User:Tamzin/340/112/16: An RfA debrief

 If you came here looking for a continuation of my RfA's political discourse, you'll likely be disappointed. I discuss that a bit in, but the focus of this debrief is on what it was like for me, personally, as the candidate. To that end, I have not referenced any opposes by name, to avoid giving off the impression that I am continuing the debate.

The point of this debrief is to explain two things:
 * 1) Why people don't RfA.
 * 2) Why people shouldn't RfA.

To the extent that this is a critique, I want to stress that it is a systemic one. No one person is responsible for the system winding up the way it has. If anything, we are all pawns in that system.

Thoughts prior to RfA
I think a lot of the discussion of RfA is built around a set of premises that haven't been true for years. Going into this RfA, I have the technical access to do most of the things that were admin-only when I started editing. There seems to be an assumption, in discussions of RfA, that people should want to RfA. Why should we?

Look at it from a decision theory point of view. Let's say that a qualified candidate has an 80% chance of passing and 20% chance of crashing and burning, without much middle ground. There are three main benefits if one passes: That altruistic desire and that egotistical one, while opposite in nature, are functionally the same in practice, both being endogenous benefits that might motivate the potential candidate to make a decision not justified by exogenous needs.
 * 1) Reduced hassle to accomplishing five specific things: blocking users, protecting pages, deleting pages, editing deleted pages, and viewing deleted pages.  We assume that a qualified candidate will have a high success rate when requesting any of these things, so, contrary to what is sometimes claimed, the benefit of adminship isn't about getting to do them. It's about getting to do them quickly and in some cases en masse.
 * 2) Greater ability to make a difference to the project.
 * 3) Power and validation.

There are two main downsides if one fails:
 * 1) More difficulty becoming an admin in the future.  (From a decision theory POV, the simple fact of failing itself has no weight, since you've lost nothing.)
 * 2) Loss of social standing.  Specifically, loss of the sense that the community trusts your judgment.

Downside #2 is the critical one. Most people inclined to RfA will already be doing work that expects them to have admin-level trust. For me, that would be requesting blocks at SPI or making complex closes at RfD. I know that personally those are two of my favorite admin-like things to do on Wikipedia, and I am guessing the same is true of many other qualified users. So assuming that 80/20 split, by running I accept a 20% chance not just that I won't become an admin now, not just that it will be harder for me to become an admin in the future, but that I will wind up with less than I started with, deprived of things I liked to do.

Benefit #1 and Downside #2 stem from the same place, and thus tend to cancel out. For someone to run, then, one of five things must be true:
 * 1) They are exceedingly confident that they will pass.  This only applies to users who have perfect track records of avoiding drama and either doing noncontroversial content work or doing very popular technical work (and still a healthy amount of content work).
 * 2) They do not do many things to which Downside #2 would apply.  This applies to a similar demographic to the above.
 * 3) They are so incredibly inconvenienced by the lack of admin tools that it is worth it for them to risk the 20% chance of Downside #2.  Due to the level of tool unbundling in the past decade, this is very rare.  The last RfA I can recall of this sort is Trappist the monk's, which happened while Requests for comment/Template editor user right was ongoing.  You could argue that there's a degree of this when non-admin SPI clerks like me RfA, but we're also perhaps the most vulnerable to Downside #2.
 * 4) Benefits 2 and/or 3 matter so much to them that they are willing to take the gamble.

Thus, among people who are already involved in administrative areas to any particular degree, RfA will tend to only attract people who have a strong desire to help or a strong desire to be powerful. That amounts to a trickle of a few people a year, whom of course we will AGF all fall under the former, not the latter—aside from the one we know for a fact was motivated by the latter. (And of course I know I have pure motivations here.)

Thus I contend that RfA discourse largely misunderstands the decisionmaking calculus for potential candidates. If we want more people to run for adminship, one of two things must change:
 * 1) The benefits of adminship could be expanded.  This seems unlikely to happen, as gatekeeping of tools to the admin toolset is meant to be done by necessity, not to create artificial scarcity.
 * 2) The RfA process could be restructured to remove or greatly reduce the possibility of crash-and-burn failures (roughly making the social impact of a failed RfA no worse than that of a failed request for a sub-admin permission). This restructuring would have to be major; the first idea that comes to mind as potentially sufficient is banning replies to votes by anyone other than the candidate.

