What is a 'ciphertext-only attack'?

A cryptanalysis scenario where the attacker has only the encrypted message to work with — no known plaintexts, no cribs — relying on statistical analysis and pattern detection An attack that only works if the attacker can see the actual cipher machine or algorithm Trying to break a cipher by only guessing the key without looking at the ciphertext Breaking a cipher by reading only the first few characters of the encrypted message

What is a 'known-plaintext attack'?

An attack where the analyst has both the plaintext and corresponding ciphertext for some messages — they can compare the two to deduce the key or encryption rules An attack where the attacker knows the plaintext but not the ciphertext Reading the plaintext directly without needing to decrypt anything An attack that only works when the plaintext is written in a known language

What is 'index of coincidence' in cryptanalysis?

A statistical measure of how often letters repeat in a text — natural English text has high coincidence; truly random text has low; polyalphabetic ciphers produce intermediate values that reveal key length How many times a ciphertext matches the plaintext by coincidence without knowing the key The index page of a codebook that lists where each code word appears The number of errors that coincidentally occur in a transmitted message

What is a 'digraph' and why are digraph frequencies useful in codebreaking?

A pair of letters — in English, certain pairs (TH, HE, IN, ER, AN) appear far more often than others — their frequency in ciphertext helps identify common letter pairs Two different graphs showing letter frequencies from two different sources A cipher that uses pairs of letters as its basic unit of encryption Two identical characters used as markers in a code — like quotation marks in text

What does it mean that a cipher is 'computationally secure'?

That breaking it would require more computing time or resources than any attacker could practically obtain — not theoretically unbreakable, but practically so given today's technology That a computer is required to encrypt the message — it's too complex for humans alone A cipher that has been officially certified as secure by a government computer agency That the cipher can be computed (calculated) — it has a precise mathematical definition

What is 'Kasiski examination' and what does it reveal?

A technique for finding the key length of a Vigenère cipher by looking for repeated patterns in ciphertext — the spacing between repetitions is likely a multiple of the key length A test to determine whether a message was encrypted with a military or civilian cipher A method of cracking book ciphers by examining the page numbers used A technique named after a famous codebreaker who examined ancient Greek texts

What makes a 'crib' useful in codebreaking?

A crib is a known or guessed piece of plaintext — matching it against ciphertext positions dramatically reduces the search space for the key A crib is a small notebook where codebreakers record their progress A crib is a shortcut or cheat sheet for memorizing complex cipher rules A crib is a copy of the enemy's cipher machine used for reference

What is 'frequency analysis' and who invented it?

The technique of analyzing how often each character appears in ciphertext, exploiting natural language frequency distributions — invented by Arab mathematician Al-Kindi around 850 AD A technique invented by Alan Turing during WWII to break the Enigma cipher A modern computer method that requires analyzing millions of cipher messages A statistical method developed by the NSA in the 1950s for modern cryptanalysis

Why is a longer key generally more secure in cryptography?

More key bits mean exponentially more possible keys — doubling key length squares the number of keys to try, making brute force attacks exponentially harder Longer keys take longer to encrypt and decrypt, making the cipher slower and harder to break Longer keys are more random — short keys tend to have patterns Longer keys encrypt more characters at once, making the cipher more efficient

What does a codebreaker do when their current hypothesis doesn't work?

Treat the failure as useful information — eliminate that hypothesis, revise their model, and systematically try the next most likely possibility Start over completely and try a different type of cipher from the beginning Accept that the message is unbreakable and move on Get a faster computer — failure always means not enough computing power