Farcaster

Data

Curating /outcasters has become more challenging due to the increase in spam on Farcaster. So I have decided to run a new experiment:

Step 1: download list of all Power Badge users
Step 2: use list to filter top 1000 casts in last 7 days
Step 3: remove spammy users from list
Step 4: go back to step 1 and repeat weekly

I've just looked at the top 1000 casts based on most likes recieved since Sunday, April 21, and it turns out that @imthedude has casted 175 of them. That's 17.5% of the top-performing casts in the last three days. If this isn't spam, I don't know what is! But somehow, this person still has the power badge.

thot leadership at /opacity --
host of vibe check podcast 
▶️ https://pods.media/vibe-check
❄️ icebreaker.xyz/dawufi

Shouldn't the user be getting nerfed on Warpcast though? What they are doing is not ambiguous

If the casts perform they perform

I've literally never seen him in my feed tho so I'm fine over here and algo seems to be working in that regard

Even if Warpcast figures out a way to hide all the spam, I wont be celebrating if other clients can't give us a similar experience. The promise of Farcaster is that you can switch clients and barely notice the difference. We are not making good on this promise yet.

Nouns-x-Farcaster

i disagree on the promise of farcaster

the promise to me is that you can switch clients and have a massively different experience. Same graph, different use case. 

TBF though, we're not making good on that promise yet either lol.  My guess is /nounsxfarcaster will be the first, but always hard to call these things

The biggest issue in social is if you ruffle the wrong feathers you get rugged. Can’t get rugged on FC but if every client has a massively different experience, then it kind of defeats the purpose no? Not saying all clients need to be clones of course, but we do need a few that have close feature parity with WC.

maybe in the looooong run, but good luck to anyone trying to keep up with v and co on feature parity.  Way better to align incentives and bring a differentiated vision instead of being antagonistic to wc attention driving expertise

I've always been of the opinion that it's not good if the protocol is captured by a single team/client. Sounds like this is happening with FC. I hope it wont be the case in the long run.

0% chance in the long run, the data is too valuable for others not to figure out novel use cases for it

but in the short run, fc growth is dependent on wc growth so i think it's all fine

they've got xmtp spam filters which could theoretically help

but not a part of the team anymore so going to need to talk to someone over there to be sure :)

Is Airstack doing anything interesting that would help with clearing out the spam? Because the data needs to be cleaned up in order for it to be truly valuable.