you are viewing a single comment's thread.

view the rest of the comments →

[–]JasonCarswell[S] 1 insightful - 2 fun1 insightful - 1 fun2 insightful - 2 fun -  (0 children)

We can't expect to get miraculous information from binary data, so yes it's more complex than upvote/downvote. But a 5x5 input grid is not really that complex. Get that working and we can add all the extra bells and whistles to the GUI. More critically is how the data is utilized. THERE is the complexity. There is where it can be screwed up, depending on how you use it. Yet even still, the original input data is just the original input data and should not be corrupted. It's the use of the data to generate other data that could be the problem if the expressions are wonky. Nothing complicated will ever be perfect on the first try. There will be hurdles.

But it is great to have far more voting options to (1) give some feedback to the poster, and (2) rank the post in the list or category.

I would love to have more stats available to the poster and for ranking too. For example, how many people even look at the post/article/image/video? How many look at the comments? For how long? How many folks vote on the title alone without even looking at it?

With this MetaVote thing and the data it would produce it wouldn't matter who was a STAB or not. There'd be no need for filters besides reporting something illegal. Users would be classified, stereotyped, and tribalized by their actions, their opinions, and how others view their content. Echochambers would form - but critically the site admins and/or the users would decide how much content they want to see that's not within their statistical scope. Say 75%-90% could be what's expected, but that remaining 10%-25% could be mixed content that's outside their norm. By marking it they would contributing to the auto-curation process by strengthening the definitions of their circles of interests. Obviously not all rap fans will like country music, but some will and they will form an overlapping sub-group.

I would like about 10 options like friends with different colour codings.

And the posts are ordered by how much points my friends gave the posts.

I like this idea a lot. Of course this kind of thing could be done with MetaVote data too.

Additionally the admins have certain friends and they determine the ordering of the default list. (A bit like "editor's choice" on )

I guess I'll have to check that out as I don't understand it.

This will make the site easier accessible for new people, and guide the site into the way the admins want.

What/however that is/works I'm all for it.

I think that a site needs some guidance to keep it from turning into a chaos of information.

Indeed! That's why since I arrived (from wiki) I wanted to improve the sub categorization and organization process.

This way STAB will not be visible on the site, even if they work together massively.

Sounds good but I don't understand how.

With your voting you grade the quality of the information, but people put bias in the voting.

Indeed and that is to be expected, and maybe even embraced. Everyone will have their own voting style. Some people will hypervote and some will barely vote and everything in between. If someone is an extreme voter their votes should be weighted to have less influence - IF it's a point system. I would reject the point system. Amazon has a simple 5 star voting system and you can see when there are polarized votes and when there's a consensus and when there's a variety. Based on that alone you may tell if you'd like the book or not depending on your politics and the topic, and even more by the pro/con reviews.

Also what if 3 people really like some music and 100 hate it. Who's opinion is valid?

I say let everyone express themselves and vote as they like. Even STABs. Their natures will reveal themselves and thus be filterable.

I think problems may arise with too many users (including new and STABs) and/or not enough adequate input to gauge and sort them properly. And/or not coming up with algorithms to sort them adequately. But again, these would be expected.

Like I think that most news is a psyops, misinfo and gaslighting. And it can be rational, well sourced, reliable gaslighting.

That's why the opinions are 1 click and the "Topic Right-Tags" you click as many as are appropriate.

Your statements are more than just keywords or metatags.


Here's something I always intended to show in GUI illustrations: manual metatag input. Why be limited to 25 2525 or 25 million squares? Also, if there are 6000 squares (I don't think there are even that many subs on SaidIt), IMO it makes sense for a big chuck (2000?) to be custom user tags - and for users to share and trade grids. Yes it's huge process to learn and a huge GUI to develop, but the data ramifications are profound.

Also, if we really had our shit together we could tie it all into Sure the sister companies are corrupt and maybe even WikiData is biased in ways, but it would be a huge leap forward by just adopting their classifications, categories, and hierarchy systems until they could be unbent.