- Overview
- Ciphers
- Substitution Cipher
- Aristocrat Cipher: Practice for Science Olympiad Codebusters
Aristocrat Cipher: Practice for Science Olympiad Codebusters
Generate a new Aristocrat, Patristocrat or Xenocrypt for Division B or C, solve it against the clock, and get it marked the way Codebusters marks it: two errors free, then 100 points per error, plus the timing bonus.
On This Page
Article sections
On This Page
An Aristocrat cipher is a simple substitution cryptogram with the spaces left in, and it is the first question on every Science Olympiad Codebusters exam. This page generates a fresh Aristocrat cipher, Patristocrat or Xenocrypt whenever you want one, built to the 2026 rules for Division B or Division C. It times you as the event supervisor would and marks your answer the way the rules do: two mistakes are free, and every mistake after that costs 100 points. When you are stuck, one click sends the puzzle to the substitution cipher solver so you can see how it was broken.
How to use the Aristocrat practice tool
- Pick your division. Division B and Division C allow different alphabets, so the tool only offers what your division can be given. More on that in the table below.
- Pick the cipher type. An Aristocrat cipher keeps the word breaks. A Patristocrat removes them and prints the letters in groups of five. A Xenocrypt is an Aristocrat cipher in Spanish.
- Pick the alphabet. K1, K2, K3 or a random alphabet, as your division allows. Choose a hint if you want a head start: one plaintext word, or three letters.
- Start the timer and solve. Type a letter under any ciphertext letter and the same guess fills in under every copy of that letter, just as it would if you were working in pencil on a real Aristocrat cipher. The frequency table above the puzzle counts every ciphertext letter and shows your current guess under each one; you can type there too.
- Check your answer. The tool counts your errors, highlights them, and scores the puzzle as a 400-point question. For an Aristocrat cipher it also shows the timing bonus you would have earned if this were the timed first question.
- Find the keyword. For a K1, K2 or K3 puzzle, type the keyword you think built the alphabet and check it. Division C exams sometimes ask for the keyword instead of the plaintext.
- Review. Show solution gives the plaintext, the author and the book. Open in solver hands the ciphertext to the solver, which shows its letter counts, its answer and the keyword it reads from the key.
Every quotation comes from a public-domain book on Project Gutenberg, copied word for word: Lewis Carroll, Jane Austen, Mark Twain, Louisa May Alcott, Frances Hodgson Burnett and others for English, and Benito Pérez Galdós, Juan Valera, Vicente Blasco Ibáñez, Juan Ramón Jiménez and others for Spanish. Each one is between 60 and 160 letters long, the usual range for a Codebusters cryptogram.
What is an Aristocrat cipher?
The name comes from the American Cryptogram Association (ACA), whose members have been trading hand-made cryptograms for nearly a century. The ACA sorts substitution puzzles by how they are written. An Aristocrat is the friendliest kind: a quotation encrypted with one substitution alphabet, with spaces and punctuation left exactly where they were. Science Olympiad borrowed the ACA's terms when it created Codebusters, so the Aristocrat cipher on a Codebusters exam is the same thing.
Three rules from the 2026 Codebusters rules decide how an Aristocrat cipher can look:
- One alphabet for the whole message. Each ciphertext letter always stands for the same plaintext letter.
- No letter decrypts to itself. If ciphertext Q appears, it is not plaintext Q. That rule alone removes a candidate for every letter in the puzzle.
- The alphabet is K1, K2, K3 or random. Keyword alphabets are explained below. A random alphabet is any shuffle that obeys the previous rule.
Because the word breaks survive, an Aristocrat cipher is solved mostly through words: one-letter words are A or I, apostrophes usually sit before S, T, D, LL or RE, and repeated patterns such as THAT (ABCA) or PEOPLE (ABCADB) give themselves away. The substitution cipher solver page works through a full Aristocrat cipher by hand, step by step.
Patristocrat: the same cipher without word breaks
A Patristocrat is encrypted exactly like an Aristocrat cipher, but the spaces and punctuation are removed and the letters are printed in blocks of five. Here is a real practice Patristocrat from this tool, a 110-letter sentence from Pollyanna (1913) on a K1 alphabet:
QFQHB VHFSF GOJLF QIVPF IJVLR
UWZRU NRKXQ SFHND QEWLJ GRRNV
WLJSF WHLJI XQIJU EFUPJ QWVFQ
IWLJK JZSDW DKXOO BXQFW WUFHW
DYJIU JVVJV
Without word lengths you cannot start from short words, so a Patristocrat is attacked through letter groups instead of the word patterns that open up an Aristocrat cipher:
- Count first. Without word breaks the counts are your main clue. Here J appears 12 times, F 11 times and W 10 times, so those three are good candidates for E, T, A or O.
- Look for repeated trigrams. THE, AND, ING, ION and ENT are the most common three-letter groups in English. In the puzzle above, WLJ appears three times and ends in J, the most common letter: a strong candidate for THE. FQI appears twice and begins with F, the second most common letter: try AND.
