User:Seresin/RfA review

tl;dr avoidance

After finishing, it seems that the answers I have meander dangerously close to tl;dr territory. So here's a (hopefully) shorter version of the below. (If it matters :D)

Condensed

Candidate selection (inviting someone to stand as a candidate)

Nominators should nominate people they believe would make good, competant administrators.

Administrator coaching (either formally or informally)

Some feedback and possible pointers about the RfA itself is good, long-term coaching preening a candidate to pass RfA is not.

Nomination, co-nomination and self-nomination (introducing the candidate)

Nominators should be either the administrator coach, or the first person to offer a nom (or someone else if who offered a nom that the candidate prefers for whatever reason). Co-nominators should be there for a specific reason, not just another person who is impressed. Self-noms are okay. I see administrator nominators (or a non-administrator I already respect) more favorably, since their judgement is vetted.

Advertising and canvassing

Most canvassing is bad because it distorts the commenting pool one way or the other. On-wiki, public and neutral notifcations of a few highly invested editors is fine with me. I would hope most people here have the integrity to evaluate candidates honestly even if they have been canvassed.

Debate (Presenting questions to the candidate)

Debate is good — this is a discussion after all. Drama discussion is not, however, and should be quickly cut. More questions is good in my mind; it gives us more information with which we can make an informed decision. Candidates should be free to refuse to answer a question without fear of prima facie opposition. But if the questioner cannot make a decision without an answer, he should be free to oppose.

Election (including providing reasons for support/oppose)

Supports do not need reasonings if they agree with the nominator, any other reasoning should be elaborated. Opposes should have reasons to facilitate discussion, but it is not always needed. Commenters should revisit the discussion at least once again after commenting, at the end so that they may re-evaluate their comment based on further developments. Everyone should thoroughly evaluate the candidate, and, regardless of friendship, vote honestly and critically.

Withdrawal (the candidate withdrawing from the process)

Withdrawal to reduce drama or save the time of the community is good.

Declaration (the bureaucrat closing the application. Also includes WP:NOTNOW closes)

Bureaucrats should have a high percentage to pass, so that we ensure they are always inline with the community. Bureaucrats should regularly go through reconfirmation. Snow closes should happen much more quickly, since people cannot refrain from posting hurtful opposes when the RfA will clearly fail. This also reduces the time the community must spend on it.

Training (use of New Admin School, other post-election training)

The NAS is useless; the technical tools are pretty self-explanitory, and what new administrators need is more guidance from experienced and highly competent administrators.

Recall (the Administrators Open to Recall process)

I think recall is an excellent idea until there are measures other than ArbCom to revoke sysop status. It is, however, a useless gesture most of the time, as nobody is bound by it. I think voluntarily entered but unvoluntarily enforced recall is a good idea.

How do you view the role of an administrator?

They have access to technically restricted tools because they have demostrated (RfA) that they are qualified to use them. They also have ability not related to technical tools; they are judges of consensus and expected to help resolve disputes, and they (should) have the authority to enforce these resolutions.

What attributes do you feel an administrator should possess?

He should be competant, speak without mincing words, knowledgeable about policy, neutral when acting qua administrator, should research situations before getting involved, and resolve drama.

Have you ever voted in a request for Adminship? If so what was your experience?

Yes. It was fine.

Have you ever stood as a candidate under the Request for Adminship process? If so what was your experience?

Yes. It was fun.

Do you have any further thoughts or opinions on the Request for Adminship process?

It's not as broken as it is often touted to be, but it does have faults. Mostly with those who comment, not the process itself.

It should be forbidden to have yet another discussion in an RfA with Kmweber about his opposes. Srsly.

Questions

When thinking about the adminship process, what are your thoughts and opinions about the following areas:

Candidate selection (inviting someone to stand as a candidate)
If a person sees someone he believes would be capable of being a competent and effective administrator, and is willing, he should offer to nominate him. Either publicly or privately is fine, but sometimes people would rather it not be offered openly for some reason, so I think discretion is sometimes best. My thoughts on the nominator will be two below.
Administrator coaching (either formally or informally)
I think "coaching" is a bad word. It creates a picture in my mind of merely grooming a candidate for passing an RfA, with purely a checklist of objectives to be completed to appease commenters. When I asked for coaching, I was primarily looking for somebody experienced who would tell me if I stood any chance, and perhaps offer a few pointers on areas in which I was deficient and which could cause problems at a future RfA. Not really a full-on guide who held my hand and gave me lots of things to do to pass, but somebody to give me an early thorough evaluation, point out any problems I might have and offer some advice on how to fix them, and perhaps a few pointers about the RfA itself.

