Below is a brief consideration of what recent successful RfAs have looked like and the criteria I think summarise the main guiding judgements I and quite a few others use.
RfA Support Percentages Don't Look Right
editRfAs are tough, numbers of applications remain low - 7 (5 successful) as of 24/02 in 2020, 31 in 2019 (22 successful) and 16 in 2018 (10 successful). Yet, amongst those that pass, the consensus percent remains extremely high.
- 95% or above: 21/37 (57%)
- 90%-94%: 6/37 (16%)
- 85%-89%: 3/37 (8.1%)
- 80%-84%: 2/37 (5.4%)
- 70%-79%: 4/37 (10.8%)
- 69% or below: 1/37 (2.7%)
On first glance, we should be having candidates scattered more evenly, 3 main reasons for why this isn't the case come to mind:
- Only 1 person has to do the research - some stats are checked by most !voters, but many more personal issues are originally located by a specific editor, and other editors (who might otherwise have supported) accept it's a valid reason and also oppose.
- Opposes beget opposes - frequently, early opposers get challenged by supporters of the candidate. This discourages breaking ranks. However, once a handful of opposes are in, it becomes more "acceptable" to also !vote oppose, so the effect snowballs.
- RfAs are tough. Contested RfAs are worse. - RfAs are heavily accepted as feeling like organised hazing. Having large numbers of respected individuals (perhaps those considered friends) analysing the details of a candidate's contributions feels tough. But if everyone is saying "Has a clue, not a jerk", then negatives are minimised and it can feel both informative and empowering. Close fought RfAs (and indeed, any unsuccessful RfAs) include having competence and personality criticised from numerous directions. Informative, but unpleasant. So many competent candidates might not apply at the first viable point but hold out for a smoother journey (or be advised by others to do so).
and?
editAs to what this means? I've put in my personal criteria below, and a selection of others can be seen at RfA essays and criteria. It is a long, reasonably high, list of requirements. That is because, at least at the moment, the community definitely does think that adminship is a big deal. Minimums are "what would be accepted as workable, rather than "just tolerable if the moon lines up right". I'll also include good and great levels for each.
But the plan is to indicate what might get you over the line as compared to what at least gives you a basis for that smooth sailing
As with anything there will be exceptions, with me IARing to both support and oppose in contradiction, but I ran it over the past 2 years candidates, and it's almost matched my !vote for 1st time candidates in every instance.
Within each category, the order is by no means a ranking system of assessed comparative importance
Criteria: Critical Aspects
edit- Tenure: 15 months is a basis to let you cover the key areas for adminship and show some experience in your chosen areas. An extra 3 months will get many more onside, and very few (though definitely some) will quibble at the 2 year mark.
- Edit Count: Editcountitis has definitely calmed down in 2019. 12,500 edits is again a base mark, if only because it's tough to generate both enough content and work in a couple of admin areas for long enough with fewer edits. 15,000 is better, 20,000 is named by a few as a personal criterion.
- Blocks/Bans I don't believe the community will accept anyone with a current limited block (IBAN/TBAN) of any type. I would generally say a clear 6 months since the end of any limited block, or a 24/31hr block (usually 3RR-related) or ARBCOM admonishment. For a longer block, I'd say a year. For an indef or a long block, I'd advise 2 years. These durations are both for the sheer act of being blocked, but also for the collateral unhappiness that likely was caused at the time.
- Content work - at RfA, everyone wants content work. Everyone also has their own judgement on what "a good amount" is; whether it be number, quality, GA/FA or mainspace %. I generally feel you need 5+ articles and 2 B-classes. I discount stubs from the article count. Getting something through GA will ease many content creators' concerns, and getting 2 through, or an FA, will ease almost all. More is always better - once you've got a bare minimum to show you can write a quality piece, I feel 20 starts shows as much content focus as a GA. 1 note, those interested in admin DYK work are generally held to higher content requirements.
- Need and Knowledge for the Tools: While no admin could regularly use the full toolkit, you must show a strong need for the tools. I, and most others, feel that at least 2 areas should be indicated (some like AIV and RFPP go together). Firm competence must be demonstrated in each of these areas. You will receive questions on them, including analysis of prior decisions and often "case study" scenarios. As your chosen area, you need to be well in excess of the base requirements that a candidate not planning to be in that field should have.
- Calmness under fire/civility: Not all candidates will have been active in hostile areas, but whether you are or aren't, I want to see a more civil individual than the low levels required under WP:CIVIL. I judge that as "swearing is okay", but "swearing at editors" isn't. Neither is being rude to them. The occasional lapse happens, given provocation, but admins have higher behavioural requirements, and thus so do candidates.
- Good RfA Competence: this is in 2 main areas: answering questions and RfA behaviour. Questions really can both lose and win support. Answer all (barring exceptional reasons), give detailed thoughts, make them your thoughts (your answers shouldn't be contradict-able by behaviour elsewhere). The "RfA behaviour" includes things like not going after those who oppose you (even if feels stupid).
- "Modesty": - perhaps "lacking arrogance" is better. Talking modestly is good. But letting your actions speak for you on this is non-negotiable. Willingness to compromise, engagement with others, and most of all...be capable of admitting fault. This can be changing your mind from another's reasoning to a straight up "yep, bad call by me". You will have made mistakes. You will make mistakes in the future. An admin who can't own their errors and doubles down is a threat to us all. The community is, despite occasional appearances to the contrary, quite good at accepting admitted mistakes.
Critieria: Very Good to Have
edit- AfD Experience: Even if it isn't a likely admin area, the vast majority of successful candidates have at least some AfD experience. Getting experience in AfDs teaches huge amounts of policy at once, while also practising discussion with those who disagree, conflict management, source assessment and others. AfD general competence (that is, for those not planning to mop in AfD) has different aspects, but the following work if combined:
- Number of AfDs: 50 is good, 75 better, 100 is more than enough
- Consensus %: if you learn your policy and guidelines, and learn how to interpret them, you should have a reasonable consensus score. If you learn to convince others, it will be even higher. But...
- Stand on your feet: if you pile on to the end of already settled discussions, you can have a near-100% result. But all that shows is you can jump the way the wind is blowing. Candidates need to show they are self-supporting, even if others might disagree. Get experience being the nom or first !voter.
- Articles not just for deletion: there are 4 common AfD results (Delete, Keep, Merge & Redirect). Use them. Be neither a radical deletionist or inclusionist. Consider alternatives to deletion.
- Community Trust Role: experience in a role that takes true community trust is a big plus. There's a few of these, the more common (but not only!) of which are Clerking, BAG, and OTRS. N.B. This is distinct from trust as a user - that's compulsory.
- Near perfect use of Edit Summaries: I am frustrated by edits without summaries, whether working on a page with them or trying to review a candidate at RfA. I'd be very concerned with under 90% and want over 95%.
Criteria: Good to have
edit- CSD Experience: have Twinkle keep a CSD log. Get a couple of dozen CSD nominations in a few different areas. Accuracy trumps volume. 90% minimum
- Good Nominator(s): I don't !vote for someone because they have a nom vs being a self-nom. However, I find having some names I know and trust beneficial if there's a negative you want to handle (in effect, they're willing to trust in reform).
- Conflict Experience: Not everyone has conflict experience. Not all admins intend to be active in that field. But it's definitely a plus as a really helpful skill.