Wikipedia talk:RfA reform 2011/Minimum requirement

Just a few thoughts here.

Firstly, we do not need a software prevention. If this minimum requirement gains consensus, any editor can un-transclude. That should be sufficient. Software prevention would be nice if at all possible, but should not scupper the proposal 09:44, 18 August 2011 (UTC)

Secondly, I've suggested a low bar to increase the chance of the proposal being accepted. If I've gone too low or too high, let me know, but I've explained my reasonings on the page.

Thirdly, there's a way out. My phrasing "users fulfilling the criteria may nominate users who do not" ensures that an editor with over 2000 edits can nominate an editor with less. And we have WP:REQUESTNOM for that!

Would appreciate feedback before taking it to the community. Worm^TT · (talk) 13:36, 15 August 2011 (UTC)

Providing this is done as a software fix then I'm happy to support it (I've tried and failed to get things very like this implemented in the past, but we need to do this and it is time to raise it again). But I would probably oppose if this was done on the basis of "anyone can close one of these as a snowfail/invalid nomination as I see tat as even more bitey than our current system. Taking Jimmy's example from Wikimania, it is really important that the Software doesn't allow people to do things that they are not allowed to do. ϢereSpielChequers 17:33, 15 August 2011 (UTC)

WereSpielChequers said a long time ago that if we were to start with a relatively uncontroversial change first rather than go for a whole bunch of changes in one go, it might pave the way for further reform. I concur with that rationale, and it seems to have had some consensus at WT:RFA2011. However, for policy changes requiring software interventions, recent experience has shown that the devs now have a policy of refusing to implement software changes they personally don't like, even ones called for by an overwhelming consensus from heavily subscribed central RfAs. So let's not imagine for a moment that any changes, however moderate, will be easy. Neverthless,

That said, I would support this, as our first proposal for change, on the basis of:

• Minimum of 2,000 edits excluding userspace, AWB and Huggle
• Minimum of six (6) consecutive months of activity since last beak of 30+ days.

Any RfC would have to be extremely carefully worded. Again, experience shows that participants in discussions of this type, particularly the detractors, will try as hard as possible to misquote the syntax or put their own spin on the proposal. Kudpung กุดผึ้ง (talk) 20:11, 15 August 2011 (UTC)

I understand WSC's reservations, and talking to the devs would be a good thing - if a software fix is possible, fine, if not it does not negate the rest of the proposal. I'd rather we got consensus that a minimum is the right way to go and then take that to the devs to implement a software fix. Having said that, is there any way to ask the devs if a software fix is possible or likely?

Regarding the provisos, Kudpung - I do agree with them in principle, but I think it's over-complicating the proposal. It's very easy to see the 2000 mark, similarly the tenure since registration. A year ago, I agree that a much more carefully worded RfC would be required, but I think that especially with Jimbo's recent "we require facts, not emotive responses" argument, it may not be as disagreeable as you may think.

I've asked Swarm to put out a shout for the masses, so we can see what the rest of the task force think. Worm^TT · (talk) 21:03, 15 August 2011 (UTC)

New Proposal (and progress on an old one)

Note: Copied from Wikipedia talk:RfA reform 2011

I've put together a proposal at Wikipedia:RfA reform 2011/Minimum requirement. I'd like to know what people think of it, feel free to copyedit it. I think that we have a real chance to make a change with this proposal, and whilst it's not a grass-roots radical change it is a significant step in the right direction. As I said, we've already got 2 steps out of this reform, and I think it'd be good to have a proper proposal or two to come out. Why do I say "or two"? Because I see Swarm has been pushing the clerks proposal so it's looking a lot closer to something we could put before the community. Worm^TT · (talk) 14:58, 15 August 2011 (UTC)

I'll support it. Of the 20 SNOW/NOTNOW closes, were any even close to 2,000 edits? If you can take the edit count lower, I'd go as low as you can so it's obvious we're just trying to cull the very new users that likely don't understand RfA yet. — Bility (talk) 16:37, 15 August 2011 (UTC)

13 had less than 1000 edits, 2 between 1000 and 2000, and 4 had between 2000 and 3200 but less than 6 months making a total of 19. I liked the 2000/6m mark because it has parallels with WP:Service awards. By the way, if we upped it to 3000, we'd take out 26 failures... but 2 were not NOTNOW or SNOW. Worm^TT · (talk) 16:51, 15 August 2011 (UTC)

Under this proposal, presumably the two that weren't SNOW/NOTNOW (or all of them maybe :P) would request a nomination and either be let through anyway or they'd be (nicely) encouraged to wait until they had a better chance. Sounds like a good system to me, but you're probably right that 2,000 edits will be more palatable for the community. — Bility (talk) 17:11, 15 August 2011 (UTC)

I was just thinking a proposal like this yesterday. This is more than fair and exactly what I would suggest. Swarm ^{u | t} 17:00, 15 August 2011 (UTC)

