Well, I'm really, really sorry to everyone who has spent time trying to tackle this puzzle. I made a HUGE mistake in the presentation. I honestly don't know how I managed to make a mistake this big, but in any case, I'm sorry to have wasted your time. What's worse is that I didn't even realize the mistake I had made and I posted comments encouraging people and telling them that they were on the right track.

The rubik's cube image is wrong. I don't really have an excuse, but this is what happened: during the development of this puzzle, I toyed with several different ways to incorporate the rubik's cube. One of the versions of the puzzle involved the rubik's cube image that I ended up using, but it isn't the one that matches the cipher text that I provided.

The correct image should feature a solved rubik's cube net numbered in the following fashion.

         [ 1][ 2]
         [ 3][ 4]

[ 5][ 6] [ 7][ 8] [ 9][10]
[11][12] [13][14] [15][16]



and the exact same unsolved rubik's cube, with no letters on it, and every square marked with it's appropriate number.

Again, I'm really, really sorry to have wasted people's time. This isn't quite the impression I had intended to make for my debut question on the site. I'm going to flag this question for closure as "off-topic," but I'm not sure that that is the correct course of action.

The intended solution is:

permute each block of the cipher text the same way that the cube squares are permuted

and then

Use the substitution defined by mapping the "in order" alphabet above the board to the first line of the cipher text.

My girlfriend teaches the third grade at a nearby elementary school. Today, I was supposed to meet her in her classroom after the school day ended, so we could go to the movies together.

However, when I got to her classroom, I couldn't find her. Instead,this message was printed on the white board:


On her desk, there was a 2x2x2 rubik's cube with some writing on it. I'm not sure if it is important, but I've reproduced the net of the cube as I found it, as well as the net of the solved state.

Solved Cube: Solved Cube

Her Cube, as found: Updated Cube

It appears that she had written the lettering on herself.

A final detail in the room struck me as out of place. Above the white board, there were cards displaying the cursive alphabet, e.g. Aa, Bb, Cc..., however it appeared that the final letters Yy and Zz were missing.

All in all, the board looked something like this:

Me looking at board

Can you help me decipher the message on the board?

Edit: I modified her cube so that one of two logical orderings is forced. This does not change the answer, but reduces the trial and error to find it.

I also added another image which might be helpful.


Permutation(s) are involved (pretty obvious because of the rubik's cube.), but just permutations will not solve this.

Bigger hint:

The first line of cipher text is not actually part of the message! It is there to help you decipher the following two lines.

  • 2
    $\begingroup$ Welcome to puzzling! This a great puzzle for a new user! $\endgroup$ Commented Oct 9, 2016 at 12:58
  • 3
    $\begingroup$ @BeastlyGerbil Thanks. Lurking pays off eventually. $\endgroup$
    – Liam
    Commented Oct 9, 2016 at 13:01
  • 3
    $\begingroup$ I'm sorry to say mate, but I don't think she wants to be found :( $\endgroup$
    – n00dles
    Commented Oct 9, 2016 at 19:08
  • 3
    $\begingroup$ She really doesn't want to go to the movies, does she? $\endgroup$
    – QBrute
    Commented Oct 9, 2016 at 21:45
  • 1
    $\begingroup$ Do you think you can fix this puzzle rather than closing it? If so, I'd suggest doing that (by replacing the erroneous images with the correct ones). Seems like some solvers were on their way to figuring it out. $\endgroup$ Commented Oct 10, 2016 at 18:28

5 Answers 5


If we solve the cube with the writing on it we get:

enter image description here

This suggest that each section is a letter in the alphabet, Liam has said also said the alphabet goes in a snaking pattern:

enter image description here

And the pattern:

enter image description here

This is why Y and Z cards are missing, there are only 24 squares

This also means you girlfriend's cube looks like this:

enter image description here

Now the cipher...

We know it can't just be from the cubes because there is a 'y' in the cipher but obviously not in the cubes, (also translating just gives gibberish)

  • $\begingroup$ This is correct and a necessary step. $\endgroup$
    – Liam
    Commented Oct 9, 2016 at 13:35
  • $\begingroup$ @Liam I have the next bit just writing it up... $\endgroup$ Commented Oct 9, 2016 at 13:36
  • $\begingroup$ Interesting, to observe that "K,.,M,N,O" are in the natural order. So one would guess the place of the letter L instantly. Although I have no clue what this would mean, if anything. $\endgroup$
    – Matsmath
    Commented Oct 9, 2016 at 13:40
  • $\begingroup$ @BeastlyGerbil I'm going to give you a hint because this is my fault. In hindsight I should have made this clearer. The ordering should be snaking AB and then next line DC, etc. I should have added letters on different lines to force a single valid ordering. $\endgroup$
    – Liam
    Commented Oct 9, 2016 at 13:48
  • $\begingroup$ @Liam oops, so I have it the wrong way round? $\endgroup$ Commented Oct 9, 2016 at 13:50

In case anyone is wondering, it says:

dear liam i ran to the bathroom please do not touch my stuff

OP gave away the solve path. What went wrong was that the letter cues in the "Her Cube" image were incorrect, and led solvers to use an incorrect reading order for deriving the permutation.

