Get rid of the question ban on Meta; suspend actively harmful users manually

Question

This user appears to have been question banned because of two unpopular Meta suggestions. (He says so in Chat)

Look at the two contributions that appear to have gotten him banned. Like those of many others, they may not be of perfect quality, but they are not actively harmful to the site, either. As it stands, you can get question blocked on Meta simply for asking questions that aren't popular with the Meta crowd.

Why does the team insist on keeping the question ban active on Meta?

Why can't real problem users be dealt with through manual moderator action? Don't tell me it's a problem of scale; as far as I can see, actually toxic users are very rare on Meta.

People will frequently complain that Meta actively discourages the expression of unpopular viewpoints and suggestions. With the threat of question bans in place, I find those complaints difficult to refute. I hate that.

Answer 1

[EDIT - see modifications (@ post end) based on Bart bringing this to our attention again.]

Today, having watched my first question (a feature request) here on Meta drop to -6 with no bottom in sight, I immediately got discouraged. Then a kind comment mentioned that downvotes meant mere disagreement about using or implementing the suggestions.

And then I found this thread.

Scary.

As a noob, having read many of these related posts, I'd say there is a large elephant in the room with 'double standard' painted on its side. (If you want to shoot a newly arrived messenger and ban me for "whining" or "complaining," or down vote me to heck, so be it.)

You got problems here. Problems that need human, not robotic, intervention, with human compassion, not bull-headed, superficial statistical analysis driving a robot. (Gads MDeSchaepmeester had TWO -12 voted questions and he was tossed? I read them - they were not dangerous or mean-spirited, not even annoying. In fact, what happened to him makes me want to Self-Ban.)

How can people be banned for "bad" (*) ideas on a site that is intended for meta-level discussions? Moreover, unless the FAQ is hypocritical and Meta is not really "run by the community," how can robotic blacklisting be tolerated? Worse, the posts I've read indicate the one banned has no recourse or appeal - other than making comments as here.

(*) "Bad" is a fuzzy logic concept, "Up or Down" votes are only binary. Therefore the existing voting methodology cannot possibly encompass meta-level questions, which by their nature must be more sophisticated than fuzzy logic.

Commentators should be can be constructively critical [if they want] and say WHY an idea is bad --- which is what happened in my case (the idea would be too heavy data-wise, it could be done better in other ways, etc.) [EDIT: searches show the sub-topic of requiring comment on down vote has been discussed to near beat-the-horse-dead status]

[EDIT 23AUG13 - side discussions lead me to withdraw the sub-suggestions of a down vote requiring a comment. Ditto a sliding vote scale (which I might suggest in the future): ]

I'd say, ***only here on META***, unless you are a moderator, or posting-god/-goddess you should not be allowed to down vote without a small comment as to why. OR the up / down voting should be changed to a sliding scale of 1 to 10 ... seriously, how can a Meta discussion be pidgin-holed into yes-no tallies?

As others here have voiced, banning should only be done by real people - and more than one ... perhaps at least three, high-level moderators - and only with due cause. And there needs to be an appeal mechanism. Otherwise "by the community" is mere double speak.

[ Edit after a day of comments below] Why no technical solution ? If there is some robo-banner banning people automagically, it does so using a threshold.

Simply replace:

if (threshold > bad) { ban_em ; }

with

if (threshold > bad) { notify_review_board ; } // clearly the mechanism is there already with flagging.

@random & Bart: by high-level or posting-god/-goddess, I mean at least a trusted user. Perhaps a good review board could be composed of a true cross sampling of the user community. An SE employee (in rotation, if need), at least two trusted users, two moderators, and two or three users below 500 - even include a near-noob like me (level allowed to up-vote and down-vote). When they get notified, they decide by equal-weighted majority, one person, one vote. This seems to me to be the only fair way to ban someone.

And if that seems like too much work for everyone, merely set the bad threshold higher. Notifying the board less would not thereby cause more bad posts from getting through, because the community can still bring nasty stuff to the proper attention via the flag mechanism.

