User talk:Dank/RFA/Archive

Endorsements

Is the goal of the endorsements phase of an RFA (see Wikipedia talk:Requests for adminship#Arbitrary break) to create some solid nominating statements and a list of solid questions and answers that would hopefully answer the criticisms that are likely to arise in the second phase? - Dank (push to talk) 00:05, 26 March 2011 (UTC)

Yeah, in a nutshell. I envisioned it as a sort of place for editors to get their bearings before being exposed to the mass community.

• A candidate has certain criteria to meet before being able to post on the page.

  People argue against having certain criteria because "adminship is not a big deal", but there are community enforced criteria. It would be better for a candidate to see that they don't meet community standards rather than see a wall of opposes shouting NOTNOW.

• A request for endorsement stays up for a maximum of 72 hours.
• A group of say, 20 interested admins are the people permitted to endorse on "WP:Adminship endorsements" or whatever we choose to call it. They can pose questions in addition to the usual three, perhaps anticipating an opposers question and answering one of their own prior to endorsing.
• The candidate presents a nomination statement similar to a self nom.
• The endorsers are required to provide a short statement (a mini-nom) to complement the candidate's statement.
• After (7?) endorsements, the clock freezes. The candidate expresses the desire to move on (so they aren't caught at work when the request is transcluded), the request is transferred to the main RFA page and the countdown is resumed at whatever point the 7th endorsement stops it at.
• RFA continues as normal from here.

Regards, MacMed^talk_stalk 02:36, 26 March 2011 (UTC)

Okay, Balloonman said: "If this idea were to pass, there would literally be 40-50 supporters on just about every RfA before anybody had the "right" to ask questions or oppose." I agree with him, I think a significant number of opposers are going to think of this as a "giveaway"; we have to find some role for opposers during the first phase. WSC said, "The difficulty with having up to three days for seven endorsements is that a proportion of RFAs fold fairly quickly after an oppose or question that brings up a serious issue. Waiting for three days before that sort of thing can be discussed strikes me as inefficient use of everyone's time." I think that argument will carry weight too, so the first phase has to have something for everyone ... it would be a chance for the candidate to be dealing mostly with the supporters (important because of the "driver's license" issue at WT:RFA), and a chance for the opposers to question what the supporters are saying. You know, this sounds like a legal proceeding to me, in the sense that one side gets to put on evidence and the other side can question and rebut but can't go off on an unrelated tangent, then the other side gets their turn. Maybe 3.5 days for each? You know, this would solve another recurring problem at RFA, the problem that both sides often walk away mad because they thought the other side wasn't taking their concerns seriously ... if you had 3.5 days where you had no choice but to focus on, and support or rebut, the other side's arguments, voters might be happier at the end of the RFA. - Dank (push to talk) 03:05, 26 March 2011 (UTC)

Maybe something like this:

• The group of (let's expand to 30) admins are split into group 1 and group 2
• Each RFE (request for endorsement), a group is assigned the "endorsement" role and the "devil's advocate" role
• For each endorsement made, a "devil's advocate" can post an "anti-endorsement", basically a statement of some sort of problem with the applicant

  The talk page is open for other editors to bring things to the admin's attention that they feel is a problem, but the admin judges if they feel the concern is substantial, otherwise the editor is told they can bring it up in open RFA

  The concern must be legitimate, no "not enough edits in a month". Something the editor can actually work on.

• 7 endorsements must still be gained to move on, but now the editor can decide to withdraw if the opposing statements give him/her problems that they want to work on before entering full RFA

Regards, MacMed^talk_stalk 03:15, 26 March 2011 (UTC)

Okay, suppose a voter is in group 1, but after a bad answer or an embarrassing diff, they won't be able to take that side in good faith any more, and they don't want to act like a lawyer and defend someone they don't believe in. Is there a way to give them flexibility? And ... if we've divided the week in half, do we really need to count to 7? No matter how many people are supporting, the crats will probably be able to tell after 3.5 days if the RFA isn't going to make it, and they can close it. - Dank (push to talk) 03:31, 26 March 2011 (UTC)

Just allow them to strike their endorsement and give a reason for it (maybe "supported by diffs" being a rule?). And I think the 3.5 days would only stick if we were very, very encouraging of outside opinions on the talk, and admins responded with reasoning for exclusion (if the rationale were to be excluded). Otherwise a lot of people might feel like they are losing some of their voice by not happening to check RFA within the three days that they could !vote on the nomination. Regards, MacMed^talk_stalk 03:55, 26 March 2011 (UTC)

