Mnemonic Slots

[coinableS]

So I was thinking... https://bitcointalk.org/index.php?topic=5309788.msg56156990#msg56156990

Mnemonic Slots

Imagine a giant slot machine with 12 reels and each reel has 2048 BIP39 mnemonic words. Each time you spin you generate a new mnemonic phrase. The resulting first address generated from the mnemonic is checked for a balance or activity. If you "find" a balance you "win".

Kinda along the lines of directory.io but different.

Live demo: https://coinables.github.io/mnemonic-slots/

[1714933604]

1714933604

[1714933604]

1714933604

[1714933604]

1714933604

[1714933604]

1714933604

[BlackHatCoiner]

Does this differ from the classic brute force way? Despite that it's a million times slower? I mean, how can this be fun? The website design is nice and I admit that it must have been a good exercise, but the entire project is useless.

BTW, are the mnemonics randomly chosen? I got abandon abandon abandon abandon abandon abandon abandon abandon abandon abandon abandon abandon on a try.

[joniboini]

BTW, are the mnemonics randomly chosen? I got abandon abandon abandon abandon abandon abandon abandon abandon abandon abandon abandon abandon on a try.

I tried them just now and it seems to be working as intended, got 12 different words. Did you try to spin it again? I tried multiple times and it shows different combinations each time. No balance found though (as expected).

[NeuroticFish]

So it's a slot machine game for free where in theory you can win big and in reality you'll never win.
Well, it can be some sort of time waster for some, but I guess that all would get bored fast since they actually never win.
(So I agree, it's rather pointless).

[o_e_l_e_o]

So... attempting to steal money from innocent bitcoin users?

Obviously, you could run this until the death of the sun and never find an address with a balance to steal from, so in that sense it is harmless, but I don't exactly agree with the principle behind it. Just because you found a wallet on the street doesn't mean the contents belong to you. Just because you find a private key/seed phrase collision doesn't mean the bitcoin within belongs to you.

[LoyceV]

I don't exactly agree with the principle behind it. Just because you found a wallet on the street doesn't mean the contents belong to you.

I like to say: "let them try!". If anything, they'll prove to themselves that Bitcoin is secure. OP does this in public, but I have no doubt many others are trying millions of private keys per second, hoping to find one that's funded. Let them, Bitcoin is designed to withstand that, and Bitcoin wouldn't be worth anything if this would matter.

Just because you find a private key/seed phrase collision doesn't mean the bitcoin within belongs to you.

On a philosophical or moral level, you are right. But on a technical level, it's much simpler:  having the key means having the coins.

[o_e_l_e_o]

On a philosophical or moral level, you are right. But on a technical level, it's much simpler:  having the key means having the coins.

Sure, but I'm not sure what it is about bitcoin that makes otherwise honest people think it is OK to steal from people. Most people in the world if they found a set of car keys lying on the ground on a parking lot wouldn't pick them up, find the relevant vehicle, and drive it home. But it seems that there are a significant number of people on this forum who, if they found a private key collision, wouldn't think twice about stealing someone else's coins.

[bitmover]

So... attempting to steal money from innocent bitcoin users?

Obviously, you could run this until the death of the sun and never find an address with a balance to steal from, so in that sense it is harmless, but I don't exactly agree with the principle behind it. Just because you found a wallet on the street doesn't mean the contents belong to you. Just because you find a private key/seed phrase collision doesn't mean the bitcoin within belongs to you.

What is worse about it is that people would have to pay to make an attempt.

You can a beautiful website, like a casino. You pay like 0.00001 btc for 1 attempt. If you find balance, it is yours.

It is an easy way to make money. If someone wins the wheel, he pays with someone else's money. It is a win-win game for him. he can even charge 20% of the "total price" ("or total theft")  before transfering to the user wallet!

New way to use clients processing power to  brute force private keys and even make money with their processing power.

lol

[LoyceV]

Sure, but I'm not sure what it is about bitcoin that makes otherwise honest people think it is OK to steal from people. Most people in the world if they found a set of car keys lying on the ground on a parking lot wouldn't pick them up, find the relevant vehicle, and drive it home. But it seems that there are a significant number of people on this forum who, if they found a private key collision, wouldn't think twice about stealing someone else's coins.

