Indeed, anyone can see that the coins might be hiding something: that is why they are tainted.  One point I didn't mention is that the coins might be recognized as tainted only after a mixing or after a transaction. Just imagine Bob steals some bitcoins and quickly uses a mixer, tainted or untainted. After an hour or so, the coins are reported stolen and whoever got them in the mixing is "stuck" with tainted coins. Now what? The fact that the coins are tainted does not give any "tips" to anyone. People knew that these coins were shady, which is why some entity marked them as tainted in the first place. The only information you gain on the perpetrator is that he used a mixer, which can be seen by observing the blockchain anyways. All this information could have been derived without this whole taint proposal.

Thus, in this case, the taint is useless.

Yes, it will sound some alarm, but whoever is receiving these tainted coins does not know, by design of the mixer, who sent them these coins. They have an address since the transaction is visible in the blockchain, but they may not know anything about whom that address belongs to.From a law enforcement perspective, the trail dies. In this case, is there anyone that benefits from the knowledge that these coins are "tainted"?

Edit: quote formatting

This whitelist/blacklist approach to criminality linked bitcoins seems very interesting due to its decentralized aspect, at least in theory. I find the Spamhaus analogy to be quite appropriate. While I am uncertain of the overall efficiency of AML laws in our societies, I agree that some compromise will probably have to be done eventually.

However, the loss of fungibility still makes me uneasy. Would it be easy/free to get a coin whitelisted? I believe the hassle of tainted coins not being accepted or causing a delay in a restaurant might be excessive. A more reasonable option would be if the restaurant accepting the payment without a word and then reporting the event to the appropriate authorities if they deem it necessary. Thus regular users really wouldn't care if their coin was tainted or not, and enterprises would not want to discriminate for fear of driving away consumers with tainted coins. The downside I see is that enterprises would have no real incentive to do appropriate reporting other than being good citizens, and I believe that mandatory reporting would be against the spirit of Bitcoin although this might end up being what happens.

Sadly, it seems to me that this whole proposal would be made useless by the use of mixers. Bad guy Bob could simply use a mixer and be done with it, having received fresh coins. Some mixers, centralized or decentralized, could adopt as a policy to only accept untainted coins, but then that might incentivize some mixers to accept tainted coins and change a higher fee. While I would think that most people who use these tainted mixers would have something to hide (otherwise why pay the extra fee and received possibly tainted bitcoins?), these mixers would still provide plausible deniability for the origin of the coins. This would make the whole proposal useless. I also believe in the necessity of mixers for the anonymity of Bitcoin and I wouldn't want them banned if that were even possible.

In any case, I believe that the Bitcoin community needs to address these concerns and discuss them instead of doing witch hunts for statist pawns. Mike's proposal might not be perfect both for technical and ideological reasons, but the community needs to seriously consider what is going to happen when governments around the world start making laws.

I understand the no-waste and no-transaction advantage of chronobit, but for anything that does transactions, this idea seems very simple and does not destroy coins. Also, it does not need a centralized service at all and no one even knows a commitment was made.


Also, I read this paper but I feel like it has no advantage over previous solutions other than being proven secure, did I miss something?
http://people.scs.carleton.ca/~clark/commitcoin/abstract.pdf

Where? Was it taken off?


Also, why isn't the unSYSTEM clearly identified as a bitcoin conference?

Afaik, similar cases were never actually enforced. Anyone remember the illegal prime number used to decode DVDs?
http://en.wikipedia.org/wiki/Illegal_prime

Songs, t-shirts and mugs were made with that number. Once it's out, nobody can stop it.

Actually, that figure, attributed to Ron and Shamir, says that 78% of bitcoins have not been moved within a 3 month period. That includes all dead coins as well as many cold wallets. Since it is impossible to differentiate between a dead coin and one that is stored, we can't really know how to interpret that percentage.

As to the amount of drugs sold on the Silk Road, this is from http://arxiv.org/abs/1207.7139
"While we cannot estimate the number of buyers, we show that Silk Road is a relatively significant market, with a few sellers, and monthly total revenue of about USD 1.9 million."

Right now, mostly just understanding the community and motivations of Bitcoin users. Eventually, if most transactions end up being small, I'd like to get feedback on the idea of putting a hardcap on the amount of Bitcoins that can be outgoin from a single address within a certain timeframe, say 100 btc for one day. I believe that would help Bitcoin eventually comply with AML laws which might eventually cause trouble for the currency, with a minimal impact on the majority of users. Getting to know what people do with their Bitcoins is the first step though.


Many thanks, I'l take a look.


