Wikipedia talk:Community de-adminship/Draft RfC/two-phase and finalised polling discussion

Does 5.4 mean we need 'two-phase' polling at the final RFC?

What do people think about the final RFC being a two-phase poll? For me this is the only fair way to present the various support for 5.4. People were expressing different things in 5.4 - not necessarily voting for the final route. As I understood it, idea of 'consensus' (above strict percentages) was central to it.

Phase One could be:

Vote 1) Consensus-based decision made by bureaucrats, with a 70% baseline

Vote 2) Consensus-based decision made by bureaucrats, with a 60% baseline

Vote 3) Consensus-based decision made by bureaucrats, with no baseline specified (or changing 80% to 90% perhaps - Bureaucrats would find it hard to turn down 80% given the 'rule of thumb', and it could concivably be too low for some admins comfort. - added Matt Lewis (talk) 00:29, 15 January 2010 (UTC) ).

Vote 4) General Oppose

Voters can 'support' as many of the four as they like (including 'General Oppose' if they want to), so we needn't have opposes for each choice, as everything is covered (including those who wish to oppose all, but want a contingency vote too). We then toll them up. A dead heat would require a round of Talk-page 'second option?' requests. Alternatively, we could request leaving optional second preferences at Phase One, (though people could complain of over-complicating).

If the opposes 'have it', the whole CDA proposal ends there. Otherwise we (We then) proceed to Phase Two, and poll the most supported option using For and Against. A number of people (like myself) would actually vote 'For' for any of the above versions of CDA (although I have my personal preferences) - others would not. The key is that we have honed it down to the most favoured options: people need a chance to specify what the final poll is, before supporting what may not be their exact preference. The opposers have a chance to 'vote out' the CDA at each phase, which is fair I think. (struck all Oppose vote detail, as an oppose vote in Phase One could prejudice the other votes)

Given the voting data we have from 5.4 (and other of the various polls too), I think it is right that we make the final public RFC as foolproof as we can. CDA is no small change to Wikipedia, and it demands a reasonably comprehensive poll, which would make the whole RFC less vulnerable to criticism too (and it will certainly be ciriticed however it is done).

Any support for this? Matt Lewis (talk) 14:57, 11 January 2010 (UTC)

No, I disagree with you on this Matt, and I'm one who argued for more range for the 'crats on the lower end. My current take, given where we are now in the process, is that it will be best if we stick to the rough 70-80% range for 'crat discretion, just as in a Rfa, for simplicity. At this stage, less is more. Again, let's keep this as much as possible an easy-to-understand reverse Rfa, and stay focused on that for the Rfc period. Thanks. Jusdafax 16:53, 11 January 2010 (UTC)

But who has said we must follow the original text of 5.4? The support for it wasn't that clear. If we do, aren't you worred that 1) The final RFA can be far too-easily criticised, and 2) It could easily fail? What will happen then, other than it being deemed as proof that the 'community' does not want CDA? I cannot be sure at all that 70/80 is what people want - so am very uneasy about proposing it without some kind of process to find out what form of 'consensus' people prefer. Basically, we are not quite at the end of the refining process. Like some other late-comers, I don't want to be rushed on this.

Also - if people really stop and think about the 'reverse RfA' idea, they will see how illogical it is. RfA is an optimistic thing, CDA a very negative one. Editors aren't that daft, and I think it is laying up something that can be easily broken down. It's just not wise wise to focus on it, and simplicity needs to be achieved in the presentation I think. Matt Lewis (talk) 17:08, 11 January 2010 (UTC)

"Any support for this?" Not from me. I agree with Jusda. Let me clarify about the reverse RfA concept. The idea is that if the community can give the sysop bit, then the community should be able to take it away. Here, that manifests as there needing to be a clear sentiment of the community as a whole supporting a motion to desysop, not just the sentiment of a small and unrepresentative group of editors. I think everyone actually agrees that CSD should work in that way, in principle. Where we disagree is as to how to define community consensus operationally for this proposed process: is it "consensus" as determined by bureaucrats, or is it "consensus" determined by bureaucrats using a non-binding <70% and >80% guideline, as, merely, a non-binding guideline. That has already been discussed by editors, and the consensus was for the latter. Matt has explained his concerns, and appears not to be persuading the rest of us, so that consensus appears to stand. --Tryptofish (talk) 17:44, 11 January 2010 (UTC)