I was thinking more like 3 months and 1000 edits just because it may be hard to set a bar where there has been none before; however, the ability to be nominated should solve the problem. I think we should be willing to back down to the lower bar if there is heavy opposition to 2000 edits/6 months. Ryan Vesey Review me! 19:23, 15 August 2011 (UTC)

Can we move and continue this discussion on the right page at WT:RFA2011/CANDIDATES to keep it all together? Kudpung กุดผึ้ง (talk) 17:14, 15 August 2011 (UTC)

I don't know how easy it is for the regulars here to follow the discussions, but for me (as a new member) it is easier to follow if all discussion is centralized. Maybe some super genius can figure out a way to transclude just a single section. Ryan Vesey Review me! 19:23, 15 August 2011 (UTC)

It's quite easy, and it's built roughly on the lines of any Wiki project. It has a master page that clearly identifies and explains the individual problem items to be resolved, and recommendations for addressing the work to be done. Its talk page discusses broad project issues and management of the project. The project main page also includes a project map (in a pink box, a list of links to the sub pages and their talk pages) where the individual problem items are discussed, and researching discussed on one long talk page. That was the problem at WT:RfA where it became impossible to locate unconnected threads and scroll back to stuff that magically gets archived. Nobody sifts through archives, and that's the main reason why the same topics get dished up over and over again with no real action being taken. With a project of this kind, active participants, especially the task force, will keep all the pages and their talk pages on their watchlists. Kudpung กุดผึ้ง (talk) 23:49, 15 August 2011 (UTC)

Ok, wikilinking a collapse box helps a lot too. Ryan Vesey Review me! 00:38, 16 August 2011 (UTC)

A corollary to this

I think that a corollary to this proposal, if it is introduced, should be that oppose rationales along the lines of "2,001 edits is not enough" or "six months and one day is not enough" should be accorded no weight, and possibly removed, because they would be forum shopping (i.e. if they really think that they should seek to have the minimum requirement increased rather than bringing up at an actual RfA). James500 (talk) 22:47, 15 August 2011 (UTC)

I disagree, many editors have different beliefs as to how many edits an admin hopeful should have. The minimum edit count is supposed to reduce SNOW and not now closes. Ryan Vesey Review me! 02:54, 16 August 2011 (UTC)

Absolutely. When setting the bar for this, it would also be prudent to take into consideration the average minmum edits required by the !voters who have published their criteria. (There's already data for that too). --Kudpung กุดผึ้ง (talk) 13:15, 16 August 2011 (UTC)

So you agree with me? Ryan Vesey Review me! 13:23, 16 August 2011 (UTC)

I agree that the bar is intended to prevent time wasters from transcluding. I agree that this project's goal of setting a bar (which incidentally most other Wikipedias already practise) is in no way intended to raise the qualifications for the serious applications from mature, experienced candidates of the right calibre. I confirm that these objectives have been the goals of this project since it was first started in user space. I have not fully understood the purpose of James500's suggestion above - it does not appear to simplify the proposals that are being made, does not seem to take into account all the research that has been made and provided in tabular form, and does not look as if it is compatible with Worm's proposal. Maybe I have missed something.
IMHO, I think it's time to reach a local consensus for the minima for edits and tenure, agree on a short, clear, unambiguous RfC statement, and let the broad community decide. I will repeat however, that great caution should be exercised in the wording of such a proposal because we are dealing with a perennial issue, and detractors will argue more on the syntax (commonly known as 'splitting hairs') than on the merits of the suggested new policy, and will almost certainly confuse the RfC by introducing their own alternative ideas that should be the subject of a completely separate call for consensus. --Kudpung กุดผึ้ง (talk) 15:41, 16 August 2011 (UTC)

I'll try and work out a way to make it clearer, but the point of this proposal is to stop "no-hope" RfAs from seeing the light of day. So, whilst 1999 edits is definitely not enough, 2001 is also not going to be enough. I would suggest that we "recommend at least double that" or "recommend many thousands of edits" - to make it clear that this is meant as a bright line, rather than an acceptable point. Also, we need to keep this proposal simple, as it's the only way we stand a chance of any change. Worm^TT · (talk) 08:38, 17 August 2011 (UTC)

Good point, we should make it clear that meeting the requirement does not guarantee a successful RfA and several thousand edits are still recommended. Swarm ^{u | t} 23:50, 17 August 2011 (UTC)

For the sake of simplicity and certainty, I would prefer a policy that said "X is the acceptable number of edits". Failing that, I would prefer an actual number of edits to be recommended rather than "many thousands" or "several thousand" (the sort of remark that I would stick a {{quantify}} tag on if it appeared in an article). James500 (talk) 06:51, 18 August 2011 (UTC)

It's a nice idea, but it doesn't play out.

1. Everyone has a different number of edits they consider a minimum before they'd support (Kudpung aims at 6000 above, I'm comfortable at 4000, one essay suggested that an editor should have 10000 edits) - There's more at Wikipedia:RfA reform 2011/Candidates.
2. We don't want to make adminship about edit count, as there's many ways to inflate your edit count (mentioned again by Kudpung in his suggestion). It will also promote WP:EDITCOUNTITIS.
3. Adminship is about experience, and my theory is that under 2000 edits, you will not have enough experience to even understand fully what adminship is about. After 2000, you may still not have enough, but the number who do not drops significantly. WP:NOTNOTNOW implies that you should not use NOTNOW on editors who have thousands of edits and many months of experience.