By the time anyone reads this, we'll know how I fit into the very small dataset that we do all our analysis based on. Either way, hopefully people will consider what I've had to say when it comes time for the next inevitable RfA RfC.

Thoughts after RfA
The RfA became exactly the last thing I wanted it to be: about something more than myself, with support and oppose votes representing sides in a broader cultural dispute as much as opinions on my fitness. As such, I'd like to start about my reframing this on me. Well, us.

A disclaimer in advance: When a mentally ill person describes a sequence of events from their perspective, some people have a tendency to conflate that with them either trying to "blame things on their mental illness" or conversely to guilt-trip others. But just as you, in describing how you handled a situation, have to speak to your own state of mind, so do I have to speak to mine. My state of mind just happens to be a bit different than most people's.

Dramatis personae
Hi. We are Tamzin—a shared identity of several members of a plural system. While we disclose this, we've only discussed that at any length on Wikipedia once before, but it's hard to write a debrief about our own experiences from the perspective of a singular identity that in reality is more an averaging of our individual identities.

Collectively, we're a 26-year-old nonbinary transfem living in Cape May County, New Jersey.

Individually:
 * C (they/she) is the more outgoing and outspoken one. If you talk to us at SPI, you're probably talking to them, and if you talk to us at AN(I) you're definitely talking to them.  They do our article-length content work, as well as content work of any sort about political or military topics. Outside of Wikipedia, they like watching action movies with their wife.
 * 5 (she/they) is the quieter, more studious, more technically-inclined one. If you talk to us at RfD, or about a technical matter, or at a particularly behavioral-analysis-heavy SPI, you're probably talking to her.  She does the rest of our content work.  Outside of Wikipedia, she codes in Python and writes fiction and poetry.
 * F (she/her) doesn't edit, but does some of our Wikimedia-related programming, usually finding creative solutions when 5 and C are stumped. She keeps to herself.
 * R (he/they) doesn't actively contribute, but plays an advisory role in some of our decisionmaking. He likewise keeps to himself.

C is currently the "main fronter", i.e. the person who most frequently speaks as "Tamzin", on-wiki and off-. The way our system's dynamics work, "the twins" (C and 5) basically always have at least a dreamlike awareness of what's going on in the outer world, regardless of who's fronting (controlling the body).

Prep
C had mostly taken the lead on getting the RfA going, combining a support offer from BDD last August with a support we'd solicited from Drmies and obtained after we met his content-contribution expectations. As RfA readiness neared, C decided that they didn't want to go through all the rigmarole that some do at RfA: no long email threads, no "shadow noms", no extensive prep. C and 5 drafted an A1-3 and prepped answers to some questions we thought likely.

The start was a bit chaotic. C had said to Drmies that they wanted to just go for it, and he took that at face value in boldly creating the RfA page after briefly discussing a draft nom but doing no prep beyond that. Far from upset, C was excited that he shared their casual attitude. The stressful part, though, was wrangling BDD, who was caught flatfooted but nonetheless managed to put together an eloquent conom within 24 hours of our letting him know the RfA page had gone up.

We touched up our A1-3, wrote a brief acceptance statement, and transcluded. I remarked to a friend, "This is going to be either 250–≤10, 250–100, or <50%".

Smooth sailing
C is good at stressful situations. They manned the front for the first 48 hours, working quickly to address the first oppose's content concerns and handling the first wave of questions.

The supports were rolling in, but I was mostly just focused on the opposes. A few friends made fun of me for that, but it's not like I wasn't aware I was doing well. It's just that I was aware any oppose might be the runaway one, the one where I and my supporters lose the narrative. That's why I was trying to nail every question as soon as it was asked.

The situation with Q13 overloaded C just a bit, though. Since things were otherwise calm, I tagged them out and took over, with the intention that I would see through the remainder of an uneventful cruise toward the first of C's predicted outcomes.

Just in time for... Q14! Three characters we never want to see in that order again.

My initial reaction was one of "We said that?" I hadn't said that; C had. I knew that C stood by the statement regarding desysopping Trump supporters, and C's political views are our political views. C is bad at admitting when they've used hyperbole, but I pressed them on the matter of the "wouldn't vote for a conservative", and they conceded it was unreasonable.