- Use the doubled letters. LL, EE, SS, OO and TT are the common doubles. The puzzle has four doubles, RR, OO, VV and WW. If W is T, then WW is TT, and the other three are worth testing against LL, SS and OO.
- Mark the word ends you find. Once a word is clear, draw a bar after it. Every bar makes the next word easier to see.
With THE and AND in place, words such as BOOKS, HANDS and DRESSES appear quickly. The plaintext is Nancy's capable hands made short work of unpacking the books, the patched undergarments, and the few pitifully unattractive dresses, and the K1 keyword is ZIGZAG with a shift of 2.
A Patristocrat is much harder than an Aristocrat cipher when the text is short. In our measurements on unseen books, the solver found the right answer for 93% of 80-letter Patristocrats but fewer than half at 50 letters, against 83% for an Aristocrat cipher of 50 letters. People feel the same difference. Codebusters limits Patristocrats to K1 and K2 alphabets, so once the start of the keyword shows up in your partly solved alphabet, you can often guess the rest and fill in several letters at once.
Xenocrypt: an Aristocrat cipher in Spanish
A Xenocrypt is an Aristocrat cipher whose plaintext is in a language other than English. In Codebusters that language is Spanish, and the alphabet is built from an English keyword. Division C exams can include up to two Xenocrypts at regional level and two or three at state and national level; Division B exams have at most one at regionals and one or two at higher levels.
Spanish changes the starting points you rely on in English:
- E and A are both very common, followed by O, S, R and N. The resource sheet handed out at the event lists Spanish letter frequencies, so you will not need to memorise them.
- One-letter words are Y, A and O (and, to, or), not A and I.
- The common short words are different: DE, LA, QUE, EL, EN, LOS, DEL, SE, LAS, POR, UN, CON. QUE is the Spanish THE: frequent, three letters, and in almost every sentence.
- Accents are dropped and Ñ is treated as N, following the ACA convention for Spanish cryptograms. An accented É in the original is just E in the puzzle.
Even if you do not speak Spanish, those few facts are usually enough to break a Xenocrypt, because Spanish spelling is very regular. Solve a few Xenocrypts here with the hint turned on, then without it.
K1, K2 and K3 alphabets
Codebusters, like the ACA, builds keyword alphabets in three standard ways. All three start from the same keyed alphabet: write the keyword, drop any repeated letters, then add the unused letters in order. For the keyword CODEBUSTER that gives:
C O D E B U S T R A F G H I J K L M N P Q V W X Y Z
The three alphabet types differ in where that keyed alphabet goes. The examples below use a shift chosen so that no letter maps to itself, as the rules require.
K1 alphabet: keyed plaintext
In a K1 alphabet the keyed alphabet is the plaintext row and the ciphertext row is the ordinary alphabet, shifted:
Plain: C O D E B U S T R A F G H I J K L M N P Q V W X Y Z
Cipher: E F G H I J K L M N O P Q R S T U V W X Y Z A B C D
MEET ME AT THE LIBRARY AFTER SCHOOL becomes VHHL VH NL LQH URIMNMC NOLHM KEQFFU.
To find a K1 keyword after solving, write the ciphertext alphabet A to Z and put the plaintext letter you found under each one. The keyword appears in that bottom row, here as WXYZCODEBUSTRAFGH…. The four letters before it are just the end of the alphabet wrapping around because of the shift.
K2 alphabet: keyed ciphertext
In a K2 alphabet it is the other way round: the plaintext row is ordinary and the keyed alphabet is the ciphertext row:
Plain: A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Cipher: W X Y Z C O D E B U S T R A F G H I J K L M N P Q V
The same message becomes RCCK RC WK KEC TBXIWIQ WOKCI JYEFFT. To find a K2 keyword, write the plaintext alphabet A to Z and put the ciphertext letter under each one, which is simply the encryption key. CODEBUSTR is readable straight away.
The two layouts are each other's mirror image. The same keyword and shift used as K1 and as K2 give an encryption key and its decryption key.
K3 alphabet: the same keyed alphabet twice
In a K3 alphabet both rows are the keyed alphabet, with the lower row shifted:
Plain: C O D E B U S T R A F G H I J K L M N P Q V W X Y Z
Cipher: O D E B U S T R A F G H I J K L M N P Q V W X Y Z C
Now MEET ME AT THE LIBRARY AFTER SCHOOL becomes NBBR NB FR RIB MJUAFAZ FGRBA TOIDDM. Neither row of the solved key is in alphabetical order, so the keyword does not appear by simply listing the key. Instead, follow a chain: start at any letter, look up what it encrypts to, then look up what that letter encrypts to, and so on. With a shift of 1, the chain from C runs C → O → D → E → B → U → S → T → R → A…, which is the keyed alphabet itself. With a larger shift the chain visits every second, third or fourth letter of the keyed alphabet, and you space the letters out accordingly. K3 alphabets appear only in a Division C Aristocrat cipher, never in a Patristocrat or a Xenocrypt.