Yes, we'd need the community to rule (and if there's a disagreement, let a crat rule) on whether a question was relevant to something the other side said or not; I don't think it will be that hard to figure out. We could fix the problem of short notice by letting candidates announce that they'll be running up to 3.5 days ahead of time. - Dank (push to talk) 04:18, 26 March 2011 (UTC)

I would suggest something not too different; here is what I had in mind. The problems as I see it are that;

1. RFA is too easy to start. Much of the drama is due to candidates who fall under wp:snow and should not have stood in the first place. Thus, making it more difficult to start an RFA is part of the answer.

   I believe this problem is already solved by the current proposal to start off with a period where the candidate is interacting mainly with supporters; that gets rid of the drama and the harm done, and the appearance of harm. - Dank (push to talk)

2. RFA is too difficult to complete successfully. Some good candidates are put off because of the very high standards demanded by the community. Thus, setting out a set of criteria, somewhat less demanding than the existing unwritten criteria, would be helpful.

   It depends on what you mean ... if you're talking about doing our best to tally and summarize RFA results so that people have a better idea what to expect, I totally agree. Most recently, Kudpung has done some of that. But if we tell some opposer "You can't vote no, that's not on the list", they're going to laugh at us, and the crats have always been willing to count all but the wackiest oppose rationales. I think that's just how it is. - Dank (push to talk)

3. RFA is too pressured, too confrontational, and too noisy. Thus, altering the mechanism so that more discussion is encouraged and both candidates and other contributors feel less pressured is desirable.

   Yes, the current proposal seems to deal with that too. - Dank (push to talk) 12:42, 26 March 2011 (UTC)

How would I suggest dealing with that - something inspired by the same ideas as McMed's suggestion, but also dealing with Balloonman's concern;

1. A set of criteria are published (I would suggest something rather less stringent than the current effective, unwritten criteria).
2. When an RFA is requested it enters a 3-day "discussion phase". During this time two things happen;
   1. Editors-in-good-standing can write "endorsements" or "co-nominations" (perhaps the latter as it is a stronger term). These should be longer than one sentence. Sensible candidates will ask potential co-nominators in advance; as a result, poor candidates will be encouraged to wait longer, but this feedback will be given in a less confrontational forum than RFA, and from people who they know and respect. Strong candidates will gain reassurance from the fact that people they know and respect are prepared to make more effort to support them. I would set the number of endorsements at a minimum of 5 and a maximum of 10.
   2. Editors who are not endorsing/con-nominating can review the candidate's contributions and ask questions. Why do this early? Questions are actually a very valuable activity on RFA. We should encourage them to be asked early so !voters have the full information. By taking the questions out of the immediate context of support/oppose !voting we would also encourage a more genuine dialogue over the content of the questions. Further, this would also give candidates more time to answer questions as they would not fear people saying "Oppose, my question has been left unanswered for a whole six hours!" reducing candidate stress levels. And finally, if an inquisitive question produces something which leaves the candidate / their co-nominators in doubt as to whether it is worth proceeding, the candidate can decide to withdraw with more dignity than if their RFA was shot down in flames by oppose !votes.
   3. As a further aside, it is very difficult to write a pile-on endorsement, or a pile-on question. Currently there are circumstances where RFAs attract many pile-on !votes of one kind or another early on, and this is harmful to the process.
3. If after 3 days' discussion, the candidate has received 5 or more endorsements, and wishes to continue, the RFA moves from "discussion phase" to "consensus phase" where the community can !vote as it currently does, just hopefully with more useful information to go on. Questions can continue to be asked and answered at this stage.

