Posted on Leave a comment

Column Ciphers

Column Ciphers, or Poem ciphers have been in use by the intelligence services for some time.

Like one time ciphers they require shared knowledge of a code – but in this case the code isn’t used to painstakingly encode every single letter but used to create a column cipher.

How do you make column ciphers?

 The first step in creating a column cipher is to decide on the word you plan to use.  Intelligence agents will often use a long word from a book, which has been agreed upon, or a long word from a memorised poem.

Popular poems of the age were often used, but as they were published, it meant the opposition could easily decode the messages.  Eventually the intelligence services created poetry of their own they would memorise.  That way the code word could never be learned as the original source poem wasn’t published.   READ MORE HERE

Once you’ve picked your word you then need to decide on your message.  In this case we’ll use:

“Sometimes I hate ciphers”


Write the message (ignoring all spaces or using a hyphen for the spaces) into the grid.  The code word dictates the number of columns in  the grid.

Any spare squares should be filled with either a hyphen or a letter – it’s very much personal choice.

Now you need to number your code word, giving the number 1, to the letter which appears first in the alphabet, and the number 2 to the second letter.  In our case, A is the first letter, and as there is no B in our code word, C is our next letter.

Once you’ve numbered your letters you can rearrange the code word into alphabetical order.

Encoding and decoding the text would be easier if we used a word without repeating letters.  But if you assume the first letter of the pair will be the first letter numbered, you can still work it out!

Once the columns are rearranged to match the rearrangement of the code word you now have two ways to provide the code to your handler.

The traditional way is to read the code horisontally, so your message would be:


But you can also be sneaky and read vertically providing the message as:


So long as both the encoder and the receiver know which way the codes are being provided then they will get the right answer.  But you may find you need to try both methods to see which works for your message!

(HINT: If it’s a vertical read, then you really need to know how long the message is in order to work out how many rows you need to write before you move to the column, so you will want to count the characters in the message and divide the number by the number of letters in the code word to give you the right number of rows.)


Extra sneaky?

It was very common for secret agents to double encode messages, using one word to encode the initial message and then another word to re-encode the message. 

This works best if the two code words are of different lengths.

So keep an eye on your coded messages.  If at first you don’t succeed, you may need to find a second code word!

Posted on Leave a comment

Vigenère cipher

A Vigenere cipher is a form of letter substitution cipher that is incredibly difficult to break.

It operates by changing the cipher shift number on each letter used.

It is very easy to understand and use, but despite this it took 300 years before anyone was able to break it successfully. 

Many more complex ciphers (like Enigma) are based on the ever changing vigenere cipher which swaps one letter for another in a specific pattern.

To encode or decode it’s helpful to have a vigenere square to hand.

How do you crack a Vigenere Cipher?

It’s probably easiest to explain how you encode text first using a vigenere cipher so you can then understand how you would reverse it.

First of all you need to decide on the code words / sentences you want to use.  These need to be known by your receiver.  So if you are transmitting in vigenere code, you also need to find a method of transmitting the code word.

In Cosykiller we provide the code words in different ways, so to save a lot of lengthy trial and error work – you may find a website like THIS ONE will help you try a few different word combinations quickly.

For the purposes of the explanation we’re using the code word “Pigsmightfly”.

To make it easier to encode, I’ve highlighted the letters I’ll be using on the vigenere square.

Then you decide on the message you want to send.  In this case “JoLovesFloss”

You write the cipher text underneath the message so you know which letters provide the combination shift.



So to encode the letter J, you will be using the “P” line.

To encode the letter O, you will be using the “I” line.

Using the grid, you then cross reference the code word with the message.

If your code word is shorter than the message you want to write, all you do is repeat the codeword so that all the letters are matched.

Your encrypted message is formed by using the cross referenced letter.

If grid squares aren’t your thing, then you can use a code wheel, setting the code letters so that the letter A matches the correct letter of your code word and then reading off the new letter from the inner wheel.

To decrypt Vigenere, you work exactly in reverse.

Write out the codeword, write out the encrypted message and then cross reference it using the internal square to read off the original message on the outer edge of the grid.

