documentaries and movies based on the Enigma codes and machines, will generally only ever seen Naval personnel associated with it so why the pronouncement at 00:28 ?
That's mainly because most of the early recovered machines were from the navy, they were used in plenty of other places.
While Turing is credited for cracking the Enigma, the Swedish mathematician Arne Beurling cracked a more complicated version of the German crypto machines (Siemens & Halske T52), using nothing but pen and paper.
Just an observation that despite the enigma machine doing all that clever encoding and decoding, it still relies on the sender and receiver having an agreed secret set of code/decode words ie keys, for each message. The convenient thing is that the key can be very short. But today, if we can agree on the secret keys then due to technology, length of key isn't really an issue. I can just have a file that for this message maps the first key press to one list of randomly generated numbers, the next keypress to another set of randomly generated numbers and so on up to however many keypresses is more than enough. This is utterly undecipherable ever, without the key. But what we can't do today is rely on the key remaining secret. But if we could, we wouldn't need the clever subsequent encoding and decoding algorithm anyway.
I love this duo !!! More please !
I’ve always wondered how much more secure the army and Air Force enigma protocol would’ve been if they’d handed out 20 sided lettered dies to the operators.
Surely this would have eliminated lot of the human error in choosing their three letter codes?
Perhaps even more effective than the navy protocol?
Wonderful video! Thank you 🍷
Along with the nothing to report guy there was also a guy who always sent a weather report and he always started it Weather Report. They also could guess what the weather might be and use that.
It's written that the German Navy personnel were much more disciplined in following their procedures and training and made fewer mistakes versus their army or Luftwaffe code clerks, where a Bletchley Park person thought of how a code clerk who was hung-over, tired, or bored and not paying attention could make mistakes and learned to search for those mistakes and then decode them (the mistake or subsequent corrective action) made it possible to break the coding via manual methods.
was this the same system the high ranking officials used? I thought I heard they had an entirely more advanced system, that could also transmit automatically. And They built Collossus to break that?
@@iWhacko They used something called Lorenz which is different to Enigma!
@@AnotherRoof Right :) love to hear more about that too
Lorenz which used 12 rotors and moved them in totally different ways. It also was not morse but tellytype
I thought, they use the word „Wetterbericht“ to break / brute force the code.
Why the hell would they repeat the secret setting? I know you said, to be sure. But that seems like an enormous flaw. Like, let's assume you received a code book and an enigma machine and were intercepting the messages. You didn't know about the secret setting. But every single message you receive starts with the first three letters repeated, then a bunch of nonsense. I don't think it would have taken any serious problem solving to just try those three letters as the machine setting. Really, it's the typing it out twice that's the problem. It's just a massive hint that the first three letters of every message are important.
Why not do something like type a random sentence and use the first, third, and fifth word's first letters as the setting? Since we would use recognizable words there doesn't seem to be a need to type the sentence twice to ensure accuracy. And this would significantly obfuscate the obvious three character setting.
"But every single message you receive starts with the first three letters repeated, then a bunch of nonsense."
Possible misunderstanding here -- the message didn't start with three repeated letters followed by nonsense. Those three repeated letters were *encrypted*. So a message might look like:
GJAOENFEQOC...
GJAOEN decrypts to ABCABC using the ground setting.
FEQOC... decrypts to the message content using the ABC setting.
So it's not as if the message setting ABC was plain to see at the start of a transmission.
Also, it's worth nothing a couple of things:
The Nazis were also kind of arrogant and believed the Enigma to be uncrackable, as a code machine in and of itself. As James mentioned, the Nazi forces stopped using this protocol early in the war, as the Polish cracked Enigma by recognising that the first six characters have repititions (i.e. exploiting a flaw in the protocol rather than cracking Enigma itself). After that, they just sent the message setting once.
Hope that helps!
@AnotherRoof No I understand that you need the base setting to get the secret code. My argument is that once you have that book, the "secret" setting is so absurdly obvious as to be completely pointless. There are so many simple ways to obfuscate that second setting.