Thanks Trypto. Matt, a number of us have been commenting and !voting on this process/proposal, originally drafted and presented by admin Uncle G, for some time, and I believe we have reached the point of diminishing returns on it. 70/80 is easy to understand, and makes the bar to desysop just as high as it is to hand over the mop. There is logic in that. As I see it, lowering the bar from 70/80 gives more ammo to the critics who want the concept of CDA shot down. I understand as a late-comer you'd like more time, but there will always be late-comers, and my own deep concern is that we are past the point of reasonable ongoing discussion, and starting to dither. I urge we finish up asap, set a date and just do the danged RfC asap. If it fails, it fails. Jusdafax 17:54, 11 January 2010 (UTC)

Jusdafax - my proposal above is not about "changing to 60/80" though is it?! It is about using a fair two-phase poll to find out what people really want from 5.4. Wikipedia needs a 'parallel' admin-removal system yes, but there is no other "logic" in comparing 'handing over the mop' to taking it away. A parallel system would be to make an equally fair one - which does not mean simply copy the margin percentages. RFA is intrinsically optimistic - CDA is to stop-dead further offences being made! The consequences of 'failure' with CDA is very serious - as we can get a bad admin re-affirmed, and other admin/editors can resign etc in disgust. It has to be based on serious decision making. With RfA 'failure' is simply to discourage a potentially good admin, rather than have them try for adminship again - which we know they most-commonly do, as it is hard to pass it first time. They are just not comparable in terms of consensus margins.

I think the comment "If it fails, it fails" says everything, and I really don't share that attitude at all. In fact, I am getting very unhappy with the rush here guys - I feel like people's hopes have been raised unfairly, and I'm personally starting to wonder if I'm not wasting my time a bit. Where is everyone? Is the CDA seen as destined to fail now? I tell you something if it is - it won't make people like myself feel any better about Wikipedia. But then, of course - who gives a shit if we are happy or not? No-one is obliged to care about worker happiness, so no-one really does. A decent CDA can actually change that - a weak CDA could be useless at best, and the illusion of fairness at worse.

Another thing I haven't said about 70/80 - I think that in some cases people will simply not feel comfortable (and secure) enough about starting an CDA, if they feel it has to get 70% to be actually debated by the Bureaucrats. Peole will be much happier with there being a guaranteed informed 'consensus-based' decision on all CDAs that get through, with an wisely liberal margin (ie 60% - or even better - none at all). After all - there are plenty of safeguards to even getting a CDA on a bad admin going! If it passes them, it HAS to be taken seriously. Or else who will risk opening a CDA in difficult cases? Lets be honest - the completely obvious cases of admin 'gone bad' will be dealt with by admin anyway. CDA is for the more complicated less-obviously-apparent matters, which have enough public support. Ultimately, this is not so much about removing bad apples - it is really about how well/badly editors 'feel' about admin's accountability in everyday Wikipedia life. It is vital for Wikipedia that its editors feel that Administrators are at least 'potentially' removable, and not over-safeguarded with a practical 'job for life'.

All I'm asking is that we poll in two phases, to make sure we are voting on what people want. If 70/80 gets to be the final vote, so be it! We could really be messing it up at this stage guys - does anyone out there support me? To not poll intelligently/fairly, and to steamroller 70/80 through, I feel certain is risking too much. This could depend on whether you really want CDA to succeed or not. Matt Lewis (talk) 19:10, 11 January 2010 (UTC)

I also think the 'community present' is intrusting us here to properly represent them, and create a final CDA proposal that not just represents them, but has a real chance of 1) being taken seriously, 2) getting through. Matt Lewis (talk) 19:25, 11 January 2010 (UTC)

This is a good time for everyone to take a deep breath. The situation is not as dire as that. I think the "if it fails it fails" comment was said in haste, and no one intends this to fail. If other people speak up for an extra poll, then good. If not, then I hope Matt will accept that. Myself, I think we've already polled a lot, and we should respect what editors already said (which is what I'm trying to do, since I'm advocating something I originally tried to argue against, but which I've been persuaded to now support). Please let me suggest, instead, that editors simply comment here on how best to construct the "5.4" part of the proposal. --Tryptofish (talk) 19:34, 11 January 2010 (UTC)

