unforum

[alexbutterfield]

Edit: this was a list of 1,2 and 3 letter words, but an updated version is posted later.

[Agent Lex]

Sorry for the lack of ciphertext last night. To make up for it, I'll post 2 tonight. Later.

[chulo333]

was working on the doritos quest puzzles the last couple of days, but i'm done now and i'll get back to working on this tonight after work. Thanks for the comprehensive list and analysis Alex!

[GreenWindmill]

One possibility regarding the two letter word with both letters the same - if these are song lyrics as has been speculated then "oo" could appear. It's not exactly uncommon in songs. Just a thought...

[Omega]

Well, I'm back from the dead (aka "out of town"). Unfortunately, I have had no revelations concerning this puzzle. With 26 unique symbols, it's bound to be a substitution of some kind, but I'm not making the proper connections somehow. I may have to reconsider and try a different approach (aka "wallow in self-pity").

[alexbutterfield]

Text 2.

23g19r3r. 13g1r6r23g24r. 22g3g18r. 20g6g8g11r. 10g10r16r. 21g6g15r16r5r. 21g3r19r.

20g6g19g. 12r1r6g18g. 15r. 21g3r10r16g. 21g6r22r. 17r14r15g20g20r.

23r10g21r21g. 15g. 23g8g15r14g. 1g. 21g6r8r9r21g20r. 20r10r24r16g11g10r26r16g.

15r17g. 20r6g22g. 18r8r1g11r. 25g16r. 14g3r. 10r. 18r10r20g20g.

-----------------------------------

1r8r4g. 10g17r. 10g. 9r10g16g9g. 17r1g18r. 20g3r22g.

15g. 4g10r4r8g11g. 7r8r14r2r22g. 21g3r19r16g. 1g1r26r9r.

6r22r14r. 15r14g. 4g6g9g17g1g11g. 9g9g1r20r20r20r. 25g1r14g14g16r9g.

10r. 23r15r20r20r. 1r20r18r8g21g17r. 24r16r16g21g. 14g19r16g. 17r8r26r9g.

-----------------------------------

20r3r10g16g. 21g6r22r. 24r6g16r16r10g9r9g. 8g1r4r. 24r6r16r16r10g16r16g.

21r3g15r16g. 20r3r19g. 23g10g11r22r. 1r21g20r. 25g20r. 22r16r8g16g14r.

20r3g15g9g. 20r6g22g. 18r19r16g1r9r15r9g16r. 18g16r9g9r. 11r3g14g9r14g19g16g16r.

20r3g15g9g. 20r6g22g. 23r19r16r1r. 23r9g9r9g. 8r2r1r9g11g.

-----------------------------------

8g8r5r. 18g19r16g1r. 1r14r. 20g1g17r11g. 15g. 24r15r1g5g. 21r6r22r.

20r6r22g16r. 17r6r1g14r. 18r10r20g21g. 17r15g21r20r. 14g19g16g. 1r15g16r.

17g10r8r11r. 15g14r. 20g3g22g4g. 24g3g. 10r. 7r8g1r. 19r16r8r16g. 21r3r19r.

26g8r12r16g. 15r11g. 9g8r17r21r. 14g3r. 3r16r. 1r9r8r16r. 21r6r19r.

3r22g. 14r19r16g. 11r19r10r1r11r17r. 20g3g19g. 4g3r. 16g8r4r16g1r9g. 21g6g19r. 14r6g. 25r16g.

3g22r. 20r6r22g. 13r8g3r18r. 15g. 18g10g20g21g.

10r. 23g15r21r21r.

[alexbutterfield]

After translating the second text, and seeing that many single-letter words are repeated, it gives me confidence that this is substitution (with a choice of 2 symbols per letter. I don't know if this is a random choice or not).

Below is a new analysis of 1,2 and 3 letter words from texts 1 and 2 cumulatively.

1 letter words

15r.
1g.
10r. x4
10g. x3
15g. x6
8g.
8r.

= at least 4 single-letter words.

2 letter words

21r1g.
10g1g.
8r8r. * weird one. Could be initials. (only 2 letter word with letter repeated is Aa, a type of lava.)
10r8g.
22g7g.
15g1r.
10r26r.
6r19r.
20r1g.
14r20g.
25g20g.
11r6g.
25r16r.
26r16g.
17r3r.

15r17g.
25g16r.
14g3r.
10g17r.
15r14g.
25g20r.
1r14r.
15g14r.
24g3g.
15r11g.
14g3r.
3r16r.
3r22g.
4g3r.
14r6g.
25r16g.
3g22r.

I don't think any of these 2 letter words are repeated. Although there are many that contain one common letter (in the same position), so it's possible that there are simply different permutations of the same word.

eg. 25g16r. 3r16r. 3r22g., these could be 3 (of 4) permutations of the same word. I can't find a set of 4 (eg. 25g22g would complete this set, and would make me think that it was almost certainly one word)

3 letter words

For 3 letter words there are 8 permutations (8 possible ways to spell THE for example.)

