16

Are there rules or guidance for moderators to help them determine minimal amount of suspension time depending on the offense? Or a rule requiring transparency if the community considers the sentencing too lenient?

Sometimes there doesn't seem to be consistency between the infractions users commit and the suspensions they receive - on some sites it seems that users get short suspensions for very problematic behaviors while on other sites they get long suspensions for actions that seem minor.

If a suspension seems too short for the behavior that led to it, do users have the right to demand an explanation in such a case? Or can community Staff impose a longer suspension if they deem it to be appropriate?

4
  • We'd generally need staff for a voting reversal - and it seems giving a short suspension pending CM investigation is a prudent thing to do. I don't know if that's the case here, but that's what I would do if there were a lot of votes, and/or any doubts. Simply, we don't have all the information to tell what's happening, nor should we. A suspension's between the moderation team and suspension, and the policy for years has been that we don't discuss suspensions - so in general unless the suspended user requests it, we don't talk about it . Commented Aug 15, 2023 at 14:37
  • @JourneymanGeek Generally or always? Are the CMs necessarily involved in reviewing the sentence - i.e. their say required on the matter? The network's moderators are generally extremely lenient, often excessively, and I don't trust their judgment. Commented Aug 15, 2023 at 15:35
  • 5
    I'd just like to point out that asking questions to have another account answer is not itself a violation of network rules. Accepting or upvoting those answers is, however, definitely against the rules. I would recommend limiting your discussion to those acts that are clearly unacceptable. Commented Aug 15, 2023 at 16:20
  • 2
    @IanCampbell Misguided non-violation. Self-Q&A's are tremendously penalized on DSP.SE - speaking as author of 20+ high effort & quality self-Q&A's, many of which received zero upvotes (example). I suspect it's the same on other small networks. It's good you brought this up, as it adds to the list of "not undone by rep reversal". I might open an MSE on this topic. Commented Aug 15, 2023 at 18:26

3 Answers 3

14

Are there rules on minimal amount of suspension time depending on the offense?

I've been moderating for a while now, and I've never been presented with a list of offenses for which the general guideline (warning, 7-30-365 days of suspension) must be considered 'not applicable'. So I'm going to say there is no such list of rules.

Or a rule requiring transparency if the community considers the sentencing too lenient?

In general, we don't discuss suspensions, see the section on "what can moderators share publicly?". So no, there is no such rule either, and in general the guidance there is now would work against having such a rule anyways: You can't have that guidance but at the same time have groups of users talking about another users suspension and what they think about the duration of it.

But, the Code of Conduct page says the following, which seems applicable to your situation of being unsatisfied with a moderator decision:

All actions will be taken on a case-by-case basis at the discretion of our moderators. If you’re not satisfied with a moderator decision, you are welcome to appeal it by contacting us directly. We review each report individually and will respond as quickly as we can.

2
  • Thanks for that last URL. Re: no transparency rule - I trust it's a valid answer, but not a satisfactory one. What if Jon Skeet's true rep was 10k all along? There simply must be a line we can draw, else what message does it send to future offenders? You've answered my question as-asked, and I'll accept it eventually, also giving others a chance. Commented Aug 16, 2023 at 12:30
  • 4
    "else what message does it send to future offenders?" > Those are probably very, very unlikely to first look and see what others got away with. They just start doing this. A suspension is meant to stop the behavior from a single user, not as some kind of deterrant 'look what happens when you do this!'. In order to use it like the latter suspensions are going to need a lot more visibility. Something like a page called /gallows-field and making that the landing page for every new user....
    – Tinkeringbell Mod
    Commented Aug 16, 2023 at 12:49
8

Purpose of Sanction

For reference, the next several paragraphs are based on a short expository paper by Jennifer Marson.

Suspending a user is a kind of sanction, akin to a speeding ticket or a jail sentence. In understanding why suspensions might be different from one site to another, it helps to understand what the purpose of such sanctions are. Broadly speaking, sanctions are usually justified for reasons which broadly fall into the following categories:

  • Retribution: In a retributive model, sanctions are meant to punish a bad actor—this is, essentially, the "eye for an eye" model.

  • Incapacitation / Segregation: In this model, the goal of sanctions are to prevent a bad actor from doing more harm, either by removing them from the group, or by restraining their ability to act.

  • Deterrence: Here, the goal is to make punishments so bad that bad actors will choose not to engage in harmful acts, because the threat of punishment is so severe.

  • Restitution / Restorative Justice: The idea of a restorative model is that sanctions should serve to make the victims or a community "whole".

  • Rehabilitation: In this model, sanctions are imposed in order to retrain a bad actor—the goal is to teach the bad actor different behavioural patterns.

