Wikipedia talk:RfA reform 2011/Voter profiles

Comments

edit

Now I need to figure out how to rub the other top 20 the wrong way! Only kidding of course. This is interesting and offers a nice level of transparency. Oddly enough, I had fun seeing the ratio of where my name was to others (I have a whopping 5) through the various sorts. Cptnono (talk) 04:26, 7 April 2011 (UTC)Reply

An interesting list. I originally thought that my count might have been distorted by my username change. However, a comparison of this and changes at User:CT Cooper/Requests for adminship over the last twelve months show this not to be the case. CT Cooper · talk 07:56, 7 April 2011 (UTC)Reply

I manually found and adjusted any users who had username changes over the last year, so hopefully your stats are correct. —SW— converse 16:22, 7 April 2011 (UTC)Reply

Thanks

edit

We are grateful to User:Snottywong for the preparation of this data. For this and other similar work he has been doing for us in the background, and on other projects, I have awarded him a barnstar. --Kudpung กุดผึ้ง (talk) 06:59, 7 April 2011 (UTC)Reply

Seconded. Good work, SW! 28bytes (talk) 14:59, 7 April 2011 (UTC)Reply
Thirded, well done SW.WormTT · (talk) 15:03, 7 April 2011 (UTC)Reply
Fourthed. Thanks SW! Swarm X 06:35, 8 April 2011 (UTC)Reply
Yup - well done SW :o) Pesky (talk) 08:20, 13 April 2011 (UTC)Reply

Worm's Analysis

edit

(More editors' analyses further down)

Right. I've done a little snooping at the numbers - and here's some things I've found.

  • 578 (38.6%) editors only !voted once.
  • 1015 (67.8%) editors !voted less than 5 times.
  • 1318 (88%) editors !voted less than 20 times.
  • 842 (56.2%) editors have never opposed. (Including NewYorkBrad with 72 supports!)
    • 474 of these !voters had !voted once
    • Only 97 of these have !voted more than 5 times.
  • 133 (8.8%) editors have never supported
    • None had more than 9 !votes
  • A further 65 (4.3%) !voters voted oppose more than support.

Which tells me that the majority of people do support. But of course, that's no suprise - a WP:SNOW or WP:NOTNOW close will receive 8-9 opposes, whilst a "snow" support will garner well over 100 supports. WormTT · (talk) 08:53, 7 April 2011 (UTC)Reply

I wonder what this would look like if you screened out all the supports for candidates who got over 85% and all the opposes for candidates with less than 50%. I suspect that many !voters would drop off the list. Taking NYB for example, he may only !vote support, but I've never known him Support any NotNow candidates. ϢereSpielChequers 10:59, 7 April 2011 (UTC)Reply
When I did a similar analysis about 2 years ago, I eliminated anybody who had fewer than 5 !votes figuring that they were voting for/against specific individuals, not as an ongoing activity.---Balloonman Poppa Balloon 22:41, 26 April 2011 (UTC)Reply

Size

edit

Interesting work, though there is something about the size that made my PC hang several times before letting me see it. Perhaps replacing the admin columns yes/No with Y/N and the date of first edit with yyyy/mm would trim this a fair bit. also looking through the !voters who have very few edits I see a number of alternate accounts. I appreciate that it would be fiddly to do, but merging those into their main accounts would make the table shorter and more accurate of the !voters. also if you are going to list editors like this it is probably best to respect the opt out requests at Wikipedia:List of Wikipedians by number of edits/Anonymous, and replace the usernames of those editors with <placeholder> ϢereSpielChequers 10:35, 7 April 2011 (UTC)Reply

I've found that large tables like this usually cause problems if you try to access them using the secure server. Not sure why. They pop right up on the regular server. I tried to manually merge most duplicate users or users with changed usernames, and it was a time consuming process. I may have missed some editors, but most of the ones that remain are probably blocked socks who voted in 1 or 2 RfA's. Also, if any users don't want to be identified on this list, they are free to remove themselves (as one has already done). I don't think there's any reason to jump through hoops to preserve the "anonymity" of various users' edit counts, especially when that information is already publicly available to anyone. If anyone would like to manually go through the list and replace those users with <placeholder>, you are more than welcome. I don't think I'll have the time to do that. —SW— chatter 16:41, 7 April 2011 (UTC)Reply

Design

edit

This approach doesn't differentiate between a very supportive editor who !votes for lots of snow candidates and only opposes those who make really egregious errors in their RFA, but never bothers to !vote in RFAs that are going to succeed anyway; and an editor who never bothers with RFAs that have less than 70% support but who has arbitrary criteria for edits or tenure that lead them to oppose many admins who comfortably pass. Both RFA !voters could have the same proportion of supports and opposes in their !voting record despite having radically different !voting behaviour. ϢereSpielChequers 10:35, 7 April 2011 (UTC)Reply

I can see what you mean. I am listed as never opposing, which while technically true in the last twelve months, is because I usually don't bother with RfAs that are going down the tubes already, of which I would probably vote oppose. CT Cooper · talk 10:58, 7 April 2011 (UTC)Reply
Obviously we can't count the votes that were not made. However, I don't vote on every RfA either. If I did I would approach a score like Boing's. There's no strict pattern to the way I vote. I sometimes skip an RfA that's obviously going to pass, and ones that are obviously going to fail. I might cast an early 'oppose' on a NOTNOW to usher in an early close in the hope of saving the candidate some pain (though I think the SNOWs are a lot thicker skinned than more established editors.) My purpose in voting is more for reasons of being among the more experienced regular voters - kinda supporting the system however broken it is, rather than for the individual candidates.
I fully expected the overall number of admins who vote to be low, so I was wrong on my private prognosis. However, on an individual RfA basis, the admin participation could fluctuate greatly. The number of votes from new and/or low edit count voters gave me a bigger knee jerk than I expected. It would be interesting to investigate the voters who stick out in the stats - how loaded are theier talk pages with warnings, blocks, and redlinked creations/uploaded images, etc.; how many are 'fan' votes, and how many are 'hate' votes. How many of the votes were accompanied with less serious/relevant comments. How many of them are drama mongers? The burning question is: Those of us who might think there should be a qualification for voting, are we right? --Kudpung กุดผึ้ง (talk) 11:42, 7 April 2011 (UTC)Reply

Kudpungs stats

edit

These are just numbers. No percentges.

  • 578 voted one time only
  • 1,015 voted 1 - 5 times
  • 169 voted 6 - 10 times
  • 259 voted 11 - 49 times
  • 54 voted 50+ times
  • 50 voters had 100 or less edits
  • One voter with less than 20 edits voted 19 times
  • 166 voters had less than 500 edits
  • 1157 voters had less than 1,000 edits
  • Of the 54 who voted 50+ times,
  • 20 were admins
  • 16 voters had less than 10,000 edits
  • 2 voters had less than 5,000 edits
  • 465 sysops voted, although some may not already have been sysops for many of their votes (e.g. esp. Boing said Zebedee & Kudpung)
  • 45 voters had over 100,000 edits
  • 33 were admins
  • 16 voted 10+ times
  • 22 voted 5 times or less
  • 11 voted 1 one time only
  • 8 one time only voters were syops

Of the 33 100,000+ edit admins

  • 8 voted one time only
  • 7 voted 20+ times
  • 4 voted 40+ times
  • 109 voters had 51,000 - 99,999 edits
  • 72 were admins
  • 16 voted 10+ times
  • 22 voted 5 times or less
  • 11 voted 1 one time only

8 one time only voters were syops

Of the 72 51,000 - 99,999 edit admins

  • 8 voted one time only
  • 7 voted 20+ times
  • 4 voted 40+ times

Registration

edit
  • 335 voters registered before 1 January 2006
  • 335 voters registered in 2006
  • 239 voters registered in 2007
  • 210 voters registered in 2008
  • 175 voters registered in 2009
  • 175 voters registered in 2010
  • 22 voters registered in 2011
  • Of the 22 who had registered in 2011, 8 had less than 10 edits.
  • Of the 175 who had registered in 2010, 26 had less than 100 edits.
  • Of the 175 who had registered in 2010, 51 had less than 500 edits.
  • Of the 175 who had registered in 2009, 22 had less than 1,000 edits.

Kudpung กุดผึ้ง (talk) 11:05, 7 April 2011 (UTC)Reply

Thoughts

edit

I hadn't spotted that approx 1/10 of voters have less than 500 edits, but approx 2/3 have less than 1000. Also about 80% have been here over 2 years. WormTT · (talk) 15:03, 7 April 2011 (UTC)Reply

That's a real thought-provoker, that one isl the number of edits thing. Surprising. I find myself wondering why that would be. Pesky (talk) 08:25, 13 April 2011 (UTC)Reply

Excellent idea!

edit

(Thread copied from Wikipedia talk:RfA reform 2011)

I particularly like the idea of minimum qualifications to apply and then qualifications to vote, also. I like it being a right like reviewers and rollbackers. Many other excellent ideas, but those two really jumped out at me. Have a really neutral, clearly defined threshold, and then an evaluation by people who have been proven not to be trolls. Montanabw(talk) 23:55, 28 March 2011 (UTC)Reply

I think it could work if it's given out similarly to reviewer: given to virtually everyone who wants it, but it still provides that filter. You know the community, though: always fearful of new user rights. Swarm X 05:05, 29 March 2011 (UTC)Reply
I'm glad there is some support for this. However, I am also very conscious of the fact that it conflicts with most people's perceptions of democracy. 99.9% of any country's electorate does not need a degree in political science to vote for the party MP or representative of their choice. Ironically, Wikipedia is not a democracy... Kudpung กุดผึ้ง (talk) 05:51, 29 March 2011 (UTC)Reply
Then again, in many democracies you still have to register if you want to vote... Swarm X 09:26, 29 March 2011 (UTC)Reply
And if you're serving time for a felony, you can't vote! In fact, in some states, even for misdemeanors. And in some other states, with a felony conviction, you can lose your voting rights for life! =:-O And besides, a lot of voting, like for cabinet appointments, is indirect, via, for example, the US Senate... (lol) Montanabw(talk) 20:21, 29 March 2011 (UTC)Reply

<outdent> And most places there is some kind of minimum criterion - like age. And I don't think people who've been certified insane can vote in many places. I think having a 'mental age' (subjective, probably!) and 'recognition of sanity' badge is not a bad thing. Pesky (talk) 20:49, 29 March 2011 (UTC)Reply

Observations

edit

By my count there are 198 editors who oppose more often than they support. Excluding those who voted fewer than three times in the period sampled, there are 76. Of those 76, 16 are admins, including one arb and one 'crat. 115 voters offered zero support votes in the time period sampled. 28bytes (talk) 15:32, 7 April 2011 (UTC)Reply

Are RfA standards too low?

edit

Judging by the voting patterns of the "top 25 opposers", at least some of our experienced editors and admins think so. Shall we solicit their opinions on how to better ensure candidates can meet their high standards before showing up at RfA? Perhaps we can turn the perceived negative of the "serial RfA opposer" into a positive by clearly laying out for potential candidates what the most demanding RfA participants want and expect from candidates. 28bytes (talk) 16:35, 7 April 2011 (UTC)Reply

