Cryptography and ULTRA (as I procrastinate for crypto exam)

[unmerged(104627)]

I write this post as I dangerously procrastinate for an upcoming cryptography exam.

First of all, in HOI2 there is a misnomer. If you may not have noticed, you can research two things: encryption and decryption. What is meant, however, is cryptography and cryptanalysis (devising crypo schemes and devising methods of breaking them, respectively). Also, from what I recall, researching each one gives you a 15% boost (researching "encryption" give you a 15% boost in secrecy, while researching "decryption" gives you a 15% boost in intelligence). This is simply not so: It would be more proper if "encryption" give you a large boost in secrecy, while cryptanalysis ("decryption") give only a modest boost in intelligence (and cost a LOT more to research).

Consider the difference in complexity of an enigma cipher and the Bombe computer used to try to break it. Enigma was broken not by Alan Turing finding a *magic forumula* to solve all engima communications, but rather by intelligence agents doing the usual espionage activities (spies, double agents, capturing machines, code books, etc...). Sure, the Bombe was useful and an incredible breakthrough, but it simple could not decrypt the volume of Enigma communications going around.

I propose that cryptopgraphy, the development of mathematical models of ciphers and their physical implementations, ("encryption" in HOI2 terms), be a relatively easy to research. And require equal amounts of mathematics and electrical engineering. It should give a large boost to protecting secrecy. (e.g, 15% boost for each upgrade)

Since cryptanalysis, the development of algorithms and electronics to break ciphers without knowing the key, is hard (and not always successful), it should give only a modest boost to intelligence (e.g., 5% boost). It should also require modest amounts of mathematics research, and large amounts of electrical engineering (it should take much longer than "encryption" to research).

Therefore, intelligence collection should be done via other means of espionage that the game will include. Cryptanalysis alone should not be a reliable method of collecting intelligence.


FYI: Consider the difference in complexity between the electronics in an Enigma
Slightly simplified diagram (just four letters)


and the Bombe...

 

[Delta107]

I agree, but all that must be checked in detail for in-game balance. And that could be a problem with HOI2. Since HOI3 uses different aproaches we'll see of course interesting techs.

 

[unmerged(41649)]

On a side note, intelligence would be MUCH MUCH more useful if your intelligence officers had a brain, or at least a memory. As it stands now, you cannot look at past intelligence reports and see what kind of divisions the Soviet Union had 2 months ago, or where certain divisions were. If you lose your intelligence network, suddenly you have absolutely no idea at all what kind of army your enemy has, even if 2 weeks ago you new about every division, you suddenly forgot all that information.

 

[Cardus]

I write this post as I dangerously procrastinate for an upcoming cryptography exam.

First of all, in HOI2 there is a misnomer. If you may not have noticed, you can research two things: encryption and decryption. What is meant, however, is cryptography and cryptanalysis (devising crypo schemes and devising methods of breaking them, respectively). Also, from what I recall, researching each one gives you a 15% boost (researching "encryption" give you a 15% boost in secrecy, while researching "decryption" gives you a 15% boost in intelligence). This is simply not so: It would be more proper if "encryption" give you a large boost in secrecy, while cryptanalysis ("decryption") give only a modest boost in intelligence (and cost a LOT more to research).

Consider the difference in complexity of an enigma cipher and the Bombe computer used to try to break it. Enigma was broken not by Alan Turing finding a *magic forumula* to solve all engima communications, but rather by intelligence agents doing the usual espionage activities (spies, double agents, capturing machines, code books, etc...). Sure, the Bombe was useful and an incredible breakthrough, but it simple could not decrypt the volume of Enigma communications going around.

I propose that cryptopgraphy, the development of mathematical models of ciphers and their physical implementations, ("encryption" in HOI2 terms), be a relatively easy to research. And require equal amounts of mathematics and electrical engineering. It should give a large boost to protecting secrecy. (e.g, 15% boost for each upgrade)

Since cryptanalysis, the development of algorithms and electronics to break ciphers without knowing the key, is hard (and not always successful), it should give only a modest boost to intelligence (e.g., 5% boost). It should also require modest amounts of mathematics research, and large amounts of electrical engineering (it should take much longer than "encryption" to research).

Therefore, intelligence collection should be done via other means of espionage that the game will include. Cryptanalysis alone should not be a reliable method of collecting intelligence.


FYI: Consider the difference in complexity between the electronics in an Enigma
Slightly simplified diagram (just four letters)


and the Bombe...

good point!

 

[Neomann]

Nice one, i agree too

 

[Jmland]

I write this post as I dangerously procrastinate for an upcoming cryptography exam.

First of all, in HOI2 there is a misnomer. If you may not have noticed, you can research two things: encryption and decryption. What is meant, however, is cryptography and cryptanalysis (devising crypo schemes and devising methods of breaking them, respectively). Also, from what I recall, researching each one gives you a 15% boost (researching "encryption" give you a 15% boost in secrecy, while researching "decryption" gives you a 15% boost in intelligence). This is simply not so: It would be more proper if "encryption" give you a large boost in secrecy, while cryptanalysis ("decryption") give only a modest boost in intelligence (and cost a LOT more to research).