I'm not sure what that percentage is, and what is the typical transaction size for these users is. The following paper is a nice analysis of The Silk Road but did not include the info I am looking for.
http://arxiv.org/abs/1207.7139
I'm not looking to shut down drug buyers in any way, just trying to understand what the typical transaction amount is and what's a typical pattern.

I couldn't find a similar thread, but if I missed it, please refer me and close this one.

I found this list of the things that can theoretically be bought with Bitcoin but I was wondering what typical users actually buy.
https://en.bitcoin.it/wiki/Trade

Do you buy actual stuff (or services) with Bitcoin? Do you keep your Bitcoins in a wallet and hope the value goes up to sell them? Do you pay debts to people you know? What is the size of your typical transaction of this type?

Ofc, if you buy drugs or other illegal items, you probably won't want to just say it here, but you are welcome to post what is the amount of btc in your typical transaction anyways.

I know I could just look at the blockchain to see the size of the average transaction but I can't think of an easy way to differentiate the sale of a product/services from an attempt to anonymize oneself.

I had though of that but somehow had not thought properly of the consequences for someone who is following the money trail. The trail does get scrambled when thought of this way. Thank you for enlightening me.

Out of curiosity, do web wallets usually create one address for every user or do they have one or a few big addresses every user dumps his bitcoins into?

Edit: answering my own question, both ways seem possible, the latter especially with hybrid wallets
https://en.bitcoin.it/wiki/Browser-based_wallet

I know multi-sender transactions with sending addresses that belong to different people are possible in theory in bitcoin, but does anyone actually use them? Many analyses of the bitcoin network assume that they are seldom used, I'm just trying to get an idea of how much that is true in practice.

I'd be interested in reading it, is the google docs version the latest? Do you have a PDF version?

Many thanks. So bitcoin discussion it is.

Is there a section to discuss scholarly articles somewhere on the forum? I'm trying to understand the Reid and Harrigan analysis of anonymity in bitcoin but I can't find a forum thread on it. http://anonymity-in-bitcoin.blogspot.ca/2011/07/bitcoin-is-not-anonymous.html

I saw that Shamir's paper was discussed in general discussion, is there a better place?

Edit: found out how to get out of the newbie forum

Point taken
Out of curiosity, how did you find it? I was able to close one leak, but I wonder if we found the same one.

DannyHamilton, I agree that entry and exit of the Bitcoin network are the most crucial parts for staying anonymous, but I believe that they are hard to secure. I am trying to find out if there is a way to make those events unlinkable, as you said. Of course, one must trust the mixing entity to not be with the government and not keep logs. However, that confidence could be increased by passing the funds through many mixing services, and it is my theory that unless all of them are cooperating with the government or careless, the link cannot be established. An even stronger guarantee of anonymity would be that even if the link is found, no one could be sure that the link actually points to your series of transactions. But if the whole world is conspiring against you, nothing can be done indeed.

That situation is interesting indeed, but that is not quite the one I was looking at. Say Mr. X, who is very rich, wants to fund (presumably illegal) organization Y. Then Mr. X can, over a few months or so, accumulate some Bitcoins by trading them through Mt. Gox for USD or through other means using different aliases and addresses if need be. Mr. X then sends these coins through a mixer (let's say the mixer doesn't keep logs more than 1 week) that sends the accumulated amount to a new address belonging to Mr. X. He then uses that address to simply send the amount to the known address of organization Y.

Once a government has seen the money arrive at the address of organization Y, is there a way for them to find out whom Mr. X is? Even if it is not the case, can the government look at the blockchain and figure out what were the original addresses Mr. X used to collect the bitcoins even though they were mixed? Is there such a thing as a mixing service that can guarantee that, from a theoretical point of view?

That is indeed the case I am interested in. I really don't care if the government can find out if Mr. X is buying a sexual toy, but rather if they can find out who is funding paramilitary organization Y with millions of dollars even though I would expect such a person to use the best privacy methods. Specifically, I am looking into graph analysis and what it enables us to know even when used with strong privacy measures.

Hello everyone

I have been browsing the forum for a few weeks and I decided to jump in and create an account.

While I do have Bitcoin addresses, I do not currently actively trade BTC so their balance is very low. I have more of a theoretical interest in Bitcoin especially in its cryptography. I am also interested in other crypto-currencies as well as e-cash in general.

I have already read quite a bit about Bitcoin, and produced an introduction to Bitcoin's cryptography talk that I will link to once it's edited.

I am currently interested in finding out just how anonymous Bitcoin is or can be against someone with a large amount of ressources (i.e. government, etc.), as well as the implications of such anonymity.

You can sometimes find me on irc in #bitcoin.