Perhaps we need to take into account their editing experience. Length of membership in lieu of edits is a controversial area. My voting behaviour and the comments I make around the board will show that I do not consider that pure length of usership with a low edit count does not count as experience. For example, I've been around for 5+ years, but I was a total Wiki Dummy until around 2 years ago - since when I probably made 95% of my edits. When we've finished looking at the table, we need to start looking at what's behind it. How many of them have published their criteria? The best measure IMO of standards is to take a mean of the standards published by editors who vote a lot and/or regularly. But standards is another discussion. --Kudpung กุดผึ้ง (talk) 17:18, 7 April 2011 (UTC)Reply
Just because someone frequently opposes doesn't mean they think that standards are too low. After all many of these opposes will have been to candidates that did not succeed. Also RFA is far too complex to reduce all candidates to one consistent scale. I have opposed candidates who have passed and supported ones that failed, generally I have higher standards re candidates understanding of deletion policy than many if not most !voters, but I'm more relaxed about tenure and editcount than a lot of RFA participants. I suspect that a comparison of oppose reasons and RFA criteria would be interesting. As for long tenure, it is a very effective screen against the return of Pastor Theo et al. Five years of little or no editing followed by a few months of being busy is very hard for a sockpuppet to plan. ϢereSpielChequers 21:46, 7 April 2011 (UTC)Reply
As the editor with the dubious privilege of having opposed candidates the most, I took a look at my track record when it comes to RfA. If you're interested, a list of all the RfAs I've opposed is here. You'll see that I've opposed 22 WP:NOTNOW/WP:SNOW candidacies and 13 candidates who chose to withdraw in advance. Of the 20 remaining RfAs I !voted in, 9 failed and 11 succeeded. In the successful ones, I opposed due to concerns regarding csd tagging (2), lack of experience in admin-related areas (4), regarding an unreassuring understanding of WP:BLP (1), concerns regarding activity and temperament (1) and wrong answer to a question (3, in 1 case regarding csd tagging). In the unsuccessful ones, I opposed tue to perceived block-happiness (1), lack of experience in admin-related areas (3, in 1 case also due to unsatisfactory answers), concerns regarding temperament (3, in 1 case also due to the candidate's block log), concerns regarding AfD closures (1) and concerns regarding csd tagging (1). All in all, I don't think RfA standards are too low; in general, to support, I'd like a candidate to be good-tempered, experienced in admin-related areas and well-versed in the deletion policy. Salvio Let's talk about it! 01:02, 9 April 2011 (UTC)Reply
Thanks, Salvio, that's some useful context. (I should put together a list like that myself.) The main thing I found interesting about the "top 25" was that there was no overlap between that list and the "signatories" here, which suggested we might not be having as broad of a set of perspectives about the RfA process as we could, so I'm glad you offered your input here. 28bytes (talk) 01:21, 9 April 2011 (UTC)Reply
The rubric for signing up says "Joining here assumes you are already firmly in favour of reform, and can invest time to regularly take part in its development.", which might filter out some perspectives. - Pointillist (talk) 14:48, 11 May 2011 (UTC)Reply
  • FYI, this isn't actually a list of the "top 25 opposers" in the sense of "who has cast the most oppose !votes". The criterion for selection seems to be the difference between the number of support and oppose !votes cast by an editor. I've added an O rank column that shows how each editor ranks in the original table sorted by total number of opposes. By this measure, the list shows only seven of the "top 25 opposers". - Pointillist (talk) 15:35, 11 May 2011 (UTC)Reply
User Votes S O N Edit count First edit Admin S − O S% O rank
Townlake 75 18 54 3 2628 March 24, 2007 No -36 24% 3
Vodello 40 8 31 1 3536 June 5, 2006 No -23 20% 7
Cirt 48 13 35 0 134878 September 22, 2007 Yes -22 27% 6
Salvio giuliano 98 36 58 4 18965 December 10, 2009 Yes -22 37% 2
Malleus Fatuorum 40 12 27 1 101751 June 3, 2006 No -15 30% 12
Cube lurker 13 1 11 1 9575 August 29, 2007 No -10 8% 49
Mono 15 2 11 2 17068 September 24, 2009 No -9 13% 49
ArcAngel 17 4 13 0 11786 September 24, 2006 No -9 24% 39
Kraftlos 53 20 29 4 12474 November 26, 2006 No -9 38% 10
Regancy42 8 0 8 0 5643 November 22, 2009 No -8 0% 81
Uncle G 10 1 9 0 42602 January 6, 2005 Yes -8 10% 70
TCO 16 3 11 2 8968 April 8, 2007 No -8 19% 49
Wisdom89 22 7 15 0 30124 September 7, 2004 No -8 32% 29
Colonel Warden 32 12 20 0 21203 February 26, 2006 No -8 38% 19
Hipocrite 8 0 7 1 18780 October 28, 2004 No -7 0% 94
Belovedfreak 11 1 8 2 68683 November 7, 2006 No -7 9% 81
Logan 13 3 10 0 30106 March 14, 2007 No -7 23% 60
Sandstein 13 3 9 1 46642 July 31, 2005 Yes -6 23% 70
Leaky caldron 5 0 5 0 6474 February 23, 2006 No -5 0% 143
Goodvac 5 0 5 0 7492 December 22, 2007 No -5 0% 143
Bali ultimate 9 2 7 0 11069 October 4, 2008 No -5 22% 94
Nakon 11 3 8 0 53334 December 13, 2005 Yes -5 27% 81
Rockfang 5 0 4 1 23045 September 24, 2007 No -4 0% 171
David Fuchs 10 2 6 2 29389 October 15, 2005 Yes -4 20% 117
Pointillist 10 3 7 0 8079 September 25, 2007 No -4 30% 94

More....?

edit

I like it, SW, is there any way your bot can see about who/how many users are asking questions at RFA? I think that would be another piece of useful information. Tofutwitch11 (TALK) 18:59, 7 April 2011 (UTC)Reply

You read my mind (again) Tofu. --Kudpung กุดผึ้ง (talk) 06:35, 8 April 2011 (UTC)Reply
Two great mind think alike, ay ;) Tofutwitch11 (TALK) 19:00, 8 April 2011 (UTC)Reply
Also... would it be possible to see how often people vote in comparison to the result? So editors who regularly support failed candidates or oppose ones that succeed. Perhaps also how people vote in NOTNOWs and SNOWs. WormTT · (talk) 07:54, 8 April 2011 (UTC)Reply
I've been a bit short on free time lately, but I'll see what I can do. —SW— yak 14:26, 8 April 2011 (UTC)Reply
I think a table on the users who pose the questions in a higher priority. The voter stats table IMO should now be used as a tool for researching how individual voters performed on RfA and what their general maturity and civility levels are like in other areas. What we are looking at are relatively new users, users with low edit counts ho vote a lot, and to question why some users with huge edit counts only rarely vote. Why does it appear that new, inexpereinced users always go straight to NPP and RfA as their prime involvement on Wikipedia? Can we help them to better understand what Wikipedia is all about, and what it is not? Can we write to each of them? Can we include more advice to voters without instruction creep? Kudpung กุดผึ้ง (talk) 21:18, 8 April 2011 (UTC)Reply
This is one that has really got me wondering, too. Why get into this side of WP if not-so-much interest in editing. No, I;m not saying that a voter should have x-thousand edits, but new accounts with little editing history suddenly turning up to vote seems ...... odd. Now I'm (sometimes) a suspicious sort of soul, and I find myself thinking along lines of 'power attracts the corruptible', and 'are they really new?' Pesky (talk) 08:31, 13 April 2011 (UTC)Reply

Threads copied from main page

edit

(Please continue these discussions here)

Voting qualification

edit

Provided it was set low, clear and automatic I would be OK with a qualification for !voting. I remember when I first checked out RFA I couldn't work out what the unwritten criteria was for voting and I left the page for months. I think I'd been editing over 12 months before I first !voted in an RFA, so a low threshold such as 200 edits would actually make it clearer more open and less cliquey. ϢereSpielChequers 13:29, 25 March 2011 (UTC)Reply

I do quite like the suggestions from Kudpung actually, basically "2m good standing". The only problem is that it makes blocks punative... WormTT · (talk) 13:39, 25 March 2011 (UTC)Reply
That's interesting, WereSpielChequers, because my 1st !vote was "Support (Feel free to remove if non-admins are not allowed to vote in RFAs)." That caused this response on my talk page. Reaper Eternal (talk) 14:35, 25 March 2011 (UTC)Reply
I agree with PhantomSteve - this is a community decision not an admin one. I think a potential solution would be to make it an indirect community decision by electing a committee to appoint admins, but that masks the real issue which is that we haven't agreed the criteria that admins should be judged against. If we set the criteria then I believe a whole bunch of editors will look at it see if they meet it and come forward. We'd have to set a criteria if we elected a committee - either explicitly or implicitly by the views of those elected. So why not set a criteria and then judge candidates against it? ϢereSpielChequers 16:29, 25 March 2011 (UTC)Reply
I like clearly defined goalposts. It's so much easier to hit a target when you're not blindfolded in advance. And having a clear standard to be judged against would rule out a lot of 'personality clash' votes or 'xe's my friend' votes. Pesky (talk) 19:03, 28 March 2011 (UTC)Reply

What we would need are some stats on voter profiles, perhaps in a sortable table like we've done for the NPPer profiles. It would go something like this:

  • Number of voters on RfA over the past 12 months
  • Date of each voter's first edit
  • Number of edits of each voter
  • Number of times the voter voted
  • Number of times the voter voted 'support'
  • Number of times the voter voted 'oppose'
  • Number of times the voter voted 'neutral'

This may help to establish a 'right to vote'. There's not much point in speculating what that threshold would be yet. We need those stats and then discuss it further.

Voter profiles

edit

We now have an excellent sortable table of who and how voted on RfA over the last 12 months. Play with it - there is some extraordinary information to be gleaned from it. Kudpung กุดผึ้ง (talk) 03:51, 7 April 2011 (UTC)Reply

Minimum qualifications

edit

My feeling is that autoconfirmed status needs to be the bare minimum qualification to vote or ask questions on an RfA. This was imposed on mine when I noticed and reported some suspicious activity in the support !votes, which turned out to be sock puppetry. I don't know what the motive was, but it strikes me that, at the very least, setting a minimum qualification would stop this.

I'd like it to be higher, as I don't really see how someone who has been on wikipedia for, say, a month or two can critically evaluate a prospective admin. Catfish Jim & the soapdish 18:41, 2 May 2011 (UTC)Reply

Minimum participants

edit

There is a temptation to simply say that chance would be a fine thing, but it is in theory better to decide rules in advance of needing them. I'm loathe to have a rule that a set number of participants are required as this could give an opposer an invidious choice. - Oppose and it passes 29 to 1 or stay stumm and it fails 29 - 0 for lack of participation. These sort of thresholds only work if you define a minimum number of supporters. So if it needs 22 supports and 70-75% support then 20 - 0, 21 - 7 and 69-31 all fail but 22-7 is a success and 22-8 and 22-9 are crat calls. ϢereSpielChequers 13:29, 25 March 2011 (UTC)Reply

I know it's better to decide the rules before needing them, and I know it is one of my suggestions, but I'm also aware that we want to avoid instruction creep - that's why it's all only at the idea stage. There's also the possiblity that it might encourage canvassing. There was a time, a long time ago by Wikipedia standards, when such low turnout RfA would pass, but times have changed, and 100 votes of all kind are common place. Even 100+ support votes are no longer as extraordinary as they once were. It's certainly worth more discussion. --Kudpung (talk) 18:34, 25 March 2011 (UTC)Reply
One of the ideas I considered is to transclude the RfA to the main RfA page but don't set the 7 day timer until a minimum number sign up. There are nuances to this type of approach which can be further defined. My76Strat (talk) 03:18, 2 April 2011 (UTC)Reply

My thoughts

edit
  • I do not approve of RfAs being votes, as I feel this is contrary to the fact that Wikipedia isn't a democracy.
  • I do not approve of banning self-nominations as some potential admins may receive little attention.
  • I approve of lowering the approval rate if voting is kept, and if voting is abolished, then the bureaucrat will decide.
    • If a user feels that his/her RfA was closed unfairly, he/she should be able to appeal to another bureaucrat or (in some cases) a steward.
  • I do not approve of applications "right to vote". While Wikipedia does not have complete freedom of speech, I feel that this should instead by relegated to all autoconfirmed users instead of requiring application, as then the applications would be an unnecessary load on bureaucrats and current admins.
  • I do not approve of a 250-word limit which would be extremely difficult to enforce (both technically and practically) and would be a limit on potential useful feeback.
  • I definitely approve of banning off-topic comments and comments based on illegitimate reasons like hatred of the user.
  • Requiring both support and oppose votes from a user is very ambiguous, and I feel that it is unnecessary.
  • Definitely ban non-autoconfirmed users.
  • If voting is kept, I propose the following:
# of words in comment Voting units Voting units (if comment is in truly bad faith)
0 1 1
1-20 3 1
21-60 5 0
60-80 9
80-120 12
120-200 15
200-250 17
250-300 20
301+ 22
    • Of course, this would not be so clear cut and the content of the comments should determine the actual # of units.
  • The # of support votes should be a minimum of 15, not 30. 30 is a little too much.
  • No limits on questions, but Bureaucrats can discount obviously insignificant questions.

