User talk:FT2/Quality user mockups

Not crazy about it. It really ignores the possibility of factions. Trust me that you can find 20 people to stamp even the worst troll. Plus you can find 20 people to filibuster the best candidate, if they have made any enemies.

I think there is room for community input. But there has to be a threshold to weed out bad applicants. I think 2 FAs is really fair, because FAs are a consensus process that requires you to understand high quality. If you understand high quality, then you understand quality in general.

I have no idea how to stop a faction from filibustering a candidate though. I guess it's the same problem as RFA, although maybe we could get away with a lower threshold here. - Randomran

The thought here is, yes, in the open community you can. But that doesn't help much, because all the open community can do is create < 50% for the user (which is quite sizeable) and provide reasons for opposes. If existing trusted content editors don't agree or thinks it's worth a try anyway, then that goes nowhere. Ditto trying to "push" a candidate, if 25% say no then it goes nowhere. Fillibustering of course is dead because it's not a "debate" so you can't "hold up" dialog.

So in terms of attacks by pro/anti user factions, the first defense is Piotrus' point, that we can hope for cooler heads and fewer cliques of this kind, in such a group, and those users will be the ones making the final decision. Why? Because they in turn got in by evidencing good conduct as a general editor as well. Sedond defense is that the communal views are visible after the community's voted but before that, so games like that will be rather obvious.

Ultimately you can't stop cliques doing stuff. But you can set up ways to make it hard for them. I think with some thought, this could be developed into a good balance between automation and quality control. FT2 ^{(Talk | email)} 05:31, 27 November 2009 (UTC)

So if I understand it, first there is an anonymous community vote for 50% approval; than there is a trusted editors only vote for 75% approval, yes? That sounds plausible and indeed should not be very easy to game, but I am confused about what votes/comments are public and which are not. And of course - how do we select the first trusted content creators? How about automatically making every editor who wrote 2+ FAs a trusted content creator (we need to populate this category to make the second round voting possible, and Randomran seems as good as any to me...)? --Piotrus 05:42, 27 November 2009 (UTC)

Yes, you're basically right. Anonymous community vote for 50% approval, then the voting and any comments are visible... and senior editors (who can now see the details and can consider any community concerns) vote -- also anonymously the same way -- for 75% approval. After voting ends their votes and comments are visible as well.

Selecting the first is always an exceptional job -- a simple answer (using the above method) is once there are 100+ senior editors, they can do the process. Until then, we need a temporary way to do the 2nd round voting... maybe use admins who also have 2 or more FAs or 4 or more GAs. We have hundreds of admins who are recognized as very good content writers. The benefit being if the FA/GA writers have also passed RFA we have a good basis to believe their general conduct, understanding of standards, etc is good, so they'll watch for those qualities in our early intake. FT2 ^{(Talk | email)} 07:39, 27 November 2009 (UTC)

Does that mean we agree that 2 FA's are a good way to test someone's understanding of quality? If so, then we really just need the second half of the skillset: someone who understands community-building (as well as quality). A vote is decent, but are there other ways to get there besides opening things up to a battleground? Would it be enough to have never been blocked? At least that would be a starting point. Then we could use a vote for error correction: to recall users who weren't blocked who shouldn't be senior editors, and to promote users who were blocked who should be. I'm brainstorming. Randomran 14:59, 27 November 2009 (UTC)

We're discussing here a slightly different question. We want to bootstrap the process, meaning, how do we run this process first few times, before we have any "senior editors" to review the nominees. This is a once-off issue because after we have enough who pass the process, they do that step themselves. We can be fussy (in two ways) -- it's a once-off matter, and also, we want the best to get it to a good start.

I don't think any number of FA's proves anything about user conduct or even editing on all articles. But we only need a temporary pool good for this temporary job of starting things going. We need users who are experienced, understand content norms enough to judge if a user is applying them to a high standard, and who have good understanding (and show it in their own conduct) of behavioral standards between users. A wide understanding of the wiki and of disruptive and problematic editing doesn't hurt either.

So I suggest we do it simply. People who are both FA/GA writers and admins, should have ample clear understanding of all the things we need to form an initial pool. In effect, we start with a subset of FA/GA writers -- not "all" FA/GA writers, but only those FA/GA writers who have also passed community peer review of their trustworthiness, approach, and understanding of wider conduct and policy norms.

This isn't difficult. RFA participants usually require prospective admins to have significant content skills (asking about these is common) so there will be many FA/GA writing admins. We don't need to look much further; this alone is probably enough to ensure we cover all bases. FT2 ^{(Talk | email)} 23:47, 27 November 2009 (UTC)

I think it makes sense to start with the pool of admins who have written two FAs. (Not GAs, because they are not reviewed by the community. They are often reviewed by someone in the same topic area, which means that they are susceptible to bias, even if they are well written and well researched.) An admin is usually a model of good behavior, and someone who creates FAs understands the many factors that contribute to quality.

To be devil's advocate, though, I see a few problems with the remaining standards.

• The first is that I think it will attract accusations that we're trying to make a cabal. Trusted users vote-in new trusted users, making it like some sort of exclusive club. I'm not sure this is a bad thing in practice. But in principle, a lot of people just hate cabals.
• Second off, I'm not sure that two votes will be enough to ensure anything but popularity. And if we add other quality requirements on top of the two votes, I think we'd be accused of creating excessive bureacracy.
• Third off, too much emphasis that first open vote will likely exclude any volunteer who has ever waded into a controversial subject area, or a controversial behavioral dispute. It's not hard for one cabal or another to stumble onto a user page, and then canvass each other off wiki or just by following each others' contributions.

Again, just being devil's advocate. I think we can address some of these challenges and make the proposal even stronger. Randomran 17:11, 29 November 2009 (UTC)