The difference is of course anonymity: a car is registered, and the owner can be found. Bitcoin is more like cash, but on top of that is on the internet. I'm pretty sure more people would return the car keys than a pile of cash they find.

[Vod]

Does this differ from the classic brute force way? Despite that it's a million times slower? I mean, how can this be fun? The website design is nice and I admit that it must have been a good exercise, but the entire project is useless.

I disagree.  All projects are not made for the experienced user to earn coin.

For a person who doesn't understand seeds, it may be interesting enough to try.  When they do not win they may read up on why.  (links will be good)

If even one person changes a repeated passphrase with a seed, the OP has made the world a safer place, right?

Stop thinking with your wallet first, and you'll create actual value.

I'm pretty sure more people would return the car keys than a pile of cash they find.

Not in Canada.

[Timelord2067]

I'm trying the link on an Android device and see one or two words in a couple of boxes then all boxes go blank. Then, after about five seconds, information is displayed in the lower grey box. Basically, I can't see any of the rolled words.

Is 2048 the total number of words possible, or, a selection?

[coinableS]

I'm trying the link on an Android device and see one or two words in a couple of boxes then all boxes go blank. Then, after about five seconds, information is displayed in the lower grey box. Basically, I can't see any of the rolled words.

Is 2048 the total number of words possible, or, a selection?

Thanks for the bug report, I appreciate it. I'm having unexpected positioning results based on browser and screen resolution.

It's just a concept/demo for now. I'm testing on Macbook Air with safari, chrome and firefox, and it works. Also seems to work on my android on chrome. But people with ios phones are experiencing what you're describing, what browser are you using?

I'll have to tinker around more and figure out the best way to achieve the positioning across multiple browsers and devices.

[Timelord2067]

Tor latest version - a couple of times the Ballance comes up, but not the corresponding details.


You could set up a modified version where users pay one Satoshi via Lightning Network and have say just four tumblers (and even say what the words are ahead of time for transparency).

Whoever rolls the correct answer wins the pot. (Almost like slots when you think about it)

[coinableS]

So... attempting to steal money from innocent bitcoin users?

Obviously, you could run this until the death of the sun and never find an address with a balance to steal from, so in that sense it is harmless, but I don't exactly agree with the principle behind it. Just because you found a wallet on the street doesn't mean the contents belong to you. Just because you find a private key/seed phrase collision doesn't mean the bitcoin within belongs to you.

Anytime you create a BIP39 wallet, with any software, you might find someone else's coins. The likelihood of getting a winning spin versus stumbling across them randomly while creating a new wallet are both practically zero %. But it's not zero. We all know it's essentially zero. But it's not. So that what makes the site fun, I thought so at least. Nobody is going to win, but they can try. Also like booblicker Vod posted above, it could be an educational tool for noobs to help understand really big numbers and how bitcoin takes advantage of them for security.

[snaiperkg]

If anyone is interested. There is a similar site http://directory.io/, but with automatic search in the browser. http://privatekeys.info/
But the chances of finding, as on other similar sites, are almost zero. But there is a chance

[LoyceV]

But the chances of finding, as on other similar sites, are almost zero. But there is a chance

You're much more likely to guess credit card number. And if it doesn't have to be mine, it's much easier. With 2.8 billion credit cards in use and only 16 digits per number, you'll find on average 1 valid number for every 3,571,429 combinations you try. I made^* you the first batch of 3,571,429 random 16 digit numbers

^* Give it 40 minutes to complete.

[BlackHatCoiner]

But there is a chance

[Timelord2067]

My mobile phone's version of TOR (9.5.4) seems to work better than the version on my Tablet (10.something).

I was just wondering how "random" are the words selected? And, (newbie question here) I thought the last two or three words were a check-sum to verify the previous words, is this not the case?

[o_e_l_e_o]

Anytime you create a BIP39 wallet, with any software, you might find someone else's coins.