It's no surprise that this issue clearly lies at the heart of the whole problem. Looking at the detailed discussion again, there are quite possibly as many who supported 5.4 who'd like 60% as there are who would support 70% (or higher). The difficulty is compounded by the regularly expressed view that being too specific with numbers is bad juju. I think the solution is to ensure the issue is spelled out clearly and ask two (and ONLY two) questions, in the reverse order suggested above. Something like:

Vote 1) General Support or Oppose

Vote 2) If enacted, 60% baseline or 70% baseline

That way round, if people give up half way through you can reasonably assume they don't care about the %age. Whether we like it or not I think 70% has to be the default setting however. Ben MacDui 19:36, 11 January 2010 (UTC)

I'm uncomfortable with that. I see what you are saying about the archived comments on 5.4, but I'm also sensitive to the archived comments on 5.3 (>50% threshold). If there was such resistance to >50%, are we really so sure it gets better at 60%? I think not. I also think there is a problem with presenting a proposal with that "vote 2" as part of it: the opposers are certain to say "look, this proposal isn't even finished, the people proposing it haven't thought it out, it should just be defeated". We all know they'll say that. Let's get this right before we put anything to a final vote. So, how many editors here really feel that we need to re-poll on this question at all? --Tryptofish (talk) 19:55, 11 January 2010 (UTC)

I don't actually think the resistance to 50% here will be necessarily mirrored in a public poll - and we are actually polling on 60/80, which has had support without the loud dissent. I also think we are far, far more likely to be accused of unfairly fixing onto a single route (ie the half-popular 70/80), than to be chastised for giving people actual options to vote on. We need to repoll this question as we simply don't have an answer to it yet. It's not up to us to just pick one. This is supposed to be about Bureaucrats finding consensus - we need to give them a decent margin to do it in. I'm sure they would find it a thankless task if they had to start at 70%, and would rather resist CDA altogether with that margin percent. Matt Lewis (talk) 20:31, 11 January 2010 (UTC)

I agree it is better to be clear, but what I hear Matt saying is that it isn't clear, and I can't abide the thought of another round of debate. Here is a suggested wording for the "Comments" Section per the project page here. I am not wedded to this, just trying to move the debate along. Ben MacDui 20:24, 11 January 2010 (UTC)

We may be at the point where it would be more productive to see what other editors say, instead of the 3 or 4 of us debating it. (Honestly, at this point, I think it is only one editor who is unhappy about the issue, but I'm prepared to be proven wrong if other comments appear.) My preference would be to have another pre-poll first, but finally to have a simple up or down !vote on a single finished proposal. --Tryptofish (talk) 20:38, 11 January 2010 (UTC)

There is not only the one editor who is unhappy about this issue - that is guaranteed! We'll certainly have to poll again for sure, and more voices are essential. I still can't see why the public can't be intrusted with non-simplistic matters, but we could refine further in here. Also, if we get another round of debate - so be it, surely? I do appreciate what you've all been through (none of you would believe what I've been through myself trying to improve Wikipedia), but are you all sure your attitudes aren't failing you at the very last? Matt Lewis (talk) 23:25, 11 January 2010 (UTC)

On the Vote 1, Vote 2 suggestion by MacDui (a little above) - I think whether people oppose CDA or not can depend on whether the baseline is 60 or 70%, and the outcome of vote 2 can effect vote 1. I think the only foolproof way is a two-phase poll of some kind. One phase polls can get very complicated when they are done fairly in matters like this. This RFA proposal of CDA the most important part - what everything has been leading up to, and I can't help feeling people are just desperate to get it over with!

POLL QUESTION: Perhaps we can run my two-phase poll above, but poll Phase 1 in here? (asking for second options, and without letting the Oppose votes end everything) - what do people think? We then take phase 2 (the baseline to be voted for/against) to the RFC - ie allow people to vote For/Against a CDA with a set baseline, and in doing so offer the community complete CDA proposal. How about that? We could advertise this final talk page poll the way the Motion to Close one was. Looking at the 'big names' in that should remind us all what we are up against. Matt Lewis (talk) 23:25, 11 January 2010 (UTC)

Well, as I said before, if we do decide to do more polling, I would hope we do the polling on the percent etc issue here as a first step, and then use the results from that to formulate an RfC with that question already agreed upon. I, for one, am not "desperate" for anything. My advice would be to see if more editors besides Matt come here (sans canvassing of course) over the next few days, and ask for more polling on this point. If they do, then let's do it. In the mean time, I suggest, again, that all take a deep breath, and focus for the moment on fixing up what we do agree upon. --Tryptofish (talk) 23:54, 11 January 2010 (UTC)

I am opposed to any further polling. There is no reason to suppose any more people would turn up than did so in the last round and (insofar as most editors care either way) I think it will be seen as putting off answering a fairly straightforward question. Re "I think whether people oppose CDA or not can depend on whether the baseline is 60 or 70%, and the outcome of vote 2 can effect vote 1." Interesting - I'd like to explore the interactions here.