The fact is that there is no figure that applies to every candidate, nor to every voter. Giving a figure is therefore misleading. Worm^TT · (talk) 08:42, 18 August 2011 (UTC)

There will always be exceptions, for one reason or another. I do notice (occasionally, when I'm in a noticing mood) that there are cases where people feel like going for RfA but have to (or feel they have to) spend some time just 'racking up edits' before they want to go for it. Adding numbers just for the sake of it doesn't always work, though, and this is possibly something that can be pointed out to candidates. Smaller numbers of high-quality edits should count for more - otherwise it's easy to save a page after each and every gnomish tweak, rather than doing a load of useful stuff and then saving once. There's always going to be a certain amount of 'gaming' on the editcountitis thing, which is why some people seem to get through on much fewer edits than others. I like the idea of minimum standards, but I'd also like some kind of 'numbers are not enough' clause (which I appreciate can't be done with software, probably). Pesky (talk …stalk!) 07:17, 26 August 2011 (UTC)

We probably need to remind ourselves - and any newcomers to his project - that the figure we are are mentioning here (which seems to be around 3,000) is designed to prevent the blatant SNOW/NOTNOW, which is the only item up for discussion. The change could be done quite easily by a software script; however, recent experience on other major policy changes has shown that even with an overwhelming consensus in favour of a change, the devs who have access to the site's source code unilaterally refuse to implement software requests that they personally don't like.

We also need to avoid creating ambiguity here by confusing this suggested 3,000 NOTNOW bar with !voters' criteria such as my 6,000 for candidates whose RfAs have a chance of staying the 7 days. It's not the same thing. Kudpung กุดผึ้ง (talk) 22:36, 27 August 2011 (UTC)

On second thoughts

If it is to be kept as simple as possible without conditions for the kind of edits, after having had another look at Wikipedia:RfA reform 2011/Unsuccessful candidates' edits and doing some quick sums, I think we could, and should increase the limit a bit. Although it's not really useful to cite averages, the average number of edits among the failed RfA for 2011 is around 12,900 per candidate, but this also includes very high edit counts for a few regular editors who failed. However, out of the around 45 first-time candidacies, 25 had less than 3,000 edits. Therefore I see a perfectly reasonable suggestion to set the SNOW/NOTNOW bar at 3,000 total edits. Any RfC proposal text must make absolutely clear that the bar is intended to prevent blatant obvious failures that waste both the candidates' and our time, and not to set a hurdle for serious candidates.— Preceding unsigned comment added by Kudpung (talk • contribs)

I disagree, since there has been an editor who had a successful RFA with less than 3,000 edits, I think that is too high of a bar. Ryan Vesey Review me! 02:52, 16 August 2011 (UTC)

Could you provide a link please, and establish whether or not it was a special case, such as a bot request etc. According to Worm's research, I seem to remember there was only one user who passed with less than 3,000 edits in a very long time. I would not think that (+/- 0.5% of recent RfAs) would be enough to make an exception. Tunning for office. No Wikipedia guidelines or policies are graven in stone, but the effort is to encourage potential candidates to read pages such as WP:RFAADVICE before wasting their own and anyone else's time. This is one of the reasons why a minimum requirement should be taken into policy where IAR could be invoked for genuine exceptions, and why I don't think a software block, as suggested by WereSpielChequers, is entirely neccessary. It's worth noting that the average number of edits from truly successful editors is actually around 12,000 - 13,000. Let's not forget that I raised my bar above because it was suggested we should not specify the kind of edits. It would probably be most inappropriate, for example, to consider a candidate whose only edits were 3,000 'Wikify' tags per Twinkle on new stubs, or even 30,000 edits exclusively to remove vandalism with Huggle. It's my view that the results of RfA should be based on some intelligent metrics of performance (5.5 minutes work per day for 6 months):, rather than 'I like him/her' or 'I don't like him/her', or, as is sometimes clearly the case with some !voters: 'I don't like admins'. --Kudpung กุดผึ้ง (talk) 13:06, 16 August 2011 (UTC)

I have no link, only the page WP:RFA Guide which states that in 2008 one candidate passed with less than 3,000. I find it hard to stomach a minimum proposal of 3,000 edits when that could conceivably keep people who would otherwise pass from running. (And yes, I am aware that people who do not meet the requirements can ask for a nomination) Ryan Vesey Review me! 13:22, 16 August 2011 (UTC)

Stats have been posted on this project for total edits, and article space edits of all failed candidates for 2011. This is the number of RfAs since 2008:

• 2011 unsuccessful 67. successful 38. Total RfA 105
• 2010 unsuccessful 155. successful 75. Total RfA 230
• 2009 unsuccessful 234. successful 121. Total RfA 355
• 2008 unsuccessful 392. successful 201. Total RfA 592

