9

On the per-site-metas, users don't have a reputation score different from their score on the site (so a user's meta.stackoverflow reputation is just their Stack Overflow reputation). But meta.stackexchange has no base site, so it just has its own reputation score. But why?

The recurring theme on meta is that votes on meta mean different things. Most relevantly, feature request voting is completely different:

On posts tagged feature-request, voting indicates agreement or disagreement with the proposed change rather than just the quality or usefulness of the post itself.

Reputation by its very nature is using gamification to create incentives. On most of the sites this incentive is clear: have users contribute quality content. On meta, it's less clear. A user who submits an unpopular feature request on meta gets the double frustration of having their request denied and losing a bunch of reputation in the process. Additionally, given how the site-specific-metas work in terms of reputation, I'm left wondering whether meta was ever supposed to have reputation in the first place.

So what is the intended incentive of reputation on meta? Is discouraging users from submitting unpopular feature requests by design or an oversight?

Note: This question is not about functionality. I understand that privileges are unlocked with reputation, but that could easily be done via a Stack Exchange-wide reputation rather than a meta-specific one. This question is just about the incentives that reputation on meta creates.

| improve this question | | | | |
  • 3
    Is discouraging users from submitting unpopular feature requests by design or an oversight? By design, who needs an website with 4500 unpopular feature requests and incentives for users to post even more of those? – rene Dec 17 '18 at 13:43
  • 3
    I very strongly agree with your perspective that one shouldn't lose rep or privileges just for expressing unpopular viewpoints. Additionally, the system offers a "clique" feeling by giving you the option to get your lost rep and privileges back in exchange for recanting your idea. However, your main question, which is why MSE has its own separate rep, is discussed in that post. – Sonic the Stay-Home Hedgehog Dec 17 '18 at 13:44
  • @SonictheInclusiveHedgehog That question related, but it is about unlocking privileges, which I explicitly mentioned is not what I'm asking about. I really want to know about the intended incentives reputation creates around controversial or unpopular posts – Zags Dec 17 '18 at 13:45
  • 3
    Are you asking why meta has rep, or trying to make a case that it shouldn't. You already know why: you state it in your Q, and there are further details in Sonic's link. So it seems to me you would are actually recommending a change. If that's the case, I think it would be better to cast your Q in that light, rather than a question of "why". – Dan Bron Dec 17 '18 at 13:50
  • 1
    @DanBron I may submit a feature request depending on the outcome of this discussion, but I would prefer to get insight into the history of the current situation first, as that both impacts whether or not such a feature request is a good idea, or even what such a feature would look like. – Zags Dec 17 '18 at 13:56
  • 2
    Since this post is not tagged feature-request, and since this is a high-quality post on-topic for this site, I'm upvoting this. – Sonic the Stay-Home Hedgehog Dec 17 '18 at 13:59
  • Also, I edited your question to make it clear that it's not a duplicate of that other post. – Sonic the Stay-Home Hedgehog Dec 17 '18 at 14:04
  • 1
    I think the reputation system has the effect of disincentivizing unpopular suggestions, but that wasn't the primary goal. Gamification is baked into it, but I find that users anywhere don't provide good content when their focus is on gaining reputation, and that focus gives a pretty rough reception. I think a larger issue is the defensiveness that askers adopt when their feature requests are negatively received. – fbueckert Dec 17 '18 at 14:36
  • I've found out through experience that, while it is extremely easy to lose reputation, it is also very easy to gain reputation. – Pika the Wizard of the Whales Dec 18 '18 at 4:29
13

The idea of negative reputation is certainly not intended to be a barrier to any earnest attempt to participate, even if that attempt is to propose an idea that's very likely to be unpopular. I wouldn't call it an oversight either, because we're aware of the problems users encounter when learning how to navigate meta.

I'd venture to say it's not wholly "rep" that causes trepidation, seeing something you put some time into writing up and explaining get decorated with dozens of down-votes is far from encouraging, and it happens frequently enough that people take on a pretty reasonable fear of it happening to them. So, the problem isn't really exclusive to just MSE, it's just exacerbated here because loss-aversion instincts have more to cling to.

However, we really need folks that come here to talk about changes they'd like to see to have some kind of stake to hold; we'd like people to at least give what we've got a shot before pitching in to help reinvent it all.

At the same time, however, we need to be receptive to talking about barriers to participation on meta for folks that have used our system and have things to talk about. If the way things work here is keeping people that actually use our sites from talking about how they could be better, that's a problem. And, if you ever have a problem where the canonical solution is get a thicker skin, well, then you have two problems.