A thing to know about us: We don't lie about important things, except to protect our or others' safety or privacy. There may be times we decline to answer a question, but that wasn't really an option with Q14. Not answering "Do you stand by these views?" is taken as a yes.

In light of that, I knew immediately that it was only ever going to be damage control. There was no way to say something noncontroversial without lying. I ran the answer by two people, looking for the gentlest way to put things, and posted it. I wrote in an email to BDD, "I workshopped A14 a lot toward what would lose me the fewest supports, but seems inevitable I'll lose at least a few. Unpleasant, but I'm no liar, so there was nothing really to do other than find the most polite way to present it."

I thought, "That should get us about 20-30 opposes." Given my previous prediction, I should have known what that would mean. There was no 250-30 in there.

The 32 hours
We went to sleep around 5 AM and woke up around 5 PM. (Horrifying, I know, but normal for us.) As we read the night's comments, C quickly retook command. Arguments, as a rule, are C territory.

Before I continue: C had the front for this next part, but I'm going to narrate it from my perspective, because I think they would just get drawn into relitigating Q14. Let me say some things about my twin: They are wonderful. I love them. They have gotten us through some very tough moments. They are a good person, kind to me, incredibly caring toward (and adorable with) their wife.

They are also an incredibly opinionated person, not prone to backing down, and prone to an attitude of "Fuck 'em, they can hate me if they want to". A big part of functioning in general society as a plural system is learning to balance out individual parts' negative tendencies, and usually I just steer C to some other focus. But the matter of RfA was all-consuming.

C also doesn't get sleepy. Obviously plurality can't override physiology: The body still gets fatigued; C just doesn't feel it. I have seen them drive a car through a partly-on-fire forest around hairpin turns, all executed perfectly; when they passed control to me, I found the body so incredibly exhausted that I got disoriented on a simple turn and crashed.

C stayed up for 32 hours, intently following the RfA. We needed to be taking a conciliatory tone, and they are not good at that. They are good at telling people that they are wrong. They are good at emphasizing just how strongly they believe the things they believe. I am good at conciliatory tones. But once C is on edge, they stay at the front until they feel the "threat" has passed.

C was convinced to stop replying by a timely email intervention from an admin who'd had a difficult RfA. Things started to stabilize a bit, particularly as people began to defend our right to object to support for our own oppression.

I did make one reply after that, with 5's blessing. It may have been the one that most contributed to raising the drama levels; I'd like to think that otherwise most of what happened wasn't our doing. But I don't know... There are some things it's hard not to respond to. If anything I think it's just bullshit that RfA candidates are expected to take every single hit lying down. I hope one important precedent we set is that you can stick up for yourself (at least slightly) and pass.

Not at no cost, though.

C and I got the chance to respond to some questions that clarified the scope of the view we'd expressed in A14, and the support-to-oppose bleeding began to come under control. Velocity was negative, but decelerating.

After about 30 hours awake, I regained the "front", and, much like that night when C drove past a wildfire, I was immediately hit with all the things they hadn't let themself feel. I cried for a very long time, profoundly overwhelmed at the day we'd just had. After a while, I was able to eat something for the first time since the night before. Around 11 PM we finally slept.

One thing 5 skipped past: This day lost us a friend. A well-respected right-of-center Wikimedian, who had supported us (never struck their support, in fact), and who was one of the people I'd had in mind as an obviously competent right-of-center editor when I conceded that the line about not supporting conservatives was hyperbolic. This person was outraged that I'd said such horrible things. They rattled off a list of offensive things I'd said, none of which I had actually said. When I pressed them on if they'd read what I said, they became evasive, then lashed out at me further, till I blocked them on Discord.

Perhaps that's a good microcosm for the day.

Watching the clock tick down
About 12 hours into C's very bad day, I had asked our mother if she could come visit. To her eternal credit, she found a way to make it from Washington, D.C., to Cape May County without a car.

By Saturday I decided to only check the names on the opposes. I went back later and did read every one, but reading them as they came in was only stressing me out.

Our mom got here Saturday evening, about 30 hours before the RfA was to close. It was an immediate help to have something to draw our attention away from the screen.