What do people think....? — Preceding unsigned comment added by The Land (talk • contribs)

I gave short answers above ... just my two cents, but I think the first step is to get some quick, minimal but decisive changes so that everyone knows that we're taking their concerns seriously (so that we can survive an RFC), and to prove that we're capable of making changes to RFA that seem to the wider community to make a real difference, something that we've never been able to do before. When we can get that going, that's when we throw a bunch of ideas on the table and try to optimize the results. - Dank (push to talk) 12:42, 26 March 2011 (UTC)

That's precisely what I've been saying, Dank - what we need are the ideas and suggestions. The work group can then evaluate them all and come to a consensus among themselves. FWIW the above nomination suggestion is a good idea, but in practice it would probably be too complex, add too much bureauocracy, and would entrain instruction creep. I believe keeping it as simple as possible so that our younger editors, who make up a majority of SNOW/NOTNOW, have no difficulty understanding the requirements. Kudpung (talk) 13:00, 26 March 2011 (UTC)

I agree with all of that except just one part. I really think that some version of what we've got on WT:RFA would survive an RFC, unlike past efforts, because we're finally taking the wax out of our ears and listening to what people have been telling us for a long time. Personally, I think it's just this simple: figure out what the worst kinds of "admin abuse" have been (and I try to carefully make the argument that it doesn't feel like "abuse" to the admin doing it ... I'm guilty of it a bit I think, and didn't realize it, it's very common ... I hope that will help to get everyone on board). Say that people promoted after the cutoff date can't do that anymore. And have the first part of the RFA be the "supporters' turn" in some sense, so that we don't get in the face of the candidate when they're trying to take the test and show us what they've done. I firmly believe that those two things, or something with the same effect, are the absolute minimum needed to deal with the objections of the wider community. But they're also the maximum that we need to do ... it's not necessary to fix every problem at RFA, it's only necessary to show that we're capable of listening to the wider community and making a significant change. Once we do that, instead of telling everyone how it's supposed to work, we should just let it evolve at RFA for a little while. People don't like imposed solutions. After we get some experience with how it works, that's when we do the task force IMO to try to maximize the benefit. - Dank (push to talk) 13:11, 26 March 2011 (UTC)

Back to original question

Since I just glanced at my talk page, I will answer the original question. My point in thinking up the "endorsement" idea was not an attempt to get RFA to pass more users—it wasn't even an attempt to get more nomination statements. Rather, it was an attempt to reduce the pile-on opposition of new users and users like Dusti. When the users would realize that at most a couple people thought that they would make a good admin, they could withdraw without being pile-on opposed and made to feel like crap. Additionally, I strongly agree with Balloonman that the endorsements should not be a chance for supporters to "jump the gun" on the opposition. Reaper Eternal (talk) 17:34, 26 March 2011 (UTC)

Agreed. My proposal doesn't let anyone jump the gun. - Dank (push to talk) 18:37, 26 March 2011 (UTC)

What about RfC style

One of the issues is that the opposers don't get a say or the supporters get a head start, if we're considering splitting the week in two, could the first half be set out in an RfC/U style, where editors could state whether they've had interactions with the candidate - how they've been, what they've seen. Other editors could endorse these summaries. After 3 days, it should be pretty clear if an RfA would SNOW, the candidate should get decent feedback without anyone opposing anything. I'm just not sure if that solves the unpleasantness. Worm^TT · (talk) 13:17, 28 March 2011 (UTC)

I have no objection and it's worth considering. At the moment, my big concern is that we're going to treat the very next candidate at RFA in the same way as always ... after we seem to have a small consensus that that's not acceptable. My preference would be to get a quick consensus, even if it's just within Kudpung's task force, that it's okay to untransclude RFAs and help them build a case before re-transcluding. This is something we don't need an RFC for, although I'd be happy to have one. I'll copy my comment over there. - Dank (push to talk) 13:25, 28 March 2011 (UTC)

Oh, I wasn't suggesting we had an RfC on this, just that the "endorsements" phase could be set out like an RfC - so that it didn't stop opposers from having a comment. Worm^TT · (talk) 13:30, 28 March 2011 (UTC)

Ah, gotcha. Before I copy my comments, I want to ask around ... do you agree that it would be a good idea to do something soon-ish (that doesn't need an RFC) that will make it easier for candidates to get started in an RFA? - Dank (push to talk) 13:56, 28 March 2011 (UTC)

Definitely. Tacking something on to the beginning and end of RfA would be a good idea - Something at the beginning to help them get started, and something at the end (assuming they are unsuccessful) to remind them that they are a worthwhile member of the community. We've lost 4 good editors already this year, and I'd rather this stopped happening. Worm^TT · (talk) 14:02, 28 March 2011 (UTC)

Thanks for the research, and I totally agree. - Dank (push to talk) 14:18, 28 March 2011 (UTC)