@TheTonyMcD Then I'll refer you to the latter two points and what James said about how capturing codebooks was very rare (and only lasted a month at most). Don't know what else to say -- the Nazis were (thankfully) bad at security it seems
@@TheTonyMcD The starting letters were used to synchronise the sending and receiving machines. At midnight, all machines were set to the same day settings: choice of rotors, ring settings, order of the rotors in the machine and plugboard settings. As time passed, machines would get "out of sync" (i.e. the rotors would move as messages were typed.) Enigma decrypts by typing in the encrypted text on a machine set exactly as the sending machine, i.e. the rotor positions when used to encrypt (because the actual rotor choice, order, ring settings etc were set for everyone and only changed at midnight - except in 1942 for submarines when they changed twice a day) So the receiver has to know what the senders rotor position was when the message was encoded. This is how they did it: choose three letters, e.g. GAF. Set the Enigma rotors to show GAF in the windows and then select three more letters, e.g. TVE, and type them in to get, say, MKP. The sender would set their machine back to TVE and type in the message. At the start of the message the sender would send GAFMKP followed by the encrypted message. The receiver would then set their machine to show GAF and type in MKP to reveal TVE. They would then set their machine to show TVE and type in the rest of the message revealing the clear text. So actually it doesn't really matter per-se about the choice of the initial 6 letters and their randomness, the Bombe would be able to break it anyway. By choosing something less random for your 6 letters, e.g. CILLIE (your girlfriend's name), and using it repeatedly just makes it somewhat easier to break - it becomes easier for the cribsters to decide on a crib to use in breaking the message due to past history of that sender. I say somewhat because although you might recognise the first 6 letters, CILRDF to really be CILLIE (i.e. you know the cleartext of RDF) you still need to know the rotors used, the ring settings, the rotor order and plugboard settings to actually break that message and any others sent on the same network.
The Germans cryptologists, who were not stupid, would know that including anything non-random would increase the likelihood of a code being broken, because that's just how it is, so went to lengths for the navy to reduce the randomness. The problem of course is you have a miltary consisting of butchers, bakers and candlestick makers so the process has to be simple to work, the drawback being humans are rubbish at random. Note that the Allies were no better off in this regard and the Germans broke plenty of their codes as well.
The reality is in breaking Enigma the Allies didn't actually need non-randomness to do so, it just gave better clues. Of course, if you had the sheet of daily settings and the Germans didn't know, then you could read all messages sent on that network for as long as the sheet lasted - no cryptology work was needed at all.
what movie do they talk about?
The Imitation Game - It is a 2014 American period biographical thriller film directed by Morten Tyldum and written by Graham Moore, based on the 1983 biography Alan Turing: The Enigma by Andrew Hodges.
The guy on the right is so passionate about this, and if I was a guy on the left, i would be nodding but not getting anything 😂
Did you know you could buy an Enigma in the 1920s. The Americans bought a couple, as did yhe Poles who pretty much cdacked it.
It was commercial machine by Arthur Scherbius. Quite good device and practically unbreakable in it's intended niche.
Super video.
The metadata in the headers was plain text, enough for graph network analysis.
Was't the plans for the bomb presented to British inteligence in July of 1939 by Polish code brakers? They build bomb with 6 stages , however Polish cod brakers handed plans to build bomb with 60 stages. What the diffrence between Polish bomb decoding maschine and Turing bomb decoding maschine? The is a lot of confusion.
Are you going to take a look at the movie "Hidden Figures "?
Awww, you both know you're goodies x
The most impressive thing about Engima is that even today it can't just be bruteforced.
Human stupidity can always be brute-forced. Alien stupidity is another matter.
Mike from Numberphile did some demos of bruteforcing it th-cam.com/video/RzWB5jL5RX0/w-d-xo.html with some statistical analysis. Colossus seems to have started that type of computer assisted analysis , albeit on the Lorentz cipher
Except for the Polish dude who reverse engineered it in his head
@@tomarmadiyer2698 Marian Rejewski.
The Poles originally created the Bombe, Turing and Welchman, and a host of others improved on it until we had the monster that is in the NSA Museum in DC.
Yep, too little credit is given to the Poles.
