This is just an FYI: I've just published the Q4 2021 Community team roadmap on the blog (and it's early, even!). I'm taking questions or comments here, or over there. :-)
asked Sep 20, 2021 at 14:07
-
1I think this should be tagged product-discovery, but I’m not completely sure.– Ekadh SinghCommented Sep 20, 2021 at 16:32
-
12Not images of text please.– philipxyCommented Sep 20, 2021 at 17:11
-
FYI the post does not appear to have the correct tags to show up across the network. It is not in the yellow box on the right hand side of this post, for instance.– LShaverCommented Sep 20, 2021 at 19:56
-
1@LShaver for this to show up in the yellow box on the right hand side it would need to be tagged [featured], and this cannot be tagged featured because there can only be two posts on meta featured at a time.– Ekadh SinghCommented Sep 20, 2021 at 20:47
-
2@EkadhSingh-ReinstateMonica I meant the blog post itself. Compare: i.sstatic.net/nzRym.png– LShaverCommented Sep 20, 2021 at 21:18
-
2@EkadhSingh-ReinstateMonica - not sure what's going on there, but I'll find out.– Philippe StaffModCommented Sep 21, 2021 at 6:41
-
@LShaver isn't MSE excluded from that?– LuuklagCommented Sep 23, 2021 at 12:50
-
@Luuklag per this answer, if the blog post had the "announcements" tag it would appear on the sidebar on all sites.– LShaverCommented Sep 23, 2021 at 14:47
-
collectives aren't even on the chart, not a community initiative?– Kevin BCommented Sep 23, 2021 at 19:07
-
3Collectives are not driven by the community team, though we are deeply involved in them. That's driven by another team within the org.– Philippe StaffModCommented Sep 23, 2021 at 19:10
-
In the blog, you showed the current status of all of the projects you mentioned. Is there any way to view what the status on those projects is right now?– Ekadh SinghCommented Oct 25, 2021 at 15:55
-
1@EkadhSingh-ReinstateMonica - I should really find a way to make that dashboard public. But I'll tell you that as of today, all of the projects that are listed for Q4 are in progress, except for the final one (evaluating chat for dependencies), which kicks off this week, I believe.– Philippe StaffModCommented Oct 25, 2021 at 16:08
Add a comment
|
7 Answers
First off, great to see this posted before the actual quarter started. So kudos on that!
I understand why you included an image of the roadmap in the blogpost. It is a great way to condense all that information into one place. But could we please keep in mind those who are less equipped visibility wise, and at least include a decent alt-text.
I must admit there is improvement when comparing it with the Q3 roadmap, there was literally no alt-text on the graphs provided there.
-
4Luuklag, what would you consider a sufficient alt-text there to be? I can certainly go back and edit it :) Commented Sep 21, 2021 at 6:46
-
8@Philippe generally an alt-text should convey exactly what is in the image, unless it is elsewhere in the text and the image is used as a visual summary of the text. In that case it is less needed. See for good guidelines this W3C article: w3.org/WAI/tutorials/images/decision-tree– LuuklagCommented Sep 21, 2021 at 7:04
This could perhaps be a new meta question by itself, but what's the meaning of Trust and Safety?
Research: Dependencies on Chat: I’ve been amazed at the amount of things that are integrating into our chat systems. So we’re going to make a comprehensive list of systems that integrate to chat. This is everything from my team’s own escalation and work handling bots up to the Charcoal spam-fighting network’s systems. This project will be run by Slate, who is temporarily seconded to the Trust and Safety team for this project (which ordinarily would be run by Trust and Safety, but we haven’t quite completed the hiring and onboarding process for their new team members yet).
I'm confused why this project would be the responsibility of Trust and Safety. According to The Community Teams @ Stack Exchange and how we work together, the Trust and Safety team "is responsible for handling user safety on the platform, including preventing harassment, PII concerns, and other abusive behaviors on the network". This task seems to be about gathering information about the workings of part of the SE system, and while it's possible that PII-related security leaks may be exposed, surely that's not the main purpose of the project?
