User:Barkeep49/RfA

On September 11, 2019, I was privileged to become a sysop on English Wikipedia. Having watched several requests for adminship in my time and now having lived through it, I have a few thoughts on the process. On the whole, I think we know a lot less about how Requests for Adminship (RfA) works than conventional wisdom would suggest. It's a bit of exaggeration to say we don't know anything but I think that is closer to the truth than to say we really understand how it works.

The observations from this essay are most true from about a period sometime in 2017 to 2020. As of January 2021 the general advice holds but RfA had a weird year in 2020 and so this essay may be out of date when you're reading it. I am going to talk in-depth about my specific experience, but in a way that is intended to illustrate broader points. While I could make the same points about someone else's RfA that doesn't feel fair to them when I can use myself, warts and all, instead.

Deciding to run

So I passed RfA easily. This ease was, on some level, really an indictment of me - if I had all this support out there clearly I waited too long to run as I could have passed at some earlier time (albeit without the level of support I ultimately ended up getting). I had some people telling me in June/July I should be running - in retrospect they were right. I ignored what they said for two reasons. First, they were my friends, so of course they were inclined to think good things about me (more on this in a moment). Second, and more importantly, I knew every mistake I had made. Every time I had a speedy nomination declined or an article I nominated for deletion end up with a keep result. Every time I had failed to be as welcoming to new editors as I could have been. The times where I didn't live up to the expectations I have for myself about how I should act on wiki (this last set of mistakes are something than I can count on one hand, but I still wish it were 0).

None of that stuff ended up being discussed. If I ran today would that stuff be discussed differently? I suspect so. This is also why the Optional RfA candidate poll is not great. I think that forum is inclined towards finding problems. And because we know much less than we think we do about RfA, the number guesses aren't useful unless there's enough that we can benefit from the wisdom of the crowd and maybe not even then.

Instead the question is does an editor have 18-36 months of work (longer tenure can work against a candidate), something substantial to point to for content creation (though not necessarily even a GA), and sysops, particularly long-established ones, showing up to their talk page or email inbox suggesting they run at RfA? If so then there's a good chance, if it's the editors first attempt at RfA, that they'll pass (the hedge is because RfA is already a small set of data and someone going through for the second or more time is even smaller and with lots of complicating factors about why the second or more RfA is occurring). When I ran my feeling was that all you needed were the sysops - I think in this changing climate you actually need more. If the sysops are not showing up and you think you would be a legitimate candidate seek out a long established sysop you've interacted with and directly ask their opinion. They might turn out to be ready to nominate you or they might give you some actionable feedback.

How supports and opposes happen (or just how much of a knife's edge is it all?)

You never know what is going to get brought up at RfA. For me this meant I worried about every every bad thing and mistake I'd ever done was going to be ferreted out and picked apart. I would never have guessed that this deletion would end up being a significant discussion point.

Thankfully I would not list my participation in that discussion as one of my mistakes. But what became a focal point of the RfA could have been one of my mistakes instead. Something like this trout from 5 weeks before my RfA would be on that list. If I had run the next day, that's obviously going to get brought up. I'm pretty amazed it didn't get brought up when I did run five weeks later. What happens if SoWhy, who didn't participate in the RfA, had turned up and opposed my RfA? At what point does a group of editors, especially if led by someone who has the respect and prominence of someone like SoWhy, decide that there is enough evidence of me being unfit with deletion (as this would build on the discussion around AfD) to garner non-trivial amounts of opposes (and cause some people who had supported me to switch)? I'm not sure what that inflection point is and I don't know that anyone else knows either. I feel confident in saying that there were people active on Wikipedia who sat out, for reasons I'll explain in the next section, and so the actual opposition was greater to me than what a look at the vote count would suggest. We all definitely do not know what would have happened if my RfA had reached that inflection point.