Yes, but wallets don't advertise themselves as tools to steal other people's coins.

The likelihood of getting a winning spin versus stumbling across them randomly while creating a new wallet are both practically zero %.

Emphasis mine. Also the quote from your site: "How do I redeem my funds if I win?". I take major issue with this wording. You are not winning a prize which has been offered up willingly. It is not a competition, and the funds are not yours to redeem. It is stealing.

Yes, we both know no one is going to find a match, but I take major issue with the way things are worded and the implication that if you do stumble across a collision then you have somehow "won" and it is morally and ethically fine to just help yourself to the funds within.

I am surprised I am the only one with an issue here.

[LoyceV]

Emphasis mine. Also the quote from your site: "How do I redeem my funds if I win?". I take major issue with this wording. You are not winning a prize which has been offered up willingly. It is not a competition, and the funds are not yours to redeem. It is stealing.

When you put it like this, it sounds bad indeed.

if you do stumble across a collision then you have somehow "won" and it is morally and ethically fine to just help yourself to the funds within.

If anyone can find my private key, I assume they'll steal my funds. And even if they don't, someone else will do it when they find my keys. None of this matters as long as you keep your keys secure.
Of course it's morally and ethically wrong, but so are many other things. That's why I lock my bicycle, because even if the first people who find it won't take it, someone will!

[coinableS]

It's the same as directory.io, no one will steal any coins from this tool. If you're offended by a joke you need to relax, or learn more about big numbers and why 1 attempt every ~6 seconds that this site provides would take longer than the life of the universe to possibly find a collision. If you're still worried, migrate to multisig.

Regarding checksum question, the spins produce validate bip39 checksums.

[userghost]

Is there a way to do this using your CPU/GPU that would be faster than the ~6 seconds that the website takes?

[ABCbits]

Is there a way to do this using your CPU/GPU that would be faster than the ~6 seconds that the website takes?

1. Convert the code to another programming language which allow high performance and GPU usage (such as C++).
2. Search similar tool which run on your computer rather than on your browser.

P.S. Read https://www.coindesk.com/bitcoin-protocol-hack-joke if you're thinking to steal Bitcoin.

[MickGhee]

This is the worst thing I read today.   Stop trying to steal and build something useful or maybe , your server goes  blank...

[coinableS]

This is the worst thing I read today.   Stop trying to steal and build something useful or maybe , your server goes  blank...

Myself and directory.io will not stop until every last bitcoin is stolen. Also it's a 100% client-side driven app that runs entirely in the browser, there is no server -- we are unstoppable.

[Noname400]

wrote a simple program.
7 currencies can be searched.
maybe someone will be interested
https://github.com/Noname400/mnemonic-colider

[Lucius]

Yes, we both know no one is going to find a match, but I take major issue with the way things are worded and the implication that if you do stumble across a collision then you have somehow "won" and it is morally and ethically fine to just help yourself to the funds within.

I am surprised I am the only one with an issue here.

We all agree that the chances are extremely slim that someone will find a seed behind someone's coins, but also that it would not be moral or ethical (perhaps even legally) for someone to steal someone's coins in this way. Maybe the whole thing should have been geared towards showing how safe Bitcoin really is in technical terms, without that part of how to claim win in case a miracle happens, and someone really stumbles upon something that isn’t its property.

I know it's not the solution, but I think everyone should use the extra word in their seed to prevent even the slightest chance that anyone will ever find your seed.

[pooya87]

I know it's not the solution, but I think everyone should use the extra word in their seed to prevent even the slightest chance that anyone will ever find your seed.

If it were possible to brute force or find a mnemonic of someone with balance in it then it would have also been possible to find a random mnemonic that generates the same address with balance that was derived using the seed + extension words.

[redwine49]

So I was thinking... https://bitcointalk.org/index.php?topic=5309788.msg56156990#msg56156990

Mnemonic Slots

Imagine a giant slot machine with 12 reels and each reel has 2048 BIP39 mnemonic words. Each time you spin you generate a new mnemonic phrase. The resulting first address generated from the mnemonic is checked for a balance or activity. If you "find" a balance you "win".