Have I misunderstood the point of the Trust and Safety team, or misunderstood the nature of this project?
answered Sep 20, 2021 at 17:58
-
8It's primarily because I had to make a decision about which team would own it, and a huge number of what I would consider trust and safety related systems (like charcoal, for instance) seem to be based on the chat system. It's an early decision based on early data. Might be totally wrong. :-) Commented Sep 21, 2021 at 6:44
Research: Weighted Close Votes
If this could be accompanied with a break down of the closure statistics from the excellent a year in moderation posts it would be great. In the sense of operational research the presented data should be as clear as possible for the community to be able to image how potential changes would impact ease and efficiency of closing.
-
9I'm working on the data from the 3-vote-close/reopen tests, too. I have all of the stuff I need but I haven't actually put it into a post yet. I'm very sorry. I was supposed to have it out last week but I didn't get there. Should be this week.– Catija StaffCommented Sep 20, 2021 at 14:32
-
9@Catija don't be sorry, your presence is more important than data and we gladly wait. Commented Sep 20, 2021 at 14:36
-
2A chart showing the CVers vs. having a silver tag badge or not; that could answer Tim's proposal and assist with this at the same time.– RobCommented Sep 20, 2021 at 15:00
-
@Catija is SE evaluating just straight up making the gold badge to apply to any close reason?– BraiamCommented Sep 20, 2021 at 16:43
-
4@Braiam No. For a few reasons - I don't generally think that gold badges are necessarily a good indicator of being great at determining whether content fits the site's scope plus, gold badges are nearly unattainable on many sites, so we need a better way to identify when people are good at judging on-topic-ness. :)– Catija StaffCommented Sep 20, 2021 at 17:38
-
@Catija that's one reason, what about the other 3? Needs details, needs focus or opinion based. All of those benefit from knowledgeable users giving accurate feedback.– BraiamCommented Sep 20, 2021 at 17:59
-
6@Braiam I think of it as two 1. gold badge != knowledge of scope 2. gold badge is not always available on smaller sites. I've been on record in the past disliking static privileges. I don't think having X rep or Y badge necessarily means you know how to use a specific feature correctly. I'd much rather something like this be based on metrics that indicate you know how to use it effectively and in alignment with the community's preferences... and potentially has a way to be lost if they show they're using it in a way the community feels is incorrect.– Catija StaffCommented Sep 20, 2021 at 19:18
-
@Catija There's a issue with that that the gold badge solves: it makes certain users essentially site wide moderators. The gold badge is more contained, and I presume that since smaller sites don't seem to have problems (this post argue that is drought of close voters rather than not enough votes) is not like doing this would also help them. Please, don't dismiss using gold badges off hand, since it would exacerbate a perverse incentive of people closing as duplicates rather than other reasons.– BraiamCommented Sep 20, 2021 at 19:42
-
3@Braiam "it would exacerbate a perverse incentive of people closing as duplicates rather than other reasons." I don't see how this makes that worse. If anything, I'd think it'd make it less likely, particularly on sites like SO where there's only 3 votes to close. A weighted vote for the correct reason seems more likely than a unilateral vote for the wrong one. But, again... I do not think that gold badges mean that someone understands when to close a question. Any gold badge user could attain this weighted vote by performing well, plus people without gold badges can, too. Less gatekeeping.– Catija StaffCommented Sep 20, 2021 at 19:46
-
@Catija no, that's moving in the opposite direction, we need more gatekeeping so misguided (but technically correct according to the rules) actions are less probable. There's an old post by Jon Ericson that basically tells you that. We don't want "drones that know the rules", we want sensible knowledgeable users to review critically the information presented. They can only do that if they understand the topic of the post reviewed.– BraiamCommented Sep 22, 2021 at 4:32
-
@Braiam You could also argue that the people who are better at asking questions rather than answering them know the site's scope better due to their experience, but tag badges ignore question scores and only consider answer scores. Closing as duplicates is different, as that's effectively answering the question with a link: if you have a lot of experience answering within the tag, you're also more likely to know which answers elsewhere answer the question. Commented Sep 23, 2021 at 6:35
-
@SonictheAnonymousHedgehog that would ignore that 50% of askers have asked exactly 1 question (sadly, this has survivor-ship bias due most posts being deleted, the number is bigger). Answeres on the other hand on avg post 12 answers. Also, remember that it is answerers, not askers, the ones that support the scope of the site. Without them, there would be no questions to ask. If the answerers are dissatisfied with the kind of questions being asked, they will stop answering altogether.– BraiamCommented Sep 23, 2021 at 10:17
About the Moderators Quick Start Guide:
Did you look at the Stack Exchange Moderator FAQ? It is quite popular, and seems (from what I can tell) to do its purpose well. Was it factored into the making of the Moderators Quick Start Guide?
