Oddbean new post about login | logout Literally me. 😎 I feel like this is an npub quality statistic waiting to happen. https://m.primal.net/Hnjn.png
Where can you see these stats? I think you completely mog me with those numbers.
nostr.band
Thanks 🙏
🫡 ma pleasure
You would have 788, Ser Sleepy.
788 ?
Score. 16551 zaps : 21 median zap That means a lot of people appreciated your stuff very often.
ohh I never seen that 788 score before where’d you see that
OHHHH lol You divided total zaps by median zap amount
I was so lost https://media1.tenor.com/m/w2pqHCw_RJ0AAAAC/confused-confused-math.gif
What’s the headache? That’s they are wrong or they can’t disprove the fact?
Flawed standard models go deep 🗝️
The numbers don't say anything on their own. Only in relationship to each other. You'd need some sort of ratio, like number of zaps : median zap Then I would have a score of 58 and you would have one of 4 because most of your sats came from one zap. Noshole would have 28 and Gigi would have 868. If my math is correct. 😏
Unfortunately, zap receipts can be spoofed so they're probably not good to use for wot. The only one that knows for sure if a zap was real, is the recipient that manages the wallet that received the zap. "This isn't really a payment proof, there is no real way to prove that the invoice is real or has been paid. You are trusting the author of the zap receipt for the legitimacy of the payment." From https://github.com/nostr-protocol/nips/blob/master/57.md
It's the same with followers in WoT. People make multiple npubs and then they can recommend each other.
True, yeah I always felt like these are both somewhat non starters.. the only things that seemed like they would work would be more like fiatjaf's pyramid, or that notabot project, where people vouch for each other, and when bots are discovered you know the invite tree (how they got in) and can prune it.
you can do that with WoT as well, because it's always one dumbass in your list that is at the root of it, cut them out, and their immediate neighbours and voila, clean again this is why i rage against wishy washy "tolerant" people
Yeah but following someone isn't the same as vouching for them
Would probably need to include length of time.
Which stat specifically on that list are you referring to, number of zaps received?
If so, i’m assuming @semisol ‘s primal algo fucker has meddled with mine 😳 https://image.nostr.build/4d7dab5c159b9e62ee8fc553da8596919bed80319da7669bfd1eb72b16a86d6f.jpg
Neat! Looks like I have a lot of work to do to reach your level 😂 https://nostrcheck.me/media/30e8cbf1427c137fa60674a639431c19a9d6f4c07fd2959df83158e674fccbaa/0bd4fb20663aa08eddced0aa9ecebd73724d7ecc05fffd053ef59f0ebbe92a3b.webp
Interesting. https://i.nostr.build/xEGmV.jpg
19 score
Wow, you got one really big zap! 👀
That was the friend who got me onto the platform. That was his incentive. It worked.
What does that mean then?
Means you've provided a lot of value to a small number of people.
Oh interesting. I think I quite like that.
That actually tracks pretty well with the rest of my life. 🤣🤣
Social credit score
Any measurement in social media is a social credit score, by definition.
“Social credit” scores (“trust” scores, whatever one calls them) are like money. Centralized: a tool for tyranny Decentralized: essential for freedom and prosperity
Fair point
@brugeman What’s the point of counting the minimal number of sats zapped? Seems like a useless statistic.
By quality meaning what?
I guess that's would be subjective. I consider npubs who have spent more time on here, having positive interactions and adding value for others, as being higher-quality.
Loud populists would win, no? Speaking for myself, I would give npubs with rare but original (self-written) content more value than the blasting spiderwebs of engagement. You will need to answer how you want to weight vertices. I don’t think that a “pagerank” (original Google approach) would work in the long run. I think “Yahoo link lists” are more stable because they work on source quality. For example: if Uncle Bob said something once a year, I might want to see it.
I think loud populists should win, to be honest. Social media expects a willingness to be social. This is a measure for Kind 01 notes. Nobody wants a SM feed full of people who "tweet" once per year. People who write seldomly, but thoroughly, or with great effort/expertise, should focus on writing long-form notes, which could be measured separately.
I see the point about being social (how this ever became something digital). But it enables professional social medians. Like professional polititians, they’re not rooted in the real world. I personally would not value influencers high. It’s the sort of hyper-centralization, the internet naturally does. And I don’t like it.
My measure is actually counter to influencing. That is the whole point. Influencers don't interact much. They receive large sums, but not frequently in small amounts. That's the sign of an interactive pleb, who is in the convo. This also helps separate "dead accounts" that aren't worth replying to because they don't interact, from simply VBAs like Odell. It's not the size, per say, that separates a pleb from an influencer.
VBA means?
Ok, then interaction would be bidirectional? This I would like. What I wouldn’t like is the gravitational field around stars.
Proxy indicators of trust can have some utility, but in general we should be thinking about replacing proxy indicators with explicit trust attestations.
WoT isn't working because so many people use follow-for-follow, even with bots and scammers.
Yup. Plan for the Grapevine: I attest I trust you (or don’t trust) in some context. If you want to sell your attestations for a few sats, you can do it, and my software will pay you automatically if my Grapevine tells me you’re worth it. This will incentivize you to provide attestations that are USEFUL; otherwise my Grapevine will tell me to stop listening to what you say. So if you follow all your follows indiscriminately including the scammers, my Grapevine will tell me your follows list is tainted and worthless so ignore it.
Better than "zap amount", which is currently the most common statistic. If you get hit randomly by one npub with 50k, you immediately top the list.
I think you could use it to weight trending list entries.
For instance, score of 1 https://nostr.band/npub1dtgg8yk3h23ldlm6jsy79tz723p4sun9mz62tqwxqe7c363szkzqm8up6m?overview=stats
For instance, score of 5 https://nostr.band/npub1g0587hzzckcncxfm78n0996qe2s58nspy29wf02tqcj5sdzcpj4q6j40hv?overview=stats
I think you could use it to weight trending list entries.
For instance, score of 1 https://nostr.band/npub1dtgg8yk3h23ldlm6jsy79tz723p4sun9mz62tqwxqe7c363szkzqm8up6m?overview=stats
For instance, score of 5 https://nostr.band/npub1g0587hzzckcncxfm78n0996qe2s58nspy29wf02tqcj5sdzcpj4q6j40hv?overview=stats
Yes, of course. But I'm interested in kicking off the discussion because it's been stalled out and I'm frustrated. 😂