The SE Model

The general philosophy of moderation on the SE network focuses primarily on using sanctions to incapacitate bad actors: suspensions are used to prevent a user from causing additional damage (e.g. when a user is suspended for "abusive behaviour", the banner on their user profile says something to the effect of "this user has been suspended for [x] days in order to cool down").[1]

Suspensions on the network also have a deterrence effect (users likely don't want to be suspended), and there are a couple of other moderator tools which emphasize rehabilitation (e.g. it is not uncommon for us to send a moderator message to a user without suspending that user—the goal of such a message is generally to explain what the user has done that is harmful to the community, and ask them not to engage in that behaviour in the future).

What this means is that sanctions are, generally speaking, progressive. In most cases, the first time that the moderation team sees a behaviour, the consequence will either be a moderator message, or a brief suspension. The goal is to get the behaviour to stop. What we hope for is rehabilitation, i.e. that the user will respond positively to a moderator message and/or short suspension by reforming their behaviour.

If, after this first contact, the user continues to act in a way that is detrimental to the community, the purpose of sanction shifts from rehabilitation and incapacitation, to deterrence and incapacitation: longer suspensions are meant to discourage a user from behaving badly when they return, and, failing that, to prevent them from behaving at all.

So What?

In this case, it looks like there is one user who, yes, was massively exploiting the system, in violation of the rules. However, no one caught their misbehaviour until new tools were introduced to the moderators. The improper votes were reversed (a form of restorative justice), and the user was suspended for a week (I suspect that the intent was some combination of rehabilitation and deterrence).

A longer suspension starts to look like retribution, which is not in keeping with the general SE model.

In fact, I would not have been opposed to a suspension of zero days—if the goal is to stop the behaviour, a moderator message putting the user on notice may very well be sufficient to stop the behaviour. If they stop causing harm, case closed.


[1] As pointed out by Wicket in the comments, this idea of incapacitation was outlined in an SO blog post titled A Day in the Penalty Box. The original SO model did not have timed suspensions, so the approach was

If we think you are reachable, and the behavior is one that we feel can change, we will try to warn you via email first when there are behavior problems — so that we can address them before they become deeper problems.

Moderation tools focused on cleaning up content, and communicating with users to rehabilitate them. This proved to be insufficient for managing all behaviour, so suspensions were introduced:

Depending on the severity of the problem behavior...your account will be placed in timed suspension for anywhere from 1 to 365 days... At the end of this timed suspension period, your reputation will be recalculated, and your account will resume as normal. We don’t hold grudges. The point of all this is to address the behavior. If the behavior improves, you are welcome back.

This is, explicitly, a model of incapacitation.

3
  • 1
    I like the old blog post explaining suspensions: A day on the penalty box. It includes a link to the Wikipedia article for "penalty box": "...sits to serve the time of a given penalty, for an offense not severe enough to merit outright expulsion from the contest". Suspensions are for not severe offenses, so it's more like retiring the driving license temporaly than a jail sentence.
    – user1359324
    Commented Aug 16, 2023 at 19:38
  • 4
    @Wicket Note that at the time the blog post was published, it was common for users to have their account permanently deleted if they did really bad things - see the final paragraph of the blog post. The common pattern was 7 days, then 30 days, then account removal. Since then, SE stopped deleting accounts as behavioral sanctions, primarily because they started anonymizing names of deleted users and deleting accounts without the user requesting it would violate the CC license. Commented Aug 16, 2023 at 19:49
  • 7
    This is a good answer: it captures the primary goal of the system and contextualizes it with other possibilities, allowing readers to both see how it might be misused if misunderstood and potentially avoid such misunderstanding.
    – Shog9
    Commented Aug 16, 2023 at 21:01
8

Most of what I'm saying is already covered well in the existing answers but I'm going to add my own spin on it.

Guidelines for suspension duration

Are there rules on minimal amount of suspension time depending on the offense?

There are not. The guidance for moderators is to start with warnings (mod message only, no suspension) in many cases but that the general recommendation when it comes to suspensions is that mods follow the 7-30-365 day progression for suspension duration. The suspension UI encourages this by having these three options as defaults while also allowing mods to choose a duration between 1-365 days. The system will also automatically choose the next of these three durations based on the last suspension duration the user received. So if the last time they were suspended it was for 7 days, the system will select 30 days if the mod indicates they want to suspend the user again.

Single site suspensions can not be given for durations longer than 365 days and only staff - generally CMs - can give network-wide suspensions (of any duration), which are usually only used in cases where a user has displayed that they are misbehaving on multiple sites.

