Lately I’ve been tinkering with address tagging/clustering again. To put it plainly, it’s pretty similar to “profiling”: it looks scientific, but the errors aren’t small. If an address gets labeled as “smart money,” it might just be someone who happened to churn a few times while claiming an airdrop; or the same batch of funds could be split into dozens of addresses, and once you cluster them back together it starts to look like a whale is controlling the market… My obsessive-compulsive tendencies make it uncomfortable to watch, so I can only treat it as one of the signals—never as a conclusion.

And not to mention how heated things are right now around privacy coins, coin mixing, and compliance. Some people think it’s “normal privacy needs,” while others directly equate it with “laundering.” Once an on-chain path is interrupted, the tagging system becomes even more like a guessing game. In any case, when I look at fund flows now, I first cross-check them against exchange net inflows, perpetual funding rate changes, and the changes in OI. Tags are only used to find “people moving in the same direction,” not to blindly worship “who is who.”

For now, that’s it—I’ll later manually go through the specific sub-addresses that received the big transfers yesterday.