On the flip side, I agree with what DHMO said in his answer: "Administrator coaching is a good idea. RfA coaching is a horrible idea." Somewhat more extensive training than what I outlined above with the purpose of coaching one on how to be an effective administrator: the details about being an administrator that one cannot learn from technical guides and such is a good thing. And we could probably use more of it.
Nomination, co-nomination and self-nomination (introducing the candidate)
The nominator should be the person who officially offered the nomination first, generally because he is the administrator coach, or is someone who noticed the candidate and was impressed. The nomination should address all aspects of the candidate that relate to RfA. Point out positives, and explain why negatives shouldn't bar the candidate from becoming an administrator. Show areas of expertise, work he has done, and why the candidate will be a good administrator. A co-nominator should have a specific reason for being there; he should not just be someone who is just generically impressed; a support/strong support will suffice for him. Too many co-noms don't sit well with me, and any more than about two will cause me personally to seriously question the judgment of the candidate (because he should be able to say no). Self-nominations are fine. I think that short, simple self-nominations (e.g. fewer than three lines) are more effective to me as self-nominations.

Who the nominator is can generally tell me a lot about a candidate. For instance, if Sarah nominates someone, I can be sure the candidate is very well qualified. Likewise, if the nominator is an editor who often does social networking, often behaves immaturely, doesn't do much content work etc., I can generally be confident the candidate is unqualified. I generally trust administrators to have good judgment, so I'll be more receptive to an administrator nomination.
Advertising and canvassing
I'm really of two minds about canvassing. The only detriment I see is that canvassing people who do not regularly comment in RfAs distorts the commenter pool, and bureaucrats cannot always discern when this has happened, and therefore cannot judge appropriately. Otherwise, I think publicly and neutrally notifying strongly invested parties on-wiki is not that big of a deal. How the person does any canvassing also matters; if it is obnoxious "vote for me!!!", it is very bad; if it is "hey, I've interacted with you a lot in the past, and so your opinions at my RfA would be helpful", I'll be okay, especially if the note is to people who have had negative significant interaction, that will reflect positively in my mind. I would hope that people who volunteer on this website would have the integrity to evaluate people based on their merits, and not just support their friends when canvassed. (I know, the lack of fair and blind evaluation even without canvassing is already a rampant problem, but eah. AGF and all that.)
Debate (Presenting questions to the candidate)
We need more debate and discussion in RfAs. When people get chafed by discussion (usually in the oppose section), I am genuinely bewildered. Isn't this (at least, in theory) a discussion? Discussions should have ... you know ... discussions. Trying to limit debate and discussion in RfAs is ridiculous. Cutting down discussion that is pure drama, however, is necessary and should happen more. We should not be afraid to bring up discussions in the support/oppose/neutral sections. As for questions, I think we need more. I was personally disappointed I didn't get more in my RfA. In the same vein, I believe that the community needs to stop opposing candidates who don't answer questions prima facie. Candidates should feel free to refuse to answer questions, but the person who asked it should also be free to oppose because, since his question is unanswered, he may not have sufficient information to allay any existing concerns. All questions should be allowed. Stupid rote questions (like the ones about cool-down blocks) can provide some benefit by letting us see if the candidate has done any research into the current RfA atmosphere, which I correlate to the likelihood of future acts where the candidate will do research on situations and background. And who knows? Maybe the candidate will answer absolutely wrongly, and we'll all oppose him, and therefore not promote a poor administrator who had all the other answers.
Election (including providing reasons for support/oppose)
Supporters do not necessarily need to provide a discourse of their reasoning. It can just be assumed that they agree with the nominator. Elucidating upon one's rationale is always helpful, as it fosters the ability to have discussion, and I think most people should do it if they have reasons other than those discussed by the nominator. Also, if there is significant opposition, supporters should address why the are not concerned enough by them. I don't think garishly laudatory supports really qualify as discussion, and would be best left unwritten and just left as # ~~~~, but it's not a problem. As for opposes, I think a rationale should be strongly recommended, and although discussion should mostly happen there, the bureaucrats should not disregard ones that lack a rationale from otherwise established members of the community. Discussion generally happens in the oppose section, and providing one's full rationale allows for discussion of the issues brought up by the oppose.

I believe that it is extremely irresponsible of commenters who support/oppose/neutral early on and never return to the RfA to view developments. If editors A–W support before editor X brings up Very Concerning Diff Y, and those editors never return and editors F–R, who would be concerned enough by the Diff to switch to oppose, do not amend comments accordingly, then the actual consensus (in an abstract sense) cannot be determined by the closing bureaucrat, because the written opinions were not updated. In simpler terms, everyone should return toward the end of the discussion so that they can re-evaluate their comment based on any further developments.