Posted on Leave a comment

Caesar Ciphers

Caesar ciphers are one of the earliest known and simplest ciphers we use in Cosykiller.

They are very common and it is probably one of the first ciphers you will have come in contact with as it’s often used in education to teach problem solving.

It is a type of letter substitution cipher in which each letter in the message is ‘shifted’ a certain number of places down the alphabet. For example, if you used A+1 then the letter A would be replaced by the letter B, and the letter B would become C, and so on.

Caesar ciphers are so easy, we often use them in conjunction with other ciphers just to make them more complex to solve.

Of course, what makes them even more tricky is when we use a language with shared lettering e.g. Egyptian Hieroglyphs where F and S share a symbol for instance (That’s a hint).

Posted on Leave a comment

Directional codes

Directional codes are probably some of the easiest codes to work with in Cosykiller.

These codes rely on you reading the text in a way that isn’t the traditional left to right pattern.

The likelihood is you will have come across these style of codes before, as this is how acrostic poetry is made.  In acrostic poems you read the letters downwards, instead of the sentences across the page to spell out new words.

Sometimes (if we’re being mean) we use directional codes in conjunction with another code, so if you work out a letter substitution or another cipher and think it doesn’t seem “right” – it may well be we’ve also encoded it using a directional code.

We use the simple form of acrostic within Cosykiller too – so do keep an eye out specifically in letters, as sometimes these will form acrostic text as well as a generally newsy piece of writing.

You’ll find acrostics crop up in our first story “An Inheritance of murder”. 

The “Curse of Humanrah” contains a few more complicated directional ciphers inspired by hieroglyphs being written either left to right, or right to left depending on which direction the animals are facing (that’s a high hint), or vertically, as in the case of some stonework inscriptions.

Posted on Leave a comment

One Time Cipher Pads

One time Cipher Pads are one of the hardest ciphers to solve, and yes – sorry – they do crop up in cosykiller.

One Time Ciphers rely on the sender and receiver having the same pre-shared key that is longer than any message likely to be sent.  Usually after a single message is sent, the key is changed, but in Cosykiller we make things slightly easier by keeping the same encryption key throughout – which means you just need to find what it is both sender and receiver are using as their shared pad.

To make them easier to work with, most agencies use blocks of letters or numbers so that less transcription errors occur.  So if you spot blocks of letters, the chances are you’re looking at a one time cipher.

How do they work?

One time ciphers work by converting letters into numbers and back again.

So… first of all you need to convert the letters of your one time cipher pad into a number.  a=1, b=2 etc.

So if your one time cipher code was going to be a message like:


Then your numerical code would be:

15 14 5 20 9 13 5 3 9 16 8 5 18 19 19 20 9 14 11

Then you need to do the same for your actual message.

So if you wanted to send a message to a friend that read:

I hate ciphers

Your Number to letter cipher would read:

9 8 1 20 5 3 9 16 8 5 18 19

Now you engage in a spot of maths!

If you write all your numbers directly above each other, then you simply add them together, which means:

15+9 gives you your first NEW letter of the encrypted message.

15+9 is 24 – or in otherwords the letter X

14 +8 gives you your second encrypted letter – V

And so on.

If you end up with a letter pair that is greater then the alphabet e.g. 20+20 then all you do is subtract 26 from the number.  so your encrypted letter becomes 14, or N.

To decrypt you just do the same in reverse, but in this time subtracting your cipher from the encrypted message.

This will occasionally give you a minus number, but all you need to do is add on 26, to get back to the number it should be and convert back into letters.

You can find electronic help  converting between letters and numbers HERE which will speed you up as you’re trying to find out fast whether you’ve found the right cipher pad.

When it comes to a shared pad – it helps to find something that both people may own.  Sometimes this might be a bible, or a book of the same imprint.  For cosykiller – just remember it’s likely to be something you’ve seen before… and something that both individuals will have.


Posted on Leave a comment

Breaking Letter Substitution codes

A Letter Substitution Cipher replaces the letters in the alphabet with symbols or another random letter.

An example of a letter substitution cipher can be seen below, and we are going to solve it!