As it dawned on me that I was on track for the second-most-controversial RfA ever by number of votes, a degree of serenity washed over me. It was an outcome so absurd that I had to just take it as something Ionescoesque.

I called my platonic soulmate, BiomatrixBackup, and vented for a few hours.

By this point I'd gotten good at maintaining that serenity. I just watched the votes roll in. The hardest part was, when I did read an oppose, knowing that I couldn't respond, even if they were misrepresenting what I'd said. Not to say that all or even most were. But some were.

More frustrating than that, though, was finding ourself too exhausted to do much of anything else on Wikipedia. I wanted to make some content improvements to List of journalists killed during the Russo-Ukrainian War, and 5 wanted to work on fixing the XFDCloser issue that had forced Steel1943 to revert our popular previous change to rfd2, but so long as we had a Wikipedia tab open, it was hard to focus on anything but the RfA.

We went to bed around 7:30 AM and slept about 6 hours, blissfully waking up to only 9 hours left in the RfA and no major change in the voting trend. Two new questions, one already struck, the other so obviously redundant with another question that we knew we'd be forgiven for not answering.

We floated through the final day of the RfA. Passing the all-time support record was at once very sweet and very bitter, as we simultaneously approached both the all-time oppose record on a successful RfA and, more stressfully, the discretionary zone.

Two and a half hours out from the end of the RfA, our mom looked up from her computer and said, "Okay, I need to stop reading this". I hadn't realized that she had also been sucked in. There really isn't a feeling like the one of yourself or someone you love being subject to so much criticism in one place. Valid or not.

Days 8 and 9 of 7
I watched the initial 'crat comments in. I saw Moneytrees' and Paine Ellsworth's comments about it being unfair to me to prolong things, and I realized, I was just glad the real chaos had ended. I'm good at waiting.

The next day I kept true to my reply to Paine and went to the beach with our mom. I brought a copy of Ringworld, which I've been lugging around for two years but somehow never get around to finishing, and watched the dolphins dive in front of the late afternoon sun, and reflected on things. I read all of two pages of the book.

As the sun set, the rising tide brought in six horseshoe crabs. A wave would push them up the beach, over a berm near the low-tide mark, and they would slowly right themselves and stagger back toward the water, just in time for the next wave to push them back.

That felt apt.

We knew that, if the 'crats found no consensus, that would just be the end of things for us on Wikipedia. It would be everything in the "crash and burn" scenario in our pre-RfA essay, amplified 100 times over by what would have been an unprecedented sort of failure. We would have lost so much more than just adminship, with no path back to a normal existence in this community, nor necessarily a desire for that path. C and I prepared a farewell statement for if it came to that.

I'd told a friend on the 29th that I'd be busy till 2:19 on 2 May UTC. On the evening of 2 May, they reached out: Them: Oi, waddup. How'd things go, m8? Me: Now approaching day 9 of 7 Them: Wot As the direction of the 'crat chat became clear, we could finally truly relax. As my RfA entered its ninth day, I went to bed knowing I'd likely be an admin when I woke up.

And I was. A sleep-deprived emotional mess of an admin.

Takeaways
It was my answer that set everything in motion. Looking back, with 20/20 hindsight, the mistake I made was answering the question in a way that supported the premise that my RfA voting standards matter. Keeping my answer shorter but still honest would have gone a long way toward that, something like "That was a hyperbolic description of my personal views of how politics interact with someone's fitness to be an administrator, said in the context of defending an editor from unfair attacks based on his past far-right views. I regret saying it, as it could give an incorrect impression of how I view politics on Wikipedia. I don't treat any editors differently based on their political views in content or conduct disputes, unless those views are so fringe as to be disruptive (namely support or apologism for regimes built on crimes against humanity, including the USSR under Stalin).  In terms of RfA voting, I don't impose a litmus test against conservatives, and shouldn't have said I would." Maybe there would have been follow-ups about some of the parts left ambiguous there, and things would have gone the same way, but it still would have been a better way to handle it. I have a tendency to be too long-winded, and it's worth remembering that a long answer, even a good long answer, can give off the false impression that there's more "there" there than there is.

