Surge in RFA promotions: a sign of lasting change?
Eight administrators were promoted in the first 16 days of 2017, following four successful requests for adminship (RFAs) in December. This is more than any month since March 2011, which saw nine new admins. January has already seen half as many promotions as did the entirety of 2016.
From a large-scale research project in 2011 to several subsequent attempts at reform, the precipitous decline in admin promotions over the last several years has drawn the attention of the community. The Signpost investigated whether these reforms are working: whether this surge is the start of a trend, or just a temporary blip on the radar. We spoke with several recently promoted administrators, past candidates, nominators, and RFA observers to explore that question.
One common theme from their responses was acclaim for the optional RFA candidate poll (ORCP), a development championed by Anna Frodesiak that began in October 2015. It provides a forum for prospective administrators to put their names forward and receive appraisals from community members on their likelihood of passing. In 2016, of the sixteen successful RfAs, eleven (69%) had used the poll. Of 2016's twenty unsuccessful candidates, only two (10%) had first used the poll.
Montanabw, who ran unsuccessfully in September 2015, points to another improvement in the RfA process: watchlist notifications.
||The watchlist notice has brought about an improvement in the tone and level of participation ... There [previously] was a free-for-all atmosphere that allowed a high level of personal attacks and unregulated debate. The lack of any broad notice provision created a situation where a significant number of participants were people who regularly monitored RfA but who were not necessarily representative of the community as a whole—many people later told me that they had no idea I was even running! I think that the process now is handled with more respect for candidates.
Generally, the sentiment that RFA has become less nasty shined through many recent candidates' reflections. NinjaRobotPirate, who passed overwhelmingly earlier this year, commented that "a lot of the nastiness has been purged" from the process. Nevertheless, he described putting himself forward as a challenge, given that "nobody likes the idea of putting their contributions on trial."
RFA's reputation as a nasty experience for well-meaning contributors has certainly not helped encourage editors to take the plunge. Sam Walton, an administrator himself who has recently begun a mini-campaign of nominations that has yielded several new administrators, offered his insight on RFA's perception problem.
||You only need to glance at WT:RFA to see the endless debates and proposals on how to fix what users see as a broken, horrible, why-would-anyone-bother process. While I wouldn't argue that these discussions aren't useful, they do feed into what I see as RfA's perception problem, which directly results in a lack of candidates and I think only fuels the debated problems.
He went on to say that when RFAs are rare and only one candidate runs at a time, it fosters an atmosphere of intense scrutiny on that candidate, which can lead people to look for more reasons to oppose. He has sought with his recent nominations to curb the notion that "no one runs because no one runs" and to address the recurring challenge of getting more people to run by simply asking and nominating qualified candidates. He believes his recruitment efforts are "doing a good job of spreading some goodwill and improving the negative perception of RfA."
Another recent change that may reduce the time-consuming nature of running for adminship is the limitation on questions to two per inquirer, which was imposed in 2015. In all, 11 proposals reached the request for comment stage of a 2015 reform effort, and four of those proposals passed. (The others were the tweaking of promotion discretionary range, watchlisting, and the notice on the Centralized discussion template; while the pre-RFA candidate poll happened at a similar time as the 2015 reform RFC, this was coincidental.)
Two editors at the forefront of RFA reform efforts in recent years have been WereSpielChequers and Kudpung. Kudpung was a coordinator of the massive research endeavor and reform effort of 2011, and WereSpielChequers has compiled statistics and has written about the issue for the Signpost (here, here, and here).
WereSpielChequers doubts that the reform efforts of late 2015—including the optional candidate poll and the lowering of promotion criteria from 70–80% support to 65–75%—explain the recent surge: "the slightly lower threshold for promotion has made almost no difference; Those who do pass usually do so with near unanimity or at least a strong consensus. I fear that ORCP has succeeded in persuading more to come forward, but then deterred them from actually running. It is too early to say whether the January surge is a welcome but temporary rally or a change.".
Kudpung also doubts that the recent spike in successful promotions is part of a longer-term trend. In a statement to the Signpost, he suggests that RFA remains an inherently broken process:
||I don’t believe the current spate of RfAs is actually bucking the downward trend. It’s probably just a flash in the pan due to the hard work of those who scour the land for possible candidates and Anna Frodesiak's initiative at WP:ORCP. While the precipitous decline in 'promotions' is giving rise to concern, the same could be said about the state of many aspects of the Wikipedia even though the actual content is definitely growing. RfA still remains the horrible and broken process as described by Jimbo Wales and for the same reasons, and it’s clearly the main cause for lack of interest by potential candidates. In spite of all the talk of reform, however, people turn around and are suddenly busy with something else when the actual standard of participant behaviour is mentioned.
The work of admins has never really changed ... Occasional unbundling of one or two tools, such as for example rollback, hasn’t made much difference, but to talk of admin backlogs is really to create an illusion to illustrate the claim that we need more admins ... The future of the number of truly active admins is predictable. It will continue to be as it is, and the number of new RfA has bottomed out. At some time in the future—but not for a while yet—there won’t be enough admins. By then there will however be better bots and better helper scripts.
The various recent reforms brought about in good faith in December 2015 ... have ironically cancelled each other out, leaving the one single major problem still completely unaddressed, but with just more participants, more unnecessary talk in the discussion section, and despite the lowering of the pass mark, more ‘crat chats and more contentiously close-run bids for the mop.
Although many deride RFA as a broken process, those who have recently experienced an RFA have differing opinions.
K6ka noted that "the drama (of reading old RFAs) was enough to keep me at bay" before he ran successfully earlier this year.
"I got very little sleep throughout the week. I'd stay up till midnight watching the !votes on my RFA fluctuate unsettlingly before getting eight hours of terrible quality sleep, waking up at roughly eight-thirty in the morning or so to scroll through my RFA page again on my phone in bed," he said.
However, both he and Ealdgyth, who recently passed with 250 supports and no opposes, described their RFAs as enjoyable, a sentiment not shared by Hawkeye7, a former administrator who ran for the tools again in early 2016:
"RfA remains a tough process. There was a vicious off-Wiki campaign against me. It deters people from running, and it deters people from contributing to Wikipedia. I was heartened, though, by the editors who supported my candidacy—a veritable Who's Who of Wikipedia."
Ad Orientem, whose successful candidacy in late December may have springboarded a wave in nominations, summarized RFA as it stands today:
"My RfA was contentious and for me personally, highly stressful. Twice during the RfA I seriously considered withdrawing but was talked out of it … The problem with those who complain about RfA is that there seems little consensus or even serious suggestions for an alternative. As long as we need Admins (and we do!) we are going to need some system for vetting candidates. Like many Wikipedians I have heard horror stories about RfA before the series of reforms that began a few years back and I would like to assure readers that abusive and trollish behavior is generally not tolerated anymore," he said. "Overall I think the process is fair if imperfect." GP