We're looking at making some changes to how meta behaves based on tags that people apply, and helping people better select which main tag (e.g. discussion, support, bug, feature request) to use. People get frustrated with people asking the same basic support questions over and over again, and that manifests in voting pile-ons. So one idea is just making much easier for people mark support questions as duplicates and more aggressively culling any that don't get upvotes. If we do that, maybe we wouldn't need down votes on support questions.

Bugs don't need upvotes, they need "I can reproduce this" buttons, and they need to tie into our back-end bug tracker (currently Azure). It's really hard to differentiate upvotes between "this was entertaining" and "this is happening to me too", or down-votes from "this wasn't written well or is a dup" from "can't reproduce". That makes bugs difficult to triage.

We (employees) need to be able to blog on meta without actually asking a question, but we need to build a more blog-ish archetype for that.

So I'd say it's more a work in progress that should have had a higher priority for a while now, but one that we haven't forgotten about. The per-site metas on 95% of our sites function just fine and people indicate that they feel good about low barriers to participate on them. The problem is here with the 'Senate' of meta sites (MSE) and similarly high-traffic parts where attempting to build consensus can feel like a pile-on.

We also need polls implemented much more deliberately.

I'm going to be proposing some ideas for some experiments in 2019 after our product teams settle in a bit more (we've got like 6 more hires to make at the time of this writing). We're not going to do anything drastic like overhaul everything in the middle of the night, but we do need to have conversations about what's not working with each area that meta tries to serve, and address that area by area.

| improve this answer | | | | |
  • I once wrote a long argument as to why losing reputation just for unpopularity is a bad thing, and emailed it to Shog. (I believe you can read it, since Jon Ericson was able to read it, but if you need, I can send you a copy as well.) – Sonic the Stay-Home Hedgehog Dec 17 '18 at 14:50
  • 3
    The problem with having the site treat questions with different tags differently is people will just intentionally mis-tag their question to get the behavior they want. People already intentionally tag their feature requests as things other than feature requests to try to get people to change how they vote on them, saying things like, "You can't downvote my question if you don't like the feature I'm proposing because I didn't tag it as a feature request." Having the system enforce different voting behaviors is just going to strengthen the desire to mis-tag questions. – Servy Dec 17 '18 at 14:56
  • 2
    I am hopeful that if you implement the "blog-ish" architecture here, you still openly and earnestly invite responses to those posts, even if they're not answers. The point of Meta is to solicit feedback; one reason Meta gets so much more traffic and attention from the community than the blog is that we feel we have a voice here (and it's also the reason we get so ... cranky when MSE is used as a executive soapbox, like the infamous "Time to take a stand" post). If you just want a blog, well, use the blog. – Dan Bron Dec 17 '18 at 15:46
  • 1
    @Servy: "The problem with having the site treat questions with different tags differently is people will just intentionally mis-tag their question to get the behavior they want." Then they will get either re-tagged or closed. Also, I imagine that in such a system, the poster would be given different "Ask Question" dialogs based on what they're trying to report or say. – Nicol Bolas Dec 17 '18 at 17:50
  • 1
    @NicolBolas Well it often doesn't happen now. That or they roll back the changes. And even if you can make the changes and have them stick, it means you constantly need to be dealing with that problem (or the consequences of not catching it quickly). At a minimum, they need to have a plan for how to deal with that before making such a change. – Servy Dec 17 '18 at 18:04
7

Practically - I don't think it was intended to do anything. It was essentially 'just' another site thrown together cause people asked for it. Stack Overflow used uservoice in its early days instead.

I've never cared about reputation here. (Well I wanted 10k to see a deleted post - but that's a story). I do care about the potential impact things can have on the communities I am part of, and well, I guess I've done well reputation wise for it.

Reputation does, at least in theory reflect subject matter expertise. On one hand, the value of being an expert on the ins and outs of the Stack Exchange Network is debatable. On the other hand, there's many folks with little or no SO experience who know the system really well and are willing to spend the time to share it.

On meta, it's less clear: a user who submits an unpopular feature request on meta also gets the frustration of losing a bunch of reputation and privileges in the process.

Which is true. And well, people who work here get both the best and worst of it. It's worth remembering that, well, unless you're systematically making unpopular features, it should even out a fair bit.

There's an essential part most people miss though - that it's essentially impossible to hit a post ban here.

So in effect it discourages unpopular suggestions by people who're primarily interested in well, popularity.

I've made unpopular suggestions to solve what I consider a real problem, and it's still there. And I've made popular requests which I probably ought to regret.

At the end of the day though I suspect meta is a place for powerusers of the site, and many of them are less oriented towards pure reputation than watching out for news and trying to influence the powers that be in good ways.

| improve this answer | | | | |

Not the answer you're looking for? Browse other questions tagged .