The toolset, released last week, comes with datasets for six new cryptocurrencies: Bitcoin Cash, Dash, Dogecoin, Ethereum Classic, Litecoin, and Zcash. These join Bitcoin and Ethereum in Google Cloud's catalogue.
Using the tools, the datasets can be analysed with Google's analytics platform BigQuery, which combs through transaction history to separate fact from fiction and reveal what's really going on inside the blockchains of the most popular cryptocurrencies: "I'm very interested to quantify what's happening so that we can see where the real legitimate use cases are for blockchain, so people can acknowledge that and then we can move to the next use case and develop out what these technologies are really appropriate for," says Google data scientist Allen Day, who leads the project.
Shining a light into the blockchain
Blockchain ETL (extract, transform, load), as the toolset is known, uses BigQuery's Machine Learning capabilities to tap into transaction flows and analyse movements around addresses.
Unlike existing blockchain explorers — like Etherscan and BlockCypher — which only track individual transactions and addresses, ETL zooms out, searching through the entire blockchain to reveal broader trends.
This has already been put to use to expose the discrepancy between the narrative promoted by certain crypto evangelists, and the underlying reality of individual blockchains. Examined with ETL tools, the contentious Bitcoin Cash fork of 2017 — which promised more decentralization and efficiency for smaller transactions — was found to be held in large quantities by relatively few wallets, suggesting the vision hadn't yet been realized.
The report stated: "Bitcoin Cash was purportedly created to increase transfer-of-value use cases through lower transaction fees, which should ultimately lead to a lower Gini coefficient of address balances. However, we see that the opposite is true—Bitcoin Cash holdings have actually accumulated since Bitcoin Cash forked from Bitcoin. Similarly, the Ethereum Classic currency was rapidly accumulated post-divergence and remains so."
By lifting data out of the confines of the blockchain, and presenting it in a traditional format, the tools also make it possible to apply traditional economic measures to blockchain analysis — like the Gini coefficient, which measures wealth distribution.
Taken one step further, this same data can be contrasted with real-world examples. Using Google's tools, the team were able to plot the Gini coefficients of various different cryptocurrencies, and then compare these with world economies.
To set a benchmark, the team looked at South Africa, which had a coefficient of 67 in 2010; Sweden, which had a coefficient of 26 in 2008; and the average Gini coefficient, which was found to be 39.6 in a 2013 study.
Most cryptocurrencies, however, were found to have a much higher Gini coefficient; indicating high levels of inequality. As shown on the histogram, most cryptocurrencies have a coefficient between 0.5 and 0.75, and although Dash bucked the trend with a Gini coefficient of 0.25, the presence of shielded transactions make this data suspicious.
A histogram of the reported data
Elsewhere, Google developer Tomasz Kolinko has used a smart contract analyzer tool to analyse the Ethereum dataset, probing 1.2 million smart contracts in twenty seconds and revealing that a feature called "self-destruct", designed to limit a contract's life span, had been left open on hundreds of contracts — potentially allowing unauthorised parties to "kill" the contract.
Cross-chain comparison
Aside from drilling into individual blockchains, the tools also enable comparative analysis — grouping similar blockchains under a "unified schema" that allows for direct comparison. All cryptos in the 'Satoshi family', for instance, like Bitcoin and Litecoin, can be analysed using the same queries, as can those relating to Ethereum, like Ethereum Classic.
This is made possible by putting blockchain data into a standard bookkeeping format: "double-entry book data structure". By doing so, not only is comparative analysis made easier, but data can also be more easily integrated with "conventional financial record processing systems."
There are currently 500+ projects on BigQuery, but with the addition of the new blockchains, it is expected this number will continue to grow.