Consider the difference in complexity of an enigma cipher and the Bombe computer used to try to break it. Enigma was broken not by Alan Turing finding a *magic forumula* to solve all engima communications, but rather by intelligence agents doing the usual espionage activities (spies, double agents, capturing machines, code books, etc...). Sure, the Bombe was useful and an incredible breakthrough, but it simple could not decrypt the volume of Enigma communications going around.

I propose that cryptopgraphy, the development of mathematical models of ciphers and their physical implementations, ("encryption" in HOI2 terms), be a relatively easy to research. And require equal amounts of mathematics and electrical engineering. It should give a large boost to protecting secrecy. (e.g, 15% boost for each upgrade)

Since cryptanalysis, the development of algorithms and electronics to break ciphers without knowing the key, is hard (and not always successful), it should give only a modest boost to intelligence (e.g., 5% boost). It should also require modest amounts of mathematics research, and large amounts of electrical engineering (it should take much longer than "encryption" to research).

Therefore, intelligence collection should be done via other means of espionage that the game will include. Cryptanalysis alone should not be a reliable method of collecting intelligence.


FYI: Consider the difference in complexity between the electronics in an Enigma
Slightly simplified diagram (just four letters)


and the Bombe...

Hmmm....Distance to "the fort" I would estimate approaches zero.....

(insider joke meant for those who understand)

 

[unmerged(104627)]

Hmmm....Distance to "the fort" I would estimate approaches zero.....

(insider joke meant for those who understand)

I wish I could either confirm or deny that.

 

[Black Lotus]

Good ideas, cryptography should be overhauled in my opinion.

 

[unmerged(7537)]

On a side note, intelligence would be MUCH MUCH more useful if your intelligence officers had a brain, or at least a memory. As it stands now, you cannot look at past intelligence reports and see what kind of divisions the Soviet Union had 2 months ago, or where certain divisions were. If you lose your intelligence network, suddenly you have absolutely no idea at all what kind of army your enemy has, even if 2 weeks ago you new about every division, you suddenly forgot all that information.

I agree, I really would love to see something like that. Maybe as a ledger entry?

 

[Sute]{h]

The trick is implementing something along this linies while keeping down micromanagement. Requiring the player to spam espionage missions is no fun. It could be solved quite simply by doing something along these lines:

Cryptography: Gives +15 secrecy per level
Cryptoanalysis: Gives +5 intelligence per level against every other country
Intelligence agents: Gives +5 intelligence per agent assigned to a specific country. The player simply sets the desired intelligence level and the computer then keeps sending agents until that level is reaches. If agents can be compromised and lost the computer then automatically send a new agent until the desired intelligence level is restored.

EDIT: Perhaps agents should also provide a +1 intelligence bonus against countries allied with their target country. They might stumble upon coordinated plans after all.

 

[unmerged(104627)]

The trick is implementing something along this linies while keeping down micromanagement. Requiring the player to spam espionage missions is no fun. It could be solved quite simply by doing something along these lines:

Cryptography: Gives +15 secrecy per level
Cryptoanalysis: Gives +5 intelligence per level against every other country
Intelligence agents: Gives +5 intelligence per agent assigned to a specific country. The player simply sets the desired intelligence level and the computer then keeps sending agents until that level is reaches. If agents can be compromised and lost the computer then automatically send a new agent until the desired intelligence level is restored.

EDIT: Perhaps agents should also provide a +1 intelligence bonus against countries allied with their target country. They might stumble upon coordinated plans after all.

Something like that would be good. Cryptanalysis could provided much more benefits, but only if they invest massive amounts of money and research.

 

[oribiasi]

Even if we adopted what VBTemp wants, how does one take into account the fact that in real life, if one knew the enemy's plans, you didn't try to show your hand too much -- if you were always in the right place at the right time, they'd be on to it, and hence change the code.

 

[unmerged(131221)]

Agree.
Cryptography should give so big bonus that a country that have researched it farther should be confident that it will take some time to break their code.
Cryptoanalisis should only give limited bonus (or be hard to advance) so it should be neccessary to go far into the tech tree and research much more. OTOH spies could help.
Germans was confident that their codes can not be easily broken. Good Cryptography should work this way IMO. Its hard to break codes so when you have an advanced system you should be more confident.

 

[Wyrm]

The german codes were not really broken directly.
For example, sets of cryptos were stolen in Italy during the north-african campaign. This was more or less the beginning of ULTRA

THAT is something that needs to be reflected some way... perhaps some espionage-mission that if successfull gives you a great boost in intelligence that degrades over time (to reflect cryptos being replaced)

 

[Balesir]

The german codes were not really broken directly.
For example, sets of cryptos were stolen in Italy during the north-african campaign. This was more or less the beginning of ULTRA

It began a lot earlier than that; the Poles did some mathematical groundwork and captured an early Enigma machine which they passed on to Bletchley Park (sort of a "blueprints" gift).

The idea of HUMINT and SIGINT with cryptanalysis applied combining is a good one, though - and from the look of the new Intel screen (q.v. Dev. Diary 20 earlier this week) I think it is an aim in HoI3.

 

[unmerged(86498)]

maybe a good way to reflect this using the hoi3 research system would be to have practical espionage experience ("HUMINT") gained in very small amounts over time and proportional to the number of operations you are running (even unsuccessful espionage ops give you some experience in that you learn what not to do in the field) (ie: number of spies abroad) and theoretical espionage experience ("SIGINT") which could be gained just like any other theoretical exp in the game: by the amount of research you do in the field...