Kinda along the lines of directory.io but different.

Live demo: https://coinables.github.io/mnemonic-slots/

the website looks good. did you make it up to make people think and learn about bitcoin security?
I agree with most of people here.  it's actually pointless to tell others about that.
most of them only want to know how to make money with bitcoin.
hahahaha let them try  

But there is a chance

hahahaha.... good memes. I laughed reading this  

[GameUnits]

This idea is good. But its not that funny, i have to agree. Maybe you can add something like a "history" to see who has won last. It reminds me to Lotto. Its a tipping game where its possible to win a large amount of money if you have tipped for the right combinations of numbers and letters. Its legally known, that lotto costs money for people to play it

[pooya87]

This idea is good. But its not that funny, i have to agree. Maybe you can add something like a "history" to see who has won last. It reminds me to Lotto. Its a tipping game where its possible to win a large amount of money if you have tipped for the right combinations of numbers and letters. Its legally known, that lotto costs money for people to play it

The thing is that you can never "win" this thing ever. And adding a "history" would require storage and dedicated hosting which I don't think there is any way you could do it using github pages. On top of that turning this to something that starts looking more like a lottery and involving tips is going to break a bunch of github terms.

[GameUnits]

This idea is good. But its not that funny, i have to agree. Maybe you can add something like a "history" to see who has won last. It reminds me to Lotto. Its a tipping game where its possible to win a large amount of money if you have tipped for the right combinations of numbers and letters. Its legally known, that lotto costs money for people to play it

The thing is that you can never "win" this thing ever. And adding a "history" would require storage and dedicated hosting which I don't think there is any way you could do it using github pages. On top of that turning this to something that starts looking more like a lottery and involving tips is going to break a bunch of github terms.

Or even if you try it hard or very often thats true. People will see that other players win in the history box so they would continue to play. At least it would take some database entries for that to realise it and there needs to be a login page with the fitting functions implemented. Isnt it possible to host it on pages like gitlab or on private storages? Anyhow, i like the idea of an own server for experimenting

[Kakmakr]

The problem with this idea is this :

Slots are designed to "reward" players on a regular interval to keep people wagering. Most casinos only profit from the wagering, so they will get Slots from third party providers that will draw in a crowd that will wager for hours on end.

People playing this Slot will quickly get bored, because they are not going to win anything for months, so casinos will not be interested in hosting this game. 

Most people play Slots for the entertainment value and for possible big jackpots, so Slot providers add features like Free Spins and bonuses to keep it interesting. (How will you keep people interested to play this game?)

[seoincorporation]

It would be nice if we can play with more than one combination, lets say we test 100 combinations on each click, that would give to the users a better chance.

This idea comes from the batch bets verifier i made for a casino the past year, at the end I gave the option to the users to decide how many bets they want with one click, and it worked fast even for 1,000 bets. It takes like 10 seconds to make the calc, and if you want to test 100,000 bets it could do it, but that takes some minutes to load.

[Betwrong]

So I was thinking... https://bitcointalk.org/index.php?topic=5309788.msg56156990#msg56156990

Mnemonic Slots

Imagine a giant slot machine with 12 reels and each reel has 2048 BIP39 mnemonic words. Each time you spin you generate a new mnemonic phrase. The resulting first address generated from the mnemonic is checked for a balance or activity. If you "find" a balance you "win".

Kinda along the lines of directory.io but different.

Live demo: https://coinables.github.io/mnemonic-slots/

It could be a good slot game if you funded a thousand addresses yourself, with different amounts, and then, knowing the seed phrases, arranged the game so that those with the biggest amounts were hardest to find. For example, to win $10 one needs to find 2 mnemonic words from the corresponding mnemonic phrase, and to win $200 - 3 words, and so on(the number of words in this example is completely arbitrary, but I'm sure you can figure out the right accordance).

There's a lot of hype surrounding Bitcoin now, so a game like this could really become popular. But in the form it is proposed in the OP it is like trying to have fun with stealing. Not good.