• Total: unsuccessful 848. successful 435. Total RfA 1,283

One out of a total of 435 successful candidates passed with fewer than 3,000 edits. If you have moment, you can continue the research and create a table for the others. This would establish once and for all whether a bar of 3,000 edits (of any kind) is too high. --Kudpung กุดผึ้ง (talk) 16:16, 16 August 2011 (UTC)

Honestly, I have absolutely no interest in most of the enormous mass of statistics that get posted in the realm of this task force, so I won't have a moment. Ryan Vesey Review me! 16:31, 16 August 2011 (UTC)

Some people like messing with stats. I don't - it's one of the most boring tasks I can think of. However, we've spent hours providing stats simply because we were told from on very high that if RfA reform is to be taken seriously at all, it would need to be based on numerical evidence. That's the way it is I'm afraid. My suggestion above was largely rhetorical - I actually feel that we have already provided more than enough stats, especially to support an argument for a bar of 3,000 edits which I would readily support. I would personally like to see it even higher (I don't support candidates who have less than 6,000), while still bearing in mind that we are still only talking about reducing the number of blatantly unlikely candidates, who should be encouraged to read the advice pages, instructions, and edit notices first. --Kudpung กุดผึ้ง (talk) 17:30, 16 August 2011 (UTC)

I mention the editor who passed with less than 3000 in the proposal, he had 1 million edits cross wikis. 1 million - that's certainly an exception to the rules. And I did consider 3000 seriously as an option. However, below 2000 we only have SNOWs and NOTNOWs, between 2000 and 3000, we have some editors who went the full length. Being able to say we can reduce SNOWs and NOTNOWs by nearly a third and not affecting other editors is a key selling point of the proposal. Oh and I'm the stats man, I'd happily fiddle with more, I just am too busy and think we have enough here, confirming that this bar is appropriate from 4 different directions. Worm^TT · (talk) 08:46, 17 August 2011 (UTC)

Ahh, I see now. The RFA was extremely unconventional. I wouldn't oppose an increase to 3000 then if that was the way the community wanted to go. Ryan Vesey Review me! 14:36, 18 August 2011 (UTC)

We should probably make an exemption for admins on other language Wikipedias, just because there's a precedent for that. Swarm ^{u | t} 04:25, 19 August 2011 (UTC)

I think 3,000 is a perfectly reasonable minimum bar - but is there some way that cross-wiki edits could actually be counted-in, by the software (if we're having software help on this)? Pesky (talk …stalk!) 07:21, 26 August 2011 (UTC)

Restricting this to self noms leaves wiggle room for special circumstances such as candidates with lots of contributions on other projects. ϢereSpielChequers 09:02, 5 September 2011 (UTC)

The proposed number of edits

As far as I can see, in December 2003, candidates who had 1100, 811, 1000, 943, 270 in the mainspace, over 1500 and 300 edits were successful. I am not convinced that the bright line should be set at 2000 edits. I suggest that it might be set at a lower level. James500 (talk) 22:07, 19 August 2011 (UTC)

Is this a serious comment? The encyclopedia was just over 2 years old at the time - most of the original posse were instant admins, there were no automated tools, the idea of having 2000 edits was impressive in itself. According to this File:EnwikipediaArt.PNG there was less than 100,000 articles. The encyclopedia is a vastly different place than then, it's very different to how it was 3 years ago, let alone 8. Worm^TT · (talk) 22:27, 19 August 2011 (UTC)

• The fact that the encyclopedia is bigger and older does not mean that candidates need to have more edits. That is sophistry.
• Your proposed figure does not appear to take automated tools into account (it appears to say that a person could have 1500 non-automated edits like one of the candidates, who, as far as I can see, is still an admin, did, or for that matter have 1900, and still be unacceptable), and the fact that there weren't automated tools then is not therefore relevant.James500 (talk) 23:31, 19 August 2011 (UTC)

In December 2004, the majority of successful candidates had between 1000 and 2000 edits. In December 2005, a substantial proportion had between 1000 and 2000 edits. So that is another two years. I haven't looked at 2006 yet.James500 (talk) 00:25, 20 August 2011 (UTC)

I'm sorry, I was trying to demonstrate how much the community had changed. If you had examples from 2008, I'd be more interested, but I think the fact that we have nearly 3 years without an administrator having less than 3000 edits is very telling. Worm^TT · (talk) 06:44, 20 August 2011 (UTC)

In December 2003 they probably had the sum total of all edits on Wikipedia. Ryan Vesey Review me! 22:31, 19 August 2011 (UTC)

It has become harder for candidates who have less than 3000 edits to be successful at RfA. Is this due to a change in the actual risk, if any, associated with appointing an "inexperienced" admin, or is it due to demographic changes in the community itself? Is it due to a change in what the community is !voting for, or a change in who is voting? James500 (talk) 09:36, 20 August 2011 (UTC)

I think that if you were to read the earlier parts of these discussions and to review some of the tables that have been provided, you might find the answers to your question. Kudpung กุดผึ้ง (talk) 10:55, 20 August 2011 (UTC)

What should the bar be?

