Re:Ask HN: Where are the 750k Bitcoins lost by Mt. Gox?
by deweller on Friday February 28, 2014 @10:51AM
(#7319386)
Here is an example of something you might find:
Say you traced these coins to 50 different addresses. And then, for example, say that those addresses lay dormant for 3 months. And then 45 of those addresses send bitcoins somewhere on the same day. That would imply that the coins are under the control of a single owner.
I agree it is difficult and beyond the scope of just writing a simple script.
Re:Ask HN: Where are the 750k Bitcoins lost by Mt. Gox?
by Carltonian on Friday February 28, 2014 @04:38PM
(#7321668)
This seems like the best way to go about it. Use statistical analysis to find similarities between wallets involved in transactions stemming from the root. When you can give a qualitative "Here's where these coins are going", you can start to identify the culprit. Since there doesn't seem to be a legal framework to prosecute such theft anyway, I'd really just want to find them to congratulate them. Sorry to anybody who lost their coins :(
Re:Ask HN: Where are the 750k Bitcoins lost by Mt. Gox?
by kordless on Friday February 28, 2014 @10:59AM
(#7319459)
I was thinking about dumping the Blockchain into Splunk. Investigating the best way to do that today.
Re:Ask HN: Where are the 750k Bitcoins lost by Mt. Gox?
by joeld42 on Friday February 28, 2014 @05:12PM
(#7321900)
I don't know much about btc, but John Radcliffe has a nice command line blockchain parser/query tool:
http://codesuppository.blogspot.com/2014/01/a-command-line-i...
We can construct a graph tree all destinations beginning from that one address, but what good would that be? Where do you make inferences to determine the identity of the recipients? How do you group them together?
At some point your trail becomes so long and so disparate that any effort to continue following that trail would be a pointless endeavor.
Sure the technology is there to follow the trail indefinitely, and (where available) you can easily check all recipient addresses from this "poisoned" tree against regular fiat exchange logs, but within the BTC ecosystem? It's an interesting intellectual challenge, but not useful.