some first thought (cipher 1)
 Monday, 08-Nov-1999 17:08:43 

      195.238.10.152 writes:

      Hi guys,

      I think we are going to have some fun :)

      I did not checked the net yet about some links we may found, but i see seeker did already, thanks :)

      anyway, here are my first thought about the first challenge. 

      - We should try to reconstitute pair of duets (for example TH gives BN), then build up the rectangle (or lines:(
      - we should find more than once the letters ST and OP or TO (from the word STOP).
      - ND should be crypted as BG or GX.
      - it is diagonalisable (not sure this word exist). I mean if AB is crypted as CD then BA will be crypted as DC and also CD
      will be crypted as AB.

      Well, I think that's all i got for tonight ... time to go sleeping now.
      Of course, I may be wrong in the above (I'm falling asleep). Sorry if it's the case.

      Have fun,

      Laurent. 

      laurent 

***************

Re: some first thought (cipher 1)
 Monday, 08-Nov-1999 17:51:03 

      62.156.30.216 writes:

      ... more than tired, but what me keeps the next 2 minutes awake : 
      1. END is at the end. How come these 'XXXX' ? 

      2. What happensif you 'fall' out of the square ? Continue at the other side ? 

      Enough for today 

      Regards

***************

Re: Re: some first thought (cipher 1)
 Monday, 08-Nov-1999 19:12:59 

      32.100.19.140 writes:

      You can't 'fall off the square'.. You plot two points in it and then establish the rectable with the shortest path possible.
      Even if the first letter of your snippet is in the top right corner and the second is in the middle, your rectangle will be in
      the upper right corner of the matrix, not possibly in the top left corner is you'd 'wrap around'...

      The order of the letters will dictacte the order to take the new letters out.. You take the corresponding letter on the same
      line as the first letter you plotted.. (and yes, it's possible that I'm too confused right now to explain properly ;-))

      For the XXXX I think they are there to act as filler for the 5 chars formatting. Without them you get 172 characters, or 44
      groups of 4 / 88 groups of 2.

      And Laurent, I came to the same conclusions regarding the diagonalistics of the whole thing (might as well come up with
      new words too ;)) Two letters will always encode in the same two other letters with a given matrix, only the order can
      possibly change. And the process is reversible since any two opposite corners of a rectangle will define the other two
      automatically...

      I just ran a little script to count occurances of the letters:

      J: 0
      Q: 1
      G: 2
      M: 2
      X: 2
      P: 3
      Z: 4
      H: 4
      I: 4
      C: 5
      N: 5
      K: 5
      W: 6
      V: 6
      O: 7
      B: 7
      T: 8
      E: 8
      S: 8
      U: 9
      R: 9
      A: 10
      Y: 10
      D: 11
      F: 17
      L: 23

      I wonder why one letter is decidedly higher than the others... could it be the one in the middle ? (wild guess, at first sight
      I'd say they all have equal chances..)

      I'm still looking at some patterns... see you later ;) 

********************

Crash geometry class :)
 Monday, 08-Nov-1999 20:51:30 

      32.100.19.194 writes:

      Take a 5 by 5 matrix (a grid):

      X X X X X
      X X X X X
      X X X X X
      X X X X X
      X X X X X



      We'll skip some details about the cipher here.. just put letters in your grid:

      A B C D E
      F G H I K
      L M N O P
      Q R S T U
      V W X Y Z



      This cipher takes two plaintext letter to give two ciphertext letters.. let's say you want to encode 'IQ'.. find them in the
      grid:

      . . . . .
      . . . I .
      . . . . .
      Q . . . .
      . . . . .



      Then complete the rectangle between them:

      . . . . .
      +-----I .
      |     | .
      Q-----+ .
      . . . . .



      What where the letters in those corners ?

      . . . . .
      F-----I .
      |     | .
      Q-----T .
      . . . . .



      "FT"... so here IQ encodes as FT. These are the rectangles we're talking about :) There are more details related to the
      cipher but that's basically it...

      Might have to copy the text in notepad or look at the page's source to have it display correctly.. no idea how it'll come
      out.. 

      -Tx 

*************************

Re: Re: Re: some first thought (cipher 1)
 Monday, 08-Nov-1999 20:58:30 

      128.113.195.133 writes:

      If two letters are in the same row, and at the right edge of that rule, by following the rules we would 'drop off' the square.
      Right? My guess in this case is that you just 'wrap around'.

      Anyways, I also looked at the letters, and i was also surprised by the ammount of repetiotion of some letters (I thought
      that F's 17 was also surprisingly higher than the other ones).

      However, I (still ;) haven't figured out the meaning of that, so I looked into letter pair's repetitions. Nothing special:

      3 pairs repeat 3 times (AR, OL and LD) and 10 repeat twice (TE, LU, AR, EW, IC, YL, LA, FZ, FY and BF).

      More news later (hopefully ;)

      Csativa 

**************************

Wrap around..
 Monday, 08-Nov-1999 22:12:26 

      32.100.19.194 writes:

      I guess that must be it for same line/col cases... wrap around the line or column...