I've seen a lot of discussion on this page regarding the bar, so I'll put up a quick straw poll. In my head it's between 2000 edits and 3000 edits, though please do feel free to add any other figures. I'd appreciate any thinking behind the number you choose too, so we can put the arguments forward when we take this before the community. Worm^TT · (talk) 07:56, 26 August 2011 (UTC)

500

• This could be acheived by nominating 250 articles for speedy deletion (i.e. 250 taggings + 250 user warnings). That is 50 more than the number of new pages patrolled specified by Kudpung's criteria. I am happy to have admins who only work in a single area and who only have experience in that area.
• If this is a bright line it should represent a bare minimum number.
• Candidates have in the past been sysopped on the strength of fewer edits than this.
• This is the minimum number of edits on the Italian Wikipedia.
• This recommendation is provisional.

James500 (talk) 08:43, 26 August 2011 (UTC)

1. Strongest possible oppose. Far too low edit count, and conditions too complex to ever gain a concensus at a central RfC. --Kudpung กุดผึ้ง (talk) 23:10, 27 August 2011 (UTC)

   From your point of view, this number would be an improvement on the present situation, where there is no minimum. The lower the proposed minimum number is, the more likely it is to gain consensus. For the avoidance of doubt, I wasn't suggesting that 250 CSD tags should be made a specific requirement, I was merely observing that it doesn't necessarily require more than 500 edits to do all that tagging. James500 (talk) 00:17, 28 August 2011 (UTC)

   500 is much too low. The speedy deletion example is far fetched and I still wouldn't support the candidate because the tools are bundled and being an admin is much more than just CSD tagging. Ryan Vesey Review me! 05:21, 28 August 2011 (UTC)

1000

1. This is the minimum on the Dutch Wikipedia. James500 (talk) 07:22, 29 August 2011 (UTC)

2000

1. As this matches the Journeyman award, there is only one successful candidate that does not meet the bar in the past 3 years. It's also high enough to remove all NOTNOWs whilst remaining lower than the user essays average. Worm^TT · (talk) 07:56, 26 August 2011 (UTC)
2. For the correlation with service awards and the available workaround of receiving a nomination. I think 2000 is a good number and I can't even imagine an editor with less than 1000 edits receiving a full RfA. Ryan Vesey Review me! 19:39, 26 August 2011 (UTC)
3. Providing we still have the safety valve of allowing nominated candidates I would be OK with anything up to 2,000; If only because I know that the community won't seriously consider self noms with this few edits. ϢereSpielChequers 17:56, 27 August 2011 (UTC)
4. Yep, this is pretty much the lowest bar I'd personally set. Pesky (talk …stalk!) 09:22, 28 August 2011 (UTC)
5. This would be a good minimum. I think there should be some clarification that the edits should be spread over at least 6 active months, with an active month being > 100 edits. — Preceding unsigned comment added by My76Strat (talk • contribs)
6. I like this. I need to look back at my work but this is close to the mean figure I lifted from user essays--Guerillero | My Talk 23:19, 6 September 2011 (UTC)
7. Sounds good. Any lower and too few poor candidates would be removed; any higher and we run the risk of losing good candidates. Nominations by other users should be at any level. Alzarian16 (talk) 13:28, 7 September 2011 (UTC)
8. As second choice. This is adequate and I wouldn't have trouble supporting, though I think it's a bit low.— Preceding unsigned comment added by Swarm (talk • contribs)

3000

1. Strong support. Let's not forget that this is a global figure and includes all the low value edits, such as for example, those to own user page construction, and automated edits. I think 3,000 is optimal, based on all the stats and analysis the active members of this task force have taken the trouble to provide, read, and review. --Kudpung กุดผึ้ง (talk) 23:16, 27 August 2011 (UTC)

   (Redacted) Ryan Vesey Review me! 23:54, 27 August 2011 (UTC)

   I apologize for the nature of my comment. Ryan Vesey Review me! 00:59, 28 August 2011 (UTC)

   Reply to Kudpung: It is not a given that there will be low value edits. A user page can be constructed with one edit and does not explain the size of this number. In any event, some users do not have one. Candidates may not necessarily use automated tools.James500 (talk) 00:32, 28 August 2011 (UTC)

   3,000 is a figure that is based on statistics. It takes into account that users of the SNOW/NOTNOW kind are generally newbies. It is logical to assume that most people who aspire to adminship will have a user page, and that early on in a user's career a large number of edits may be made while creating a user page. Please provide stats over at least the last 12 months that show the number of user page edits and automated edits for all failed RfA , so that we can examine this more closely. Kudpung กุดผึ้ง (talk) 02:34, 28 August 2011 (UTC)