Within this framework, mods have a significant amount of flexibility in how they opt to correct the behavior they see on the site. We do not force suspension durations in the UI, though some mod message reasons may come with a default suspension, generally only in cases where the user is currently causing an impact to the site, such as the deletion of useful content, which comes with a default 1-day suspension, which is intended to halt deletions until the user responds and gives moderators time to reverse any damage done. That said, even these defaults can be overridden.

Theory of moderation

I wonder if your question stems in part from a misunderstanding of why suspensions exist on this site. Our general moderation theory centers on the concept that suspension is used to give the user time to consider the guidance given and change their behavior, rather than being used as a punitive tool where the specific offence would require different durations as in many legal systems. This is why we recommend starting with warnings and gradually increasing suspension duration for most cases.

While many years have passed, this is pretty much in line with the original blog post from 2009 announcing the suspension tooling being added to the site. While, yes, the post does say that duration can be based on the severity of the behavior, it also says that it's at moderator discretion.

If someone doesn't know that their behavior is wrong - or if that behavior hasn't been corrected yet - reaching out and letting them know should be the first step. If they cease that behavior and begin participating in alignment with the guidance, there's no reason to force them off the site for long periods of time unless they show that they will not change. If they don't change, they'll quickly get a month-long suspension, followed by a year.

Yes, this requires a bit more time investment from the moderators but we believe a user is much more likely to change their behavior if they get a warning or a week away than if they get a year - and it can be worth it to give them the chance to show they will change.

Transparency in moderation actions

Or a rule requiring transparency if the community considers the sentencing too lenient?

As stated above, moderators are given control of the duration of suspensions and whether to suspend or not. We encourage moderators to make these decisions in consultation with their fellow site mods, the other members of the moderation community, and with any needed support from the Community Management team.

While moderators can act individually, it's worth keeping in mind that they often do not, particularly in very complicated cases. There may have been much discussion in private moderator spaces that led to the final decision of how to respond to a specific situation. I've actually seen moderators collaborate over the text of the mod messages they've sent before reaching out to users. Because we know how much thought goes into these decisions, we generally trust the mods act in accordance to the theory explained above.

Part of the theory, however is that moderation of specific users is private between that user and the moderator unless the user chooses to share it publicly or gives the moderators permission to do so. This can lead to difficulties for moderators as the actions mods take won't always be visible to community members or, when they are, the reasons for those decisions won't always be clear.

While their actions may have harmed the community and we understand that many people want to know what happened in these cases, we ask that community members trust that moderators are handling the behaviors and avoid publicly calling out users who have misbehaved, particularly if you're doing so simply because moderators haven't responded in the way you feel would be more appropriate.

If you believe that a user is acting in a way that is negatively impacting the community or needs investigation by moderators or staff, there are several options available to you, some of which I've listed in order of escalation but any specific case may skip prior steps if they don't apply. These are also covered in the MSE FAQ - https://meta.stackexchange.com/a/28869/284336 along with some other options. The list below is intended to specifically relate them to your general question about suspension duration rather than the FAQ's focus on more broadly addressing moderator actions.

  1. Start by flagging something you think is harmful or where you think something isn't right, even if you think moderators should be aware already (e.g. if you think voting on a post is out of norms).
    • While your question assumes that moderators are aware of the entire situation, that may not be the case.
    • This ensures the mods are aware of the issue and can investigate it to see if there's anything actionable. If the mods can't investigate it themselves, they can escalate it to the CM team for review.
    • Moderators can privately respond to flags and may use those responses to give the flagger some info about how the flag is being handled but you should expect these responses to be general rather than giving you details into moderation choices You can see these responses on your flag history page.
  2. If you don't understand the outcome of your flags or if you feel there's an issue that goes beyond a single user's behavior, you can ask a question on the child meta.
    • Such questions should focus on how the flag itself was handled or the broader issue.
    • We strongly discourage using such posts to name and shame users or accuse moderators of ineptitude or inaction and posts that do so may be edited to remove those references or deleted outright.
    • While mods may respond to these questions, they should avoid going into too much specific detail or discussing the moderation of specific users in response to the question.
  3. If a meta post isn't a good fit or if the outcome of a meta post doesn't address your concerns or if you want to have a specific action reviewed by staff, you may contact us using the link in the footer.
    • Please note, since we consider moderation decisions regarding specific users private, the response you get to this will likely be a brief statement saying we have assigned the request to a CM to investigate but you will not get any additional details or specific outcomes of any investigation.
    • Staff do not investigate moderator actions based on a meta post, so if your interest is in getting attention from staff, go straight to the contact form.

What we specifically want to avoid is finger pointing and accusations. While single cases of this may lead to post deletion or removing the username of the person, repeated attempts to pillory someone in public or take moderation into your own hands can be considered abusive behavior on its own.

You must log in to answer this question.

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