Motivation

RfA voting is in no small degree about motivation. Opposing a candidate has a marginally higher cost - because of the questions and responses you'll receive - than supporting a candidate, where your signature can suffice. It has a higher cost because to do this effectively you'll likely need to do research to find diffs and because you'll need to be willing to stand your ground when people disagree with you. This imbalance means that people only need to be mildly motivated to support a candidate but need to be a little more committed to opposing a candidate and for the reasons discussed in this section there are lots of reasons why someone may not cross that motivational threshold to taking the step of opposing. Further, some people will be motivated enough to write at some length in positive ways, so very motivated people cuts both ways for the candidate. A person is going to be more successful if they have a group of motivated supporters behind them. For me this core of motivated people came from the relationships I'd built on IRC and through New Page Patrol. Have a base of support from a project or other wikigathering place means that even if some opposes come in you'll still have a steady stream of people excited to support you. At least for me, it was humbling to read of people who I couldn't remember interacting with who had definite feelings about me based on interactions we'd had. In retrospect, my friends, some of whom also fit the sysop category I defined above, were better judges of me as an editor than I was of myself.

Prepping for the RfA

Knowing what I know now, I feel confident I would have passed if I had done a self-nom. However, running on a self-nom is a very risky proposition, only more so now, and it's easy enough to find a nom and so it's worth finding some who you would be proud to have represent you. I know some people supported my candidacy because of my noms. More importantly, I would speculate that some editors with concerns about me choose to sit the RfA out because they trust the judgment of my two noms, Melanie and Premeditated Chaos (PMC). Having good noms meant that people were less likely to reach the threshold where an oppose seemed worth it.

Having noms is also great because you can get feedback on what you write before you write it. For me this meant not only the standard three questions, but also five questions that I had seen regularly at the RfAs proceeding mine and to which I crafted answers well in advance. I think the time I spent thinking about these questions was well worth it and, as discussed below, came in handy during the RfA. I would encourage all candidates to do this in advance of any potential run.

During the RfA

Unlike with who will run and pass an RfA I think the conventional wisdom about how to act during an RfA is pretty solid and hasn't changed from when I ran. I believe our current convention of not having candidates, and to some extent noms respond to people, outside of the questions asked of candidates, is a good convention. I had it pretty easy, and it was still tempting at a couple of points to jump into the conversation. But in the end, I wasn't needed because there were several people, including many who I don't know particularly well, who were willing to see my point of view and express it to others. Depending on the issue there will be differing amounts of people on each side, but I think it speaks well of our community that some faith in others is rewarded at RfA and the candidate doesn't have to be the one jumping into the fray. A potential sysop showing a little restraint is a quality we should desire in someone entrusted with the toolkit.

Of course, when it comes to questions formally asked, restraint is a double-edged sword. People really do want you to answer questions. In my case there was a bureaucrats noticeboard discussion about one of the questions; when that discussion looked to be having mixed results, I just decided to answer the question and let the chips fall where they may. The person who asked that question did end up opposing me, but I also received generally positive feedback for my response, so I'm not sure it was a net negative in answering rather than ignoring it. However, I also had a fair amount of confidence in my ability to answer questions. If a candidate was less confident about their abilities, this is where high-quality noms can be a vital sounding board during the run - even with confidence in my skill to answer questions I got some useful tips I was able to use when answering.

Regardless, a candidate should take care when answering questions. Precision matters - it ensures you don't make a "slip of the tongue" (I was worried I'd make a ban/block reversal) or muddy the waters with our numerous policies, guidelines, and essays that hold more weight than some guidelines. Checking applicable policy and guidelines, no matter how well you know them, before hitting publish in response is good practice. It's also not bad practice if you become a sysop. Well-answered questions are the single best way the candidate has from stopping an RfA from reaching the "tipping point" of passing easily and passing with difficulty or not passing at all.

Four of the five questions I prepared for were asked and with a little tweaking to reflect the way the question was phrased, I had a reply I'd given plenty of thought to ready to post quickly. For the questions I couldn't predict, a big surprise for me was that a question phrased in a challenging manner didn't mean that the person was inclined to oppose me. This unpredictability played out most clearly with the question from Lordes, which was accompanied by 15 diffs. It was posted right before I left for work, so I had my commute to compose an answer in my head. I was worried even with a reasonably prompt response that some people would be scared off by a concern supported by so many diffs. This anticipation of opposes turned out not to be the case, and, in fact, Lordes ended up supporting. While some others who asked challenging questions did oppose, it was not as correlated as I had expected. Some people just want to test what potential sysops are made of, and, to that, I say, "Fair enough."