23g19r3r.
22g3g18r.
21g15g16g.
21g3r19r.
21g6g19r.
21g6r22r. x2
21r3r19r.
21r6r19r.
21r6r22r. x2
20g6g19g.
20g3g19g.
20g6g22g.
20g3r22g.
20g1g22g.
20g6r22g. x2
20r6g22g.
20r6r19g.
20r3r19g.
20r6g22g. x2
18r8g24g.
17r1g18r.
14r16g11g.
14r3g11r.
14r19r16g.
14r9r11g.
14g19g16g. x2
14g19r16g.
11r22g16r.
10g10r16r.
8g1r5r.
8g8g4r.
8g8r5r.
7r8g1r.
6g19g11r.
6r22r14r.
3g15r11r.
1r21g20r.
1r15g16r.
1r8r4g. x2
1r8g4g.
1g8r4g.

You can see from these that there are some exact repeats. There are also many common symbols (again leading to believe that this is direct substitution). I can't just at a glance spot a set of 8 using 3 red and three greens - if you can let me know.

[alexbutterfield]

Here is a list of freq. from text 1 and 2.
in brackets is the frequency of each letter as the last letter in a word.

1g x24 (5)
2g x4 (1)
3g x17 (1)
4g x10 (7)
5g x4 (2)
6g x20 (2)
7g x3 (1)
8g x25 (4)
9g x26 (9)
10g x15 (3)
11g x21 (11)
12g x1
13g x1
14g x18 (3)
15g x19 (6)
16g x31 (17)
17g x6 (4)
18g x7 (1)
19g x14 (6)
20g x22 (6)
21g x19 (6)
22g x22 (11)
23g x7
24g x11 (2)
25g x6
26g x1

1r x39 (5)
2r x7
3r x24 (5)
4r x6 (4)
5r x3 (3) (possibly Y)
6r x20
7r x4 (1)
8r x25 (2)
9r x22 (6)
10r x19 (6)
11r x24 (9)
12r x3
13r x4 (1)
14r x22 (10)
15r x15 (1)
16r x32 (12)
17r x13 (4)
18r x11 (3)
19r x19 (5)
20r x25 (9)
21r x15 (3)
22r x15 (6)
23r x8
24r x8 (2)
25r x2
26r x5 (1)

[Agent Lex]

Okay, exam revision is finally over since the exam was yesterday. I'll get at least 2 ciphertexts up tonight, promise promise. Also, because of my being away, the prize has been upgraded to a stroopwafel and a brownie!

[alexbutterfield]

I just wanna check. I'm not at the point of begging yet. But each text is using exactly the same encoding right. Not a variation on a theme?

[Agent Lex]

[chulo333]

Using Alex's frequency numbers, I did some frequency analysis graphs and found some interesting things. I'm fairly sure that the symbols can be rotated 180 degrees. I'm also fairly certain that red or green doesn't necessarily change the letter.

here's some graphs with comparisons. The first graph is half of the red symbols. The second graph is their 180 degree counterparts. The third graph is the same set of symbols as graph 1 but in green. The fourth graph is the green 180 degree counterparts. the last graph I counted both of the 180 counterparts and both colors together all as one symbol.

The corelations are pretty close throughout. I'm not sure what the colors are for or why the characters rotate, but I'm convinced that a symbol can be the same english character whether it rotates or changes color.

so maybe 8r, 8g, 1r, 1g all = i
and 10r, 10g, 15r, 15g all = a

However, that means that there are only 13 distinct symbols so far. Each symbol would have to be able to have 2 english character options.

so 8r, 8g, 1r, 1g = i OR n (example only. I'm not sure that the symbols actually are assigned these particular characters)

this would allow for that weird 2-letter double character word.

8r 8r in code1 could = "in"

I'm sure the colors and rotation come into play with this, but I'm not sure how yet.

just a theory so far. Let me know if you guys have any thoughts on this.

[alexbutterfield]

pretty interesting stuff chulo.

Though i dont think the cumulative graph is appropriate in promotng the hypothesis, as it is by definition going to fit any pattern- the pattern is undeniable.

I dont know why I fi d it harder to reconcile with half an alphabet (13 characters) than with two alphabets (52 symbols), I guess I think that if there are 4 sywmbols per 2 letters then there MUST be a system for choosing. My previous thought that there were 2symbols (one red one green) per letter could have relied on random choice. I have to reluctantly admit that a systematic approach seems likely and so a 4per2 system seems more likely than 2per1 - especially in light of these graphs which show a logical system for grouping symbols.

U got me thinking now. What's the best approach for finding letters for each group of4 symbols. If we are adding 2 letter freqs together as shown in the 5th graph, then there is no way to analyse by frequency.

Eg. Even if I could find a set of all 8 permutations of a 3 letter word, in 2per1 system, in a 4per2 system it would be meaningless.

While your hypothesis has a lot going for it. It elimntes all the ways of decodng I can think of. So I might have to continue checking for a 2per1 system if only to try and eliminate it.

If u have any ideas of where to go from here let me know.

[Agent Lex]

I wanna give you guys something to make up for lost ciphertexts, however I'm at work and I don't have the cipherfont. So instead I'll throw you some clues:

chulo has made a surprising discovery - surprising (to me) that he's on the right track, that is. It is a substitution, however your scope is still a bit too narrow.

The colours are, in fact, randomly determined. Nothing else about the cipher is random, though.

Spacing is as it was in the plaintext, so you can keep worrying about that 2-character word that's the same character twice

There is, in fact, a strong link between the two ciphertexts, and the future ciphertexts.

I am surprised that none of you have noticed something I noticed while drawing up the second ciphertext.

Also, I have chosen my words very carefully for the first hint.

I will give at least one more ciphertext tonight, and hopefully more over the weekend.

[alexbutterfield]