I personally rarely comment in RfAs where I am not familiar with the candidate; there needs to be a very compelling reason for me to be comfortable enough voting one way or the other in an RfA where I have not reviewed the candidate extensively. And that's the issue: I lack the patience and dedication to review a candidate as thoroughly as we should when supporting/opposing an RfA. That's one of the biggest dangers at RfA: people who support without doing any significant review, or for their friends, and we therefore get incompetent administrators. We need the people who spend significant amounts of time and effort reviewing candidates to be the ones who comment, not RfA regulars who do drive-by commenting.
Withdrawal (the candidate withdrawing from the process)
Withdrawing is best when it saves the community's time, such as when it will clearly fail, when it will prevent unnecessary drama, or other reasons (such as DHMO's RfA, although I'm not talking about why it was ultimately withdrawn).
Declaration (the bureaucrat closing the application. Also includes WP:NOTNOW closes)
I'm not sure what the former part of the question is asking, and answers by others do not help me, and they primarily deal only with the SNOW/NOTNOW closes. I'll just ramble breifly on bureaucrats. I think that generally the new bureaucrats know what is expected of an RfA close, and several of the older ones do as well. We should have a high (90%) pass threshold to ensure this. We should routinely retire (or forced reconfirmation with same pass percentage) old bureaucrats so that we only have bureaucrats who are up to date on RfA norms. Reasoning behind any RfA that isn't clear-cut is something I appreciate, and would prefer those be recommended. Now for SNOW/NOTNOW closures. They should happen much sooner. In RfAs where the candidate will absolutely fail quickly, we should not allow the RfA to remain open past one or two opposes, since apparently many of our RfA regulars cannot restrain themselves from heaping on (often hurtful to newbies) comments in the oppose section to the already large number of opposes. In addition, they waste the time of the community.
Training (use of New Admin School, other post-election training)
To be perfectly honest, I think the NAS is useless. I've reviewed the NAS, and while it's a bit better than I remembered it to be when I first looked through it and deemed it useless, it offers very little information that is truly helpful. Many people have used administrator tools on a test wiki elsewhere, and so need little to no assistance in learning how to actually use the tools (and they're pretty intuitive anyway). New administrators need guidance on when to use the tools, and appropriate ways to use them. For instance, the NAS page on granting rollback gives seven steps to granting rollback. One is devoted to when granting rollback is apropos; and it doesn't even explain how to determine this, other than "would make good use of it". That doesn't explain anything. Maybe it would be helpful to note that there isn't actually any decided metric to use? Just whatever arbitrary standard the administrator that happens to see the request has? (The removing rollback steps are even better: "for whatever reason"). There is a page devoted to resolving disputes qua administrator, which is what is really needed by new administrators. So good job there. Perhaps finding seasoned administrators (especially ones who have reputations for being particularly good in certain areas) and facilitating discussion between them and new administrators would be a good idea. (Although I do find this comment extremely interesting considering what I think a NAS should be.)
Recall (the Administrators Open to Recall process)
Until the community decides it is truly sick of incompetent, or abusive administrators, but can't send them to ArbCom to get them desysopped, and so creates a way for the community to remove the flag without the ArbCom, I applaud the recall idea and those who subscribe to it (and abide by that decision should they be called on it later), and am in it myself. My only quibble is that it is truly voluntary: even if there is enormous support for the removal of an administrator's flag, he is fully free to refuse to listen and not ask a steward to remove it, and the stewards will/can do nothing. With respect to desysopping, the stewards (as far as I am aware) only listen to the administrator himself and ArbCom on enwiki — they will not/can not make a decision on consensus themselves. We should work with the stewards (and I suppose with WMF since this would modify steward policy) to allow sysops to voluntarily create binding, objective recall rules that the stewards will recognize and enforce, even if it later goes against the sysop's will.

Although this isn't asked in the question, I'll talk about it here since it's important. I think administrators should have to go through forced reconfirmation every so often. As someone said (I can never recall who, no matter how many times I reference this), an administrator should be theoretically be able to pass another RfA at any time. I agree with this sentiment. We should not have administrators retain their status if the community does not trust them any longer, but has no cause to particularly revoke the status through ArbCom.

When thinking about adminship in general, what are your thoughts and opinions about the following areas:

How do you view the role of an administrator?
Administrators have access to technical tools that are restricted due to the potential for damage because of incompetence or abuse. They are allowed to use these tools because they have demonstrated (via a successful RfA) that they understand the policies governing them, and demonstrated sufficient judgment, maturity, clue and competence to use them well. They also have a bit of authority; they are the ones entrusted to settle disputes, and (should) be able to enforce these decisions, as well as general sanctions as needed to ensure a smooth running of content-building.
What attributes do you feel an administrator should possess?
He should be forthright, speak frankly and effectively, but be diplomatic. He should be, above all, competent in what he does; while mistakes are natural, they should be few and not stem from serious ignorance. He should therefore be thoroughly knowledgeable about most policies, and know where to go to find out (the hopefully few) answers he does not know. He should defuse unnecessary drama, and never start it, nor participate in it (unless that participation is directly aimed at defusing it). He should respect community norms and decorum. He should research histories and situations thoroughly so that he is fully familiar with situations in which he already is or about to become involved. He should be willing to discuss his actions, and in a civil manner. He should maintain objectivity when using his authority.

Finally, when thinking about Requests for Adminship:

Have you ever voted in a request for Adminship? If so what was your experience?
Yes. It was fine.
Have you ever stood as a candidate under the Request for Adminship process? If so what was your experience?
Yes. I thought it was fun.
Do you have any further thoughts or opinions on the Request for Adminship process?
I personally don't believe it is as broken as many people claim it is. There are, of course, flaws, but I do not believe they are as serious and enormous as they are made out to be, and nobody has proposed anything better. If anything, the problem is with the commenters, not the process.

I also think it should be forbidden to have yet another discussion in an RfA with Kmweber about his opposes. Srsly.

(Although it seems I was beaten to posting this sentiment. I had written my opinion in my draft before he posted, honest! :D).