Recently, people have been saying "On-chain data got stuck for a moment, is someone causing trouble"… Basically, many times it's just the indexer/Subgraph/RPC rate-limiting or rebuilding, the node isn't down, it's just that the data retrieval path is congested. Especially when you're watching large transfers, exchange hot and cold wallet movements, those "smart money" narratives, if the data source changes or there's a delay of a few minutes, the scene can shift from "just transferred out" to "still in place," and emotions immediately start to skew. My habit is: when I see anomalies, I first compare two different data sources before drawing conclusions.