The bombe essentially was 12 enigma in 1 machine using cribs to break the days settings
This video better not somehow be spoilers for the talk in two weeks
Thd navy enigma a had four rotors greatly increasing the number of possibilities.
Marvellous video ❤
Why does everyone only ever talk about breaking Enigma?
I don’t think I’ve ever heard anyone even mention the inventor of Enigma… even less make a movie about them.
Because Enigmas existed since the 20's and kept being improved.
hey, gunther! a rot wamz! a rot wamz!
A German trying to speak Chinese? 😄
But ve haf no rot wamz!
That’s why they screwed up without a pre-code. They would become redundant and start repeating the same code over and over and over again, which made them vulnerable.
Add to that the laziness of the operators under duress. Of course they reused settings and code strings.
Alan Turing was amazing
If the British didn't have the navy pairs table, would they have to decrypt each message individually? Since they couldn't work out the general day code?
Not having the pairs table means they would have to repeat their work next day. But being able to decrypt one message means they already had their settings of the day.
Did the German engineers who designed and built the Enigma system ever get any credit for their brilliant creation? It was human error/habit which eventually defeated it, not the concept or the performance. I'm British, but I still hugely respect the superb engineering which went into Enigma.
@@strummergr I agree and I'm fascinated by the device itself. However there was one flaw the Allies could exploit, which is that a letter never gets mapped to itself when encrypting, and this turned out to be a massive weakness!
Actually, it wasn't human error that broke it, Enigma was fundamentally flawed. The 26 combinations that couldn't exist out of a possible 159+ quintillion was all that was needed. Human error gave clues to help with obtaining daily settings but the allies were still breaking these even when those errors didn't exist.
I suppose in fairness you could argue that human error in putting rotors into the Polish postal system helped a great deal because without knowledge of rotor wirings you are stuck. But day to day operational errors didn't really contribute.
Isn't the enigma machine the first portable computer.
Ummm no there was no processing or memory it was just hard wired wheels
Alan was of the "Officer Class" , Tommy, not so. Therefore, Alan gets the credit.
If I ever have stupid lottery money, I'm going to have an Enigma Machine made for me from scratch. (while an actual original would be cool, I'd rather an identical replica because authentic ones belong in museums)
Maybe even hire someone to fix the flaws of the device (like a letter can never be itself) and make additional code wheels beyond the 8 that the Navy had.
There are online ones
When you win the lottery contact me, I'll build you one.
the german art of making things extraordinarily complicated, put to effect better than ever
Re The Protocol -- wouldnt capturing and interogating an operator have given them that info ?
that would mean capturing a navy operator alive (unlikely), identifying that person as an operator and then getting him to talk.
managing one of those? unlikely but possible, doing all three? i doubt that.
Lost opportunity to make plaint ex rotor setting to be HIT and encrypted rotor setting to be LER
But what was the likelihood that by the time the other side has decoded a message, it was already too late and the others had already carried out that plan, so it was too late to intervene or whatever?
Yes, and that was the case. Thousands of messages were copied down. So many that they would be still decoding them today.
At the start of the war this was the case - essentially, the work being done was useful in identifying in how it could be better exploited by breaking messages even weeks old. However, once the Bombes came on stream in early 1940 and the codebreaking became more streamlined, Enigma networks were more rapidly broken. It wasn't long before most Enigma networks were broken every single day and in many cases in the first 8 hour shift starting at midnight!
I don't understand how the recipient can decipher the message secret without already knowing it..
@TheLucnicLord Essentially, the "secret" was sent encrypted with the message along with the clue to unencrypt it. I explain in a comment above how it works.
The singing banana himself!
So fascinating, thanks!
And I’m delighted to see Grimes here! And a bit jealous 😂
The engima originally had 3 wheels not 5, and then they added a 4th wheel.
@@kevingeier3385 There are three rotors in the machine but five options for which to use. Since we're talking about the start of the war we didn't mention the fourth rotor added later. Hope that helps!
