What is frequency analysis?
Frequency analysis means counting how often each letter appears in a text and comparing that with how often it normally appears in a language. English uses E for about 12.7% of its letters and Z for less than 0.1%. A simple cipher swaps the letters and leaves those proportions alone, so the pattern of the original language shows through. That makes counting the usual first step against a Caesar cipher or a substitution cipher.
How to read the chart
Each column is one letter. The light bar is your text, and the darker column behind it, with a line on top, is what the chosen language expects. When the two roughly agree, you are probably looking at ordinary text or a cipher that has not moved the letters. When tall bars stand over short columns, something has moved them. Select a column to read its exact numbers. The tool counts A to Z only, so accented letters are left out.
Spotting a Caesar shift
This page opens on a coded sample. Its tallest bar stands at H, which appears 35 times in 195 letters, while the English profile peaks at E. H is three places after E, and the second peak has moved the same distance, from T to W, so the likely key is a shift of 3. To check, run the text through the Caesar solver, which tests every shift.
Working on a substitution cipher
When the letters are mixed and no shift fits, switch the order to Most common first. Your letters line up by rank above the language's letters, and each pair is a first guess: your most common letter may stand for E, the next for T. Expect some to be wrong, because neighbouring ranks often trade places, so test them on short words and repeated patterns. Once you have worked out a full key, put it into the substitution tool to read the whole text.
When counting stops working
Short texts mislead, because counts settle only after a few hundred letters. Unusual subjects, lists of names and mixed languages push the numbers around too. Vigenère and other ciphers that change alphabet as they go spread the counts out, so the chart looks flatter than any language. Modern encryption such as AES leaves no pattern of this kind, which is one reason it protects real data and these classical ciphers do not.
Most common letters by language
These are the eight most common letters in each profile the tool uses, from the most common down. The values are approximate. Accented letters are not counted, so the Polish and French profiles add up to a little under 100%.
| Language | Most common letters |
|---|---|
| English | E, T, A, O, I, N, S, H |
| Polish | A, I, E, O, Z, N, R, W |
| German | E, N, I, S, R, A, T, D |
| Spanish | E, A, O, S, R, N, I, D |
| French | E, S, A, I, T, N, R, U |
Where the method comes from
The oldest known description comes from the 9th-century Arab scholar Al-Kindi, in his manuscript on deciphering cryptographic messages. He advised finding a long sample of the same language, counting its letters, then matching the most common letter of the coded text to the most common letter of the sample. The tool on this page does the counting half of that work.
Frequently asked questions
Which letter is most common in English?
E, at about 12.7% of letters in the profile used here, followed by T at 9.1% and A at 8.2%.
Can I use this as a plain letter counter?
Yes. Paste any text and the table lists how many times each letter from A to Z appears, with its share of the total. You can copy the table into a spreadsheet.
How many letters do I need?
There is no exact limit, but a few hundred letters give a far steadier picture than a sentence or two. The tool warns you below 100 letters.
Will this tool solve my cipher?
No. It counts letters and compares them, and the conclusions are yours to test. For a Caesar cipher, the solver tries every shift for you.
Are accented letters counted?
No. Only A to Z are counted. Letters such as é, ü or ł are skipped, both in your text and in the language profiles.
Is my text sent to your server?
No. The tool on this page runs in your browser, so your text is not sent to our server.