My thoughts.Jasper Deng (talk) 00:32, 26 March 2011 (UTC)Reply

To be clear, you're advocating that longer rationales should be weighted more heavily? Useight (talk) 21:30, 27 March 2011 (UTC)Reply
Eeeek! Those who waffle best / shout longest are more important? Pesky (talk) 18:13, 28 March 2011 (UTC)Reply
Stop the trolling. Did I not say that these are subject to what the comment actually contains?Jasper Deng (talk) 00:26, 29 March 2011 (UTC)Reply
This is going to go nowhere fast if we can't refrain form accusing each other of trolling every time there is a criticism. I don't think you quite understand what trolling is if you think those comments are examples of it. Please, everyone let's try to keep this civilized and not let it become another snake pit like the one we are endeavoring to fix. Beeblebrox (talk) 00:34, 29 March 2011 (UTC)Reply
I used the term because of "My76Strat would certainly like this idea. :P", but maybe it's wrong. I think most did not read that the weighting is subject to the comment's content.Jasper Deng (talk) 00:37, 29 March 2011 (UTC)Reply
I'm still missing where you specified that, link? Useight (talk) 00:47, 29 March 2011 (UTC)Reply
Right below the table.Jasper Deng (talk) 02:05, 29 March 2011 (UTC)Reply
Oh, there it is. I was reading above the table. Thanks. Useight (talk) 00:08, 31 March 2011 (UTC)Reply
It was just a joke about Strat's well known OTT wordiness. Sorry if it somehow offended you but I think "trolling" is a little exaggerated. Swarm X 17:06, 30 March 2011 (UTC)Reply
I do not see how I am considered a TROLL. It is a scarlet letter which I resent. If I put an effort forth, it initiates from good intentions. I hope to participate in this task force, and it would be helpful if people who believe I am negative, would give me a second chance, or AGF or whatever might allow me to function as a colleague. Please. My76Strat (talk) 04:21, 2 April 2011 (UTC)Reply

No one was saying you were a troll Strat. — Oli OR Pyfan! 05:50, 2 April 2011 (UTC)Reply

Qualifications for voters

edit

I really don't like the idea imposing any minimum requirements for eligibility to !vote, as it goes against the entire equality ethos that in general has worked so surprisingly well. Generally, limiting !voting to autoconfirmed registered users is enough, I think, but I would probably support one small tweak to that. If we had a !voting right that was automatically granted once an editor has been registered for 7 days (just like the 4 day autoconfirm one), that would stop new SPA registrations after an RfA has started.

I'm not sure whether this goes far enough. To my mind, we should require !voters to have at least some familiarity with Wikipedia. How can we reasonably expect someone with a week's experience of Wikipedia to be able to judge whether someone would make a good admin?
I don't think equality is an issue here. We already have different levels of permission given for different purposes. I'd like to see !voting reserved for those with Reviewer rights (easy to get) and above. This would cut out at least some of the junk !votes I've seen, and at a guess, most !voters are of that level already. Catfish Jim & the soapdish 13:51, 5 May 2011 (UTC)Reply

Voting conditions

edit

I'm generally in support of the spirit of that whole bit - but would just caution against making it look too much like we're trying to make Support voting easier than Oppose voting.

In reality, I'm against the idea of RfA being a vote when so many other things like XfD aren't. Wikipedia is not a democracy. An Oppose vote would also be easier.Jasper Deng (talk) 19:53, 15 April 2011 (UTC)Reply
Difficult one , I admit. I think the RfA process is in fact very much a discussion, and that the commented !votes are numbered and put into sections for ease of parsing the results of over a hundred participants. I think this is very effective, and as we see often enough, it can help participants change their minds, which they often do. The main concerns are the maturity and civility with which the commenting/!voting takes place. --Kudpung กุดผึ้ง (talk) 04:31, 16 April 2011 (UTC)Reply

End Weak and Strong

edit

Kudpung is suggesting ending the use of Weak and Strong and giving the !voters just three choices, Support, Neutral and Oppose. Me I'm not sure what the point is of posting in neutral and rarely do so, nor do I think Strong is helpful (and if it did mean anything it would be open to abuse), I'd rather use the electrons to give an extra reason to support or oppose a candidate. But I do think that weak is useful, life is complex and Virtual Life even more so, I feel that three choices is insufficient to cover the range of candidates that I come across at RFA. I appreciate that prefixing my vote at RFA with weak is an invitation to the closing crat to give it less weight then normal, but some candidates I support or Oppose less strongly than others. Also I don't see any benefit to RFA in getting rid of this, and some disbenefit. ϢereSpielChequers 23:10, 28 March 2011 (UTC)Reply

Strong is certainly not helpful. As has been mentioned it is often added to arguments that are in fact quite weak or at the least no better than any other. Although I admit I have used it myself a few times, usually if I really think a candidate would be a great admin. Weak, on the other hand, is a useful self-identifier. If someone marks their comment as weak, I expect the closing 'crat to give it half weight compared to others who did not do so, but I certainly wouldn't expect them to give someone who self identified their position as strong any extra weight. Not sure we could get rid of strong without taking weak off the table as well though. Why do we even have bolded votes at the beginning of comments? They are already organized into sections for support or opposition. Maybe we should just get rid of that tradition altogether. People could still identify their position as weak in the text of their remark if they felt it was so. Beeblebrox (talk) 00:47, 29 March 2011 (UTC)Reply
The bolded text does increase readability, in my opinion. Useight (talk) 01:07, 29 March 2011 (UTC)Reply
I was thinking on the lines that doing away with the qualifying adjectives would help provide a clearer overview of the consensus that is developing, and also help the crats make up their minds in close calls. . As Beeb says, it would force voters to make their rationale ore detailed. I must admit I have used the 'strong' and 'weak' about twice in my RfA voting history, but only when the outcome is already going to be blatantantly obvious. However, we really need some crat input on this.
I personally see no objections however in continuing to use the traditional bold text as used in in all debates. Kudpung กุดผึ้ง (talk) 02:31, 29 March 2011 (UTC)Reply
I seem to recall a crat citing numerous 'weak opposes' in their rationale for promoting a <70% candidate in at least one instance. 'Strong support' is unhelpful (though it does no harm), while 'strong oppose' can be taken as uncivil. I certainly think 'weak' rationales are given weight by crats in close RfAs that might go either way. Swarm X 05:01, 29 March 2011 (UTC)Reply
My personal opinion is that the modifiers help convey nuanced positions; and that removing the modifiers brings RfA closer to pure voting. Whether or not that's a Good Thing is open to debate, but I doubt it would solve RfA's more pressing problems; perhaps it might be worth concentrating efforts elsewhere. bobrayner (talk) 10:48, 30 March 2011 (UTC)Reply
When I'm looking at close RFAs, I don't assign half-weight or double-weight depending on the adjectives used. But I do look at how many of the supports were 'weak' and how many of the opposes were 'strong' and take an extra moment to mull over the rationales used. I feel like a 'weak support' is pretty much a "meh" and a 'strong support' is an "over my dead body!", dependent on the other contents, of course. A "Strong Oppose [sig]" is particularly unhelpful. Useight (talk) 00:15, 31 March 2011 (UTC)Reply

I don't particularly like the !voting system... personally, if we follow this path I doubt meaningful change will occur. That being said, if we must have !voting, the strong/weak does help. Imagine a hypothetical RFA closing at exactly 70% support/oppose. Looking at the adjectives, you see all the supports are "Strong" and all of the opposes are "weak." Assuming rationale reasoning, closing the RfA as a pass is a lot easier. Similarly, same closing percent, but most of the supports are "weak" and the opposes are "strong Oppose." Suddenly, that RfA looks less likely to pass. The "strong/weak" help the reviewer determine how strongly the poster feels about the rationale they present. Two people can look at the same information, have the same overall "support/oppose" but for one the rationale is important for another it isn't.---Balloonman NO! I'm Spartacus! 22:23, 12 April 2011 (UTC)Reply

Bolded votes also make it much easier for bots to reliably parse your vote later, for statistical analysis. Just saying. —SW— confabulate 22:56, 12 April 2011 (UTC)Reply

More breakdown

edit

I wonder if you guys could break down the data on this page into smaller chunks. What I'm wondering is 1) do admins and non-admins vote differently (and thus could you guys make two tables divided into admins and non-admins) and 2) do editors with ≥1 year's tenure vote differently to newer editors? This could make for fascinating reading (or it could be just as inconsistent). HJ Mitchell | Penny for your thoughts? 13:37, 13 April 2011 (UTC)Reply

The answer to that would probably be to have an 'and/or' sortable table, but I guess this would need some complex regex built into the script. Snottywong's the expert. --Kudpung กุดผึ้ง (talk) 14:47, 13 April 2011 (UTC)Reply
There has been analysis on this in the past, my memory is that if there is any pattern it is that editors who have not yet run at RFA tend to be more supportive, editors who have run unsuccessfully tend to be less supportive and usually admins and non-admins split the same, but there are some "interesting" candidates where the RFA doesn't follow this pattern and admin non admin voting is different. Also there are some !voters who differ wildly from the above pattern. ϢereSpielChequers 17:18, 13 April 2011 (UTC)Reply
This is a very interesting thing, WSC: the split between editors who have and haven't run the gauntlet themselves. It parallels some kind of "hazing manhood test" - "I went through hell to get this, so you should, too" kind of approach. Like some of the barbaric tribal rituals associated with coming-of-age. Is this mentality (trial-by-ordeal to be confirmed as an adult) just inherent in human societies, and does it always need the equivalent of a human rights intervention before it ceases? Pesky (talk) 07:30, 17 April 2011 (UTC)Reply
I suspect there are a couple of different processes at play here. I'm aware of the theory that those who've undergone a hazing ceremony expect the same to be inflicted on others, but the pattern I'm seeing here is that the people who were rejected by RFA are the harshest !voters. I fear that it is the nature of hazing ceremonies to get harsher over time, and suspect that is a major reason why RFA is broken. I think that another process is that as editors get more experienced so they start working out tests to screen out candidates who they think would make bad admins, I certainly pay more attention to deletion mistakes than I did a couple of years ago. ϢereSpielChequers 22:26, 18 April 2011 (UTC)Reply
Perhaps indeed the very reason why it appears that so many admins do not come out in support of reform is because they don't believe any one else should have an easier time than they did. Perhaps others simply can't be bothered with RfA stuff since getting the bit themselves, while still others who passed with 100+ flying colours dropped their edits from 1,000s a month almost immediately to a trickle of an average of 1 a day. Some clearly only come out of the woodwork to !vote with a vengeance, and others seem to enjoy RfA as being the one place on Wikipedia where they can be uncivil with impunity. Kudpung กุดผึ้ง (talk) 02:40, 19 April 2011 (UTC)Reply
While I'd agree that most admins are inactive, I'm not sure that I've noticed many who go inactive almost immediately after RFA, if anything I suspect that becoming an admin tends to extend people's wiki careers. As for the incivil !votes at RFA, my experience is that the most incivil ones are the ones who are not admins and have no intention of ever running. ϢereSpielChequers 16:20, 19 April 2011 (UTC)Reply