I agree with notPekka - more human intervention. Consider it this way: when a human being's free actions are to be hindered, judgment should be passed by his fellow humans. If the speed trap's radar catches my car, photographs my license plate, then generates a ticket mailed to me, I still have the right to contest it in front of real people. We will rue the day when the speed trap sends us fine notices that are incontestable.

---

[ EDIT 16 SEPT 13 ] - With Bart resurfacing this issue and comments since I posted this answer, my suggestion warrants an addition.

This improvement is also simple, extending the earlier version.

Merely add in another threshold variable to prevent valuable contributors on StackOverflow (or any Stack Exchange site) from being "auto-banned" while still bringing "questionable" posts to the appropriate attention without the need of users' flagging.

As an example, the version below - compared to the previous version - is rendered in a stricter fashion on the low side and more permissive on the high. In other words, (auto)-watch newer users closer, but normal ethical users get a free pass from auto-banning. HOWEVER, if any post has "excessive" downvoting, then bring it to Moderator attention, just don't auto-ban willy nilly. Clearly such simple rules could be tailored easily to find the right levels.

//  Quality of activity thresholds 
//
//    First If block = a simple "high-pass" filter
//    Second If bloc = high value trigger  
//

SE_REP_THRESHOLD = 500  // community or SE owners pick appropriate Rep value.

//
// Modified version of previous suggestion
// 

// Get user's highest rep across all the Stack sites the user participates in. 
// This info is already in our profiles.

usersHighRepScore = get_UsersHighestRepScore (allStackExchageSites) 

if (downvote_threshold > bad) AND ( usersHighRepScore < SE_REP_THRESHOLD ) { 
    // next could be optional 
    ban_em ;             // LOW BLOCK - auto-ban prevents spam too 
    notify_moderators ;  // notify also optional, if we want to be " nicer " & unban later
}

// *Optional* High trap filter - high down vote AND high rep 
//  = post is "provocative enough" - someone should see what's up.

if (downvote_threshold > bad ) AND ( usersBestRepScore > SE_REP_THRESHOLD ) {   
   notify_review_board ; 
} 

Answer 2

First off, an explanation of the two status tags:

• Declined: We won't remove the q-ban entirely. See below.

• Completed: We already suspend actively harmful users manually, and will continue to do so. In addition, we've modified the score threshold for q-bans (again) based on a fresh analysis of the data - at this point, it is low enough that it should be extremely difficult to hit "accidentally" - if you're able to contribute to the community here in any consistently constructive fashion (for instance, providing helpful answers to support questions), you can get away with posting the odd unpopular feature-request. In a sense, you can earn the right to be occasionally annoying by being generally helpful. In no cases will expressing an unpopular viewpoint block you from further participation, unless you do it repeatedly and that's the only thing you do.

Community service vs. imprisonment

This site serves as the "nation's capital", and as such there are many people here from many different backgrounds and with many different concerns - concerns that deserve to be heard and addressed. It is important that dissenting viewpoints can be expressed, but it is also not fair to give anyone with a bone to pick an unlimited soapbox on which to stand and shout.

As I said, we do and will continue to issue suspensions for users who are abusive or disruptive to the communities here. If you're being overtly offensive to others - or even just an overly-disruptive nuisance - you'll get at least a short time-out.

However, suspension is a much more blunt instrument for dealing with folks who, for whatever reason, simply cannot seem to ever post a useful question. Unlike suspensions, you can "work your way out" of a q-ban: just post some helpful or insightful answers. The vast majority of people who use this site are in no danger of ever being blocked from posting a question because they've taken the time to participate constructively in the past, but even for those who start off on the wrong foot and... follow up by jumping up and down on that foot... it's not too late to reverse the trend: take the time to learn about how this network of sites work, share your knowledge with others, and you'll be automatically granted permission to post questions again. No need to wait out a suspension or try to bargain with a moderator - earn the privilege and it's yours to use.

