unforum

[Guin]

just a thought having been tinkering with the new find - could it not be UHUFFLED X FOURTH PRIME? Or rather SHUFFLED x 7 as the key. Have tried SHUFFLEDSHUFFLEDSHUFFLEDSHUFFLEDSHUFFLEDSHUFFLEDSHUFFLED and UHUFFLEDUHUFFLEDUHUFFLEDUHUFFLEDUHUFFLEDUHUFFLEDUHUFFLED but to no avail using standard 108 characters - I am now working on variations of the 108.

EDIT - seven times solitaired /EDIT

Another thought - again so far to no avail - wouldnt an outward rifle be more appropriate given the clue? I have used teh deck as presented in the new motor pack (JJDCHS)

so this is 53, 40, 28, 3, 17, 45, 33, 8, 22, 50, 38, 13, 1, 15, 43, 31, 6, 20, 48, 36, 11, 25, 54, 41, 29, 4, 18, 46, 34, 9, 23, 51, 39, 27, 2, 16, 44, 32, 7, 21, 49, 37, 12, 26, 14, 42, 30, 5, 19, 47, 35, 10, 24, 52,

So far nothing that provides anything of interest but i do feel we are close

I am however still concerned about the rotting to get the anagram to work it sort of feels wrong in some ways - not as clean as I had hoped. If it is correct it is truly an incredible moment of genius so good find that man (based on the hours ive played with this it wouldnt suprise me that something like this was staring me in the face all this time!)

But until the large lady sings I am happy to keep trying any new avenues of ideas

[themandotcom]

This may or may not help, but a quick anagram search revealed a word that seemed too good to be true.


DE SHUFFLE FOR TRIUMPH
Any help? I hope so!

[Monkeynavigated]

I wrote a program to compute the distance between various options of cipher text, and a uniform distribution of letters. KL is Kullback-Leibler distance (or relative entropy). I then compared that to 10000000 (overkill, but hey it's a computer) random sequences of letters of the appropriate length. Given this output is correct, there is about a 7.7% chance that the even numbers are a random sequence of letters (as would be generated by solitaire for instance). This is unfortunately a bit too high to declare the even letters a statistical anomaly but it still lends some strength to the argument that the even numbers are what we should be focusing on.

My biggest issue with declaring the even numbers non-random is that it means that the solitaire cipher is not used to solve the puzzle. This makes the star on the joker either a red herring or it is used in some other way. As a big fan of Cryptonomicon I am still hoping that solitaire is part of this card.

I have been looking at ways of systematically unscrambling the disappearing text to get a meaningful message but so far have come up with nothing. If it is a sequence of in and out shuffles (or unshuffles) then it is quite complicated, requiring at least 6 shuffles (as best as I can tell).

I am going to change the program to measure the distance from the distribution of letters in the English language to see if the fact that the even letters appear to be very close to that distribution is a statistical anomaly or not. If the even letters (or some other subset of letters) can be shown to be non-random that has huge implications for finding the solution to this card, namely that solitaire isn't used.

Here is the output from my program for those who are interested (all KL, Delta and X^2 distances are distances from a uniform distribution):

54 Characters, 10000000 random distributions:

BMHFBXQWZLTPLHYTWLOYNYBOIUMXHUAXNHSAJFTMMOMBAPVWOMKPLL
KL distance: 0.368505
Delta distance: 0.159044
X^2 distance: 0.430463
Samples with higher KL: 771646
Samples with higher delta: 1662664
Samples with higher X^2: 1776443

WBCGILCYEFIHJFWEKYJTYGJIGFVRIGRGHQRXWUJASMSVAKGVXVYZLU
KL distance: 0.293102
Delta distance: 0.124832
X^2 distance: 0.344096
Samples with higher KL: 3196737
Samples with higher delta: 5782451
Samples with higher X^2: 4493550

Random Sequences:
KL mean: 0.264221 and KL stddev: 0.0701647
Delta mean: 0.131927 and Delta stddev: 0.0282511
X^2 mean: 0.342693 and X^2 stddev: 0.1015

108 Characters, 10000000 random distributions:

WBBMCHGFIBLXCQYWEZFLITHPJLFHWYETKWYLJOTYYNGYJBIOGIFUVMRXIHGURAGXHNQHRSXAWJUFJTAMSMMOSMVBAAKPGVVWXOVMYKZPLUL
KL distance: 0.114821
Delta distance: 0.0764411
X^2 distance: 0.151868
Samples with higher KL: 5549038
Samples with higher delta: 8776607
Samples with higher X^2: 8681199

Random Sequences:
KL mean: 0.122971 and KL stddev: 0.0354518
Delta mean: 0.107811 and Delta stddev: 0.0275172
X^2 mean: 0.216996 and X^2 stddev: 0.0601189

Press <enter> to continue . . .

Sorry my thoughts in this posts are kinda fractured…

[cheesey]

just a detail that I hope will make a huge difference, and will be the whole basis of solving this card!

You have missed out an L at the end of the full WBBMC...

It should end YKZPL LUL

nice work tho!

