Bitcoin Mixing: economic analysis and design guidelines | Hacker News Reader

Bitcoin Mixing: economic analysis and design guidelines

bitcointalk.org

34 pointsdeepblueocean18 comments

> This means any anonymity service requires a large population of participating users, ideally with different adversaries, so that one can’t simply assume all coins passing out of the mixing service are “tainted”

This is why something like DarkWallet [1] is so important. We need a wallet that automatically and trustlessly mixes your coins in the background (via CoinJoin [2]), when you're not using them.

The wallet needs a very user friendly UI and solid integration to reach a wide and disparate audience. With mass adoption of wallets like these, it will add anonymity for everyone. With very little work on the users end.

http://www.indiegogo.com/at/darkwallet

https://bitcointalk.org/index.php?topic=279249.0

Can anyone explain on a conceptual level how mixing can possibly work? I'm not thinking about implementation details, but more like boot prints in the snow. You can stamp a big mess in the snow, but no matter how much you stamp about, there is still one set of your tracks entering and one set leaving. When all the transactions are clear to see in the blockchain, all the boot prints are right there.

Imagine a bunch of people traveling in different directions in the snow. Rather than going directly towards their intended destination, everybody goes first to one particular spot, where there is one pair of boot imprints.

You walk toward the bootprint, step into it, step out of it, and then continue on your merry way.

When lots of people do the same, you see a bunch of tracks going into one bootprint, and a bunch of tracks going out. But because everybody was careful to place their feet into the one bootprint indentation, the person following your tracks won't know which of the many forward paths you took after you stepped out.

But what about the part where every boot has a distinct print? Or is that where the metaphor breaks down.

they trade/change boots at that convergent stomping point (i.e., use a new wallet address)

Everyone's wearing identical boots.

Well, we already have desired situation with a different kind of snow -- cocaine.

Right now, most bills are (supposedly) tainted with cocaine. And yet holding bills with detectable cocaine is not admissible as evidence against you.

That's what these pools are going for: so that every bitcoin has touched every transaction. So yes, you can prove that a bitcoin previously touched a crime at some point ... but that fact has zero evidential value.

Dollars are not an analogue though, because we don't have a record of every time a dollar bill changed hands. Or at least so my thinking goes.

They're close enough for purposes of understanding the dynamics.

For example, cash bills weren't well-mixed, then finding cocaine on your money would indeed be evidence that you were one or two levels removed from drug handlers, and may (legally and rationally) justify a look at the people you've been in contact with.

Likewise, when it's not regular practice to mix up your bitcoins through pools, then holding a marked coin would be evidence of wrongdoing.

(And I've argued elsewhere that that's also a reason for prosecuting mixing pools as money launderers!)

Nope, there are N entering and N identical leaving (in this case) so there's no way to know which outgoing transaction belongs to which person.

Does that actually matter?

The way i understand it, Bob wants to send 1 BTC to Alice, and Charlie wants to send 1 BTC to Dave, and everyone wishes to do so anonymously. Through the wonder of mixing Bob's coin actually ends up with Dave. Unfortunately for Dave, Alice was a drug dealer and Dave gets arrested for receiving stolen property or proceeds from crime or some such.

Except in reality the mixing is done with smaller parcels, and a larger number of users. Whether or not the feds want to go after you for receiving 27c worth of drug money is another matter entirely, but i can quickly see there being some negatives to using a mixing service.

It depends how you write the blacklisting/tainting rules, which may depend on money laundering regulations and legal standards of proof. You can say that transaction outputs are all tainted if one input is tainted (an or rule), outputs are tainted only if all inputs are tainted (an and rule), or outputs taint equals the average input taint. Mixing can defeat two out of three of these rules.

I don't understand the practical need for Bitcoin mixers vs. simply mixing via existing exchanges. If Bitcoin goes into an exchange and Litecoin leaves a week later, nothing in either the Bitcoin or Litecoin blockchain would link the transactions together, especially if the funds are chopped up and sent to multiple wallets so the totals are never comparable. If the worry is that the exchange itself might be subpoenaed, one can just hop through many exchanges in many different countries. It's only necessary that one exchange be opaque for the mix to be secure.

If you're moving a unique sum of money, then it's not too hard to see X BTC in, Y LTC out, L YTC in X BTC out.

Additionally, you must trust the exchanges. Leaving your coins in a 3rd party for a week means at any point during that weak, a hack or other problem might mean you lose all your money. Needing to fully trust a third party is a failure.

On top of that, as you note, you'd need to use multiple exchanges, with significant time delays. This is not user-friendly, and will lead to people making mistakes. Not to mention, it's hard to verify if a given exchange is really operating under certain law, and not compromised. I'd be surprised if various LEAs haven't or aren't considering opening exchanges pretending to be safe havens.

"If you're moving a unique sum of money, then it's not too hard to see X BTC in, Y LTC out, L YTC in X BTC out."

Really? If I were to deposit BTC into BTC-e, send you the blockchain.info link for the deposit, and then in the next hour withdraw it to 5 separate wallets in any of the half-dozen altcoins they trade, how would you identify even one of those altcoin transactions as mine?

It'd depend on the volumes involved. I don't think it's out of the question to look for any grouping of outbound transfers around the amount in question and follow them as possibilities. This assumes you eventually want to do something with the mass of money. If you've got 10000 BTC and only want to cash out 100, yeah I suppose that's nearly untraceable.

If you use an exchange, the exchange can be subpoenaed for your identity.

Exchanges are potentially (depending upon host jurisdiction) not required to know who you are if you are only dealing in cryptocurrency.

Reply on news.ycombinator.com