Alt accounts and renames

edit

We have some !voters with more !votes than edits, these are the result of renames, including editors who have exercised rtv. There are also some alt accounts, I've removed five so far adding them into the !votes for the main account. ϢereSpielChequers 17:21, 13 April 2011 (UTC)Reply

Update. The anomalous votes where RFA !voters appear to have !voted in more RFAs than they have done edits all seem to relate to accounts being renamed, presumably the stats were collected on the signature of !votes and not updated per RTV etc. There are a number of participants who made a !vote very early in their wiki career, and notall of these have been blocked as socks. but many have. I think that the number of participants is at least 1% less than we thought, and the entire drop will be of non-admin involvement as all the occasions of alt accounts and renames that involve admins no more than one of their accounts was listed as an admin one. ϢereSpielChequers 13:27, 18 April 2011 (UTC)Reply

Interesting, but

edit

How is this a proposed policy or guideline? I doesn't seem to propose anything. Tijfo098 (talk) 05:24, 17 April 2011 (UTC)Reply

It was necessary research in orer to clarify some issues and theories. Pesky (talk) 07:08, 17 April 2011 (UTC)Reply

It clearly says: The proposal is definitely still in development and under discussion, and has not yet reached the process of gathering consensus for adoption. Thus references or links to this page should not describe it as "policy" nor yet even as a proposal. --Kudpung กุดผึ้ง (talk) 07:54, 17 April 2011 (UTC)Reply

Moving forward

edit

The research seems to have been done. At any rate, barring research into how the !voters actually !voted on passed/failed RfAs and what their rationales were, we have the basis for further extrapolation if needed. It might now be time to decide whether or not we want to propose a threshold for RfA !voting participation, per all the pros and cons that have been discussed here and elsewhere it, and if so, what levels. --Kudpung กุดผึ้ง (talk) 08:00, 17 April 2011 (UTC)Reply

Having looked through a lot of the participants with very low !votes I would say that there is little need for such a threshold. Far less than might have appeared when the stats were first run as at least twentytwo of the lowest !votes relate to alt accounts, RTV or renaming. I suspect from the proportion of the remainder who turned out to be socks that newbies who vote at RFA in their first 100 !votes should be looked at and if they aren't a declared sock they should be checkusered. In total 48 of the RFA !voters are currently blocked, thats 3% of all participants, though they tend to be low !voters so far less than 3% of the !votes will have been by blocked voters. ϢereSpielChequers 19:14, 18 April 2011 (UTC)Reply
That worries me. Let's say the average turnout (all votes) is 100, that's 3 on every RfA that are going to be suspect. hey are probably more likely to be fan club !votes in the support section, but if they were all in the opposition, that could be catastrophic for the candidate. --Kudpung กุดผึ้ง (talk) 16:44, 19 April 2011 (UTC)Reply

'Low' turnout passes (under 100):

edit
2011 (out of 23 passes)
  1. feezo - (47/1/6) (54)
  2. Valfontis (63/1/0) (64)
  3. JaGa (83/3/0) (86)
  4. Peridon (64/3/1)(68)
  5. Neelix (69/14/12) (95)
  6. ErrantX (85/2/4) (91)
  7. Rami R (2nd) (66/12/8) (86)
  8. Acdixon (82/5/6) (93)
  9. Ponyo (75/1/2) (78)
  10. Gimme danger (77/11/6) (94)
2010 (out of 75 passes)
  1. Grondemar (86/2/5) (93)
  2. PresN (70/10/3) (83)
  3. TheCatalyst31 (68/12/13) (93)
  4. Magog the Ogre (2nd) (65/4/3) (72)
  5. BigDom (2nd) (72/16/5) (93)
  6. HelloAnnyong (88/1/0) (89)
  7. Mandsford (59/16/4) (79)
  8. Michig (80/0/2) (82)
  9. Amatulic (80/2/3) (85)
  10. WOSlinker (83/4/1) (89)
  11. Jujutacular (78/0/1) (79)
  12. Joe Decker (78/7/1) (86)
  13. NativeForeigner (67/10/5) (82)
  14. 7 (2nd) (92/2/4) (98)
  15. Waldir (76/1/2) (79)
  16. Barek (65/0/2) (67)
  17. Eustress (71/1/2) (75)
  18. Father Goose (58/6/1) (65)
  19. J04n (2nd) (63/0/1) (64)
  20. Calmer Waters (81/2/4) (87)
  21. Taelus (69/5/2) (76)

High turn-out 'unsuccessful': 100+ !votes

edit
2011 failed with 100+ !votes
  • GiantSnowman (76/36/10) (122)
  • Ctjf83 (2nd) (55/38/12) (105)
2010 failed or withdrawn with 100+ !votes

(some may have since passed)

  • Richwales (58/44/9) (111)
  • Ling.Nut (113/63/7) (183)
  • The Thing That Should Not Be (2nd) (123/59/21) (203)
  • Alansohn‎ (2nd) (39/56/11) (106)
  • MZMcBride (4th) (56/124/21) (201)
  • DeltaQuad (65/34/12) (111)
  • Connormah (2nd) (88/30/11) (129)
  • Herostratus (2nd) (78/48/21) (147)
  • Blanchardb (2nd) (54/38/7) (99)
  • MichaelQSchmidt (87/60/11) (158)
  • Kingoomieiii (67/36/3) (106)
  • Ironholds (4th) (64/38/11) (209)

Kudpung กุดผึ้ง (talk) 16:50, 19 April 2011 (UTC)Reply

minor comment here, but the numbers for Ironholds don't add up. 64+38+11=113, not 209. Am I missing something there? PrincessofLlyr royal court 13:33, 3 May 2011 (UTC)Reply
Probably not missing anything. The exeercise was to demonstrate that the suggestion of a quorum of 100 to validate an RfA probably isn't necessary. History appears to show that most successful RfAs had around 100 participants or more during the period that was analysed. Kudpung กุดผึ้ง (talk) 17:57, 3 May 2011 (UTC)Reply

Restating some things

edit

(Note: Thread copied from WT:RFA2011. Please continue the discussion here.)

I wanted to restate a few things said elsewhere which should be commented on here. In one example I asked how viable it would be to transclude an RfA but not set the timer until 100 participants sign up as jurors. This implies that perhaps all RfA should be from the same size jury. Additionally I suggested that these participants should not actually vote until around day 5, leaving the first 5 days for questions and comments. Are there any valid points to glean from these considerations? My76Strat (talk) 01:01, 16 April 2011 (UTC)Reply

There is clearly an imbalance in the number of participants from RfA to RfA. This is one of the reasons that RfA is often criticised as being a popularity contest, and where it is also possible for an RfA to succeed with very low participation in all !voting sections. The suggestion above for having a minimum to effect a quorum has been touched upon previously, and I think it's a valid point, but as a radical change that would give the green light for an RfA to go ahead, but not as one that might improve RfA participants' behaviour, do we want to discuss it right here and now? That said, I think a 'I will !vote on this RfA' call is not a bad idea, and would be a possible solution for preventing the NOTNOW, but it would prolong the process while the 'jurors' are being gathered.--Kudpung กุดผึ้ง (talk) 03:38, 16 April 2011 (UTC)Reply
I agree and only want to see a collective best effort. I did state some of this in line with the openness to "radical suggestions". While your concerns are truly valid, they can perhaps provide some answers unto themselves. If a quorum of 100 was established, it could also sign up and seat participants even before an RfA was transcluded. Under such a consideration the 100 members could be ready even before the next candidate. And yes, these are all radical considerations, but they may have some useful purpose. My76Strat (talk) 11:29, 16 April 2011 (UTC)Reply
I've got several concerns about this, RFA has many problems and I don't see any merit in a radical change that doesn't address any of the problems at RFA but does add some major new ones.
  1. Why do you think we need 100 participants for an RFA? Mine was one of those that achieved that, but many don't. I can see an argument that a successful RFA needs at least a dozen support !votes - the number that a jury has to have here in the UK. But why should more be required to appoint an admin than to convict a murderer?
  2. What are you suggesting should happen to the many admins appointed by less than 100 !votes? Remember these are often the uncontentious RFAs.
  3. Not !voting for the first five days but just having questions and comments would drag out the whole process from 7 to 12 days and snow fails from hours to over 5 days. What benefit would that give us to outweigh the obvious harm?
  4. Having to sign up to participate and then return several days later to do so is a lot of extra bureaucracy for no discernible benefit. It would also reduce the number of participants, probably making 100 participant RFAs a rarity.
  5. The community is dwindling, setting a participation threshold that we don't always currently reach means designing a system that will fail if current trends continue.
  6. If RFA had a problem with lack of participation I could see an argument to change the rules to require a minimum number of supports, but minimum participation is a different and flawed idea. Saying that an RFA with only 11 supports had insufficient participation would be a workable rule. Saying that you need 100 participants would mean that an RFA with 60 supports 25 neutrals and 15 opposes would be a success, but an RFA with 96 supports, 2 neutrals and 3 opposes would fail if two of those opposes struck and withdrew from the RFA. A minimum number of supports would add one extra failure mode, but minimum participation creates two extra failure modes and puts opposers in the awkward position that an abstention might cause an RFA to fail whilst an oppose would make it succeed.
ϢereSpielChequers 08:06, 17 April 2011 (UTC)Reply
What WSC said. Tweakage that adds extra layers of rules and restrictions is vehemently unwanted tweakage. In fact, all tweakage is vehemently unwanted. The page says somewhere that desysop should not be discussed, but folks, desysop is the problem, not RfA. RfA is indeed a brutal and evil process, but evil RfAs are solely and only a symptom of glacial, painful desysop. The reason adminship IS a big deal is because desysop IS a big deal. No more. No less. Done.• Ling.Nut (talk) 08:43, 17 April 2011 (UTC)Reply
I agree with WSC's comments above, and Strat's suggestion was at least a reasonable idea, even if it is not viable. I think our desyoping system is adequate for the moment, but I'll admit that I don't know everything that goes on around here. If anything, there should be sterner measures to give some admins a hefty slap on the wrist sometimes, but while radical reform of RfA is not totally ausgeschlossen, desysoping is not on the agenda of this particular project. I've said before, that I don't believe it's one of the reasons why potential candidates won't come forward, nor is it much in the mind of the !voters themselves. Kudpung กุดผึ้ง (talk) 10:32, 17 April 2011 (UTC)Reply
Is there actually a need to get into this? I can't say I remember any RfA that suffered from lack of participation. Most RfAs will attract more than 80 participants even with candidates who have kept a low profile. This year, the successful RfA with the lowest participation level by far was Feezo's with 54 voters. Hardly a shortage of participation. Swarm X 00:20, 18 April 2011 (UTC)Reply
WSC all of your considerations are prudent. When I say 100 participants, I mean for that to be an example. It is very likely after the most thoughtful consideration, some different number would likely be more appropriate. I would not be the one to object if the number were 12. The point I wish to consider, if at all valid, is if there should be a set limit to the jury like pool or not. Nothing in my suggesting anything is meant to imply anything about a current admin. If they have the flag, they deserve it! To the extent, additional burden could be imposed upon the participant who says they wish to sit in judgment, that additional burden can be mitigated, and would not rise to a thing greater than the vote, which they intend to cast! I have stated somewhere that a participant who states TLDR, could be removed from the user group with the right to participate, IMO, because to state such a thing, discounts your sincerity to participate in this kind of decision. So I wouldn't consider it an unmanageable burden for someone who signs up to vote, to follow a format which suggests they herd the request, entirely, and then voted. I would also not insist that any thing I suggest be implemented, only considered, which I am glad to see it apparently has. But they are only meant to be ideas, as we hammer out what might be an RfA reform. My76Strat (talk) 03:58, 18 April 2011 (UTC)Reply