[Guin]

[BBuck]

Been thinking a lot about Kian's method, and (risking severe ridicule), I don't think that it is going the right way - well, not quite.

Like a couple of others who have posted, I am a little bit wary about the "-2" on one of the "U"'s and "G"s. This is fairly contrived.

However, I do think there's a good chance that anagramming the 20 letters is the way forward. This would tie in with the clue in the Sentinel and with the whispered message on the card. As I've posted before, I don't think the message solely relates to the heat-sensitive letters, as they could be discovered by accident.

Anagramming 20 letters is not easy, especially when we don't know what we're looking for. With a certain other card, there were some likely cribs that led to the answer.

One possibility (though I can't get anywhere with it yet) is that we should read the whispered message almost literally. "Entropy" = anagramming the 20 letters, to leave a message that in some way links to "cold". eg Fridge can be made from the 20, but not much else sensible.

Another is to look for instructions/description. The X might be a null, but MIX also comes out. So it might be something like:

MIXED UP THROUGH FFFLUR

Where the last six, in some order (120 possibilities? Or fewer because of the repetition) is the keyphrase.

I thought I had it when I came across

MIX UP THROUGH RED FLUFF

but REDFLUFF or FLUFFRED doesn't work as a key.

[manleym]

[Guin]

even though I am sceptical about the prime anagram I am trying something for completeness.

So far I have tried just about every conceivable method I can rationally think of using both standard and actual deck orders as well as keying the deck in a many and varied manner:

Shuffling the deck in sevens, using the word SHUFFLED 7 times as the key, solitairing the text 7 times using a as many possible ideas I can find / remeber, distributing the text in sevens and currently (and possibly my most insane idea!) every possible distribution by shuffeling EVENS (120 possibilities including duplicates) to see if thats the key - although I am certain it wont be - but I am too curious by nature not to try it and it was in my dreams - yep they are back this card has a lot to answer for

My only other concern is that once I get into the routine of doing these mad things I will find an answer and not remmeber which process I had used lol

Well the quest continues

[e_nygma]

[Monkeynavigated]

I finished a run of the program comparing the even numbered letters to the distribution of letters in the English alphabet. Using different measures of distance there is between a 3.5% and 5.5% chance that a random sequence is as close to the distribution of English letters as the even numbers letters are. This is compared to the fact that the full sequence of 108 letters is much farther away from the English distribution.

To summarize:
Chance the even numbered letters are random when compared to a uniform distribution is: ~7.71%
Chance the even numbered letters are random when compared to the distribution of letters in the English language ~5.48%

Chance the odd numbered letters are random when compared to a uniform distribution is: ~32.0%
Chance the odd numbered letters are random when compared to the distribution of letters in the English language ~33.0%

Chance all 108 letters are random when compared to a uniform distribution is: ~55.5%
Chance all 108 letters are random when compared to the distribution of letters in the English language ~80%

Although this is strong evidence that the even numbers letters are statistically significant it is in no way conclusive...

Chance here is a somewhat ambiguous term I realize, it basically represents how often a randomly generated sequence (using a uniform distribution) has features similar to different subsets of the cipher text. In this case the features I am using are the KL distances to different distributions.

[kian]

Encrypting SHUFFLED with the standard bridge order yields WEEDNKWJ, which has a beginning very similar to WBBMC... I wonder if the message might start with the word SHUFFLED and the correct deck order be a slight variation on the standard bridge order. I've tried swapping the positions of the four sevens, but only for bridge order.

Feel free to trout me on this but... With a known ciphertext and a guess at what might be the start of the message, we can work backwards to get the keystream and then try to figure out what deck order would generate that. Though it's difficult to get beyond the first handful of letters, at least by hand. Similarly, we can make a guess at the end of the message (maybe by a dictionary file) and work backwards to try to figure out a deck order that might work.

I've also been wondering about the apparent runs of letters in alphabetical (or reverse) order. There are sequences of HGF, IGH, RS, X__Y__Z, VWX___Y_Z... Is this a normal consequence of the solitaire cipher? Or coincidence? Or maybe was there an alphabet or two somehow shuffled in to mask the letter frequency (not sure about the lack of D's).

[brokasaphasia]

[Guin]

Just a point on the random distribution of even letters - in a solitaire you key the deck each time so

AAAAAAAAAA with no passkey will generate DWJXHYRFDG
AAAAAAAAAA with the passkey SHUFFLED will generate FVDAPQDHON

Unlike a substitution cipher you cant try to work out what letter = what and I think its the same for the distribution of the letters in a solitaire as you are always dependant on the top card of the deck each time when you key the deck - if that makes any sense

Having re-read a lot of the thread last night to find some further pointers 2 things about the Sentinel article that stand out:

Core and Reframe - they seem odd words to use. I can see how tey would be justifiable - get to the heart of the puzzle - rethink your methods - but these 2 words seem somewhat contrived - they fit but they seem important - perhaps its the semantics of it and im reading too much into it.

[FranG]

A fresh look

[manleym]

Re: A fresh look