What should I use it for?
In case of posts - should I downvote things that I don’t agree with? Or things that I don’t want to see in my feed (posts with low amount of votes fell more to the bottom)? Or things that I consider harmful/stupid/younameit?
In case of comments - should I downvote something that I don’t agree with? When should I use this button in case of comments. Or maybe I should not - cause what’s the point, actually?
If everybody uses a system in a different way, how useful is it to display the calculated total of votes?
For example, if people are using the 5 star rating system on Amazon differently, how much worth does it have to the users? What does 3 stars mean? What does 4 stars mean? Is 3 stars a normal and regular product with satisfying quality? Or is anything less than 5 stars bad?
Even if there were rules for using the voting buttons, it’s impossible to enforce.
Of course are rules like that not possible to enforce, but that doesn’t mean they’re useless. A community can agree on these things and mostly follow it, even when there is no way to literally and 100% enforce every vote to be exactly how the rules state.
Fair point, but I think in practice far too many people would still use the voting buttons however they want for it to work, a problem which scales with the community.
The thing is that a vote is a single binary data point with no additional context, and you can’t determine with complete accuracy what the voter’s intentions from the vote alone.
If everyone uses them as they like, and nobody can know how the majority uses them, and nobody can know how the ratios of different usages are for any given post or comment, then what exactly would the votes mean? To phrase it differently: I see that a post has 34 upvotes and 9 downvotes. I have no idea what any of those mean, none of them votes could be done with the same intention. Or all of them are done with the same intention, but which one? I can not know, so what meaning do the votes have? Can I even know that?
If there are no rules, and everyone uses votes differently, in what way are they useful?
If you have rules, and take care in informing the user, create the UI accordingly to the rules, use tooltips to remind the user about the rules, and so on… there will be a certain percentage of the community doing it in a certain way.
If you have no rules, then the above applies.
Fair enough, but I still think this depends on the community. A community focused on professional discussion about programming for example would probably have members that follow the voting rules more closely than a humour/meme community.
I see what you mean, but still… what I said above is still true for both of those communities, and such rules and their implementation would benefit both of them.
Scores aren’t about objective truth, they’re about comparisons. If someone gives less stars than others on amazon, it pushes all the scores down, but the comparisons will still be the same. This is an old and refuted critique of range voting, of which approval voting (which this uses), is a subset.
That’s a very good point but it assumes that contexts are equal which not true at all. If you compare two identical products that might work but for complex communications to imply that they can be compared on a subjective basis is a bit silly.
What do you mean with “comparisons”?
Exactly. That’s the problem at hand. This happens even when the person giving less stars actually intended to rate the item positively, but he simply interpreted the system differently than others. The same applies vice versa.
Can you supply sources for that? The only sources I’ve read a few years back detailed the problems of such systems. Maybe something different was found out since then?
And there are many critics of that system. I’ve laid out some arguments against this system in this thread, and the other thread where I’ve asked a similar question a while back. The main argument would be the formation of echo chambers and an erosion of quality contribution.