Something that has always bothered us about the new user experience was the amount of guidance we've got out there telling folks to flag rude or abusive comments, but not allowing new users to do it, because .. it's complicated.
Let's un-complicate it and let new users flag any comment on any post that they own. Or, essentially, if you're allowed to comment on a post, you're allowed to flag comments on that post, too.
There are some caveats to get around because:
- We gate the feature for a reason, and that reason is it takes a little experience to know how flagging works. We're different, it's just a fact of life.
- Duplicate close votes leave comments that shouldn't be flagged, which new users will probably want to dispute, and a 'flag' dialog seems like a reasonable place to start to most
- We're not sure how big of a tide the current restrictions are holding back.
So, here's what we propose:
New users see a modified flag UI on comments they can flag:
- Rude or abusive
- This is not a duplicate (as-appropriate) (probably churns back through the close vote queue)
But here's the thing, we'd let moderators actually see who flagged in these instances, so moderators could gain context and know what else to look for in the situation.
Theorized outcome (isn't theory great? everything works there!)
Any of the following would indicate success:
- More rude comments surface faster
- Engagement from new users (going past 1 - 2 questions) goes up
- Frequency of people contacting us to report snark they could have easily flagged if they had a bit of rep goes down
Any of the following would indicate more work is needed, or possibly a general failing of the idea:
- Moderators note an increase in noise they can't really act on from 'other' flags
- Engagement goes down
- Nothing really happens, at all
The third item in the second list is actually a distinct possibility, because the flag feature is (as it stands) a little difficult to find. We'll be making it more obvious in the future, as well as mentioning it when we launch our revised code of conduct for new users to accept, but there's a strong chance new users just won't notice the feature.
We're thinking about putting this out for testing soon and wanted to get any feedback folks might have. This isn't a permanent implementation (yet), but we really do need to let new users raise flags if they encounter snark / condescension / overt rudeness - we've found too many instances of it that just evaded other folks with privileges seeing and flagging it.
Thoughts? What we hadn't thought of yet are rate limits, and probably a few other things. Shog is working on some research and I suspect he'll post a pretty comprehensive reply, but we want your take early so we know much sooner than later if we're not seeing obvious problems.