But seriously, instead of follower counts an engagement metric calculated from the amount of replies and whatnot that your posts receive would probably be easier to do in Nostr and more valuable for end users.
Discussion
Langchain LLM AI models would love this stuff
I got more engagement on a photo of my foot than any post where I try to write something smart, not sure I want to see those metrics 😆
Something I've been wanting to look into is a metric based on follows and NOT ALL of the comments, reactions, boosts and zaps. A note would have a metric that would change based on the relative perspective.
#Attention is the ultimate metric. Social media devs seem more focussed on maximising the number of posts scrolled when the ultimate metric is how much attention -- proxy 'time' -- is spent on any one post because ultimately as humans that's the experience we are all looking for: singular focus/#theZone/being happy.
I would say, more value per time: high signal, less noise.
Would this be calculated by relays or clients?
Also slow but long running conversations would be a good indicator for genuine/interactive users vs. microshitposters.
Only problem with that is you'd end up identifying and thus attracting unwanted attention to old threads.
We do have the most relevant metric, Zaps. All other platforms have to find substitute metrics, we do not.
Money is the ultimate metric. The time is a proxy. All the marketing is just a scheme to make money.
I think more valuable than reactions are trees of reaction an counter-reaction. One-way engagement as in „reaction“ is worthless because it‘s not a conversation and it also includes spam.
I do prefer something like damus’ connection count, that changes dynamically based on what relays you connected too. But it’s still just potential. I think possibly a who’s online could be interesting, or just a count of who’s directly connect the the relays your publishing directly too… but also possibly aggressive towards people who are happy to mostly passively consume.
nostr:npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6, a few years back i got this idea from watching #BruceLee:
How can a post reflect true engagement when there is a bot farm of 1000s promoting it?
Accounts that act together should be counted as 1 so if 1000 likes for nostr:npub1ahxjq4v0zlvexf7cg8j9stumqp3nrtzqzzqxa7szpmcdgqrcumdq0h5ech and those *same accounts* also like a post by nostr:npub1acg6thl5psv62405rljzkj8spesceyfz2c32udakc2ak0dmvfeyse9p35c afterward then, as of liking Michael, they become counted as 1, revising Natalie's by -999.
I call the cluster of similar likers 'bubbles' after the UK governments social bubbles during Covid (it's when i got the idea).
Why Bruce Lee? Because Bruce appeals to a wide audience, and so getting engagement from many bubbles does the same.
What do you think?