Please see the stats below. Kudpung กุดผึ้ง (talk) 04:09, 18 April 2011 (UTC)Reply

'Low' turnout passes (under 100)

edit
2011 (out of 34 passes)
  1. TParis (48/4/4) (56)
  2. Jimp (75/4/0) (79)
  3. January (91/1/0) (92)
  4. F (78/16/eezo (47/1/6) (54)
  5. Valfontis (63/1/0) (64)
  6. JaGa (83/3/0) (86)
  7. Peridon (64/3/1)(68)
  8. Neelix (69/14/12) (95)
  9. ErrantX (85/2/4) (91)
  10. Rami R (2nd) (66/12/8) (86)
  11. Acdixon (82/5/6) (93)
  12. Ponyo (75/1/2) (78)
  13. Gimme danger (77/11/6) (94)
2010 (out of 75 passes)
  1. Grondemar (86/2/5) (93)
  2. PresN (70/10/3) (83)
  3. TheCatalyst31 (68/12/13) (93)
  4. Magog the Ogre (2nd) (65/4/3) (72)
  5. BigDom (2nd) (72/16/5) (93)
  6. HelloAnnyong (88/1/0) (89)
  7. Mandsford (59/16/4) (79)
  8. Michig (80/0/2) (82)
  9. Amatulic (80/2/3) (85)
  10. WOSlinker (83/4/1) (89)
  11. Jujutacular (78/0/1) (79)
  12. Joe Decker (78/7/1) (86)
  13. NativeForeigner (67/10/5) (82)
  14. 7 (2nd) (92/2/4) (98)
  15. Waldir (76/1/2) (79)
  16. Barek (65/0/2) (67)
  17. Eustress (71/1/2) (75)
  18. Father Goose (58/6/1) (65)
  19. J04n (2nd) (63/0/1) (64)
  20. Calmer Waters (81/2/4) (87)
  21. Taelus (69/5/2) (76)

Compiled by Kudpung กุดผึ้ง (talk) 03:04, 18 April 2011 (UTC)Reply

What about Wikipedia:Requests for adminship/Vejvančický (90/1/4) from October 2010? Alzarian16 (talk) 20:06, 2 June 2011 (UTC)Reply

High turn-out 'unsuccessful': 100+ !votes

edit
2011 failed with 100+ !votes
  • GiantSnowman (76/36/10) (122)
  • Ctjf83 (2nd) (55/38/12) (105)
2010 failed or withdrawn with 100+ !votes

(some may have since passed)

  • Richwales (58/44/9) (111)
  • Ling.Nut (113/63/7) (183)
  • The Thing That Should Not Be (2nd) (123/59/21) (203)
  • Alansohn‎ (2nd) (39/56/11) (106)
  • MZMcBride (4th) (56/124/21) (201)
  • DeltaQuad (65/34/12) (111)
  • Connormah (2nd) (88/30/11) (129)
  • Herostratus (2nd) (78/48/21) (147)
  • Blanchardb (2nd) (54/38/7) (99)
  • MichaelQSchmidt (87/60/11) (158)
  • Kingoomieiii (67/36/3) (106)
  • Ironholds (4th) (64/38/11) (209)

Complied by Kudpung กุดผึ้ง (talk) 03:24, 18 April 2011 (UTC)Reply

See also

Kudpung กุดผึ้ง (talk) 03:38, 18 April 2011 (UTC)Reply

We may seem a very long way from this being a problem, !votes per RFA having actually gone up in recent years. But this is a side effect of the fall in the number of candidates, RFA !votes per month have fallen rapidly, just not quite as rapidly as the number of RFAs. If we fix RFA and get a large increase in candidates then we may find that there is a shortage of RFA !voters, so setting a minimum number of supports may be worthwhile. I have no objection as long as the threshold is low and is measured in number of supports, not number of participants, I suspect the crats may feel they already have discretion to relist an underconsidered RFA, but if not I would have no objection to adding a phrase such as "crats have discretion to relist if there has been insufficient consideration of a candidate - this may apply if there are 12 or fewer Supports". This is probably a superfluous but uncontentious reform, however rules are best set before they are needed. ϢereSpielChequers 13:03, 18 April 2011 (UTC)Reply

Additional analysis suggestions from TCO

edit

1. Can you do a cluster analysis? See who tends to vote with who? One of those programs that will do PCA and sort of spread it out into a 2-D map that shows who has similar voting patterns to who (I think you have to go by how they voted on same candidates). I wonder if you can then see the differnt "camps".

2. Be interested in seeing how the FA flag correlates to voting patterns. Is there a difference with the broader population.

3. I guess any other intersting flags we can thing of (not sure...we don't have that much data on demographics...and this thing is pretty stag.)

TCO (talk) 05:08, 18 June 2011 (UTC)Reply

Enough stats are available in these pages - do feel free to help by extrapolating a cluster ananlysis if you feel it is necessary. By and large however, the sortable tables will tell you all you , need to know already. There is a small but strong core of regular !voters, mainly admins, bureaucrats, and some very experienced users who don't want to be admins, and there are the great many one-time !voters, many of whom have as few as 100 edits and probably are not sure what adminship is really all about, and just do pile-on !votes. --Kudpung กุดผึ้ง (talk) 05:15, 18 June 2011 (UTC)Reply
That's additive, seriously. Didn't know there were those two groups. What is an exclamation point voter? (is that a support without rationale...btw, I count NYB's one liners as essentially pass without rationale...certainly not an explained one...more like a pontification). and then both groups are !ing? Who would be some people in each group (pick ones that aren't touchy, so this does not become a drama...honest...I'm just trying to understand.)TCO (talk) 05:48, 18 June 2011 (UTC)Reply
The word !vote with an exclamation mark is a traditional IT term that means it means something else, such as in all Wikipedia discussions, things are supposed to be a discussion rather than a straight vote. It is generally accepted that any uncommented 'support' !votes are simply in support of the nomination statement with nothing else to add. 'Oppose' !votes should of course be supported by a rationale - if they're not, there is a possibility they will not be taken into consideration much in the case of a close call. I do understand the difficulty in joining a project like this without being in it at the start, it may well take an hour to read all the pages here to get up to speed. Thanks for your comments, and keep them coming! Kudpung กุดผึ้ง (talk) 05:58, 18 June 2011 (UTC)Reply
I'm a quick study. It is not "generally accepted" on the supports. Sandy Georgia (or Malleus, I might have lost track) do not accept it. And rationally, there are a lot of junk pile-on supports from inexperienced voters, are weak nominations, etc. I think understanding at a minimum, what aspects of the nom statement and factors were convincing is helpful. But I am not even trying to argue it. Just saying, there are a reasonable amount of people on the other side of this debate. I STILL don't get what you mean with the ! though. Did you mean an unexplained oppose? And who would be some people who do that? If it is too touchy, point me to the list and you don't have to name them!TCO (talk) 06:11, 18 June 2011 (UTC)Reply
Then I'm not quite sure what you mean about the (!). The main sortable table lists nearly 1,500 !voters - the opinions of of Sandy and Malleus are only two of them, but they are nevertheless highly experienced Wikipedians. If you were to follow WT:RfA (archives) over the last 18 months, you will see that there is a weak consensus for accepting uncommented 'support' !votes, because they are in support of a statement already made. This is normal also for all Wikipedia debates. In any debate, you either support the motion statement, or come up with a good reasons for opposing it; a motion for debate generally does not start with a negative proposal proposal. One of the most contentious RfA in recent times was my own. We sometimes have extremely controversial RfAs, but that's not quite the same thing - see Wikipedia:Requests for adminship/GorillaWarfare. Among other problematic !voters are those who come out of the woodwork to oppose with a vengeance, and unfortunately they are often neither truthful nor civil. The primary goals of this project are, rather than search for an entirely new system, to find ways to 1). prevent incivility altogether, and 2). to encourage more objective !voting, and 3). to discourage, as nicely as possible, candidates who don't stand a cat in hell's chance and from wasting their time and ours, but to encourage them to stay with us as regular editors. Kudpung กุดผึ้ง (talk) 06:59, 18 June 2011 (UTC)Reply

1. This is just a definitional thing: does ! mean "oppose"?

2. Well, 30% against someone can fail them. So if 70% of people think unreasoned supports are fine, that does not really mean anything...perhaps it is just the basic difference of low and high standards.

3. There are plenty of processes (for instance hiring) where this sort of "default is accept" would NOT be how things are evaluated. Many things are much more choosey. I realize you don't agree or like this...but a substantial fraction of the community disagrees. And really a non-miniscule minority can sway things at RFA, given you need 3:1 to pass.

4. Yeah...I definitely get the impression that a big driver of this initiative is that some of the people who want more admins or easier standards are not happy about things being tighter and want to either change the threshold or hush up the opposers. No wonder, I am pleased with how things are going lately and you are displeased.  ;) P.s. I really do find the opposes to be more thoughtful in their remarks. You get some random malcontents. But there are some braniacs like Sandy and Malleus. And some people with significant life and work experience that influences their stance on how organizations work. (Plus they are really good looking, too.)

TCO (talk) 07:10, 18 June 2011 (UTC)Reply

From WP:!VOTE:

The terms "!vote", "!voting" and "!voter", introduced in 2006, are sometimes used in discussions to indicate that taking part in a straw poll is not voting, but rather engaging in an act of consensus-building. These terms serve as reminders that while we do vote on things, votes without reasonable accompanying rationales receive little consideration unless you also explain why you are voting the way you are. Votes without rationales sometimes are ignored. ¶ The exclamation mark in "!vote" is the symbol for logical negation and can be read literally as "not vote". It serves as a cute little reminder that it is "not the vote" that matters, but the reasoning behind the !vote that is important.

