Inscriptions are done in 2 stages, a commit tx that has Taproot outputs and a reveal tx that includes the inscription data in the witness for an input that uses one of the TR outputs of a commit tx. A commit tx often has many TR outputs. When I’ve looked at daily averages, it runs about 11 reveal txs for each commit tx. There are small minority of transactions that are both commit and reveal txs and in those cases they are categorized as commit txs here.

I identify reveal txs by looking at the witness data for inputs and decoding the scripts and looking for data that conforms to the ordinal “spec”. Once I find a reveal tx, I look at the outputs that are being referenced by the inputs with the inscription and mark that transaction as a commit tx.

The Normal category here is anything that isn’t a commit or a reveal tx. They could still be related to inscriptions/ordinals because they could be transferring to other addresses etc, but they aren’t explicitly a reveal or a commit tx.

I’m not yet tracking outspends of inscription txs and I assume that represents some of this new activity.

Plus, who knows how much of it is people moving sats into ordinal wallets so they can participate, etc, etc.

Presumably a lot of the new activity is related to this crap, but it’s impossible to know exactly how much of it is except by inference.