A loosely moderated place to ask open-ended questions
Search asklemmy π
If your post meets the following criteria, it's welcome here!
Looking for support?
Looking for a community?
~Icon~ ~by~ ~@Double_A@discuss.tchncs.de~
I saw this discussion in the comment section of some post, and it was about what the vote should be used for and how there is no official guideline for how to use it. I thought this was such a good question, that it deserved a post of its own. Related question would be if there should be multiple scores with different meanings, through which posts can be sorted.
I think that most of lemmy users are veterans of reddit, even lemmy is clearly based on reddit. There, they have really weird rule that it is suppose to be votes on how good and meritorical the post is iirc.
In reality, on reddit, here and everywhere else, it's just mostly "agree/disagree" button with all its consequences.
It's sadly the reality. The original Reddiquette is a really good idea, though. Encouraging well written arguments, and sorting them by quality, would help a lot of topics. Emotional gut reactions on the other hand are not really helping.
Probably yeah, but i cannot even fathom why they thought it's gonna work, the agree/disagree button is happening absolutely everywhere where similar systems were established, and the more mass platform the more prevalent it is - like what happened to youtube not long ago.
They could have done a way better job at communicating this idea via the user interface. The arrows suggest that the two voting options are opposites of each other, but according to the Reddiquette, they are not. Calculating the upvotes and downvotes to a simple number at a later time made it even worse. Originally, you had seperate counts for both vote options.
I could work, if you encourage the correct usage and educate your users about it. Also, a better user interface should communicate the intention and meaning.
If you ask me, if most users would use the system like that, it would be a very strong encouragement for everyone to participate in a quality discussion, writing well thought arguments and voting on who had done this, so everyone can see the comments with the highest merit at the top.
You don't even have to agree with the user in order to vote the comment up. If someone writes up a very good argument for atomic energy, I should acknowledge that and feel encouraged to come up with an equally well written argument against it, ideally touching the topics the user used as arguments, and enriching the discussion.
It's like a lot of things: It would be good if everyone would be doing that, but many don't, so most give up... and here we are. I would really love if we would at least try.
I could work, if you encourage the correct usage and educate your users about it. Also, a better user interface should communicate the intention and meaning.
Humans are just as often driven by emotions as they are logic, if not more. It doesn't really matter whether you design a perfect system or not, if it doesn't account for human behavior.
A quick look at how language evolves, the history of human behavior, how people will regularly vote against their interests, or any other plethora of examples out there of humans being human and you should realize that even when you give a vote a very specific label such as "malicious content" people will still use it to convey any negative emotion as well as use it to control what others say or as an emotional reaction to the content of a message.
The best you can get is an approximation, and you have to understand that people will ultimately use the system differently than you expect or designed them to.
I agree with everything you said. But for me, it's an argument FOR trying to create an useful system for everyone, instead of not trying. We all would have benefits.
It doesnβt really matter whether you design a perfect system or not, if it doesnβt account for human behavior.
I agree. That's why the system should account for that, which is the core of the reason for the system. It's not something that should exist "despite" human behavior. It should exist because of it. For me, this is the reason for the system.
The best you can get is an approximation, and you have to understand that people will ultimately use the system differently than you expect or designed them to.
Some people will ignore the information. But if we come up with a good and clear user interface, and the rules are simple and actively encouraged and explained, then hopefully more and more people will use it - and create a benefit for everyone.
Reddit sadly didn't do that - for their benefit, because they sell data, and gut reactions and one-liners are worth more than informed and well written discussion.
Fair enough, I thought you were arguing that it could be perfect and was just offering a word of caution. I agree these systems can be improved upon, and should.
Yeah agree, it's just the usual case of capitalism perverting technology (or really, mostly everything it touches, commodification is really tragedy).
There is no guidelines for the vote / preference buttons. Use it however you see fit, in whatever way you like!
Lemmy's sorting algorithm is here: https://join-lemmy.org/docs/en/about/ranking.html
Some possible meanings: high/low effort, good/bad argument, agree/disagree, funny/lame. I more or less expect all of those depending on context and the type of post/comment itself. Often combined.
What some people do is to vote based on the relationship with the author. Some people always downvote any comment by a person they hate. The opposite also exists.
Hm.. Yes. I am guilty of this. I would not say I always upvote/downvote, but the bias exists.
Add Lemmy points plz
I personally would love if voting was restricted to members of a specific community. That would truly help augment the signal/noise ratio. Practical example: it's not uncommon on /c/anarchism to have stalinist fanboys come and mass-downvote all they can find... except our forum is not intended for them to consume/judge.
Instead of blocking the votes, you could count them differently. But I think this makes more sense as a parameter for federation between instances, not something specific to communities. In your example you could still federate with the people that behave badly in your instace, but assign a low value to their votes.
Interesting idea. How can you automatically decide who the intended users are? Maybe a mix of having the community subscribed, and having a certain amount of comments, or a certain amount of upvotes within that community?
Interesting idea ! Then the votes would have the meaning that the post/comment is (not) appropriate in the community in question.
espcially all the tankies from the lemmygrad instance
There is no meaning to the vote numbers here on Lemmy, yet they are used to rank comments and posts. Doesn't make sense to me. Any metric that is the result of an unknown process shouldn't be used to sort comments.
Example: The 5 star rating at Amazon. Imagine if everyone would use it however they see fit? Some use a 5 star rating for simply receiving a non-damaged product. Some use a 4 star rating for a broken product, but the return process was okay. Some use a 3 star rating for receiving a non-damaged product. Some use the stars to rate the movie, some to rate the visual quality of the release (which can differ), some because their language wasn't included, some for the delivery, some for the packaging (which can differ), and so on.
As long as the people who feed the system use it for completely different reasons and by different rules, the resulting metric is... useless.
There is no meaning to the vote numbers here on Lemmy, yet they are used to rank comments and posts.
The fact that they are used to establish a ranking does give them a meaning : one votes posts and comments depending on whether they think they are worth showcasing.
A lot of people seem to vote posts and comments based on whether they agree with their opinion, and I think this is some (mild) form of propaganda.