The correct permutation is derived like so:

  1. Write the alphabet onto the solved, flattened Rubik's cube in normal reading order (left to right, then top to bottom)
  2. "Unsolve" the cube to get it into the color configuration shown in "Her Cube". Ignore the letter cues.
  3. Flatten the cube again and read the permuted alphabet off of the cube in normal reading order.

The letters on each completed image should look like this:

KLMNOP        "Solved"

NLMPVB        "Her Cube"

This permutation key is used to reorder all three lines on the blackboard:

S. Key:   vmrdtonhilxwagfsecjkbupq
Cipher 1: oowaoijhbhjxcdbrrccbsjdo
Cipher 2: jwoiyaljrtjccmpgdtpxalcr

S. Key:   osmirlqdhkvwaxjgebtcpfnu
Cipher 1: irobwhoahboxcjcdrsocdbjj
Cipher 2: agwrotrijxjccjpmdaytcpll

The first line on the blackboard is a substitution key. After applying the substitution:

Cipher:  osmirlqdhkvwaxjgebtcpfnu


This is not an answer yet, but I couldn't just keep on posting all my observations in comments so I will do it here.

The first line have 1 of each of the 24 letters which seems to indicate another level of permutation or something. I have yet to find out how to use it but it will be important to figure out if the first level of permutation must be used on it before it is usable.
I also noticed that there is a y in the message, which can be a hint toward something or simply meant to be left as a y.

I tried many many things which all let to pure giberrish. First, since the first line has 24 different letters, and the message has 25(with the y), I assumed that the whole cube puzzle was meant to decode the first line, and then we could use the first line to solve the message, but it has not worked.

  • $\begingroup$ Your first observation is very important, especially given that I tried to hint (via proximity) that that first line might be used with the alphabet string above the board. $\endgroup$
    – Liam
    Commented Oct 9, 2016 at 14:50

Now that the rubik cube has been solved, I tried various permutaions (spoilers have some weird interaction with other formatting, but this is not the solution yet)

sorting the three lines:
abcdefghijklmnopqrstuvwx <= same as on the cube but permuted differently
abbbcccddhhijjjoooorrswx <= not all letters present, many duplicates but not enough vowels


writing them on the cube in the solved snakified order:

  vm        oo        jw  
  dr        aw        io  
tonhil    oijhbh    yaljrt  
sfgawx    rbdcxj    gpmccj  
  ec        rc        dt  
  kj        bc        xp  
  bu        sj        al  
  qp        od        rc  

...then rotating the cube to the original position


  ox        ij        aj
  mi        ob        wr
dlprhj    ahdwhc    itcojp
vfgska    obdrbc    jpmgxc
  eb        rs        da
  tc        oc        yt
  qw        ox        rc
  nu        jj        ll

writing them on the unsolved cube then solving it


  sd        ra        gi
  th        oh        yj
kvqiro    boobwi    xjrroa
afgxum    cbdjjo    cpmjlw
  ej        rc        dp
  wl        xh        ct
  cp        cd        tc
  bn        sj        al

treating the first line as a cube permutation and [un-]solving the other two lines


  ca        cs
  it        ac
dpmjrp    rbdhbc
calwtx    dijohb
  ro        ow
  yg        or
  lj        jo
  jc        jx


  jc        lc
  ac        it
bbdhbx    xpmjrc
cijodo    paljcr
  ow        yo
  jh        jt
  os        wa
  rr        dg

none of this seems to produce any meaningful words even if reading in any direction on the surface of the cube

if the second and third line are both part of the message, there must be an additonal step to replace the letters and not simply permute them. There is not enough data for frequency analysis, but roughly mapping the top 6 "cjoradb" to the most frequent letters "etaoins" makes some of the messages almost readable (with phrases like "awaiting you at the temple" scattered around if you squint). Mapping the alphabet, first line and cube to each other, however, keeps it unreadable.


My immediate thought is

an XOR cipher where the first line is the key, the second and third line XOR to each other, and give the plain-text out.

But it just gives you nonsense.

ALL of which are system characters. (Used this to decode this)

19 02  05  05  1b  06  04  00  0b 04  12 0f 02  03  04  01  17  00  09 09 11  1f 14 1e

Which is nonsense, because you have a text that never ends (1st 02), no SOT's for the 3 EOT's, and the device turns on AFTER a bunch of instructions, which leads me to believe the whole sequence of bytes before that DC1 instruction (190205051b0604000b04120f0203040117000909) is a start up sequence.

if you say the first line is needed to encode the message, then it must be a Vigenere or Variant Beaufort, but they don't do a thing.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.