Cipher 1

Now there are 2 facts about the English language which will help considerably when solving a Cipher.
1 – The most common letter in the English language is E
2 – The most common 3 letter words are ANDTHE

Using the first fact have a look at the cipher above and see what the most common letter here is used?

Cipher 2

We can see that F is the most common Cipher letter, let us initially presume that F equals (This isn’t always the case, so you may need to try other common letters if you find it doesn’t work)…

Cipher 3

Then let’s use this information on the second fact (The most common 3 letter words in the English language are both ANDTHE)

I’ve highlighted below where 3 letters are repeated (MKe is repeated twice and so is ZVC)

Cipher 4

We can make a guess that the MKe one is ‘the‘ and we could also try replacing ZVC with the word ‘and‘.

Cipher 5

and we can now fill in the newly discovered letters and replace all the Z V and C‘s with a, n & d‘s (As per our guess that ZVC is actually the word ‘and‘), and replace all the M & K‘s with t‘s and h‘s (As per our guess that MKe is actually the word ‘the’)

Your Cipher should now look like the below image.

Cipher 6

Now I would say that the assumptions we have made earlier, have come to be true, as this is starting to look like English.
We still have a while to go but we are on the right track, so lets continue!

Now here comes the fun part. We have to study the Cipher and make assumptions.

The easiest one to guess is the Cipher letter that is on its own ‘E

Letters that are usually words on their own are A‘s and I‘s as in ‘I am going to the shops for A drink…’
We can rule out A on the basis that it has been used within the word AND. So we can guess E is actually I
Also we can see there are two 3 letter words ending in ‘he‘ – these are most likely either ‘the‘ or ‘she‘ and seen as we have used ‘the‘ earlier, we can assume it’s ‘she
also we have a 2 letter word beginning with ‘t‘ which is definitely going to be ‘to

Cipher 9
So, lets start replacing those highlighted letters…

Cipher 10

And then we can go through the Cipher replacing the letters with the newly found ones, as below…

Cipher 11

Now we can study and make assumptions, focusing on certain letters which look easy to solve…

Cipher 12

Cipher 13

Focus and assume

Cipher 15Cipher 16

Now we can guess the sentence based on the context…

Cipher 17


Congratulations, you have solved your first Cipher! It wasn’t that difficult was it?

Cipher 18


Posted on Leave a comment

Knock Code

Knock Codes, sometimes called Tap Code come up intermittently in our Cosykiller boxes.

The example below is a telegram written with a classic Knock Code cipher. (We used this on a cruise we did in 2018).

Knock Code ciphers encode messages using just 10 letters rather than the usual 26, which makes them quite easy to spot. If you find a message using just 10 letters, then the chances are it’s a knock code cipher created using a Polybius Square.

Polybius Square sounds super impressive, but in actuality it’s just a 5 x 5 square with the letters of the alphabet placed within it.  Some cryptographers use C/K in the same square, others use I/J in the same square.  It depends on the encoder as to their preference, whether it be phonetic, or visual.

The above shows a typical Polybius square.  It’s created by the five letter words Mango and Fruit.

To code your message, you would replace the letter you want within the grid, with the pair of letters on the XY axis.   e.g. If you wanted to encode the letter A, your code would be MF, if you wanted to encode the letter S your letter pair would be GU.  Decoding would be the same process in reverse.


It IS possible to decode Knock Codes without creating the square, though you’ll often find our 5 letter code words show up in other ways within the boxes to give you the code words.  Unfortunately this still won’t necessarily help you decode the messages until you know how the alphabet has been placed within the grid itself.  (And that’s something we don’t tell you – I know… we’re meanies).

Once you recognise the code however, all you need to do is solve it in the same way as a letter substitution code.

If the lettering confuses you, type it all into a word document and then use “find and replace” option and replace the two letter combinations with a symbol.


Letter substitution codes are hard if there are no gaps between words.

However we often use the knock code in telegrams as this gives you a hint on how to solve them.

In telegrams a full stop is transmitted using the word STOP which means if you find an 8 letter repeating section, the chances are the letters will be STOP.  This gives you 4 letters immediately.  After than you’re back to the usual method to decode substitution ciphers by looking at letter frequency and double letters.