Q42$m02B$&
that is to say, “passwords”
For ease of remembrance, use one word/term with at least 8 characters in it in every password (see below), with a word or term (at least 4 characters) that easily identifies the access for which the password is being used.
Substitute for
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
the following:
4 d ( b [=/e] k 9 # ! ? f 1 w u 0 q p 2 $ + n 5 m 10 z y
(see below).
For ‘e’, the most commonly used letter in English, alternate = and e in a given password if it appears more than once.
Capitalize the first, then every other, actual letter.
When a letter is immediately repeated (as with the ‘t’ in ‘letter’) write 2 followed by the substitute character: so, ‘1=2+E2’ for ‘letter’ (as it happens — and E equals -1/2, I think). ‘Runner’ would be ‘2NuU=2’.
For, say, the bank account, it could be ‘D4uF’ for ‘bank’ followed by that constant word/term — here, ‘password’: so, D4uFq42$M02b.
Have one letter/number/other character to attach to the end of the password (in that order, if more than one of them is needed) if less than two different examples of any of those appear in the password: for instance, a, 3, and &. [correction (9/23/24): characters not in the code (had erroneously used both z and 9 in that “for instance” in this article initially)]
The password for the bank account becomes ‘D4uFq42$M02b&’ for ‘bankpassword’ (with the & added because there was only one ‘other character’ in the sequence).
There, two different numbers do appear; if it were only one number, no matter how many times it got repeated, a 3 (or whatever always-used number) would also be added — before the & (or whatever always-used other character). Two different ‘other characters’ is the reason for the & in the title of this article; ‘runner’ would be ‘2NuU=23&’. Adding characters at the end in order to have at least two letters, numbers, and ‘other characters’ that are different isn’t technically necessary, but it does generate greater length/complexity randomly without having to think creatively about it — then remember what it was.
Re. the part about having one word/term that is always used . . .
People hacking passwords use computers to generate random combinations of characters until they happen upon one that works: so the longer, more complex the password, the better. Even if, using this algorithm, one site were successfully hacked, the same random process would have to be used to hack another site to get the characters that were repeated. Even then, the different characters representing the words/terms accompanying different sites would still have to be randomly found (so the longer that part of the password for any site, the better). Also, the (encoded) word(s) preceding the constantly used (encoded) word(s) would randomly affect the capitalization as well as the possible presence of an extra character (or more) attached to the constantly used word(s) — which is why it should be the second part of the password, not the first. With any such change the hacker would be back to coming up with a whole password using randomly generated characters all over again. In short, even with a constant word/term being used in all passwords, this approach will provide excellent security for all sites. Having said all that, having something as obvious as ‘password’ as the constant word might not be such a great idea.
For the code:
4 looks like A; d swaps with b; ( looks like C; b swaps with d; = for e; k swaps with f (because those were the last two letters left that were not messed with — with a certain four-letter word helping with remembering that swap); 9 looks (sort of) like a g (when a person writes that letter); # for H, ! for i, and ? for j because, of course; f swaps with k; 1 for l, of course; W swaps with M; u swaps with n; 0 [zero] for o, of course; q swaps with p; p swaps with q; 2 for r because . . . (sort of?); $ for S and + for t, of course; n swaps with u; 5 for V (Roman numeral); M swaps with W; 10 for X (Roman numeral); z swaps with y and y with z (just to mess with those two letters).
Including ‘e’ and ‘=’, that’s thirteen letters, six or seven numbers (depending on if you count ‘10’ substituting for x as a separate number, given that it repeats 1 and 0), and six other characters.
Of course, no potential hacker could possibly know if anyone were using this — now published — algorithm. If interested in using it, just save this ‘article’ in your reading list for access, if needed.