litphoria

That would be ideal, yeah.

This sounds really good!

I'm ok with everything except up-votes.

Actually, Up-votes and showcasing. That'll just lead to people lording about having a showcased RP.

Why would they lord on about a showcased RP when this is about articles to help people improve? I think useful articles should be in a very visible place, where they can be found easily.

I misunderstood, I thought you were meaning that RP's would be used as examples to be upvoted/showcased.

Oh. Not at all! Roleplays shouldn't be showcased at all.

Agreed.

The "showcasing" here sounds more like finding some way to have "featured" articles, presumably ones that are upvoted a lot.

Yes, featured articles that are easy to find because they've brought a lot of positive attention and/or site staff think they're useful.

I like the ideas here.

I think voting on articles would be good. However, I think being able to downvote as well is critical to a good quality control system. Voting can be an excellent measure of usefulness and quality in a guide: if I think a guide is useful and written well, I can upvote it. But... what if a guide is poorly written, not useful, or actively endorsing and suggesting practices which I consider very poor or downright harmful? I should also be able to do something. Getting downvotes in which can not only cancel upvotes and send an article downwards in score or positioning, and in fact send it into the negatives, is an important part of being able to effectively communicate over the quality of contributions.

Sometimes it'll just be me who considers it poorly written or not useful or so on. That's alright, because if it is just me, the number of upvotes will outweigh my lone downvote. However if it is poor quality, as judged by the majority, it's going to go downwards. It should go downwards.

The argument can be made that poor articles will just sit at 0. The issue is, so will relatively unviewed articles. And a frequently-read long-standing poor quality article might have amassed enough votes over time (from the 10% of people who thought it was any good or were too generous with their voting) to sit at a score of 20, just like a decent relatively-new article which is fantastically loved. Being able to distinguish an article with 1 upvote and 0 downvotes, from an article with 100 upvotes and 99 downvotes, is kinda important - one's new or languishing, the other's extremely contentious.

Down-votes aren't necessary and can do way more damage then good. "But... what if a guide is poorly written, not useful, or actively endorsing and suggesting practices which I consider very poor or downright harmful?" Then don't up-vote it.

Not having downvotes in a review system is also liable to do quite a lot of damage, due to not being able to give effective negative feedback reflective of problems with the article in question.

Perhaps a way to flag posts for moderation? If it's flagged for moderation by one (or more) users, one of the staff can check it out, and if it's more harmful than good, mark it as such or remove it outright.

It creates a situation that will suffer from vote brigading. "I don't like this so I'll get all my friends to down-vote it" which isn't helpful at all and is just a system that can be abused. If anything allow comments on them rather then down-votes. Comments with criticism or feedback are actually useful and require someone to think before mashing the "I don't like it" button.

Mood: I'm fine with that if the moderators are willing to act effectively as editors maintaining a certain quality and content bar, but I figure the ability to downvote is going to be more beneficial for people overall for the way it can collect from a variety of opinions and stances (not, for instance, just those of the editors).

Down-votes aren't opinions though. Hence the suggestion to have comments rather then down-votes. An article with 40 'likes' or up votes, and a 30 post thread where it's topic is debated with arguments for, and against is vastly more useful then the votes alone.

It might be even better if votes are private, regardless of up or down, and only visible to the mods who will select what gets showcased. MAYBE the author. They'd act more as a flag for "I think people should see this" than a rating.

If we're discussing abuse scenarios, the inverse is also quite feasible. A couple of examples with positive upvote abuse: "This article gives bad advice, I'm going to get all my friends to upvote it for a joke" and "I gave questionable advice but I have maybe twenty friends willing to upvote it if I bug them enough".

Comments suggesting constructive changes are going to be a helpful mechanism. They aren't mutually exclusive from upvotes. Sometimes what's necessary in a system that has any care about article score is the ability to say "this article is very low quality or has significant problems, and its score needs to go down slightly."

aren't mutually exclusive from downvotes* (or upvotes, for that matter, like I originally wrote)

So the consensus is that votes shouldn't be public. (viewable by the moderators and the author possibly) Or just not exist and instead use comments. No matter how you do voting, it can be easily abused.

Personally I'm fine with a public score if there's both upvotes and downvotes. I'm fine with the concept of voting overall, downvotes are just an important part of it.

Mood: Downvotes are also important even if the score is private and only viewed by moderators who are gating "featured" articles, because of the ability to quantify the people suggesting "this is not good content and needs to not be featured" in that scenario.

I think there's benefit to having voting (with score, with upvotes, with downvotes), and that those benefits outweigh the potentials of abuse, and that potential concerns abuse aren't worth worrying about enough to pull the whole voting system and lose its benefits.

Abuse definitely is a scenario we should consider. For example, Stack Exchange has a recognised abuse pattern of voting rings where a group of authors agree to unconditionally upvote each others' material, and they've built pattern recognition checks into their software to identify vote-circles. There's also a recognised abuse pattern of serial downvoting, where someone just upvotes or downvotes absolutely everything another user posts, because of some favoritism or grudge against them. The thing is, these abuse patterns generally find a hard time withstanding the general majority of voters because both downvoting and upvoting exist to point out good and bad quality articles (bad contributions still sink despite vote-ring attempts or serial upvoting attempts, good material still rise despite serial downvoting attempts). Then while the community tide is busy uncaringly overwhelming a comparatively small unhealthy/abusive voting practice, automated checks and moderation bring the abusive practices to a halt. So, even in abuse scenarios: ability to vote both ways is good, and downvotes are healthy to have available.

Voting sucks and leads to groupthink, just look at imgur or reddit. Nominations for outstanding articles which actually require people wanting to nominate to put some effort in (and preferably these nominations get checked by an editor) seem like a much better system.

Written nominations sound cool too tbh, if the whole intent is to identify the most outstanding articles for featuring or showcasing, and sorting articles among each other isn't so important.

Down-voting isn't necessary. The number of up-votes alone is more then enough for moderators to get a feel for how people view a certain topic. Without having both up and down-votes the viability of vote brigading nearly vanishes. A 3 month old thread with 5 up-votes? Obviously not many users care/agree. A 3 day old post with 40 up-votes? Ok that needs to go to the front page. A 2 week post with 20 up-votes and 15 comments, most of which are against and are backed up with solid reasoning? Ok that one needs more debate. This is how the current feedback systems works, and it seems to work well. There's little to no reason to being about a totally different system for a similar feature.

In practice, both downvoting and upvoting leads to people downvoting stuff they don't like and upvoting stuff they like. This works ... kind of well for sites that aggregate content that lives or dies based on how likable it is - but I don't see the objective here as offering articles that cause the least amount of dislike and the most like. We want to find the highest-quality and most useful articles for the community, even if those articles touch on sensitive subjects that would otherwise attract a lot of "no no no" downvotes.

That's actually a really good point.

So, what we most want is nominating for showcasing, rather than voting?

I think that's a good way of putting it yeah.