pull down to refresh
379 sats \ 13 replies \ @siggy47 25 Feb \ parent \ on: Capital Formation Weekly #1 mostly_harmless
Bias is something I worry about a lot too. That's why I'm relying more on zaprank to get a general stacker consensus. Maybe I don't understand it fully, though. What do you think about it? It takes zaps into account, obviously, but also number of stackers who zap, I think. Does it also take into account zapper's trust level? I really want to fight my own bias in boosting visibility of content.
Hot take: there are times when you want to be brutal in rooting out bias; and there are times when bias is exactly what is wanted.
For this post (and whatever variants that follow) I want to be maximally biased -- this is the sense I've made of this little slice of the world, and that's what I'm giving to you. If people value the elvismercury bias, it will be useful; if they don't, no shame.
I try really hard not to fool myself with self-soothing messages, though. That's a kind of bias that chokes the life out of you, eventually. Or out of a community. But if you buy a can of Coke, you bloody well expect Coke to be inside.
reply
I try really hard not to fool myself with self-soothing messages, though. That's a kind of bias that chokes the life out of you, eventually. Or out of a community. But if you buy a can of Coke, you bloody well expect Coke to be inside.
Yes, we need more authenticity and less conformity. Great example with the coke. I noticed I was very unauthentic for the most part of my life. I was unauthentic for so long, I actually thought that is me. I just am an unauthentic person, lol.
Turns out that's not true, I was just trying to conform too much. And that did choke the life out of me until I realized that's what's happening.
reply
Maybe I don't understand it fully, though. What do you think about it? It takes zaps into account, obviously, but also number of stackers who zap, I think.
Yes, zaps are taken into account but the first sats have the most influence. So for example, if someone zaps 1000 sats instead of 100 sats, their influence is only 2x, it's not 10x.
This directly means that the number of stackers indeed has an influence since 100 sats from 10 stackers is more signal than 1000 sats from one stacker even though both equal to 1000 sats for the post.
Does it also take into account zapper's trust level?
Since personalized feeds, it takes into account your trust to the zappers. So there actually is no unbiased view while you're logged in anymore. You'd have to log out to get a consensus view. The consensus view has trust seeds (@k00b, @kr and me) and then we mix that PoV with everyone else's PoV. That's also how Meme Monday is graded.
So everything sorted by
hot or top (posts or comments) is now ordered according to your preferences based on what you zapped in the past.1 For example, if A zapped a post and then you zapped that post afterwards, we will now tend to show you more stuff that A zaps in the future since your zap showed us that you might trust A with showing you good content now.2I hope I didn't get anything wrong. @k00b is more knowledgeable about this stuff.
I really want to fight my own bias in boosting visibility of content.
That's very virtuous of you. My strategy is kind of embracing my own biases though. I am biased towards content that makes me think (or at least try to be). And I think that's not a bad thing if I let that show in my territories or synthesis posts. Maybe your biases are similar.
But I also should consider that I am not aware of all my biases. Maybe being aware of your own biases is even an oxymoron?
Footnotes
reply
Thanks for this explanation. I guess I'll stick with emphasizing zaprank, but maybe as more of us synthesize a little bias might not be so bad. One of my goals is to shine light on good content that was missed or ignored by most stackers. There can be many reasons for this. I can't prove it, but it seems that it depends when you post, for instance, both as to day of the week and time of day. I think that because so many stackers are from the US, posts made during business hours in North America do better. I would like to check this out empirically.
reply
I can't prove it, but it seems that it depends when you post, for instance, both as to day of the week and time of day. I think that because so many stackers are from the US, posts made during business hours in North America do better.
I think the same. Didn't @davidw do an analysis on that in the past?
reply
I still haven’t run the numbers again with a bigger sample size yet (@Natalia is plotting my punishment as we speak) but I did post some graphs in this SN comment a month or two back.
reply
reply
reply
reply