answered Sep 20, 2021 at 17:32
-
5
-
@Tinkeringbell I’ll have to take your word for it, but I would assume that FAQ would have at least some of the important stuff :) Commented Sep 20, 2021 at 18:06
-
5Thanks for pointing this out! One of the first steps in the project will be to assess what information (that's useful to new mods) is located where. The resources linked from that post are definitely a great start!– V2Blast StaffCommented Sep 20, 2021 at 19:34
-
@Tinkeringbell Argh, don't mention the secrets! They'll need to make you an ex-parrot then :-P Commented Sep 20, 2021 at 20:11
Research: Dependencies on Chat: I’ve been amazed at the amount of things that are integrating into our chat systems. So we’re going to make a comprehensive list of systems that integrate to chat. This is everything from my team’s own escalation and work handling bots up to the Charcoal spam-fighting network’s systems. This project will be run by Slate, who is temporarily seconded to the Trust and Safety team for this project (which ordinarily would be run by Trust and Safety, but we haven’t quite completed the hiring and onboarding process for their new team members yet).
I'm kinda curious about this for a couple of reasons.
One of the 'interesting' consequences of having a very stable chat, with no proper API is - a lot of things function based on essentially screenscraping tools or glorified, and somewhat jerryrigged userscripts. Basically the entire bot ecosystem relies on chat being significantly unchanged over the past decade or so :(. ... There was talk from an ex-dev of a API for chat, which would be/have been the right way to do things to start with:)
Its also worth thinking about the fact that historically - aside from big projects like charcoal, and iirc various comment monitoring scripts - there's quite a few smaller and less known bots. SObotics is probably a useful starting point - chat and otherwise, considering many utility bots probably are underpinned on projects there.
And of course, what is the end goal here in research - so what's the expected takeaway from the research? Working out 'what's' in use or how to better facilitate those tools in future?
answered Sep 23, 2021 at 5:24
-
2Initially, it's just to create an inventory of what stuff exists. That's it. That inventory could later be useful for a number of purposes, among them the ones that you suggest. Commented Sep 23, 2021 at 18:27
Project work like this ideally takes up about 20% of the team’s time. Most of the remaining 80% is taken up in handling tickets from mods, some regular ongoing work (the mod survey, onboarding new team members, interviewing, etc.), and providing internal consulting to the rest of the organization.
Thanks for sharing this detail, this is really helpful to know. I'd be curious to hear more about what that "internal consulting" role looks like -- is this the standard part of any job where employees help each other out, or something more formal, perhaps across departments/teams?
-
2Why do you think "internal consulting" is a role and not just something pretty much every employee does in a company as part of their normal work when they know more about something than another group does?– ColleenVCommented Sep 20, 2021 at 15:39
-
7It is somewhat more formal - it's when, for instance, I assign a liaison to another team to provide feedback on ideas or deployment plans. Commented Sep 20, 2021 at 16:34
Will weighted close votes be optional for each site, or will it be applied to every site?
The reason I ask is because in my main community (Politics) IMO we have plenty of people voting to close and reopen questions, and making that easier would just lead to questions getting bounced between closed and open more often.
answered Sep 20, 2021 at 16:25
-
Note: I know that weighted close votes might be beneficial to politics, but I’m asking whether or not we will be able to choose. Commented Sep 20, 2021 at 16:26
-
10That's exactly the sort of thing that we will be investigating. We have no answers at this point. Commented Sep 20, 2021 at 16:35