Who decides what you see on the fediverse? A look at voting patterns
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
the beauty of the fediverse is everyone can chose how they connect. I subscribe to piefed/lemmy communities from my hubzilla instance and receive chronological posts so there is 0 influence of anyone on what i see. -
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
I am indifferent because anything but Sort by New isn’t enough of a dopamine drip (and that’s not a Lemmy-specific problem).
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
Having most voted content at the top allow me not to see much of communities where there is a lot of automatically published content but low engagement.
Engagement is a proxy for quality. Not the best proxy but as I block communities I’m not interested in but which have engagement I discover slowly more and more niche communities with a little pool of active users but proper human-made content and engagement. -
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
No matter how many votes they make, they can only influence each post by one point, so it seems unimportant.
Unless you have evidence of vote botting?
Also, does this include the automatic upvote made when posting? If so those could just be the feed following bots that some instances have.
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
I think don’t watch the most upvoted content because i prefer chronological order
Edit :
So i understand your point with the vote quota, it’s a good idea. -
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
This looks like a Zipfian distribution. It’s a very widely observed phenomenon in many different fields.
-
No matter how many votes they make, they can only influence each post by one point, so it seems unimportant.
Unless you have evidence of vote botting?
Also, does this include the automatic upvote made when posting? If so those could just be the feed following bots that some instances have.
I think submitting over 750 upvotes a day is a pretty clear sign that a bot is involved in one way or another.
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
Would rate limiting solve the problem?
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
I guess I had made an assumption that most of the people on Lemmy and piefed were the type who chose to be here because they didn’t want votes or algos choosing what they see. As such, I figured that basically everyone here sorted by some chronological order rather than by votes.
Are you tracking this metric?
-
Would rate limiting solve the problem?
Sort by new and manipulate what people see.
-
I am indifferent because anything but Sort by New isn’t enough of a dopamine drip (and that’s not a Lemmy-specific problem).
Same. Probably means I am only of those top people who does influence posts as I vote on everything I haven’t blocked.
-
I think submitting over 750 upvotes a day is a pretty clear sign that a bot is involved in one way or another.
I’m not a bot; just a loser.
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
I feel like I cannot make a decision on this without knowing who the top 10 users are.
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
That is interesting data, thanks for presenting it.
I have to say I’m not that concerned about it, as long as there is no vote manipulation via multiple accounts involved. Everybody has one voice in each post and I don’t see a problem with some people using that voice more often. Hell, I am trying to upvote every post that isn’t something I don’t like in order to increase engagement. So I might even be one of the top 5 percent or so. And I’m certainly not doing it out of malicious ness or trying to manipulate what others can see (besides the normal up and down vote effect)
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
I wonder where I appear in these graphs. Honestly, I have no idea.
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
Each account has one vote per post/comment and if they up voted everything then everything would just have one more vote and same for the reverse. Unless most of them are voting nearly the same way it will just get lost in the noise.
It also matters when they vote. Early votes have more impact than later votes. If early voters are a representative sample of the general voting trends then they are doing a service by raising what people in general want to see and filtering out what people in general don’t.
As long as they are manually voting with a single account I don’t really care.
-
Same. Probably means I am only of those top people who does influence posts as I vote on everything I haven’t blocked.
Reading the comments here, I am amused at how many people sort by new, completely ignoring the mechanism in question. Why do we have up/down votes on posts again? lol (might be more useful for comments)
-
Reading the comments here, I am amused at how many people sort by new, completely ignoring the mechanism in question. Why do we have up/down votes on posts again? lol (might be more useful for comments)
You are reading a post that is 2 hours old. Naturally you will see a lot of people who sort by New. Come back in a couple days and see how the discussion has changed.
-
Out of the 37,000 people who voted for posts or comments in the last month, the 10 most prolific voters (0.02% of us) cast as many votes as the bottom 59%. Here’s how that looks, visually:
As you can see, a lot of people didn’t cast many votes. Someone cast 23k votes, with a group of 13 each casting at least 10k votes.
“But of course most people aren’t really engaged, most of those 37k people are just NPCs who don’t really matter” you say, “Rimu you’re just including them to make it seem worse than it is”, you might say. Ok, cool, let’s pretend the bottom 85% of us don’t matter and just look at the top 5000 voters. Here’s how the distribution looks among them:
Still super unbalanced. Let’s analyze this a bit.
Among those 5000, the top 147 (2.94%) cast as many votes as all the others (4853 people) combined. Among those 5000, the average number of votes cast in a month is 1142. Among the top 147, the average number of votes cast in a month is 6868.
How do you feel about a tiny group having this much influence over what news you receive?
I’m somewhat interested in these megavoters, mainly just in regard to if they’re real people or not and the personality of someone that seemingly votes on everything they come across. I’m curious if they interact otherwise or are just silently voting and nothing more.
I’m not really concerned as voting doesn’t seem to mean all too much here. I sort Top 6 Hr, but there’s a small enough pool of posts that I generally end up scrolling down to single digits as it is. Also, from the handful of times I used other instances, the vote totals don’t sync up anyway due to differences in federation details, so instance would still have an impact on what difference these votes have.
If anyone is also using a social media platform as their sole source of important information, I feel some potential vote manipulation is the least of that person’s problems with gathering reliable data.
Very cool stats though, I really enjoy posts like these.