If we overlooked something in our pre-RfA thoughts, it was that classic shortcoming of decision theory: failing to account for emotion. We dismissed the obvious downside of a failed RfA—failing one's RfA—as meaningless from a decision theory perspective. But that ignores the toll it takes on a person to go through that.

I think the community also ignores that toll. The community simultaneously creates at RfA a sufficiently hostile atmosphere as to drive many editors away, acknowledges the existence of that atmosphere without particularly taking responsibility for it, bemoans the fact that editors do not want to RfA, and every year has a discussion about fixing RfA that results in no improvements to RfA atmosphere.

RfA won't get better because the community doesn't want it to get better. The community, on a deep and pervasive level, enjoys RfAs like ours. Basically every active admin sat around and watched as we were subjected to a variety of personal attacks and misrepresentations, and the few who tried to enforce civility norms that are considered common-sense anywhere else were instead accused of badgering or harassment. All this even though most of those personal attacks and misrepresentations made no attempt to tie back to the question of suitability to serve as an administrator. Because RfA isn't about suitability to serve as an administrator. It is, and has always been, a virtue test.

[Minority opinion] We shouldn't have RfA'd. It was not worth it. We learned a lot of things about this community that I'd have rathered we not learn firsthand.

I don't regret that we RfA'd. I do think it establishes a clear lesson for others, though: Don't RfA. You don't know what random thing will turn out to be the single aspect of your work that hundreds of people judge you by. More importantly, though, if you RfA and succeed, you delay what is long overdue: The scuttling and wholesale replacement of RfA as a system. That is what is needed. That is what future RfA candidates deserve. But the only way it will happen is if people say publicly that they're not comfortable running under the current system, and stick to that.

This creates a bit of a catch-22: To announce you won't RfA under the current system, you have to take a controversial stand, and that's something RfA candidates are usually discouraged from doing.

But I hope if we've accomplished one thing, it's to show that you can have opinions on shit and still pass RfA.

Who knows, you might even break our record.

Acknowledgments
I would like to thank:
 * My mother, for schlepping 16 hours to come keep me company as things got bad
 * My father (Z''L), for teaching me to stand up for what I believe in
 * My found family—BiomatrixBackup and her fiancée, who is too cool for Wikipedia—for letting me vent for hours
 * Drmies and BDD, for being the two best nominators a ! gal could ask for
 * TheresNoTime, for pushing me to run, and so many other things
 * Firefly, my would-be flightmate if not for some slowness on my part, a faithful ally before, during, and after this RfA
 * Tavix, Ritchie333, Trialpears, RoySmith, Vanamonde93, and one admin I spoke to only by email, for nomination offers I was forced to decline due only to a surplus of options. Maybe next time, y'all! (Joke. Please no.)
 * Vami IV, for providing me with the anthem for this slow-motion trainwreck, "Fear Is Not My Guide"
 * theleekycauldron, both honorary younger enby sibling and honorary Jewish mother
 * Floquenbeam (even though he jynxed everything), FeydHuxtable, Celestina007, casualdejekyll, Ixtal, Elli, Endwise, BusterD (also an honorary Jewish mother), zzuuzz, ezlev, Paradise Chronicle, Zippybonzo, ARoseWolf, WereSpielChequers, ScottishFinnishRadish, Doug Weller, Paine Ellsworth, Dreamy Jazz, Mythdon, I dream of horses, Serial Number 54129, and everyone else who reached out to me, publicly and privately, during the RfA, especially those who talked me out of withdrawing
 * Everyone who's congratulated me since the RfA passed, whom I'd name but it would exceed the ping limit
 * Each and every one of my supporters. In particular:
 * Nableezy, for getting it
 * MastCell, also for getting it, and for in doing so giving me the only good laugh I had all week
 * Every oppose and neutral that gave me constructive material to work with. There were many, and I will keep them in mind.
 * The entire English Wikipedia bureaucrat team, for their careful reading of consensus. At first I was grumpy when it went to a 'crat chat, but now I can say that 9 bureaucrats agreed that there was consensus. How many admins can say that?
 * In particular, Primefac, for consistently making sense of the chaos during the RfA
 * Every editor who makes this encyclopedia great
 * And of course, the cetaceans who swim in the waters off Cape May