ryan

gave up on trying to make it in the cool kids club.

Digital Media Producer. I like particles and waves too. I like art.

So basically yes, except i’m more focused on why the algo is excluding fids. if you look at the bottom of the image, you can see the number of lines (>200,000). If you look on the right side of every line of text you’ll see the date of last review . I removed all fids that had been reviewed in the last month, and the rest haven’t been “seen” by the spam algo in over a month, some much longer ago. Google says Warpcast has 828,000 users, if so then ~1/4 of them seem to have been skipped by the algo.

Okay, so just making sure- you’re saying that the dates listed in the json file/dune dashboard are the dates that the user was set at that label, not the date of last review by the algo?

hamst.art $hamster /hamster 
martinvr.eth 
deploy.degen.eth

The algo is not excluding afaik. You only get a record and a date in the dataset when you have a label or when it is updated. It's looking at everyone regularly

it’s definitely within the realm of possibility that a churn list could be  getting used. So that when somebody is inactive for x amount of days, they get set to 0 or null/excluded.Which if true, makes sense- but that must not be the only list being used.

Woah. Interesting do you think that might be because those accounts haven’t casted or something?

Pi-chi.eth (she/her) Artist /animeoutcasts
Hosting: /sakura /kyoto /momiji  /anime-manga /farcasterunion

https://portrait.so/pichi

Could be they are excluding review of fid groups based on labels like "power user" or "active user" or... ?

Not sure but maybe @pichi or @mvr has some thoughts?