Philosophically, I like this sort of merit-based restriction a lot better than suspension. Even if in practice both mean than someone can't post questions for a few months, the former puts this decision squarely in the hands of the participant themselves and the community they wish to be a part of.

Big thanks to everyone here for their input, and to Anna Lear, Tim Post, Jon Ericson and Grace Note for taking the time to sanity-check the data and talk over the philosophy behind this internally.

Answer 3

As the person whose question prompted Pekka to put a bounty on this, I can say that it IS frustrating when you spot improvements or errors and are not allowed to contribute this information.

By nature, I'm a helper. I spend HOURS a day on this site in the MS-Access, VBA and SQL portions of SO answering questions. I received the Tenacious badge because I continue to answer questions even when some people would have gotten frustrated. No "thank you", no points, no checkmarks... and still I answer questions because I know what it's like to be a n00b.

Once in a while, I get frustrated too, and I construct a poorly worded question on Meta. It wasn't harmful, it wasn't offensive, it was just maybe too much of a "b*tch session" and got hammered. So now I can't ask questions? The system is broken if that's the current way to handle it. And I don't know how to, or if there's even a WAY to, get un-banned?

Answer 4

While I agree with and have upvoted this proposal (this site is different and low-traffic enough for it to work), perhaps it's a reasonable idea to consider an alternative as well, in case this turns out to be a no go. Primarily because Meta is also a place for support questions, having a couple of bad feature requests or not constructive posts should not exclude you from help or reporting bugs.

In that light, perhaps only ban users from question which do not have the support or bug tags. This would still allow them to report issues and get support, perhaps giving them an additional option to redeem themselves.

Of course, if users were to abuse this privilege, they could be put on a suspension from the site. Maybe even through an automatic moderator flag were one of their posts to be retagged to something other than the aforementioned tags.

Answer 5

I think this phenomenon is a case of the MSO design telling us it has a problem. In particular, I think the use of downvoting for disagreement as well as quality is inconsistent with various aspects of SE, including the automatic question ban. If we work around this with manual intervention for the question ban matter, we'll just be putting off dealing with the underlying issue.

Answer 6

I agree. A lot of times, voting on Meta is about opinion, even in discussion questions, which can lead to warnings about question bans.

Since Meta SE (as of the split) is really is about opinion, we don't want to block people from asking questions if they are legitimate requests and follow the guidelines, even if not well received.

Take for instance my show vote buttons question. At the time (I don't now), I thought it was a good idea to show the vote buttons on these posts. It wasn't well received, and got downvoted a few times. Same thing with the home page question. Now I have a warning on questions that I might get question banned.

It doesn't seem like it'd be that much work to ban the occasional troll/flamer/whatever. They'd most likely troll the main site, not the Meta.

Answer 7

I'm giving my small contribute on Stack Overflow from some years and I really love this community. I come on meta for the first time to complain about some behavior in which I saw some contradiction. I just wanted to express my point of view.

I did a big effort to detail my opinion very deeply, referencing also books of eminent psychologist, because my opinion were against the tide. Dale Carnegie said that to increase your popularity, increase your influence, your prestige, to achieve all this objective he explains that:

Fundamental Techniques in Handling People
Don't criticize, condemn, or complain.

So it's really easily to get downvoted when you try to go against the tide, unless you are a great communicator. And I'm not.

In meta people use to downvote to express disagreement. So a lot of disagreement, which can easily be achieved expressing unpopular ideas, will lead to a ban. This makes also hard not only to express opinions, but also to put the light on issues in meta.

For sure I could express better. I got upset when I started to see very hurried downvotes. But what happened is that I got largely downvoted and I was banned.

I've documented all the happening in a chat: https://chat.meta.stackoverflow.com/transcript/693

In that chat I've tried to document the contradiction I see in some of our guidelines and the issues that are the logical consequence of these.

Meta SE is doing a great work. I hope I will be able to contribute to this community.