Topic: z - page 4. (Read 58382 times)

But as long as a new address is used, you don't know who that belongs to, you can't be sure, you can't prove anything. It could be a payment or other legitimate transaction.

And the minute the original owner decides to use a proper mixing service, the output address on the other side of the mixer can't be linked at all. (no taint.)

Do you care to share the interesting part ?

looks like we have picked up the scent  

znort, as you stated the closure won't include addressees that don't engage in joint sending transactions.  But if someone performs single address to single address transfers these escape the closure.  Perhaps very conservative approach of inducing addresses that receive all their funds from already closured addresses AND when that amount represents a significant portion (say >50%) of all the coins in the closure we can be very confident that this is really an address owned by the closure owner.  This might catch slightly more addresses then the closure and taint methods as we would continue to see mining income to the wider closure which would improve the taint analysis.

Also another possible tool idea, I call it coin-transaction-age, simply a count of now many times an individual coin has been transacted.  It should then be possible to create a mean-transaction-age of someones coins and tell roughly how circulated their coins are.  A newly mined 50 value coin would have 1 transaction (the one that mined it) and the first time its transferred it becomes a 2 transaction coin and so on and so on.  If coins merge just take the weighted average.  So your earlier statistic about these 'abandoned' coins would evolve into a full-spectrum of the whole coin base from 1 transactions to thousands, I suspect we would see a power-curve with a small number of coins performing huge numbers of transactions.

Now what gets interesting is do to a mean-transaction-age for a wallets balance, this would effectively tell us how close an address is to mining (giving a low count), or if they are very far (in a transaction sense) from mining activity and engaging with older more used coins (like a merchant for example).  Could provide an interesting angle by which to analyze addresses.

Haha, didn't think of that. That's so true. Imagine how many BTC's have been 'lost' due to this.

I can't BEAR to look from envy.  

Znort, how about running the closure and 'taint' tools on this address, that would at-least give a broad-brush generalization of if these coins were scattered to the wind or stayed consolidated, I suspect it would be the former as people were much loser with BTC back then when it was really worth nothing.

What is the largest wallet that can be linked to an owner with confidence?

http://blockchain.info/fb/12hi

I believe this address would be high on the list if it wasn't "cleared out".  I believe this one belongs to Satoshi because the oldest transaction was 3 days into bitcoin's existence. It seems a lot of mining rewards from the very early days were sent here. I tried to follow the money, but bitcoin did its job and led me nowhere.

Will that buyer pay market price?
If the answer is no, don't even bother

Needless to say, I am not on this list.

Does anyone know?

i wish i was on that list   

Ahh good point. So richest wallet is $22M.

Not necessarily. The richest BTC wallet is, but the richest person could own several wallets.

So the richest BTC person is worth $22M?

And some are cursing themselves for losing their private keys.

Some of these people must be cheering right about now.

Good luck trying to sell 100k coins, there's not enough liquidity in the system to handle it.

Such collaborative TX should be very rare and they are almost certainly between individuals who have similar characteristics and would end up in the same classification anyways (miners, hoarders, speculators etc) and thus from a high level macro analysis a few over-aggressive closures should have virtually no effect on determining Macro flows.

I'll have a look at your code see if I can offer any improvements, but unfortunately I don't run Linux so may need to port it first if that proves doable.  What if any improvements do you currently plan to make?

Ok you make some good arguments but lets dig into the data and try to actually get a handle on these macro-flows of BTC.  While miners might seem to be insignificant given Mt.Gox volume, some coins are certainly freezing out (hoarded) after going through Mt.Gox, or else their trade volume would be growing over time and it's fairly stable now for nearly a year.  Thus if miners really are the only major inflow of coins into a relatively small sub-section of coins that turnover very fast but also leak-out rapidly then the miners liquidation rate could have a significant effect.  Or maybe BTC-Dollar exchanged coins really are an island that neither gains nor loses much and we simply have a plain old speculative bubble driven entirely by 'animal-spirits' and an influx of more people competing over the same number of coins.

Now to do this kind of deep analysis you probably need to get all 'closure' address groups computed.  Is that remotely feasible and could the resulting closure lists be archived such that you can start doing 2nd order analysis on it such as identifying know entities and grouping them by characteristics like transaction volume, age and obviously balance.  Then you can calculate the aggregate flows between the groupings and get a macro picture of what the whole BTC economy looks like.