Constantly voting not like others in review queues

Question

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.

Answer 1

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.