28bytes (talk) 07:19, 18 June 2011 (UTC)Reply
I am still not getting it. Hold my hand. Is a ! a non-explained oppose?TCO (talk) 07:24, 18 June 2011 (UTC)Reply
! is just decoration, essentially. 28bytes (talk) 07:29, 18 June 2011 (UTC)Reply
As 28bytes says.
I'm not displeased about the way things are going, and there is certainly no consensus for lowering the bar, nor for raising it, because there is no bar - the bar is set by the !voters themselves for each new RfA, based on their own criteria, and not all the same !voters turn out for each RfA. However, if the last 5 or so RfA have been reasonably civil, it's because unlike me, most of them have not been around long enough to have upset anyone by caching them for doing something naughty or abusing their power. Ironically also, the longer one has been around, and the more one is well known, the RfA attracts more participation so there is a slightly fairer chance that overall it might be reasonably objective - take my RfA for example with its particularly nasty opposes and personal attacks, but passed with flying colors. We need to look at the bigger picture and not start imagining any trends in improvement yet. For one thing, we are already almost at the end of Q2/2011 and still only 34 promotions. That's too low for natural replacement, and plenty of mature experienced editors have told us quite clearly why they won't step into the snake pit - I took a chance and it worked out, but probably because I've been heavily involved in policy work, I don't know, it was certainly a gamble considering the my long campaign for improvement of RfA since long before I was ever asked to run for office, and then finally agreed to. Do remember also, that some of the brainiacs might not really be opposing the candidate directly , but globally opposing adminship as a system for regulating the quality of the Wikipedia. A support !vote from Malleus (when he makes one) is worth twenty supports, but unfortunately cannot be counted as such! Kudpung กุดผึ้ง (talk) 07:47, 18 June 2011 (UTC)Reply
I am happy with a low pass rate. I am more choosey. Have high standards for moderators. They can do a lot of damage when they are stupid or mean. You were a borderline candidate as I recall. TCO (talk) 07:53, 18 June 2011 (UTC)Reply
I take a "have to prove you can hack it" approach. So if there is less history, it is a negative for me. I am more "default fail" than "default pass"TCO (talk) 07:55, 18 June 2011 (UTC)Reply
I'm happy with less of you all as well. I think it is a view of a reasonable fraction of editors. TCO (talk) 07:59, 18 June 2011 (UTC)Reply
Mine was not a borderline, it passed with a very healthy margin. Please take a moment to review some old RfAs and the tables we have provided here on this very talk page higher up. Kudpung กุดผึ้ง (talk) 08:03, 18 June 2011 (UTC)Reply
I didn't mean numerically borderline, I mean my personal evaluation. What do you want me to get from the looking at the RFAs. Surely I have participated in enough to see the ways they go down.
Actually, I think I had you confused with Carabane-man (a very questionable FA, translation with problems) Neelix. Sorry, guess you guys went through the meat grinder at the same time. Don't recall if I voted in yours. I did look at the stats, and you were definitely above 70%, but still not a January-like love-fest. It's fine though. Gotta crack some sku...eggs to make your breakfast. ;) TCO (talk) 17:17, 18 June 2011 (UTC)Reply

What the other Wikis do

edit

!Voters on the French language Wikipedia:

2.2 Who can vote?

All registered contributors can give their opinion on the choice of a sysop. However, in practice are recorded only the opinions of the contributors who have at least:
  • 1 week of operation before the opening of the vote;
  • 50 significant contributions to his credit at the time of transclusion.

Intéressant, hein? --Kudpung กุดผึ้ง (talk) 15:03, 19 June 2011 (UTC)Reply

Italian Wiki
It.wiki is much worse...  
Per it:Wikipedia:Amministratori/Sistema di voto/Requisiti
  • 60 days of operation before the opening of the vote
  • 500 contributions to one's credit at the time of transclusion
But it.wiki also has weird requirements to be eligible for adminship and yearly reconfirmation RFAs... Salvio Let's talk about it! 15:47, 19 June 2011 (UTC)Reply

I think reconfirmation or term limits are not a bad idea. It would really change the two-class dynamic that we have. Right now, it sorta feels like the admins are the "made men" from Goodfellas. I would be much happier to loosen the requirements or say "admin is no big deal" if it were not permanent. Problem admins like RHE need to get culled. He would never have passed an RFA with all his drama...but it was a bunch of bother to pry the mop from his cold dead hands (his words). It would also clean out the inactive admins.TCO (talk) 16:55, 19 June 2011 (UTC)Reply

I don't think we should be discussing here our opinions on individual candidates or how their RfA were closed. Pointing to an RfA that was a close call, however, would be fine, and we have the stats for all that. --Kudpung กุดผึ้ง (talk) 17:18, 19 June 2011 (UTC)Reply
German Wiki

50 users must vote 'support' within two weeks, with at least two-thirds of the total votes cast will be Support-votes. Neutrals do not count as votes cast. Voters must have a minuum of 200 edits.

Certainly interesting that it lasts two weeks rather than one. I don't remember any calls on the English Wikipedia to have RfAs last two weeks, and I think seven days is sufficient. The needs for two-thirds support (~67%) rather than 75% is also of interest, and the experience at de should be looked at carefully if there is any serious discussion on lowering the threshold from 75% to 65%. CT Cooper · talk 18:50, 19 June 2011 (UTC)Reply
nooooo! Keep the percentage high. I want more people like January-gurl. not more teen-aged boys with a hard-on for vandals.TCO (talk) 18:53, 19 June 2011 (UTC)Reply
Within the framework of this reform project that is mainly intended to make RfA a more pleasant prospect, our effort is primarily to improve the !voters' participation. Our bar for candidates is fine and I don't think it's within the remit of this project to discuss it, in fact there isn't a threshold for serious applicants, and generally the right candidates pass - there are few really borderline cases and most of the fails are NOTNOW, or withdrawn. The problem is that there aren't enough candidates coming forward for all the reasons stated. What is interesting is that other Wikis have minimum qualifications for !voters. Anything that concerns a 'bar' for candidates is to prevent obvious non-starters from wasting their time and ours. Hence possible suggestions such as, for example, 3,000 edits/6 consecutive months. See: Wikipedia talk:RfA reform 2011/Candidates and WT:RFA2011/PP. Kudpung กุดผึ้ง (talk) 19:25, 19 June 2011 (UTC)Reply
You really have not demonstrated to me, how we don't have enough admins. cluebot takes a lot of the load off.TCO (talk) 19:28, 19 June 2011 (UTC)Reply
I've already provided you a link to a very detailed report (and its comments) in the Signpost. However, it's not the brief of this project to prove it. Our goals are clearly stated, and that's not one of them. Kudpung กุดผึ้ง (talk) 20:56, 19 June 2011 (UTC)Reply
Note: this seems to be a reference to Wikipedia Signpost 2010-08-09 Admin stats article, in which WereSpielChequers tabulates the number of successful RFAs on enwiki by month. However, the article does not consider how many admins are required, the extent to which automated tools can reduce the need for admins, whether changes in policy might reduce vandalism, etc., so it doesn't demonstrate whether we have "enough" admins. I'm just pointing this out for the benefit of any lurkers here—I do appreciate that active participation in this initiative is only for those "already firmly in favour of reform"! - Pointillist (talk) 23:07, 19 June 2011 (UTC)Reply
Giving consideration to lowering the bar is listed as an objective at Wikipedia:RfA reform 2011#Objectives. I don't actually have a strong opinion on it personally, and have neither endorsed nor rejected a change; the point I was making is that the experience on de Wikipedia should be looked at if the bar was lowered, as conveniently they already have a lower bar. CT Cooper · talk 21:02, 19 June 2011 (UTC)Reply
They may appear to have a lower bar, but having lived there for nearly 20 years, I can understand why this may be so. We would have to take into consideration many factors pertaining their stats for unsuitable articles, the level of user discipline, the amount of disruptive editing, and the cultural inference that the de.Wiki is mainly limited to Germany, Austria, and Switzerland. Personally, I'm OK with the bar as we have it on en.Wiki, but it's a floating bar anyway, because the majority of voters change at every RfA - only a small core of regular voters apply any set standards. RfA here is a horrible process, and I've recently gone through it and I can understand why many of those who would pass are not coming forward. My next exercise will be to read through a sample of French and German RfAs. Unfortunately, anything I glean from it will only be OR. Kudpung กุดผึ้ง (talk) 21:19, 19 June 2011 (UTC)Reply
Yes, the cultural issues between the projects would need to be taken into account if any comparison is to be made. CT Cooper · talk 10:06, 20 June 2011 (UTC)Reply

@Pointillist: There is no cabal - anyone and everyone is heartly welcome to comment, it's only expected that the task force members are in favour of moving things forward for reform. It would be a bit pointless if they weren't :) Kudpung กุดผึ้ง (talk) 13:42, 20 June 2011 (UTC)Reply

It's a bit late to say that now, given that you've said the opposite for two months! - Pointillist (talk) 14:38, 20 June 2011 (UTC)Reply

Actually even though I'm "on the other side", I do think it is reasonable for those broadly in favor of "reform" to consolidtate their thinking and limit the broadening. So at least you can make a definite proposal. TCO (talk) 19:59, 24 June 2011 (UTC)Reply

Forgive me if this has been said before, but...

edit

"The total number of !votes cast in all RfAs has been decreasing since reaching a peak in 2006. The large percentage decreases show that participation in RfA as a whole is falling at at quicker rate than other areas of the encyclopedia."

Unless I'm looking at completely different chart, this is completely, utterly, and entirely untrue. Unless one is referring to fewer nominations, the average number of !votes per candidate has steadily increased since 2006. The wording makes it seem like there are fewer people vetting candidates. Again, I apologize if this is noted further up on this talk page, but this wording should be changed ASAP. -RunningOnBrains(talk) 10:14, 15 July 2011 (UTC)Reply

I produced the study that led to this conclusion, and both interpretations are correct. The total number of !votes in all completed RfAs has fallen in every year since 2006. The average number of !votes per candidate rose is every year except one (2006-07), when it fell for successful candidates and rose for unsuccessful ones. This situation is explained in the bullet point one below the one you cite, which states that "The mean number of !votes on successful and unsuccessful RfAs for each year are higher than the previous year in all but one case. The percentage increases in recent years show that participation in individual RfAs is increasing at a quicker rate than other areas of the encyclopedia. This apparent contradiction is explained by the falling number of completed RfAs." Alzarian16 (talk) 17:26, 16 July 2011 (UTC)Reply
Thanks Al, and your explanation is exactly how I had interpreted it. We have never imagined that the road to RfA reform is going to be quick and easy and what we need now is for all the editors who have contributed stats that are being used in this project, to maintain them and keep them up to date. Although the WMF generally has no influence over how the regional Wikipedias run their affairs, they have stated on numerous occasions that highly efficient statistical support is required for major policy changes. --Kudpung กุดผึ้ง (talk) 04:27, 17 July 2011 (UTC)Reply

Eligibility to vote

edit

A script is available here that could easily be adapted to check on users' eligibility to vote. Slightly more complicated sorftware-wise, this script could also be automatically triggered by an attempt to edit the 'support', 'oppose', and 'neutral' sections. --Kudpung กุดผึ้ง (talk) 09:06, 4 September 2011 (UTC)Reply

New tool

edit

For a complete breakdown of an editor's RfA votes, see RfA Vote Counter. Courtesy of Snottywong. --Kudpung กุดผึ้ง (talk) 00:18, 10 September 2011 (UTC)Reply

Time to do something now

edit

See here. --Kudpung กุดผึ้ง (talk) 08:30, 8 November 2011 (UTC)Reply

The above discussion is important. It has been closed and may soon be archived, so please look for the archive, then see the new follow-on discussion here. These events are vital references in support of proposals to introduce measures to clean up the environment at RfA. RfA has become a safe haven for those who seek a venue where they can abuse Wikipedia core policies with impunity, or treat the process as a joke, only by putting an end to the pollution, can any changes to the stagnation of nominations for adminship be expected. --Kudpung กุดผึ้ง (talk) 18:04, 8 November 2011 (UTC)Reply

