Blockchains itself are public and immutable. Every action leaves a trace in the blockchain, so there is an opportunity to follow your every step. Thanks to data analysis (heuristics, smart-contract parsing and analysis, user submissions, web scraping, machine learning) information on wallets’ ownership (both past and present) is available. At the moment most of wallet labels are algorithmically inferred.
Nobody has control of what users decide to publish on them. So there is many scam and fraud tokens, especially in DeFi space. Attacks on DeFi tokens’ liquidity and prices are reported every day. Most users have no idea what blockchain is and how it works. Due to the super profitability of cryptoassets, a large number of newbies come to the market. Due to the lack of central intermediaries like banks and funds or any control institution, there is a high probability that the user will lose his assets (loss of keys and seed phrases, phishing malware, ignorance of the network principles etc.).
Absolute numbers in blockchain say little. Some of them can distort information. Relative and dynamic indicators are more appropriative. While analyzing crypto market data, psychologically, users pay more attention to liquidity than to the product (token) itself. Unsecured stablecoins, vampire mining, manipulation on CEX lead to reverse effect in the industry: for example, the capitalization of ERC-20 tokens can easily exceed the same indicator of their ancestor, but without him they have absolutely no cost.
One of the primary things that separates a good data analytic service from one just building market models is the ability to think carefully on things like endogeneity, causal inference, and experimental and quasi-experimental design. Data scientists must understand and think on things like data generating processes and reason through how misspecifying them could influence or undermine the inferences they draw as a result of their analyses.
Despite the fact that the blockchain sphere is at least saturated with absurdities, simulacra and other tools for the mind, thanks to the existence of data analysis services (like Covalent), it is possible to structure all of this, study and draw useful conclusions (utility).