   I am willing to collect the statistics, but they are probably not relevant. What I am saying is this: If a candidate transcludes who does not in fact have a user page (or who in fact created with one edit) and who has not in fact made any automated edits, is this minimum number, which you are telling me is based on assumptions about the likely number of user page and automated edits of an average user, going to apply to that candidate? Because it shouldn't, because that would be ignoring relevant facts. If you are worried about trivial edits of one form or another, the thing to do would be to actually exclude them from the equation altogether, as you suggested above. James500 (talk) 05:06, 28 August 2011 (UTC)

   I wouldn't oppose 3,000, but I think 2,000 is a much easier number for people to swallow. Ryan Vesey Review me! 05:22, 28 August 2011 (UTC)

   People have got through with barely more than 3,000, a candidate who used the preview screen a lot and had made far less than 3,000 quality edits could still pass, so this would be way too high. It would also make the proposal unsellable to the community - maybe in a year or two if expectations have risen to the point where the community will no longer consider such candidates, but this isn't needed now. ϢereSpielChequers 09:33, 28 August 2011 (UTC)

2. Support if edits in other wikis could be taken into account, at this level, I think. Pesky (talk …stalk!) 09:24, 28 August 2011 (UTC)
3. Support: The enwiki is the most important wiki in the world, the policies are complex and the minimum bar should be at least 3,000. No need to waste people's time with RfAs which will not succeed, because of SNOW/NOTNOW. There may be some wikidragons, who will be able to build 10 GAs in 100 edits, and in the other 800 do kind of admin work, but I still have to see them. The bar can be lowered if the candidate is already an admin in a non-English wiki: 1,000 would be sufficient for that. With this proposal everyone in this list below would have been qualified. Divide et Impera (talk) 14:34, 7 September 2011 (UTC)
4. Support per precedent. The community will not accept a <3,000 RfA candidate ever. There's no need for a bar lower than this. Swarm ^{u | t} 18:38, 7 September 2011 (UTC)

   Are you sure? They accepted one with only 76 edits more just 18 months ago. Worm^TT · (talk) 19:04, 7 September 2011 (UTC)

   True, but those situations are outliers, and even if a similar candidate came along and they didn't meet the bar, it wouldn't kill them to accumulate 3,000 edits on ENWP. I personally feel 3,000 edits is below the minimum amount of experience an admin needs, but aside from that, this seems like the most reasonable level based on the statistics we've gathered. I must say, I don't understand the sudden wave of support for a significantly lower level. Swarm ^{u | t} 20:12, 7 September 2011 (UTC)

   Well, there's no reason there can't be further outliers, and this is meant to be palatable to the community AND stop the NOTNOWS, it's a bright line, not a "what we expect". Worm^TT · (talk) 19:02, 8 September 2011 (UTC)

   Since January 2009 (at least), only one user with less than 3,000 edits passed (extremely unusual but we can easily address sysops on other projects). RfA history tells us <3,000 edits = , does it not? Why then, should we allow <3,000 candidates to run? (Not a rhetorical question... :P) Swarm ^{u | t} 19:52, 8 September 2011 (UTC)

   Just because it hasn't happened doesn't mean it can't. The average for essays is below 3000, there is little added benefit to increasing to 3000 (an extra 5 candidates would have been denied - 36% vs 40%). I see this as stopping candidates who are are complete no-hopers, not just people we "don't expect to pass" Worm^TT · (talk) 15:20, 9 September 2011 (UTC)

   What about this? ~3,000 edits is the level I "don't expect to pass", but I would be more than willing to let them run. The precedent for the past several years has been that 2000-2999 candidates are complete no-hopers, hasn't it? (P.S. I don't know if you noticed my comment above, but I would still support level of 2,000, just to be clear.) Swarm ^{u | t} 18:00, 9 September 2011 (UTC)

5. Support per Divide et Impera and Swarm (including Impera's exception for candidates who are already an admin elsewhere). A reasonably high bar may deter some !voters from objecting to a candidate who makes the cut, on the grounds that he/she still has too few edits. I fear 2,000 would be too low to accomplish this — it would keep out the green newbies, but it wouldn't do a thing to address people's concerns about other candidates with "only" a few thousand edits. (Stated another way, if we propose that 3,000 edits is enough and that "too few edits" objections are out of order for anyone who has this minimum, that's got a reasonable chance of being accepted — whereas no one, IMO, will go along with such a pronouncement if the bar is only 2,000.) Richwales (talk · contribs) 20:02, 8 September 2011 (UTC)
6. Support: The best filter. ~~Ebe123~~ (+) ^talk
   _Contribs 18:28, 10 September 2011 (UTC)
7. Support: I completely agree with Swarm. We might as well kill two birds with one stone here and try to avoid the "you don't have enough edits" comments. Mato (talk) 12:49, 19 September 2011 (UTC)
8. Support. Seems a reasonable number, but should be rubbery according to circumstances, not hard and fast. 09:16, 20 September 2021 (UTC).

