Really need some help...

[nearlost]

This might be the wrong place to ask but I've asked everywhere and I remember this forum in the very early days to have some really good insights. I basically wrote a 12 digit seed-phrase incorrectly that I need to access my bitcoin. I basically have 12 words but I'm sure I wrote one of them down incorrectly. Is there a way to make a script that brute forces my current 12 words that i have but instead keep trying and replacing the 4th word with the 2000 word database until I get into my wallet again.

I'm basically trying to break back into my own wallet if anyone is able to help. I've lost all other options and would really love some insight on how I can possibly code this myself or guide me in a direction to make this

[1715534169]

1715534169

[1715534169]

1715534169

[1715534169]

1715534169

[DaveF]

Do you mean you have the 12 words out of order or some of the words are wrong, as in you wrote art instead of arm?

If its just in the wrong order take a look at:
https://btcrecover.readthedocs.io/en/latest/Usage_Examples/2020-05-02_Descrambling_a_12_word_seed/Example_Descrambling_a_12_word_seed/

If you wrote art instead of arm take a look at:
https://www.blockplate.com/pages/bip-39-wordlist
And see if you can figure it out. There are other ways but if you have a basic idea of what you wrote wrong or if the word is just not in the list you can probably get it done.

Standard security thing, don't send your words or anything to anyone offering to help online, scammers are everywhere.....

-Dave

[PawGo]

Hey,

You may check my program: https://github.com/PawelGorny/lostword
What you need, is to configure it for worker "KNOWN_POSITION". See examples how it should be configured.

I do not exactly understand what is your problem - I mean, do you know position of wrong word? Or not? then You should launch program 12 times...
It would work if you know your address (on any derivation path).

If not, and you are sure that 4th word is incorrect, you may use worker 'PRINT_SEEDS' which will give you all the correct seeds. For 12 words it should give 128 seeds to check.
Let me know if you need any other help.

[dkbit98]

I'm basically trying to break back into my own wallet if anyone is able to help. I've lost all other options and would really love some insight on how I can possibly code this myself or guide me in a direction to make this

Making mistake with one of the seed words is not such a big problem as you may think and it is possible to fix this, but are you absolutely sure that you those twelve words in correct order?
That is something much harder to correct, and I would recommend writing numbers in front of words, and in future always double check seed words after you wrote them down.
This should apply for all wallets, and some hardware wallets even have option to check if all words are correct without importing them.

[o_e_l_e_o]

If you know which word is incorrect, then I would think the fastest way to do this would be to just download and run The FinderOuter with the unknown word replaced by an asterisk. This does require you to know at least one address derived from your seed phrase and its derivation path, though.

If you don't know which word is incorrect, or you don't know an address, then I would use btcrecover. You will need to set up an address database if you don't know the address.

[NeuroticFish]

a way to make a script that brute forces my current 12 words that i have but instead keep trying and replacing the 4th word with the 2000 word database until I get into my wallet again.

I also suggest FinderOuter with * instead the 4th word, if that's the wrong one.
Make sure you select from the combo box the correct setting - whether the seed is standard BIP39 or Electrum, also watch the examples for an overview on how to use it. It's a nice tool.

[n0nce]

This might be the wrong place to ask but I've asked everywhere and I remember this forum in the very early days to have some really good insights. I basically wrote a 12 digit seed-phrase incorrectly that I need to access my bitcoin. I basically have 12 words but I'm sure I wrote one of them down incorrectly. Is there a way to make a script that brute forces my current 12 words that i have but instead keep trying and replacing the 4th word with the 2000 word database until I get into my wallet again.

Are you sure that it's the 4th word or are you using this just as an example to explain what you're trying to do?
It is possible to recover wallets, but it highly depends on how much information you have and how much you don't. Like, it matters a lot a lot. Sometimes people come in here asking like 'can a wallet be recovered if I don't have the last word' and it turns out they actually lost more than 1 word and don't even know the position; which in their mind might be mostly equivalent, while in reality making a huge difference in time required for bruteforcing.

If you don't give proper information, estimated cracking times can be off by multiple orders of magnitude - like 'it would require 100 million years instead of 7 minutes' is what I'm talking about.

So it's crucial:
1) How many words are missing?
2) Do you know the locations?
3) Are all other words definitely correct?

[BlackHatCoiner]

I basically have 12 words but I'm sure I wrote one of them down incorrectly.

This is interesting.

Writing down a seed phrase incorrectly seems kinda hard, even if you aren't so careful, and you say that you're sure you don't have the correct 4th word. You could have done a human mistake, obviously, but most wallet software require you to re-write it after it's generated. Which wallet did you use? You were able to either bypass this or you wrote it correctly somehow. Neither of those makes much sense.

[n0nce]

I basically have 12 words but I'm sure I wrote one of them down incorrectly.

This is interesting.

Writing down a seed phrase incorrectly seems kinda hard, even if you aren't so careful, and you say that you're sure you don't have the correct 4th word. You could have done a human mistake, obviously, but most wallet software require you to re-write it after it's generated. Which wallet did you use? You were able to either bypass this or you wrote it correctly somehow. Neither of those makes much sense.

