Part 1: Blockchain Analytics is More of an Art Than Science

By Heidi Wilder, Senior Associate, Coinbase Special Investigations Team


Bitcoin and lots of different cryptocurrencies are sometimes called pseudonymous. Everyone can view data on a public ledger, however not essentially know who’s behind every tackle or transaction. But what does pseudonymity seem like in apply? How are cryptocurrencies tracked? And can you actually unmask somebody on the blockchain? Let’s discover out.

The public nature of blockchains permits for a sure diploma of predictive evaluation, enabling researchers to affiliate addresses and transactions with entities and typically people. Anybody can take a look at blockchain, however what makes a distinction is the correct interpretation of this public knowledge, in addition to corroborating it with different forms of info gathered externally. Once mixed such knowledge can be utilized for blockchain analytics.

Blockchain analytics is extensively used for market intelligence, development evaluation, and investigations, amongst many rising areas. The most important goal of blockchain analytics is attribution — linking particular property and occasions to explicit entities and even people.

Attributing possession, nonetheless, is usually nuanced as a result of outdoors observers can solely infer it relying on elements akin to availability and high quality of the proof. Evidence means proof that certainly an tackle belongs to a person or entity. Unless you personal an tackle your self, it is extremely troublesome to say with absolute certainty who an tackle is owned by. This is why it’s extra becoming to think about blockchain analytics extra of an artwork than science.

Let’s perceive the fundamentals of blockchain analytics and be taught why attribution is usually extra difficult than it appears to be like.

Attribution Basics

Can you inform what entity this tackle belongs to:


Is it an alternate? Is it a darknet market? Or perhaps a non-public (in any other case generally known as an unhosted) pockets? To reply this query we have to dig for some floor reality.

1. Ground Truth Evidence

A seek for reality usually begins with plain googling or crowd-sourced websites like

Websites like can be utilized by anybody to anonymously report BTC addresses linked to suspicious exercise. Sadly, the reliability of such info might be very low. According to, our tackle of curiosity acquired over 767 BTC. implies this tackle is linked to a big offshore cryptocurrency alternate, which is corroborated by industrial blockchain analytics instruments.

Indeed, industrial blockchain analytics instruments establish this tackle as belonging to a big offshore cryptocurrency alternate.

So what concerning the nature of the exercise? Is the alternate person concerned in ransomware?

Further analysis connects this tackle to an exchanger referred to as

Coinguru permits customers to swap between numerous cryptocurrencies, offering nothing greater than an electronic mail tackle.

At this level you’re most likely asking your self: so who does this tackle belong to?

  • the BitcoinAbuse crowd-reported ransomware operator?
  • A big offshore cryptocurrency alternate?
  • Coinguru?
  • …all the above?!

Well, the reply is difficult.

We have first-hand proof of 1JxXMEbYX6juuEK7QPe6CxGXywQ91ZB5mZ being utilized by Coinguru, an alternate service working an account on a big offshore cryptocurrency alternate. Exchangers like Coinguru usually use greater platforms’ infrastructure to cut back prices and get entry to liquidity. We refer to those as nested companies. These additionally cater to customers who may not need to go to the difficulty of making their very own accounts on an alternate. In truth, some nefarious actors could use these companies to money out of illicit funds.

For labeling functions, it could suffice to say that is an exchange-owned tackle. If a regulator or a regulation enforcement company investigating ransomware associated transactions decides to investigate concerning the particulars, the cryptocurrency alternate will refer them to Coinguru who could be finest positioned to supply additional info on particular transactions.

2. Evidence high quality and commonplace of proof

Evidence can differ in high quality and blockchain analytics is not any exception. Sometimes you would possibly come across a “smoking gun”, however it’s extra doubtless you have to to spend time corroborating incomplete, circumstantial, fragmented or straight out deceptive proof. Nevertheless, even the weakest proof can trace on a selected exercise or entity behind it.

As we’ve already witnessed, crowd-reported sources akin to BitcoinAbuse stand on the underside of the reliability ladder. Not that they need to be totally discounted, however proof resulting in attribution of crypto addresses is finest gathered straight from the supply. In the case of alternate companies, the supply could be their web site displaying a deposit tackle.

The final attribution comes from the flexibility to work together with the service, incomes such proof the very best confidence rating. However, that is usually prohibited, particularly when investigating actions akin to terror funding (TF). In instances like these, analysis shifts into the world of open supply intelligence (OSINT). Much might be realized from aggregator web sites, on-line boards, discussion groups, cell communication platforms, hidden domains on the Tor community and knowledge scraping in an automatic vogue by third occasion distributors. But even the very best proof isn’t useful with out correct investigative instruments.

3. Deconflicting misattribution

Blockchain investigation instruments embody blockchain analytics software program, non-public and open supply databases, search engines like google, and many others. The finest investigative apply is to mix a mixture of these instruments, together with commercially obtainable software program, and corroborate proof utilizing unbiased sources. Sometimes, nonetheless, these sources can provide conflicting info.

For occasion, take into account this tackle: 1N9SxKeNvFoBFuFKEDU8yFCwPwoeHqgmhu.

Imagine an investigator receiving intelligence linking this tackle to the sale of Child Sexual Abuse Material (CSAM). Attribution of this tackle will differ relying on which blockchain analytics software you seek the advice of: some don’t have it labeled in any respect, whereas others attribute it to a service provider service. Open supply analysis confirms this explicit service allowed customers to add recordsdata and promote them for numerous cryptocurrencies. Addresses just like the one above had been generated for each person and had been all related to several types of exercise, relying on what a person person was shopping for.

While some uploads to this service provider service have been benign, some had been recognized as illicit, based on the Internet Watch Foundation (IWF), a non-profit combating the distribution of CSAM. Reportedly, the identical service provider service was additionally used for ransomware decryptor key uploads. So, can the tackle of curiosity belong each to a bootleg vendor and to the service provider service? Yes.

The right technique to attribute this service in a blockchain analytics software could be to take all the identified addresses related to the service and label them accordingly. Then, on account of investigating particular person addresses and their associated actions, particular labels needs to be utilized in accordance with documented findings. Labeling the entire service as illicit could be a misattribution. It can negatively influence instruments and companies that depend on blockchain analytics knowledge, akin to transaction monitoring methods or regulation enforcement subpoenas, resulting in elevated false optimistic alerts and faulty leads.

4. The unknown unknowns

Back in October 2019, a medium article was printed with a flashy title — “Huge Ethereum Mixer”. A Russian knowledge scientist analyzed ETH flows between February and September 2017 claiming that “…68% of total Ethereum transaction value [is] controlled by one system… Funds come and leave within one hour, and addresses are never used again.” The researcher spent a substantial amount of effort analyzing the conduct of the “mixer”, its transaction patterns, and share of whole transactions throughout Ethereum over time. At the middle of the article was this diagram:

Notice how most giant exchanges on the time are current: Kraken, Poloniex, Bitfinex, and many others. Can you guess which one(s) are lacking?

Hopefully, at this level it’s pretty evident that an exterior observer can not presumably achieve a full image or declare 100% confidence in attribution. Keep in thoughts, in the case of blockchain, everybody is an exterior observer, excluding addresses you management.

Stay tuned for the second half, the place we’ll dive deeper into examples of how blockchain analytics can each enlighten and confuse.

Part 1: Blockchain Analytics is More of an Art Than Science was initially printed in The Coinbase Blog on Medium, the place persons are persevering with the dialog by highlighting and responding to this story.

Add a Comment

Your email address will not be published. Required fields are marked *