Oppose any method of restricting who can run

I thought I'd leave this here to be safe and figure out if there was any opposition of the proposal. (Note that I am not opposing) Ryan Vesey Review me! 19:41, 26 August 2011 (UTC)

Comment: Irrelevant section. --Kudpung กุดผึ้ง (talk) 22:55, 27 August 2011 (UTC)

I do agree with Kudpung, the proposal can be shot down by the community, and any view here should be brought up there. The idea of this page is forming the proposal using feedback. Worm^TT · (talk) 17:42, 30 August 2011 (UTC)

1. I realize that this addresses a perceived enormous problem, but I don't think the problem is that big. The RFAs that everyone gets upset about are not the ones that this would prevent, I think. On the other hand, this requirement would have prohibited MGA73 from becoming an admin, which—as far as I can tell—would have been harmful. This seems to me like a net negative. ErikHaugen (talk | contribs) 17:35, 30 August 2011 (UTC)

   No, it wouldn't Erik - because he could have asked any user to nominate him. With over 1million edits on other wikis, I'm sure someone knew him. Even so, we'd lose one good admin (as an admin) and keep an extra half a dozen or more editors who quit after an RfA - I disagree this is a net negative. Worm^TT · (talk) 17:42, 30 August 2011 (UTC)

   Yes, pardon for being so dramatic with "prohibited". But my real point is that I don't think this is that much of an issue. It seems unlikely that many editors with 100 edits who quit after a failed RFA would have stayed around much longer if their RFA was met with "you don't have enough edits to run" instead of NOTNOW, which links to a page saying essentially that they don't have enough edits to run. ErikHaugen (talk | contribs) 20:07, 30 August 2011 (UTC)

   True. And I would be pushing for a software fix to stop editors who do not meet the criteria from doing it - but that's the next step, got to take one at a time. Worm^TT · (talk) 20:16, 30 August 2011 (UTC)

Discussion

Just to remind people that this designed is a "bare minimum" or "bright line", to stop editors who have no idea about how wikipedia works from applying. However, it's worth considering that the average number of edits of active edits is likely going up, so we need to consider what is a reasonable bar. Worm^TT · (talk) 07:56, 26 August 2011 (UTC)

I wouldn't be concerned that the average is going up - that's somewhat skewed by some very high edit count characters. The key thing to look at is the number of edits below which the community won't consider someone, as per your table below. If we still have successes in the 3,000 - 4,000 range then the community is clearly willing to look at such candidates. ϢereSpielChequers 18:01, 27 August 2011 (UTC)

Also, for relevance, the 11 editors who were promoted with less than 4000 edits since 1 January 2009 are:

Candidate	Number of edits	Date
RHM22	3894	19 April 2011
Leyo	3000	25 October 2010
MGA73	1358	6 July 2010
Jamesofur	3552	22 April 2010
Taelus	3076	2 February 2010
Floquenbeam	3565	19 January 2010
Maedin	3393	28 August 2009
Fribbler	3393	13 August 2009
Colds7ream	3523	9 June 2009
Laser brain	3493	11 May 2009
Nja247	3683	14 February 2009

Some of these were not "one offs" and for example, Taelus, 18 months ago, was a self nomination with 3076 edits. Worm^TT · (talk) 08:33, 26 August 2011 (UTC)

Thanks to some hard work by Kudpung, we now have 12 months of data on failed candidates. A quick glance through the 114 failed candidates stats shows that we would stopped 21 (18%) candidates if the bar was 1000/3m, 41 (36%) candidates at 2000/6m and 46 candidates at 3000/6m (40%). Bearing in mind the fact that the closer we get to candidates who have passed in the past, the less likely the proposal will get through, I see 2000/6m is a sensible choice. Worm^TT · (talk) 15:01, 30 August 2011 (UTC)

Implementation

It seems pretty clear to me that most candidates who fall short of the minimum requirements being discussed here are not reading the warnings designed to dissuade candidates who have no chance. If these warnings/etc are simply amended to say that those with fewer than 2000 edits/6months service are not allowed to run, then that will likely not change any behavior, since it seems reasonable to conclude that such a rule would not even be read. If this proposal is implemented, how will the experience for the new editor be different? Will it simply be that anyone can close the RFA immediately, sparing the would-be candidate a few "oppose: you don't have enough edits" votes that they might have otherwise been subjected to? Is there some other way that we can even prevent transclusion? A bot? Fully protect WP:RFA and make self-nominators request transclusion on WT:RFA? Thanks! ErikHaugen (talk | contribs) 20:17, 30 August 2011 (UTC)

Good question. I would say that the warning should be amended, with policy to back things up, transclusions could be undone easily. The theory is, if a user has a specific number to reach before RfA, they'll stick around to hit it. By the time they hit it, they'll actually understand about Wikipedia, and what RfA really means and won't just go for it. In the future, I'd be looking to get a software implementation of some sort, but I'm not sure how easy that is - would need discussions with developers. If 28bytes had a policy to back him up, say, here, today's NOTNOW wouldn't have happened. The candidate even asked for criteria! Worm^TT · (talk) 20:27, 30 August 2011 (UTC)

To me this is fundamental. If all this just means that people's RFA's get summarily discarded by other editors without any consideration other than edit count and tenure, then I see RFA as just as snarky, but slightly more efficient in its snarkiness. On the other hand I think this would be very positive if it was implemented by the devs so that you can't do what you are not allowed to do, and an attempt to submit an RFA brings up polite message saying "I'm sorry but only editors with at least 2,000 edits and six months tenure can nominate themselves or others for adminship". ϢereSpielChequers 23:56, 3 September 2011 (UTC)

A script is available here that could easily be adapted for users and candidate's to check on candidates' eligibility. This script could also be automatically triggered by an attampt to transclude.