It almost sounded intentional to me. Like: 'I remember I purposely put in a wrong 4th word & memorized it' (and obviously then forgot it).
Some people do these kind of weird things, overestimating their long-time memory in an attempt to increase their security.

[NeuroticFish]

It almost sounded intentional to me. Like: 'I remember I purposely put in a wrong 4th word & memorized it' (and obviously then forgot it).
Some people do these kind of weird things, overestimating their long-time memory in an attempt to increase their security.

* I've played a little at some point with keeping the first 4 letters and make up, where possible new words off that. And obviously I've destroyed one of the words.
* For somebody not native English errors like battle => bottle or fine => fein can happen without noticing at first

I guess that we have to tell more often that people should try to use the recovery phrase after writing it down.

[PawGo]

It almost sounded intentional to me. Like: 'I remember I purposely put in a wrong 4th word & memorized it' (and obviously then forgot it).
Some people do these kind of weird things, overestimating their long-time memory in an attempt to increase their security.

In high school I had a friend who asked for his phone number was saying digits and when other person wrote them down, he was adding "and one is wrong".

So yes, I think it was a way of 'securing' the seed. I only wonder if seed with replaced word still validates (if checksum is correct), I think chances are small.

[o_e_l_e_o]

I only wonder if seed with replaced word still validates (if checksum is correct), I think chances are small.

For a 12 word seed phrase the checksum is 4 bytes, so 2⁴ possibilities with one of those being correct. Therefore, there is a 1 in 16 chance on average that replacing a word will still result in a valid checksum.
For a 24 word seed phrase, there is a 1 in 256 chance on average as the checksum is 8 bits.

[DaveF]

I only wonder if seed with replaced word still validates (if checksum is correct), I think chances are small.

For a 12 word seed phrase the checksum is 4 bytes, so 2⁴ possibilities with one of those being correct. Therefore, there is a 1 in 16 chance on average that replacing a word will still result in a valid checksum.
For a 24 word seed phrase, there is a 1 in 256 chance on average as the checksum is 8 bits.

1 in 16 or 1 in 256 then what is the point of a checksum when you can be dealing with a hell of a lot of money?
Yes I understand there are limitations and all of that, but then whoever was doing it should have spent more time and found something better.
One of the engineering programs that one of my customers use uses a checksum on every single line drawn and it is 1 in 64000 to be wrong and the engineers still complain about that.

-Dave

[n0nce]

I only wonder if seed with replaced word still validates (if checksum is correct), I think chances are small.

For a 12 word seed phrase the checksum is 4 bytes, so 2⁴ possibilities with one of those being correct. Therefore, there is a 1 in 16 chance on average that replacing a word will still result in a valid checksum.
For a 24 word seed phrase, there is a 1 in 256 chance on average as the checksum is 8 bits.

1 in 16 or 1 in 256 then what is the point of a checksum when you can be dealing with a hell of a lot of money?
Yes I understand there are limitations and all of that, but then whoever was doing it should have spent more time and found something better.
One of the engineering programs that one of my customers use uses a checksum on every single line drawn and it is 1 in 64000 to be wrong and the engineers still complain about that.

-Dave

I mean, it's probably a different domain. We're not talking about a 1 in 256 chance of a writing error (like 'o' instead of 'a') not being flagged. We're talking about someone mistakenly writing down a completely different word than what is shown on screen, while still being one of the 2048 in the Bitcoin seed word list and then there's a 1 in 256 chance of the checksum algorithm not giving a red flag. We're also dealing with words here, which makes it harder to misspell and easier to find errors (e.g. 'bcttle' is not a valid word but 'bottle' is. or: 'the 5th letter really looks like a 1, but I know it can't be, because it wouldn't be a word, soo even though I read bott1e, I know it's bottle').

[pooya87]

1 in 16 or 1 in 256 then what is the point of a checksum when you can be dealing with a hell of a lot of money?
Yes I understand there are limitations and all of that, but then whoever was doing it should have spent more time and found something better.
One of the engineering programs that one of my customers use uses a checksum on every single line drawn and it is 1 in 64000 to be wrong and the engineers still complain about that.

-Dave

One of the main purposes of the checksum in BIP39 is padding. When we produce entropy, it is always a factor of 2 for example 128 bits or 16 bytes. But this can't be split into 11 bit chunks, so we have to add some more bits at the end. What better than a checksum.

Checksum is also not to meant for recovery, but for a quick way for wallet software to recognize invalid inputs. For example if you enter your Electrum or Aezeen seed phrases into a software expecting BIP39 seed, it can quickly figure out that it is invalid instead of deriving keys and confusing you why they have no balance.

[o_e_l_e_o]

-snip-

If you don't want a checksum, there is absolutely nothing stopping you from turning a seed phrase with an invalid checksum in to a wallet (except maybe your wallet software refusing to accept it). You just pass the words through the usual PBKDF2 function. Indeed, BIP39 only says that if you enter a seed phrase with an invalid checksum the software should "issue a warning", but nothing about refusing to accept it. So you could quite easily generate and use a seed phrase from 132 bits of entropy, instead of 128 bits entropy + 4 bits checksum (or 264 instead of 256 + 8). I only see downsides to doing so, though: Your private keys will be no more secure than 128 bits, and you'll almost certainly encounter problems in the future when you come to recover from it.