One thing I was grateful that happened during my RfA, was that Mz7 asked me the question he did. I knew heading in that one of the potential weaknesses was the quality of the case I could make around needing the toolset. It was something I went on at length about when Melanie initially emailed me saying I should consider running. Of the three standard questions, my response to question 1 evolved the most as I tried to more forcefully present the affirmative case I could make for the toolset. However, when it came up, as a concern for some neutral and oppose participants, it was a fantastic relief to get an open-ended question that let me address those concerns - concerns that some were already defending me about but valid concerns none-the-less. As an open-ended question, I retained a lot of latitude in how I responded to it. I think this is an opportunity all candidates should get and am thinking about trying out a question along the lines of "What question do you wish you were asked and what is your answer to it?" at some upcoming RfAs to provide others the same chance that Mz7 provided me after several concerns in a given theme in order to let the candidate have a chance to head-off the inflection point.

One thing I don't recall having read advice about is how to act on wiki during the RfA. Wikipedia's a large place, and it's 100% likely that if you're active while the RfA is going on you'll interact with people who have no clue it's going on and while your contribs might get a little more scrutiny than they would otherwise you're also probably not as under the microscope as it can feel. During the first day, I was terrified of doing anything besides responding to questions. This inaction was probably more restrained than I needed to be. As the RfA went on, I resumed something closer to regular wiki activity, if still being a bit more conservative in where I participated. I think I got it right closer to the end of the RfA - conservative - than I did at the beginning - terrified.

Stray personal thoughts on running for RfA

While the above is designed to make some universal points through the lens of my experience, this part contains some personal thoughts.

I slept really poorly the night before I launched the RfA and ended up being up super early in the morning to launch it. I was worried that my tiredness would impact my answers to questions but fortunately by the time exhaustion really set in the question rate had slowed way down. As mentioned above there was a bureaucrats noticeboard discussion about my RfA and thus about me that I was never pinged in. It was weird watching others discuss me and my RfA in that way. I think the "what areas are you worse at" question that comes up is a bad question. You want a sysop who is self-reflective and knows their limitations. This question doesn't give that answer. It just produces unsatisfying answers. If this information is important for a voter - and I think it should be - there are better ways of getting at it, not the least of which is other questions they answer especially about actions they'd done. I was prepped when this question came up for me, even with a variant of giving two areas, but that doesn't mean it adds any value for the hazing it does. I'm glad I'd written WP:MUSHROOM because it gives me a good guide for how to act now that I have the sysop toolkit. This is especially true because it's at once immensely gratifying and also really bizarre to have so many people saying so much nice stuff about you. I think it's a little tacky when a candidate uses the thank button for those who support (or participate). So I didn't do that. Instead, I wrote a whole bunch of thank you notes, beginning with the opposes and neutrals. I really am thankful for those who expressed concerns or had suggestions however they ended up voting. I always want to be a better editor and hearing about ways I could improve, or at least ways my actions are or could be perceived is feedback I need to keep in mind going forward. If you'd laid me even money odds that I'd have ended up on WP:200 I'd have bet against me. I honestly expected to pass in the mid 80s with around 150 or so supports. I say this not to humblebrag, but to put concrete numbers on points I made above. And because I believe in holding myself accountable when I get something wrong. It was fun learning about the administrator menagerie and gaining a place on that. I'd never heard of it before I was told I had been put on it.

After the RfA

I was asked to provide some thoughts on my first year as a sysop. I will attempt to do that formally at some point after the first year, but for now some stray observations:

• Since my RfA, my talk page gets a lot more activity.
• I also get a lot more responses in discussions - I'd been used to putting a lot of thought into some topic and then being ignored. Since my RfA I get ignored a lot less.
• Neither of the two proceeding bullet points are surprising in thinking about them, I just hadn't known it was going to happen in advance.
• I think I've used the toolset a bunch so I continue to think that line of opposition silly.