The auto messages don't have to be snarky though. If accompanied by a link to WP:RFAADVICE which the cndidates almost certainly won't have read (if indeed they have read anything at all), they'll then realise that even with 2,001 edits adminship still isn't a fait accompli . Kudpung กุดผึ้ง (talk) 02:05, 5 September 2011 (UTC)

I like that idea; if we can word the message gently but firmly, it will greatly help deterring clearly unqualified candidates without discouraging them. And by the time they get to 2000 edits, they should have a sense for whether they're truly qualified. The Blade of the Northern Lights (話して下さい) 02:02, 7 September 2011 (UTC)

Implementation (again)

My personal support for this proposal would likely rest largely on whether or not it was able to be implemented by the software. Is there a chance that a decision could be made as to whether or not this is possible, and following this, that the choice of implementation is included in the proposal before it is put before the community? Mato (talk) 12:43, 19 September 2011 (UTC)

I expect there is a chance, yes - but I'm not exactly certain of the burocracy around getting the devs involved. In fact, the only way I do know is to go to the village pump, and in so doing, we'd effectively be putting it before the community. I'd appreciate any suggestions as to how to get a dev on board directly Worm^TT · (talk) 12:46, 19 September 2011 (UTC)

The procedure is supposed to be that you get consensus for a change before filing a bugzilla request. I'd be surprised if it was difficult to implement and as it fits in perfectly with what Jimbo was saying in Haifa about getting the Software to not let people do what they'd be told off for doing it really shouldn't be difficult to get coded. ϢereSpielChequers 14:04, 19 September 2011 (UTC)

I see. I just think maybe it should have a mention in the proposal, because I mean if for whatever reason there wasn't able to be a software fix, the end-result would be quite different. The proposal as it stands doesn't mention the implementation at all, so if it were to be put to the community as is, there would surely have to be a further (community) consultation about implementation. Mato (talk) 15:40, 19 September 2011 (UTC)

Will be tweaking the proposal soon based on some feedback anyway, and will include mention of the intended software fix as part of the implementation. By the way, WSC, did Jimbo's comment get documented anywhere to WP, or was it just something he mentioned at the conference? Worm^TT · (talk) 15:47, 19 September 2011 (UTC)

Ok great, thanks Worm. Mato (talk) 16:14, 19 September 2011 (UTC)

My understanding is that Jimbo suggesting something is good is not necessarily enough to get the devs/etc to do something. Unless a non-dev-involved implementation approach is possible—edit filter? I don't know—I would suggest running this by them before spending any more of your valuable time working on this (or any) policy change. ErikHaugen (talk | contribs) 17:35, 19 September 2011 (UTC)

Time to revisit

In the light of Wikipedia:Requests for adminship/Σ and its talk page, perhaps the time is now ripe to rekindle some interest in reform. Gernarally, the idea we discussed for introducing minim for candidacy was to reduce the number of NOTNOW/SNOW, however, that seems to be no longer the major concern. What the linked RfAZ demonstrates however, is that voters are now reaching far further back into candidates' histories than hitherto has been customary. If it is to be allowed to drag up dirt on a reformed editor that is older than 12 moths or so, why should we be allowing 6-month, 10,000-edit candidates with less on-Wiki history to check, to run for adminship? Kudpung กุดผึ้ง (talk) 09:24, 8 October 2012 (UTC)

I don't have the time now to read all of the previous discussions, and don't expect to have much time to get involved here either. However, I note at Wikipedia:RfA reform (continued)/Candidates#Comparison RfA other Wikipedias that de-wp requires 1 year of contributions, while here on en-wp we appear to have no official consensus for any such minimum requirement. I agree that introducing such requirements is highly likely to reduce the number of NOTNOW/SNOW outcomes, as proposed.

• At this time, I'd support a proposal for 12 months' editing history in any 18-month period (window to cater for any perceived penalty incurred by potential candidates taking time off for Wikibreaks/illness). Even for potential candidates who are practically 24/7 Wikipedians from the start, I agree with the principle that 6 months is probably too short a period of time: I believe that non-contact time is an important aspect in personal development. Such development may be particularly valuable in the case of younger candidates, while the older ones are (stereotypically) more likely to have the patience to wait that little bit longer anyway.
• I'd also support a proposal for a formal minimum edit count. The actual figure and how this could be prescribed in terms of content creation/anti-vandalism/user interaction etc. will likely pose some issues, although I'm sure they've already been discussed previously.

-- Trevj (talk) 13:27, 9 October 2012 (UTC)