We need to lower the standard, by force.Jasper Deng (talk) 18:09, 8 November 2011 (UTC)Reply
*Loads shotgun* Swarm X11|11|11 18:14, 8 November 2011 (UTC)Reply
Jasper, lowering the standard for adminship is not one of the objectives of this project, and should not be. The standard is something set by the community - and I'm not convinced that the standard is wrong. I think there are many users who would meet the current standard but refuse to run because of the atmosphere at RfA - that is where the issue lies. WormTT · (talk) 19:19, 8 November 2011 (UTC)Reply
I don't understand, why didn't anyone propose an RfA ban for that user? A "civility block" is a borderline unspeakable thing on Wikipedia, but, as you say, we're at the point where we seriously need to eliminate the pollution. Swarm X11|11|11 18:31, 8 November 2011 (UTC)Reply
I am also surprised, but any one of us could have done so - I didn't, because I stir up enough commotion on Wikipedia already as it is ;) The actual outcomes of the AN/I and the RFC/U are not as important as the factual impact they will have when cited in arguments for reform of the process. However, and this is only my opinion, I do seem to detect over the years that there appears to be a certain attitude from several editors of all user groups to keep defend RfA as a venue for the worst of Wikipedia incivility that would certainly get them blocked or at least topic banned anywhere else - to the extent that some people might even be inclined to find any reason at all to torpedo any good faith attempts to bring about reform of RfA and redress intractable behaviour. Kudpung กุดผึ้ง (talk) 21:18, 8 November 2011 (UTC)Reply
I think we need to find a way to invalidate incivil !votes in order to encourage more objective feedback.Jasper Deng (talk) 23:34, 8 November 2011 (UTC)Reply
What if a vote is both incivil and at the same time objective? The two aren't exactly mutually exclusive. —Jeremy v^_^v Components:V S M 03:06, 13 November 2011 (UTC)Reply
If the vote is truly objective (the majority of incivil votes aren't), the user should be warned in the usual way for incivility. If the user has a significant history of incivility anywhere on the site, and if it persists, and attempts to encourage civility have failed, then measures should be undertaken to prevent further disruption. In cases of repeated gross incivility, and/or after three warnings, immediate block might be considered by some admins, otherwise a RfC/U would probably be preferable. The result could be an eventual block, or if incivility was contained to RfA, a topic ban from the voting process. A current example of such a discussion can be found at Wikipedia:Requests for comment/Badger Drink. --Kudpung กุดผึ้ง (talk) 04:55, 13 November 2011 (UTC)Reply

See here for my suggestions on this. Feel free to move that to this page if it's better over here! Pesky (talkstalk!) 07:01, 15 November 2011 (UTC)Reply

Reflections on voting by Jéské Couriano

edit

(Coordination note: Copied from WT:RFA2011. Please continue the discussion here.)

I'd go a step further and say undereducated pile-on opposes are an issue. In my last RfA, everyone latched onto one of two incidents (one me baiting a troll in order to get him to run through his IPs, the second me blowing my top due to a situation that had real-world ramifications for an editor) that took place within the two months prior to my run. That's the main reason I can't read it anymore without seeing red - everyone focused on a small aspect rather than the overall picture. The problem is, everyone has some issue in their past if they wanna run for RfA, and if they don't, I wouldn't be two-faced if they got opposed for not taking a chance. It's hypocrisy, pure and simple (to say nothing of Malleus Fatorum opposing for incivility and then taking that opportunity to rail on Young Earth Creationists, but I digress).

I'm sorry if it seems like I'm ranting; I'm just quite bitter because I feel I got shafted by everyone involved in the RfA process. I'm not going to say whether or not it worked or failed (as I am in no state of mind to bring that up when mentioning my second RfA torques me off), but I would say that, at least in the RfA of today, due diligence and civility are out the window in favor of extreme-short-term history (I understand short-term, but realistically two months isn't that long a time) and snarking on the candidate's fur color and whose shirts he wears. —Jeremy v^_^v Components:V S M 03:57, 13 November 2011 (UTC)Reply

There was gross incivility, deceit, and unresearched pile-ons to dubious opposes on my own RfA. Fortunately it passed, but a worthy candidate with a lower history of contributions, and less well known might not have fared so well. Granted, RfA is not (supposed) to be a popularity or unpopularity contest, and we've seen candidates pass in March and June this year with as few as 47 support votes. --Kudpung กุดผึ้ง (talk) 05:15, 13 November 2011 (UTC)Reply
It really is almost certainly our biggest problem at RfA. The combination of incivility and cherry-picking one or two glitches that an editor may have (and I bet we've all got the odd kick in our gallop over one or two things), and then blowing them up out of all proportion to an editor's over-all potential as an admin. Pile-on opposes can never be good, especially when they don't seem to mention anything at all in support of the candidate. The atmosphere at RfA is one of the things that first made me decide I nevver want to be an admin; but to be totally honest, since then, I've realised that adminship isn't for me anyway (it's not the kind of area I'm interested in, and I'd oppose myself on the basis that I have no need for the mop!) We do have any number of good, sound, experienced editors who would make excellent admins, but just don't want to run the gauntlet because of precisely these issues - which means we are losing out, and losing out for abhorrent reasons. Pesky (talkstalk!) 08:05, 13 November 2011 (UTC)Reply
The single most totally distasteful area of Wikipedia is that of RfA. It seems to be the one place where some individuals go to be unpleasant to fellow editors with impunity while hiding behind the anonymity that the Internet affords. Being an admin is most definitely no big deal and the RfA process to obtain that 'no big deal' is far too often a disgusting arena of controversy, contempt, contentious behaviour and deceit. I had committed myself to do something about it long before there were rumours of admins being a dying breed, or even dreaming of joining the brigade myself, and I am distressed to see users at the current RfC/U displaying a clear lack of good faith in suggesting that the discussion has been perpetrated and supported by those who were dissatisfied with the outcome of an RfA they were involved in. The behaviour at RfA (many of them) is but the tip of an iceberg, and if the recent RfA was a catalyst, then so be it, and it should be a further stimulus for reform of the system. --Kudpung กุดผึ้ง (talk) 10:24, 13 November 2011 (UTC)Reply

The poisonous atmosphere (again) - courage and topic bans required

edit

(Coordination note: This thread has been moved from WT:RFA2011. Please continue the discussion here.)

As far as I can see, there really is only one way to deal with the "poisonous atmosphere" at RfA, and that's to decide on an acceptable standard of behaviour with respect to civility at RfA - and in my view this should be a higher standard of behaviour than, for example, users' talk pages - and to stand firm and enforce it. This means people having the guts to say when something is badly wrong with a voter's behaviour on a regular basis, follow it through, and, where appropriate, have a (for example) three-month topic ban (RfA) imposed. There is currently no effective deterrent for bad behaviour, and whenever there is no effective deterrent for any unwanted behaviour, you're going to get the unwanted behaviour. It's how humans work.

And, yes, you're going to get a few people whingeing about "Civility Police" - louts, yobs and vandals threatened with the possibility of ASBO's are always going to mouth off about the presence of police. All Wikipedia's various forms of vandals mouth off about it when they're taken to task - nothing new there. Pesky (talkstalk!) 06:40, 15 November 2011 (UTC)Reply

Corollary: and, if they come back to RfA after a topic ban and cross that borderline again, their vote is struck (discounted), topic-ban again, with an extra month tacked on for good measure. Three such bans and it becomes a permanent topic ban. Something along those lines would rapidly clean up the atmosphere. Moral: if you want to be allowed to play at RfA, you have to stay scrupulously civil. Pesky (talkstalk!) 06:56, 15 November 2011 (UTC)Reply

'nother one: I'm sure many of us are aware of persistent offenders - pro-active move would be a collection of all relevant diffs pointing to a pattern of uncivil behaviour over there, (like Kudpung's list in the current case) for the known offenders, and have them ready to present for an immediate vote-strike and topic ban when they do it again. If persistent offenders are aware that evidence is likely to be being collected about them, then if they want their next (and future) !vote(s) to be counted, they will have to behave from now on. This could start on the cure for this problem immediately. Pesky (talkstalk!) 07:13, 15 November 2011 (UTC)Reply

Question: Would it be feasible to bring a case to ArbCom (not on a named individual, but on the over-all subject of cleaning up RfA), with sufficient evidence (if needed!) to get them to rule that from now on the problem of solution to incivility at RfA will be rigorously enforced in this way? If so, better for someone / group of people with higher 'standing' and more street-cred than I have, to do it. If we can get a structure in place which has the instant weight of AE behind it, it would also prevent "fan clubs" from removing topic bans. This might then (conceivably) have the knock-on effect of raising our civility standards right across the wiki; if someone has, for example, an AE-weight topic ban for incivility at RfA, they would have an incentive to be more civil everywhere. Pesky (talkstalk!) 07:24, 15 November 2011 (UTC)Reply

Arbcom does not make policy, or decide what it's role should be. The roles and functions of user groups are decided by consensus of the community. The Wikipedia:Arbitration Committee arbitrates on individual cases and applies existing policy on a case-by-case basis. There are recognised steps within the structure of the Wikipedia disciplinary system, which should generally escalate only when one level has failed to reach a satisfactory conclusion or solution:
  1. an appeal to order on a user's talk page
  2. AN/I or appropriate disciplinary notice board (sanctions can be imposed by an administrator following consensus)
  3. RfC/U - a discussion on a user's behaviour. Recommdations for solutions are made, usually towards a favourable agreent by the offending user to abide by the rules. The conclusions of RfC/U are informal and non-binding.
  4. Arbcom - generally the highest instance, especially for Amin misbehaviour and other cases where all else has failed.
It is generally not recommended to escalate to another level or to move to another kind of dispute resolution board until it is clear that a current discussion has clearly failed. Some discussions are expected to remain open for a minimum period of time. Moving early from AN/I to RfC/U may not be the most suitable option. Moving to Arbcom is usual after 30 days at RfC/U, and an Arbcom case must be opened before the RfC/U actually expires or has shown no movement for a significant period of time. Escalation or moving horizontally from one process to another should preferably be done with consensus of a community if that community is currently and actively discussing an issue. As with most processes on Wikipedia, these are guidelines, not rules, but they are generally accepted by the community as the best way to go.
As regards this WP:RFA2011 project is concerned, there is little it can do to intervene in such procedures. The goal of this project is to promote necessary reforms that participants feel could be useful to the RfA system, and to propose them for discussion by the community. At sometime or another since the creation of this project all possible flaws in RfA have been examined and discussed, and it is hoped that the project will move forward to proposing some of the suggested reforms. This takes time, as do all new Wikipedia proposals for changes to policies and site systems. Everyone on Wikipedia is considered equal, so users with 'street cred' have no special powers. This flat hierarchy ensures that decisions are made through consensus. Individuals can of course bring their own proposals to the table, but it would be highly recommended that they ensure they would have a minimum backing first. Generally, such individual proposals begin in an appropriate department of the the Village pump, where a trial gathering of views may be conducive to moving to a more broadly exposed central discussion, and that's where Wikipedia projects can be of greatest assistance.Kudpung กุดผึ้ง (talk) 08:38, 15 November 2011 (UTC)Reply
I probably didn't make what I was thinking clear enough (either that or I've totally misunderstood you); I was thinking that ArbCom could take on the "case" of "incivility at RfA" as a whole, and stamp it out. Ignore me if I'm being brain-dead here - it's that Real Life stuff again! Pesky (talkstalk!) 13:19, 15 November 2011 (UTC)Reply
I've numbered the bulleted points above in order (more or less) of priority. I was just attempting to outline the various resources and steps to address behavioural problems to save you hunting round the site to find out (even I didn't know much about all this before I became an admin). Incivility 'as a hole' would have to be addressed by the community at large, by proposing some kind of metrics for it - which would be extremely difficult to define. Hence each occurrence of incivility has to be addressed on a case-by-case basis. The irony is that not all AN/I discussions are populated by the same people and admins, so the threshold for incivility decided by consensus iwould be different in each individual case - plus the fact that other circumstances are taken into account (although I'll never understand why). The only other alternative would be to introduce a common law system. Everyone on Wikipedia knows that AfD is a snake pit whether they admit it or pretend it isn't , and it's not necessary to relist the reasons here. What we are doing here on this project is to examine the cause and effect, address the suggestions for changes, and ultimately propose them to the community for debate and implementation. .
Put another way, empirical studies of why RfA is broken appear to point mainly to incivility and fraudulent voting, and I'll hasten to add - lest I am misunderstood by some as being an advocate of limiting the rights of the oppose voters - on both sides of the fence. Our goal here is to clean up that process by encouraging voters to act reasonably and vote accurately, and if they don't, apply some measures to preventing abuse of the system - bearing in mind that blocks and bans are for preventing further disruption and not to punish, although there is not always a perfectly bright line between prevention and punishment (take prison systems for example). This may have to be achieved by limiting the right of all users to vote such as they are on other Wikis. Two analogies: A bad driver will get points on his licence and when has enough, he gets banned, but we don't punish the entire community. If however an entire stretch of road is so straight that drivers regularly drive too fast on it, we impose a speed limit for every one. That's then a rule, and if the limit is reasonable most drivers will stick to it; and those that don't will get points (a warning), and a fine (punishment), and repeat offenders will be banned (prevention). Kudpung กุดผึ้ง (talk) 14:28, 15 November 2011 (UTC)Reply
Every editor I've tried to convince to run for admin in the past year has declined specifically because they didn't want to go through the inevitable attacks and incivility. I would suggest making it clear that all participants are required to adhere to a high standard of civility, and that any uncivil comments will be subject to removal. I would also suggest prohibiting participation by anyone who has been blocked in the past month or two. If we do end up having an RfA clerk, they could be tasked with moderating the discussion and removing uncivil comments. Kaldari (talk) 07:37, 16 November 2011 (UTC)Reply
Out of curiosity, how many potential candidates have we lost through this? I've only approached 2 people on being an admin (the rest have approached me), and 1 declined for that reason. Kudpung, Kaldari - I'm curious to know roughly how many people you've approached since the beginning of the year? WormTT · (talk) 11:18, 16 November 2011 (UTC)Reply
I've approached 11 since Sept last year. Most declined for the usual reasons. One said they they were quite happy to continue to contribute articles (and excellent articles they are), and one said they would not want the flak that using the tools would attract. They would all have been excellent admins - people who have no controversial issues at all. I've had several requests since the 'Request-a-nom' page was started, but none of them would have passed. --Kudpung กุดผึ้ง (talk) 11:36, 16 November 2011 (UTC)Reply
Update: By coincidence, I've have another user decline today. Kudpung กุดผึ้ง (talk) 23:46, 16 November 2011 (UTC)Reply

