Wikipedia talk:Requests for adminship

Archives

RFA discussions prior to June 2003 took place on a mailing list

Archived RFA votes
- Archived unsuccessful applications since April 2004
- Archived successful applications
Archived discussion about RFA
- For discussions from June 2003 to present see /Archives
- RFA discussions may also be at the Bureaucrats' noticeboard

[[:|WT:RFA]]

Predicting RFA

Being curious, I have been gathering data on the evolution of RFAs during the process. The figure at right is one pretty way of looking at that data. It shows the empirical probability that an RFA will succeed as a function of how many days it has been listed and what its support percentage is at the moment. As one can see, after 1 day an RFA in the 80-90% range is pretty much a toss-up. This narrows over time so that most RFA outcomes are pretty certain by the 5th day (or so). A sharp promotion threshold at 75% also appears to be indicated. The line at 60% is entirely a byproduct of the Carnildo 3 promotion (which stands out very starkly here).

So, if anyone is in the midst of running an RFA and wants to know there odds of success, feel free to point them at this. Dragons flight 04:47, 26 February 2007 (UTC)[reply]

PS. I forgot to highlight the (obvious) huge swaths of zero. There are large chunks of parameter space where it is fairly easy to predict that an RFA will fail. Dragons flight 04:51, 26 February 2007 (UTC)[reply]

(edit conflict) From this graphic, I only gather that Carnildo would have earned a nice $$$ bonus had he bet for himself :-) -- ReyBrujo 04:53, 26 February 2007 (UTC)[reply]

Very nice. Two questions on the data source - it says 'a year's worth' of RfAs, but how many is that? (Not too large if one outlier is so obvious, I guess?) And does the probability calculation include RfAs that were closed early/withdrawn? Opabinia regalis 07:21, 26 February 2007 (UTC)[reply]

Nice data report. This also corroborates that there is a certain amount of bandwagon effect/groupthink at RFA, especially from the less experienced users - people tend to pile-on whichever side seems to be majority view. I noticed this a long time ago - whatever the early majority says shapes the rest of the vote; sometimes early minority voters will even change their mind to go with the flow. It's only human - I admit I myself have a tendency to change my mind when I see everyone having the opposite view from my initial impression so I'm not just saying this to "pick on sheep voters". It's unlikely that there are any easy solutions to this though. —Quarl ^(talk) 2007-02-26 07:31Z

There's a correlation v. causation issue there. There may be some pile on effect simply because the initial few voters are likely to bring up most of the major concerns that will be raised in the RfA. JoshuaZ 16:34, 26 February 2007 (UTC)[reply]

Do you notice how the shape of the graph at 7 days end corresponds almost exactly with RFA guidelines? Also there's a thin vertical line around 0 that appears to correspond closely with WP:SNOW, which means maybe Badlydrawnjeff deserves some more attention than we've been giving them.

Basically, this graph shows how specified guidelines can shape a system. Do we have enough data to simulate how changing the system might shape the graph?

--Kim Bruning 16:29, 26 February 2007 (UTC)[reply]

Interesting graph. I surprised how wide the light red range at the end of the 7 days is - you need to have over 82% support to be sure of passing, it seems. Which RfA's have failed with over 80% support? Also - what kind of smoothing did you use of the data? I imagine the prominence of the line caused by Carnildo's RfA is because there aren't many RfA's with support in the high 60's after a few days (most will have moved one way of the other by then). I'm guessing you used some kind of constant smoothing (fixed width groups), maybe a variable smoothing (fixed population groups) would get a better graph, without the anomalies. --Tango 17:22, 26 February 2007 (UTC)[reply]

Since June of 2003, the highest raw support percentage RfA that was not passed as successful was Wikipedia:Requests_for_adminship/Kafziel_2 which closed with 77.08% raw support. That RfA was in November of 2006. Kafziel passed unopposed on 7 February 2007 on the 3rd RfA. --Durin 17:35, 26 February 2007 (UTC)[reply]
- I could certainly have used that fact when I wrote his 3rd nomination. :) Newyorkbrad 18:25, 26 February 2007 (UTC)[reply]
  - Wise man say, those who do not ask a question get no answer. :) --Durin 19:10, 26 February 2007 (UTC)[reply]
- That's a dubious distinction, but I'll take it. :) Kafziel ^Talk 21:11, 26 February 2007 (UTC)[reply]
  - Well, you have another dubious distinction, being either (1) the first candidate to be cheated out of a chance at WP:100 by an early favorable SNOW closing, or (2) the candidate on the RfA that proved we have a bureaucrat who can't count to 7. :) Regards, Newyorkbrad 21:18, 26 February 2007 (UTC)[reply]

Interesting... can I have a copy of the raw data you used? I'd like to run some tests of my own. Veinor ^{(talk to me)} 21:38, 26 February 2007 (UTC)[reply]

This might be useful. Majorly _{(o rly?)} 21:45, 26 February 2007 (UTC)[reply]

Perhaps it may be best to not have the votes displayed while an RfA is open then (if such a system is workable on Wikipedia)? It would require users to actually peruse the contributions of a nominee and arrive at their own decision instead of being influenced by popular choice. I will not lie, on some occassions I have been a little influenced by the vast majority of popular choice when I see many people voting to Support giving brief reasons, yet while there are a few Oppose with greatly detailed and supported reasons for deleting. --Ozgod 04:18, 28 February 2007 (UTC)[reply]