      Interesting tidbit: a letter can't encode itself... a cipher E will never represent a plain E....

      with this, here's a list of starting indexes where you can encode REDPENGUINFRENZY (0-based)

      0, 4, 6, 8, 9, 13, 14, 15, 16, 19, 24, 26, 27,
      33, 34, 36, 37, 38, 40, 41, 43, 44, 45, 47, 48,
      49, 50, 51, 53, 56, 58, 60, 62, 64, 68, 69, 70,
      73, 76, 79, 80, 81, 83, 84, 88, 91, 93, 94, 95,
      97, 99, 101, 102, 104, 105, 107, 108, 109, 112,
      114, 116, 117, 118, 121, 122, 124, 125, 126,
      127, 128, 130, 133, 137, 139, 141, 142, 145,
      147, 148, 149, 150, 152, 158, 159

      the same for BEWAREICEWEASELS:

      4, 7, 9, 13, 14, 16, 17, 20, 21, 23, 25, 26, 28,
      41, 43, 47, 49, 52, 54, 55, 61, 63, 64, 69, 71,
      76, 78, 79, 82, 84, 85, 86, 90, 93, 95, 96, 97,
      101, 102, 105, 111, 113, 114, 115, 116, 117,
      119, 120, 121, 122, 123, 124, 125, 126, 127,
      128, 132, 133, 135, 138, 146, 147, 149, 150,
      151, 154, 155, 156, 157, 158, 159

      That's still quite a lot of possibilities... 

      -Tx 
*********************************

Duh..
 Monday, 08-Nov-1999 22:16:00 

      32.100.19.194 writes:

      Didn't account for the string's length.. so take out the last few indexes in each case.. since the string won't fit because
      it'll run out of cipher ;) 

      -Tx 

********************************

Re: Wrap around..
 Monday, 08-Nov-1999 23:09:47 

      128.113.195.133 writes:

      Very interesting Tx.
      Now here this might help lower the possibilities... of course, it only applies if redpenguinfrenzy is there (but, well, if they
      said it... i bet it is there).

      anyways, if we divide in two letter sets, starting from the first one, there would be two same sets of encyphered 'EN' three
      sets apart... there are none of those :(

      but at least now we know that the redpenguinfrenzy starts in an uneven letter... thus your list cuts down to the uneven
      ones... still a lot but at least less.

      let's crack it open,

      Csativa 

*****************************

Re: Re: Wrap around..
 Tuesday, 09-Nov-1999 01:48:17 

      195.238.21.128 writes:

      Great work Tx :)

      applying the same princip as Csativa did, we should find a pair 'EW'-'WE' from 'beware ice weasels'. Unfortunately, i did
      not found it :(

      so, same conclusion as Csativa 'beware ice weasels' should start on a even position (even because 'EW' is on a uneven
      position in the phrase).

      Got to go to work now :( ... I'll try to make some free time there :)

      Regards,
      Laurent. 

**********************************


Strategies for deciphering

fished from the web http://www.achiever.com/freehmpg/cryptology/strateg.html

The first things you should do is

get organized!

Compile a list of known codes and cipher methodologies that you can use on secret messages. Once you have tried all the methods you know of, use a hit and miss method, trial and error. The structure of the English language enables cryptologists to see patterns that emerge from normal messages. Just by looking at this paragraph of text, or playing Wheel of Fortune, you can figure out which letters are most common.

Vowels and spaces are the most frequent in the English language. A few constants, such as "S, T, N, etc." are more common than others. Finally, a few known bigrams (2 letters), such as "ll, ee, ss, etc," are more common than others. By recognizing patterns in ciphers you can often guess what the letters are decipher the message. The computer application "Cryptogrammer" teaches you how to decipher this way. Ask your instructor if you are interested in playing Cryptogrammer.

Most Common Letters

In order of most comon to least common:
1. E
2. T
3. A, O, N, R, I, S
4. H
5. D, L, F, C, M, U
6. G, Y, P, W, B
7. V, K, X, J, Q, Z

Bigram Frequency

In order from most common to least common:
TH, HE, AN, RE, ER, IN, ON, AT, ND, ST, ES, EN, OF, TE

Bigram Same Letter Frequency

In order frommost common to least common:
LL, EE, SS, OO, TT, FF, RR, NN, PP, CC, MM, GG

Trigram Frequency

THE, ING, CON, ENT, ERE, ERS, EVE, FOR, HER, TED, TER, TIO, VER

Intial Letters

T, A, O, M, H, W, C, I, P, B, E, S


Second Letters

H, O, E, I, A, U, N, R, T

Third Letters
E, S, A, R, N, I

Final Letters
E, T, S, D, N, R, Y, G

*More than 50% of English words end with "E."
*More than 50% of English words start with T, A, O, S, or W.

Frequency Tables for Deciphering


It's a good idea to

create a table of the counts

for letters of the alphabet, bigrams, trigrams, and initial and ending letters to see how many times a certain letter(s), number(s), or symbol(s) from an encrypted message occurs. The highest occurance of a single letter/number/symbol is most likely to be "E" as seen in the frequency table above. The most common 2 letter/number/symbol bigram is likely to be "TH." If you can place a few likely letters you will often see a few short words, like "THE" appear.

Once words start to appear you're on to something. See if there is an easy pattern to get the rest of the letters. If not, continue

analyzing frequency tables

to decode the rest of the letters.

Use trial and error

and replace letters/numbers/symbols with different letters until more words begin to appear. Try this method on the simple cipher below. Can you figure out what the encrypted message says?


Wpxfmt bsf uif nptu dpnnpo mfuufs