These days I've been looking at address profiling with those tags/clustering, basically just piecing together a story that groups a bunch of addresses as "the same person/organization." It can be used, but don't trust it too much, especially when on-chain funds are moved around, crossed multiple bridges, and then go into CEXs, what you see as "smart money inflow" might just be a market maker adjusting positions... My eye-rolling point is this: the data looks very scientific, but the conclusions are often very mystical.

Recently, retail traders complaining about MEV and unfair ordering is also quite real; as validator income rises, that "fund flow" on-chain seems more like being pushed around by others. My approach remains the same: use profiling as a reference, not as an oracle; keep positions smaller, and when volatility picks up, use options to hedge the tail risk, don’t let a few tags set the rhythm.