The substitution cipher solver tries all three layouts and every shift on the key it recovers and reports the keyword it finds, or says the alphabet looks random.
Division B and Division C
The two divisions use the same cipher types with different alphabets. This summary follows the 2026 rules manuals for both divisions:
| Cipher | Division B | Division C |
|---|---|---|
| Aristocrat | K1 or random | K1, K2, K3 or random |
| Patristocrat | K1 | K1 or K2 |
| Xenocrypt | K1 | K1 or K2 |
| Keyword asked instead of plaintext | No | Possible for an Aristocrat cipher on a K1, K2 or K3 alphabet |
Both divisions may give a hint with a monoalphabetic cipher, and in both, no letter can decrypt to itself.
How Codebusters scores an Aristocrat cipher
The scoring rules are short, and knowing them changes how you work under time pressure.
Plaintext answers. Two or fewer wrong letters count as fully correct. Each extra error costs 100 points, and a question cannot go below zero. A 400-point Aristocrat cipher with five errors earns 100 points; with seven errors it earns nothing. So once a quotation is readable, it is usually better to move on than to polish the last rare letter.
Keyword answers. When a question asks for the keyword or key phrase, every error costs 100 points from the first one. There are no free mistakes.
The timed question. Question one is always an Aristocrat cipher, and it is timed from the start of the event. A team member signals when it is solved, and if the answer has two errors or fewer the team earns a bonus of 2 × (600 − seconds taken). Solving it in 6 minutes is worth 2 × (600 − 360) = 480 points; in 435 seconds, 330 points. A wrong answer has no penalty and you can try again, but after 10 minutes the bonus is zero. The practice timer on this page shows the same bonus so you can track your speed from one session to the next.
The rules also allow up to three Class I or Class II calculators but no other resources; everything you need for an Aristocrat cipher has to be in your head and on the resource sheet.
Tips for solving an Aristocrat cipher quickly
- Start with the shortest words. A lone letter is A or I. Two-letter words come from a short list: OF, TO, IN, IT, IS, BE, AS, AT, SO, WE, HE, BY, OR, ON, DO, IF, ME, MY, UP, AN, GO, NO, US, AM.
- Use "no letter decrypts to itself". If a one-letter word is the ciphertext letter A, it must be I. If it is I, it must be A.
- Look for word patterns. THAT, THERE, PEOPLE, LITTLE, SHOULD and EVERY all have distinctive repeated-letter patterns. Learning a dozen of them saves minutes.
- Watch apostrophes. A single letter after an apostrophe is almost always S or T; two letters are usually LL, RE or VE.
- Fill the frequency table as you go. Writing each guess under its ciphertext letter in the table shows at a glance which plaintext letters you have already used, which stops you from giving two ciphertext letters the same answer.
- For a keyword alphabet, use the keyword early. Once you see the start of a keyword in the partly solved alphabet, guess the rest of it and fill in the remaining letters at once.
Practise each type a few times with the timer running, then check your answers against the solver. The aim is to make short words and patterns automatic, because on the timed Aristocrat cipher every second saved is worth two points.
Aristocrat cipher and Codebusters questions
What is the difference between an Aristocrat and a Patristocrat?
Both are simple substitution ciphers. An Aristocrat keeps the spaces and punctuation of the original, so you can see word lengths. A Patristocrat removes them and prints the letters in groups of five, so you have to find the words yourself. Codebusters also restricts Patristocrats to K1 alphabets in Division B and K1 or K2 in Division C.
Is the K1 alphabet the same as a keyword cipher?
Close. Both start from a keyed alphabet: the keyword without repeated letters, then the rest of the alphabet in order. In a K1 alphabet that keyed alphabet is the plaintext row and the ordinary alphabet, shifted, is the ciphertext row. A classic keyword cipher puts the keyed alphabet in the ciphertext row instead, which is the K2 layout.
How is the timing bonus calculated?
Only the first question, which is always an Aristocrat, is timed. If your team solves it with two errors or fewer, the bonus is 2 × (600 − seconds taken). Six minutes gives 480 points; after ten minutes the bonus is zero, although you can still answer the question for its normal points.
Do I need to know Spanish for the Xenocrypt?
Not really. Learn the one-letter words Y, A and O, the very common words DE, LA, QUE, EL and EN, and that E and A are both frequent. Spanish spelling is regular enough that those clues usually carry you through. Accents are dropped and Ñ is written as N.
Where do the practice quotations come from?
Every quotation is copied word for word from a public-domain book on Project Gutenberg, published before 1929: English authors such as Lewis Carroll, Jane Austen and Mark Twain, and Spanish authors such as Benito Pérez Galdós and Juan Ramón Jiménez. Show solution names the author, book and year.
Can I use a solver or this site during a Codebusters event?
No. The 2026 rules allow only the materials the event supervisor provides, plus up to three Class I or Class II calculators. This page is for practice beforehand; use the solver afterwards to check your work and see where you went wrong.