Are there any rules against constantly voting not like others in review queues?

If yes, do the rules apply only to sites where they are enforced by the engine, in the form of review audits*, or are they universal and apply to all sites in the Stack Exchange network, including beta sites where review audits are not available?

* There are different audits, but most of them rely on relatively reliable opinions of others, e.g. if a user votes against closing a question which others closed and heavily downvoted, something is likely wrong, according to the logic of the engine.

Context: I have noticed a user on Stack Overflow in Russian who votes against closing in 85% of cases. Often, he is the only one voting against closing, especially on questions like, "I've downloaded NT 4.0 from torrents, how to compile it?" (quoted in full) He would have likely been review-banned as a "robo-reviewer" already.

However, review audits are disabled on Stack Overflow in Russian, so I reported the behavior in chat, but it caused a heated discussion. Several users expressed their opinion that voting not like others is totally fine, is a form of free speech and must not be suppressed. Furthermore, some users said that rules enforced by the engine are not actually rules and cannot be applied.

So I want to know what official policy on this behavior is, whether the rules can be applied if the engine does not enforce them, and how moderators are expected to act on it, especially on beta sites.


There are two different types of review bans, depending on what triggered them:

  1. algorithmically by failing honeypot review audits
  2. manually by intervention of a human moderator

I believe that the rules for the first class are set in code, but that since no code is involved in those from the second class, they are not. At the very least, they are not public if they do exist.

Because you need a human exception-handler for cases that code can’t, won’t, or doesn’t detect, this always requires latitude in judgement, and so it is always possible that two moderators will approach the same situation differently. This is especially noticeable on SO proper, but in principle it is applicable to any site at all.

I cannot speak for your moderation team and I do not know their thinking, but they should not make public their exact criteria for these things particularly when it is about a specific user not a more general matter.

It always helps to build community consensus by going to a site’s meta as you have done. Different sites have different sentiments, and a beta site in particular is still a community-in-the-making.

If a given user consistently goes against the bulk of the other reviews by community members, it could mean any of several completely different things, and little good could come from speculating which if any might apply in any given instance. The moderation team can look into situations like this in greater depth in private. That privacy includes just what they look at it, how they look at it, and any general guidelines considered or specific actions taken.

I think that is the best you can hope for.

However, if the moderation team does decide to apply a manual review ban to a particular user, they are bound by the moderator agreement never to reveal personal matters of moderation such as this would be. So you might never know. This policy of protecting a user’s privacy is very important on all Stack Exchange sites, and this very much includes most aspects of moderation activity with respect to any given user.

On a personal note, I have often observed various sorts of correlations in review actions by different pairs of reviewers. Some of these are very strong correlations and they run both ways. So for example, users X and Y might vote the same way more than 90% of the time, but users X and Z might vote the opposite way more than 90% of the time. So if you know how reviewer X voted, you can predict with sometimes surprisingly high confidence how reviewers Y and Z will vote. This is not collusion, nor is it something that the system or a human moderator need look into. It simply represents different attitudes held by different community members.

Original Answer

Perhaps in Russia it is still illegal to vote differently from how others vote, but in most of the world voting represents free choice.

If it does not, then it is not properly called voting, but rather something more on the order of rubber stamping.

  • 1
    Voting not like others results in review-bans on Stack Overflow and other sites, so "in most of the world" it's illegal to do this. – Athari - Make SE Awesome Again May 3 '15 at 14:24
  • 2
    You can vote how you want. It needs 5 close voters or 3 leave open voters. Where did you find that a users get banned for not following the majority vote? I have never been banned and I must have some close votes on SO where I didn't follow the pack... – rene May 3 '15 at 14:29
  • @rene If you fail many review audits (i.e. vote not like others), you'll be banned from reviews. Disagreeing on disputable posts is fine, sometimes disagreeing on less disputable posts is fine, but constantly disagreeing with others will get you banned from reviews. – Athari - Make SE Awesome Again May 3 '15 at 14:32
  • 4
    Voting in review isn't free. You need to understand the policies and enforce them – that's the point of review. If somebody is reviewing poorly, that's exactly what a review ban is for. – bjb568 May 3 '15 at 14:33
  • 1
    @bjb568 There is a difference between failing a deliberately devised audit-test (which is not a real review at all but a trick to catch robo-reviewers) and simply voting differently than others have voted. I read the OP as complaining not of the former but of the latter. Failing audit reviews is not merely voting differently. – tchrist May 3 '15 at 14:55
  • 4
    @tchrist I'm talking about a user who votes against closing obvious offtopic questions — exactly the questions which the system uses for review audits. He would have likely be review-banned as a "rovo-reviewer" already, if review audits were enabled. The question I've provided as an example is quoted in full, it's not just a title. It's already deleted, by the way. – Athari - Make SE Awesome Again May 3 '15 at 15:26
  • @Athari Thank you for clarifying, because I didn’t quite understand you entirely the first time around. I hope my revision better addresses your issue. – tchrist May 3 '15 at 16:38

You must log in to answer this question.

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