You didn’t lose the password. Your brain just filed it away.

The reason password-cracking works at all is that human passwords are not random. They are assembled from a surprisingly small library of personal ingredients. Once you accept that, the problem changes from “I’ll never remember” to “which ingredients did I use?”

The anatomy of a human password

Nearly every human-chosen password follows the same skeleton, in some order:

• A base word — a name, a place, a pet, a hobby, a word in your native language.
• A number — a year, a birthday, an age, a house number, a lucky number, or 1/123/007.
• A symbol — usually one of a small set: ! @ # $ . *.
• A capitalisation rule — first letter, every word (camelCase), or the letter the site forced you to capitalise.
• Substitutions — leetspeak: 3 for e, @ for a, 0 for o, $ for s, ! for i or 1.

That is it. The universe of “personal password ingredients” is tiny compared to the universe of all possible strings — which is exactly why recovery with even one clue is realistic, and why a genuinely random password is not.

Ask yourself the right questions

Memory is associative. Instead of “what was the password?”, ask questions that let your brain answer sideways:

Keyboard layouts and physical habits

Your fingers remember what your conscious mind forgot. Try:

• Typing it “blind” — put your hands in position and let your fingers trace the old rhythm. Muscle memory is a real, documented recall channel.
• Patterns — many passwords are keyboard shapes: qwerty, asdf, 1qaz2wsx, diagonal or zig-zag runs. If the base feels like a “shape”, name the shape.
• Adjacent keys — if your memory says o but the password “feels” one key to the right, try the shift: sites that demanded “an uppercase letter” forced thousands of people to write Password1! with a capital P.
• Old keyboard layouts — if you switched between AZERTY/QWERTY or a phone keyboard since, the password may now type differently. Try both interpretations of every character.

Build your candidate list

Now turn recall into a candidate list — a plain text file of every string that could plausibly be the password. This is the raw material for everything that follows:

1. 1

   Free-write, then filter

   Spend 20 minutes writing everything: names, dates, words, numbers, symbols, old passwords. No judgement. Then underline the ones that “feel” right.

2. 2

   Generate the variants

   For each candidate base, write the 20–50 obvious mutations: with/without capitals, with each symbol, with each year, symbols at the start vs end.

3. 3

   Test the shortlist by hand

   Try the top 30–50 candidates directly in the wallet app (carefully — see the note below about repeated attempts).

4. 4

   Keep the full list

   Save everything in your password journal. If the shortlist fails, this list becomes the token file for offline recovery tools — the next guide explains how.

When memory is not enough

If your password was generated by a password manager — 16+ random characters — no amount of memory work will reproduce it. That is not a failure of your method; it is a different problem, and it is why searching every stored copy comes before memory work in this guide series. And if only a few characters are missing, or you have fragments, that is still a solvable case — the clue set you just built is exactly what the tools need to run a targeted offline search.

Next in the series

Once you have a candidate list, the real question becomes: how fast can each wallet type be tested? That decides whether your case is a weekend project or weeks of GPU time — see is recovery technically possible? And before you run any tool, make sure your wallet files are safe: protect your wallet files.