Don't forget, this is just a hobby

I must admit that it is very frustrating looking at the answers that some people give as to why others do not rate to be admins. They say in order to be an admin you "MUST" know about this, this and this. I say B.S. This should be a vote on fair and trustowrthy editors who have shown a true dedication to the project. It is safe to say that people did not spend months to years editing, reverting and discussing just to become dicks the first day they get extra tools. If they do, then there is a system in place to deal with that 0.001%. So what if they have not been active in AfD or CfD for the last few months or ever. If they are good editors then in time they might migrate there, learn the rules of the road and help out. Will anyone here not deny that their editing habits change over time depending on where they are bouncing around? We should be voting on an individual's performance, fairness, etc... Not the fact that they don't spend months going through image copyright tags. Give them time as admins and almost all will flourish in their roles. Vote the person and their performance and let them grow. They do not need to be experts at everything from day one.--Looper5920 08:38, 28 February 2007 (UTC)[reply]

I can't speak for others, but I judge based on their experience in the areas which they say they will contribute in. So while an admin doesn't have to participate in AfDs (this RfA seems to be going very well, for instance), if they then say they will start closing them, I think there's reason for concern. For me, it's a two-stage decision: firstly, do I trust this user enough to know that they won't abuse the tools; secondly, has this editor demonstrated competence to show that they won't (accidentally) misuse the tools. Trebor 18:57, 28 February 2007 (UTC)[reply]

Can we remove the trick question?

One of the questions says "What sysop chores do you anticipate helping with? Please check out Category:Wikipedia backlog and Category:Administrative backlog". Candidates who actually check out Category:Wikipedia backlog and base their answer on that will automatically attract oppose votes ("those are not sysop chores!"). Can we stop asking people to check out a backlog when they'll be opposed if they say they want to tackle it? What the question tests is mostly whether the candidate has lurked at WP:RFA and knows what people expect; we should try to encourage honesty instead. Can we remove the non-admin backlog from the question? Kusma (討論) 09:47, 28 February 2007 (UTC)[reply]

Go for it. Viridae ^Talk 10:14, 28 February 2007 (UTC)[reply]

Huh, I've never actually clicked on that link - I always assumed it linked to the admin backlog. Kill it! ;) Glen 10:48, 28 February 2007 (UTC)[reply]

I always assumed it was either a nod to the people who think adminship should be about editing, or a hint to candidates about how they could demonstrate that they should have the tools. It's pretty inappropriate anyway, though; it would make a lot more sense on the Community portal (I check... it's there already, on Template:Opentask). --ais523 10:52, 28 February 2007 (U T C)

I removed it. Kusma (討論) 10:56, 28 February 2007 (UTC)[reply]

It's not intended as a trick question. When the question was added, the "regular backlog" was the only one in existence. The "admin backlog" was split out later. Aside from that I find it rather silly to oppose users because they want to do cleanup>Radiant< 11:32, 1 March 2007 (UTC)[reply]

It's asking what they want to do as an admin though; any user can do cleanup. Majorly _{(o rly?)} 11:36, 1 March 2007 (UTC)[reply]

Amlder20

Hi, I just wanted to know if you guys think my work is fit enough to for me to become an admin eventually? I prefer getting opinions first before I decide. Amlder20 18:49, 28 February 2007 (UTC)[reply]

With an edit count of only 299, you would be opposed as being too inexperienced for people to be sure that you wouldn't screw up when using the tools. Read WP:GRFA, keep editing, and after a while you might be in a better position to request adminship. --ais523 18:52, 28 February 2007 (U T C)

If you want to see how you're doing so far, you may be interested in getting an editor review. Trebor 18:58, 28 February 2007 (UTC)[reply]

I've changed my opinion. Thanks for the input anyway guys. Amlder20 19:33, 28 February 2007 (UTC)[reply]

Minimum Edits

What is the minimum edit count needed to apply for Adminship and what do you look for in an Editor to be chosen as an Admin??--Cometstyles 13:06, 1 March 2007 (UTC)[reply]

There is technically no minimum edit count needed and anyone asking that is probably unlikely to pass. Most people look for 2000-3000 though. – Chacor 13:19, 1 March 2007 (UTC)[reply]

The last lengthly discussion was here. Cheers!--Kchase T 13:22, 1 March 2007 (UTC)[reply]

Have a look at Wikipedia:Guide to requests for adminship. --Durin 13:39, 1 March 2007 (UTC)[reply]

This chart puts my Success rate at O.8% which I believe is Quite good for a novice..--Cometstyles 13:45, 1 March 2007 (UTC)[reply]

First, that chart is a year out of date and other charts I have down show that standards slowly rise over time. That said, there isn't any edit count based reason why someone with 4,000 edits should be opposed. Which leads to the second point; don't focus on edit counts, either for yourself or others. Focus instead on how you help the project. If the ways in which you want to help the project are appropriate to an administrator, you've demonstrated the ability to be trusted with the additional tools granted to administrators, and you've demonstrated knowledge of Wikipedia policies, there isn't any reason you should not be an admin. However, there are going to be edit countitis addicts who will insist that your current 191 edits to the Wikipedia project space is too low. Focus on the positives. --Durin 14:04, 1 March 2007 (UTC)[reply]