16:10 so… you wanted to get demonetized or…?
tbh, I would have prefered they stuck to 'nothing to report' 😬
Should've just given the operators dice 😂 (36 possibilities, more than enough for the alphabet)
How and why are different words and sometimes they are inappropriately used. The hosts are discussing HOW they differ, not WHY. Why they were made differently? Only those who chose to do so can answer that and most of them are now dead.
26:3=78 if anyone knows what 1 million is let me know that would be ZZZ
The fact is that Turing really acted on a thought that many knew. Codes and other information needs to be processed and the resulting information discovered very quickly. Although many of the enigma codes were being broken and other information about the codes was known, the solution of the codes came after the textual information they contained was uselessly old. German businesses were using the Enigma device before the war to discourage corporate espionage. Polish mathematicians were decoding some of the messages back then. But it would have been the same result finding out after the information was useful. Turing had designed a device he thought would work. But it didn't. Someone at Bletchley Park knew the other piece of the puzzle. He or she knew about a British Telecom technician named Fellows (I believe). Once that person knew what Turing was trying to achieve he went home and returned a week later with a schematic for the circuitry that did what Turning said was needed and at an blinding speed. So it wasn't a single person who solved the problem, it was a single person, Turing, that put them on the right track. But, even more importantly, Turning set humanity on a whole new road where today we rely on electronic circuitry to solve many of the question that can be asked. George Bernard Shaw ( in "Back to Methuselah") had it right when he had the snake in the garden of Eden say, "Some men look at things that are and wonder why? I think of things that never were and say, "Why not." (or words to that effect)(also, I believe, an expression used frequently by Robert F. Kennedy, who also likely got it from the same place, Shaw being an Irish author and all.)
Never break German cryptography for the British government if you want to keep your balls
If you believe Peter Wright (autobiography: spy catcher) there was more to it than that. He implied Turing talked too much... and to the wrong people, although he never clearly stated exactly whatever happened
A letter would never be substituted with itself, so the odds were only 1 in 25.
@@intentionaloffside8934 But a letter can become itself (and did) after three substitutions.
@@AnotherRoof It's not clear what you mean here: Enigma cannot physically encrypt a letter to itself (A will never, ever, encrypt to A)
@@andyj2106 We know. James is talking about *three substitutions* at around 20:10.
A -> 1 -> 2 -> 3. 1 is not A, 2 is not 1, and 3 is not 2. James is saying that there's a 1/26 chance that 3 is A.
@@AnotherRoof I guess your comment isn't very clear to me. I think the OP here is saying that in any particular position, there is a 1:25 chance of a particular letter being lit when a key is pressed. Maths isn't my strongpoint, but I think the 1:26 combination comes from the fact that James is talking about loops and that in any particular position, there is a 1:25 chance of a letter appearing on the lampboard PLUS the key pressed giving 1:26. In other words the probability is derived because it is calculated across the cleartext and the enciphered text? In fact at *any* position in the message there is a 1:26 chance of a letter appearing when including cleartext and enciphered text?
Can I surmise that when you say that "a letter can become itself (and did) after three substitutions" that was not a definitive statement and you were referring to the example James used? This is the bit of your comment I'm not clear on, or indeed A->1->2->3. At any given starting position, a loop could occur from position 2 onwards (e.g. using HEIL as the cleartext, H -> E; E -> H would be a possibility) and that at any position after the first key press there is a 1:26 chance of a loop arising?
@@andyj2106 The only time a probability of 1 in 26 or 1 in 25 comes up in the video is in the aforementioned section, so I assume that's what OP is referring to.
I think your understanding is correct. If we did find a loop after two keystrokes, then we can use the methodology James discusses. I used three keystrokes because that's the example in the video, but the loops could be any length greater than 1.
Nope.. A Team did it not Alan Turing alone.
But Turing contributions were critical to the team success. It’s what the people that were there and on the “team” said.
You guys left out the Polish who broke the code first
Marian Rejewski did the real work.
Turing just copied the Rejewskis work.
@@richardgrego We didn't. James mentioned this when we discuss the army protocol and how the Polish method relied on that. Turing and the others built on their work but cracked it in a way that didn't depend on that protocol and thus could be applied to naval messages etc.
Hope this helps!