Reviewing flagged posts

I never volunteered to administer this Forum, and nobody asked me if I wanted to. I believe this responsibility was thrust upon me by some Discourse algorithm.

When I reject a post, it’s typically because it’s way off-topic. In theory, I might also reject a post for being inflammatory, but I was only tempted once, and I decided not to. I sure hope it’s not my job to censor posts that resemble generative AI content!

Do we have a written policy or guidelines on which flagged posts should be approved and which should be rejected?

2 Likes

I don’t think we have a specific policy. Most of the time it’s pretty obvious. Sometimes (like the post earlier) it’s a bit grey.

I was also somehow thrust into the role of approving posts. Perhaps this thread will be good so we know how many people were similarly thrust into the role and actively do it.

I ban about one spammer a week. They are getting sneakier, though.

There have been a few where I’ve had my suspicions because of how it was posted, where, age of the thread, etc… but the post didn’t itself have any issues. So I left it open and sure enough within a day or two the post had been edited to include spam links and I got out the ban hammer.

4 Likes

if my memory does not trick me, i had aprooving ability as long as i had “admin” privileges on the forum. at the point i asked to be removed from that group due to inactivity i have never seen such a notice again.

1 Like

I was given admin rights to the VR forum once, so I also get those flags shown. It doesn’t happen that often, though. Maybe I frequent the forum too seldom? :slight_smile:

For me, it happens maybe once every 2 months? (Although lately it felt a bit more)
Usually it’s very clear though, e.g. posts by bots linking to some webshop or ads in general.

1 Like

I see that several of the users in this topic - including @sgold - are, in fact, members of the @moderators group. (They all have a little shield icon next to their username.) These are users who were actually manually granted their powers by an admin.

@pspeed, OTOH, got his powers from the Discourse automated system. He’s a Trust Level 4, AKA “Leader” (orthogonal to the JME-specific group “Engine Leader”) Basically, continual high levels of engagement that get positive feedback from other users, and a tendency to show good judgement when flagging content. (Many/most flags end up being confirmed/agreed with by other moderators.)

  • As a note: @sgold would also qualify for this criteria if he wasn’t actually in the moderators group.

The idea behind the automated system is to give some of the more active community members “Moderator Lite” privileges to triage some of the more obvious moderation tasks, such as deleting/blocking spam(mers), spreading the load out a bit.

I did notice that several used the phrase “Approving Posts”. Do these tend to be posts that were flagged by other users, filtered against by the automatic spam filter, or are they from new accounts that are still sort of “probationary”? Knowing where the work is coming from might allow some adjustments to reduce how much of it needs to be done.

1 Like

From my experience, “all of the above”… and so far it’s been 100% accurate from the perspective of “no false flagging”… so settings are probably fine.

I don’t really mind doing it… and I’m not afraid to aggressively shit-can obvious spammers. I had just always wondered who else was doing it. The forum seems pretty clean for how busy it is… and two or so ‘approvals’ a week is really not bad if I’m mostly the only one doing them.

2 Likes

Flagged posts come in bursts. I see a couple dozen a year.

The volume of flagged posts isn’t problematic, and I understand why flagging and reviewing is necessary. However, I resent people second-guessing my decisions. This seems to happen every time I decide to approve a flagged post.

Since other people don’t mind doing the work, I should simply leave all flagged posts for them to deal with.

Discourse statistics suggest my estimate of “a couple dozen a year” was about 3x too high:

Is that just flagged posts or does it include flagged users? Is this forum-wide or just your user?

…numbers seem low to me, too, but perhaps I’m also misremembering.

1 Like

I looked for documentation that would clarify, but didn’t find any. Source is available, but it’s mostly in Ruby: GitHub - discourse/discourse: A platform for community discussion. Free, open, simple.

The “Category” dropdown suggests to me this report counts posts, not users.
And if you see the same data I do, then it’s probably site-wide, not per-admin.

Here’s the report URL: https://hub.jmonkeyengine.org/admin/reports/flags?end_date=2023-11-29&mode=table&start_date=0021-11-29

Do you see the same data?

That report seems to be broken now.

It’s likely that it’s site-wide and I guess 3 in November fits my feeling of “about 1 a week” at the time. But I feel like I shit-canned some things in August and September, too… but maybe that was another forum.

Some poking around on the discourse support forum shows that this report is for actual, cast flags on specific posts. There doesn’t seem to be a built in report for the approval queue, (New users on probation) but there’s supposed to be a way to save an SQL query if the data explorer plugin is installed. (relevant query here)

The moderator activity report might have some more information.


The admin reports tab and is sub-pages are only accessible to actual staff users, not the TL3 ‘Moderator Lite’, so @pspeed won’t be able to see them.

1 Like

Thank you @sailsman63 for those explanations.