As a moderator, should I re-flag spam that's already flagged, or just delete it?

Question

I've been an elected moderator for about a month or so, and I'm still a bit confused about the correct way to handle spam that others have already flagged as spam, but that hasn't got to six flags yet.

Note that I'm talking about cases where I've decided it definitely is spam. If I don't think it is spam then I would explicitly clear the spam flags to make sure the user doesn't get hit by the relevant penalties.

I'm aware that if I come across spam that isn't flagged as spam at all, then I should flag it to make sure that the system does treat it as spam rather than something deleted for another reason.

But if it has already been flagged as spam, does it make any practical difference if I just delete it rather than adding a (binding) spam flag myself? I've heard other moderators say that the right thing to do is to add the spam flag, but I haven't heard a definitive explanation as to why.

I've noticed that when I delete then I get marked in the mod dashboard as having handled the existing flags, but when I spam flag I don't. In both cases the previous spam flags do get marked as helpful and are dismissed.

I also know that I can destroy a user if the only thing they've done is post spam, which is the common case. But I don't always want to do that immediately because I may want to check the user out a bit more carefully first but in the meantime I want to get the spam removed.

So, is one approach preferred, and if so why?

I know there are at least a couple of concrete consequences from a post being treated as spam:

• The IP/user are treated as spammers for the purposes of blocking future posts

• The post itself is "scrubbed" so that 10K+ users who can view deleted posts don't see the text by default, but instead see:

This answer was marked as spam or offensive and is therefore not shown - you can see the revision history for details.

I want to know if the route I choose makes any difference to those or any other consequences.

Answer 1

If this is a clear spam post from someone who only ever posted spam (so, something that has only been posted in order to sell a product or try to use Stack Exchange as a link farm), the correct thing is to destroy the user. This is under the mod menu of the user page - use the first destroy reason (user only used for spam).

By all means, also flag the post as spam, as this would ensure +10k users will not be exposed to the spam on the deleted post (it gets replaced with "This question was marked as spam or offensive and is therefore not shown - you can see the revision history for details."). Destroying the user before the post has been flag-deleted as spam means the original spam will be visible on the page to these users (not usually a big problem, given one needs the URL to access the page, which isn't linked anywhere once deleted).

The reason for that is that this flag reason feeds our anti-spam measures and is instrumental in ensuring we get less of the same across the network. Destroying a user in this manner will also delete their posts - this is the fastest and most effective way to minimize damage from spammers.

---

In less clear cut cases, you need to use your judgment. This is why you were elected moderator.

If this is a user who simply overly self promotes, you can mod message them about their behavior and escalate to suspension as needed.

I suggest you read through the different moderator only help center articles - in this case, the spam moderator abilities article (moderator only link - you need to be a moderator on the site to see it) seems most relevant.

Answer 2

When we rolled out the spam system, we were putting a huge amount of faith in folks to flag things appropriately. Spam flags are often misused to say "I don't like this for some reason that wasn't presented in the list", so the system was built around a graduated set of weights.

Something getting spam-flag deleted increased a weight, a user being destroyed for spamming increased a weight, someone hitting a honeypot increased a weight - it was built so that no single action would (in theory) trigger a block.

Thus, when a mod saw a blatant spammer, the spam-flag-then-destroy combo assured a blocking weight.

Turns out, people really got this right, and continue to get it right. Now that I have a whole lot of data saying that it's safe to train the system faster if:

• A post is spam-flag deleted
• A mod destroys a user for spamming

... moderators now don't need to spam-flag something to add extra weight, destroying the account alone will deal a blocking weight in the system. Likewise, if the community spam-flag deletes something, it's an instant blocking weight.

Note, this doesn't prevent anyone from reading the site, it just prevents 'new-ish' accounts with not much history from posting until the block expires (which happens not long after we stop seeing abuse from their location).

I need to get some other knobs turned and tuned for that system to make it easier to manage and 'ramp up' when we see globlal activity, so this change will go out tomorrow.