What the Bureaucrats are saying

edit

- about policing RfA. It's interesting to note that WBscribe has also had some editors decline his proposals for adminship. Interesting reading for anyone who has 10 minutes to spare; many points covered, and it may have some influence on how a possible proposal for clerks might be addressed. The discussion is still ongoing but may be archived soon.--Kudpung กุดผึ้ง (talk) 14:35, 3 December 2011 (UTC)Reply

The Miller test

edit

At the main talkpage, Kudpung has made reference to an RfA vote that, in the eyes of several people, was considered invalid. In addition, we also saw another RfA vote a few days ago that was so bad it had to be oversighted. This brings up a topic which has been of much concern; the manner in which people conduct themselves at RfA. Through the many discussions, we've established that 1. everyone has a different standard for civility and 2. RfA is, by its nature, somewhat confrontational (if you're opposing at RfA, you're opposing the person, unlike everywhere else here where you (or at least you should) oppose the action). However, the fact that it was another opposer who e-mailed oversight in the second example above demonstrates that, even now, there is a certain threshold for what we're willing to take; there is at least some semblance of common sense. To make RfA a less distasteful experience for most people, it will require a more liberal application of said common sense. There is no way to completely quantify it, so we have to know it when we see it. The two ways I can think of to go about enforcing it are the following. If a substantial number of editors (I'm talking 4-5) agree, with no opposition, that something is completely beyond the pale, an admin/bureaucrat (doesn't matter which access level we choose) can strike and indent it, or if it's bad enough remove it altogether. The second would be what was suggested above; to let the bureaucrats police RfA votes. I won't make any grand proclamations that this will solve all the problems of RfA, but it's a way of combating one of the problems we've identified. The Blade of the Northern Lights (話して下さい) 01:20, 6 December 2011 (UTC)Reply

Because it's obvious which !vote is referred to in one of the above examples, and I made a few comments myself - I'll weigh in briefly:
  • Yes, it seems there do need to be "rules" about what rationales are "acceptable".
  • Inevitably, applying these rules will need liberal doses of "common sense" and judgement.
  • There seems also to be a need for a procedural way to "enforce" this. Clerking sounds good.
I ended up commenting far more than I should, and probably making a messy situation look worse - but, if I had known that procedure would deal with the issue, I would have found it a lot easier to "stop arguing". I feel bad, now, that a wish to comment on what I saw as a bad "advert" for RFA just probably resulted in it looking like an even worse "advert". I could have confined myself to talk page comments - probably I should have (I notice Kudpung was able to do this) - but it's hard not to want your refutation of a poor rationale to be visible when you feel strongly, and once 2 or 3 people join in - there's your "mess". Begoontalk 11:57, 6 December 2011 (UTC)Reply
My opinion? It was a perfectly valid vote, with a perfectly valid reason behind it. The reason was weak, but the badgering the opposer received on the RfA was plainly disgusting, and a lesser editor would have taken it much worse. I'd like to commend BlueRaspberry for how well he took the criticism. Clerking, whilst a great idea, should just be doing what the community already does, so unless a decision from the community clearly shows that it was inappropriate and can define why for future situations then it should not be removed by anyone, let alone clerks. WormTT · (talk) 12:14, 6 December 2011 (UTC)Reply
You may be absolutely correct, and if you are, then it was comments like mine that needed to be removed, prevented, or at least properly discouraged. That would serve to clean up some of the "mess". Begoontalk 12:22, 6 December 2011 (UTC)Reply
Indeed. And I don't blame you for it at all, wanting to rebuke an oppose you disagree with is perfectly normal. I do wonder if we should automatically put all chatter after an oppose on the talk page with a link to the section... WormTT · (talk) 12:28, 6 December 2011 (UTC)Reply
Well, holding the thought, it would not have offended or irked me one bit to have it all moved to the talk page. I wish I'd thought to be bold enough to do that myself. In this scenario, I become the latest in a long line of users to be the "walking proof" that we can't "just trust" even well meaning !voters to do the right thing in the heat of RFA. Begoontalk 12:35, 6 December 2011 (UTC)Reply
I think that anything more than 2 or 3 comments after an oppose should be moved to the talk page. One oppose and the badgering that follows shouldn't be able to take up my whole screen. —SW— squeal 16:08, 6 December 2011 (UTC)Reply
That I can agree with, and I suspect being quicker to do that would receive wide support. In addition, I'm not going to comment on the specifics of that example, and whether or not that's a valid oppose or not is up for debate; however, there are some instances where everyone, even other opposers, agrees the oppose is completely ridiculous. I'd argue that in such a case (i.e. I saw one once that was "I have to oppose for all the other people who would have" with nothing else) it should be moved to the talkpage and not count as a vote at all; if it contains unfounded insults (i.e. "[It doesn't seem you sufficiently value the time of project volunteers who aren't operating on your wavelength.") it should be removed altogether. Again, this will require knowing it when we see it, because it's not possible to come up with a hard list of things you can't say, but it's usually easy enough to see when it happens. That we haven't done anything about it in all instances doesn't mean the vast majority of us don't already. The Blade of the Northern Lights (話して下さい) 03:01, 7 December 2011 (UTC)Reply
I believe that clerking would lead to more propriety in the process. To what degree beyond simply maintaining the format of the page, checking for duplicate votes, socks, and banned users, or moving long theads to the talk page, etc., remains open to discussion (a list of suggested tasks is at WP:RFA/C). Some votes, (and some are even based on lies and deceit) are clearly not acceptable but if nothing is done, they can - and do - lead to pile-ons. and such votes are even far more toxic than entirely uncommitted ones. A participant who justifies their vote by expressing an opinion that something on a practically empty user page is blasphemous, clearly needs to be at least drawn to the importance of objective voting if they intend to take part in RfA more often; it's how we do it that will have effect. Oppose votes, for example, for reasons of a user's religious affiliation (or lack of it) are also out of order and totally irrelevant, although the community has been unable to reach a consensus. In the current climate of RfA, badgering is therefore a necessary evil in many cases, but as I have previously mentioned, whatever clerks actually do, the fact that RfA would be being watched more closely would almost certainly have some psychological impact on the behaviour of the voters. Kudpung กุดผึ้ง (talk) 03:49, 7 December 2011 (UTC)Reply
I've meant to respond to this for a while now; I completely agree. I don't know who else would be up to the task, but if the bureaucrats don't want to do it we'll need to find people. I'm not quite sure how we'd go about doing that; the one idea I can think of that wouldn't involve tedious elections is to have the crats select 4-5 people who they think are capable of handling it. But even that I'm not sure they'd go for, in which case I'm (for the moment at least) out of ideas. The Blade of the Northern Lights (話して下さい) 18:11, 19 December 2011 (UTC)Reply
I think we need to review everything we've discussed at WP:RFA/C and see if there's something there than can be proposed. However, that said, although I've never been in favour of radical alternatives, after the recent Arbcom elections I'm wondering now if a similar secure poll system would be worth trying, perhaps as a trial for a while where candidates could choose one or the other system. Admittedly this system amounts to little more than simple vote counting, but it would force noms and self-noms to make strong nom statements, and may just force participants to do some proper research of their own before voting, and cut out some of the trolling (which is now actually getting worse than ever). Kudpung กุดผึ้ง (talk) 04:56, 20 December 2011 (UTC)Reply
I doubt SecurePoll will work. It was tried for the 2010 Oversight/Checkuser elections, and it wound up, rather infamously, only getting one CU and no OS promoted. ArbCom was forced to intervene. —Jeremy v^_^v Components:V S M 10:03, 20 December 2011 (UTC)Reply
I don't think the 2010 election was much of a precedent either way for using SecurePoll at RfA because (as was discussed in the RfC afterwards) that election lacked RfA-style discussion and open commenting. I think there are two questions to be asked. First, should we consider having separate stages for commenting and for voting? This would mean that no votes would be cast until all comments were closed, and votes would not include comments. The question of using SecurePoll is secondary to that and raises different issues, e.g. as Risker said in the 2010 post-election discussion, "Voting, particularly secret balloting, tells us who is most popular. We should not care who is popular." - Pointillist (talk) 14:59 & 23:09, 20 December 2011 (UTC)Reply
Badgering the voters does not seem to be enough. Kudpung กุดผึ้ง (talk) 08:12, 20 December 2011 (UTC)Reply

A New Year's resolution

edit

Some people who don't appreciate all the hard work and discussion that's been going on at this project for the last 7 months are complaining that RFA2011 is nailed to its perch. What we have at least done is create the first ever focused set of discussions on RfA reform, confirmed what's blatantly wrong with the process, and come upwith a lot of ideas. I suggest we now finally get round to drafting some concrete proposals and throw them to RfC and see what happens. If one fails, we follow up with another, and so on every 30 days until we have a clear idea of what the community wants (or doesn't want). --Kudpung กุดผึ้ง (talk) 03:48, 21 December 2011 (UTC)Reply