Assuming everyone answers both questions:

if I am opposed to the principle I'm likely to !vote oppose/70%

if I am supportive of the principle I'm likely to !vote support/60%

there will be those who genuinely prefer support/70% - hard to know how many

if I am tending towards a conservative view of the %age (i.e. higher number) and see lots of support/60% I might well oppose.

if I am tending towards a radical view of the %age (i.e. lower number) and see lots of support/70% I might well oppose.

It is very hard to predict - which may be an advantage as it is hard for any us here to game it, consciously or otherwise. I do see an advantage of the two question approach is that it reduces the numbers in the fourth group from the outset and if that increases the numbers in the third group, surely that is an honest outcome.

An intriguing puzzle. Ben MacDui 17:58, 13 January 2010 (UTC)

And a difficult puzzle (at least for me)! It's taken me a lot of thought to wrap my mind around it, but I think I understand it now. My take is that you actually point out a potential problem we could run into. If we go with two questions, people may look at the 60/70 question results-to-date, and try to game their overall support/opposition depending on which way it's going. The two-question approach just begs for unintended (bad) consequences. I really want to push now for a simple, single-question RfC. Something that has occurred to me over the last day or two is that the 60% version raises a question that could be difficult to answer: why is the number 10% lower for CDA than for RfA? Even though I started out as a fan of a much lower threshold, I definitely see now (based very much on what editors have said in the polling we've already done) a very attractive argument that the language about percents in CDA is the same as the language in RfA. Unless more editors comment here in the next couple of days that they want 60%, I would really advocate going with single-question 70%, and without needing another poll first. --Tryptofish (talk) 20:29, 13 January 2010 (UTC)

CDA has nothing to do with WP:RFA! PLEASE UNDERSTAND THAT! It is vitally important. CDA is not a de-election, and it is a real con to suggest that it is. CDA is for the community to advance the removal of a clearly destructive admin! A 70% baseline couldn't be more different for each one. Also, basic logic shows that you need at least two phases to poll three questions (60/70/oppose?). Sorry guys - but that should be obvious to both of you, but you are trying to fit square pegs in round holes in all manner of ways to end all this. The rushed attitude in here is really awful to me - sorry if that sounds rude, but it's genuinely from the heart. We MUST poll on this matter - we CANNOT just 'run' with one. We can do the first polling phase here, and the next (ie the answer) to the public at the RFC, or we can poll both phases to the public at the RFC. Matt Lewis (talk) 22:50, 13 January 2010 (UTC)

Then poll. Don't do an RfC that is badly designed, designed to fail. --Tryptofish (talk) 22:56, 13 January 2010 (UTC)

I'll prepare it now, and poll it tommorrow. I agree that a simpler RFC is better in terms of encouraging votes. Matt Lewis (talk) 23:35, 13 January 2010 (UTC)

Please be sure to check with MacDui. --Tryptofish (talk) 23:42, 13 January 2010 (UTC)

Don't understand how "(60/70/oppose?)" is 3 questions.

I'm also not sure where the "rushed attitude" issue comes in. I've said that I am not in favour of another separate poll and whilst I believe we need to launch the RfC by the end of this month, we have already extended the draft discussion and no date has yet been set. Part of my concern is that in another "poll" there is no reason to suppose someone won't propose 40%, 90% and/or 67.45% and it is unlikely the same group of people would participate in three different discussions on the same subject, making it increasingly hard to bring coherence to the process. Ben MacDui 17:01, 14 January 2010 (UTC)

Proposing 'final phase' of this draft RFC

• "70%", "60%" and "Oppose CDA" don't strictly have to be "three questions", but you certainly have to poll them in two phases. The matter is just too complicated to poll them in a single phase. Asking "60% or 70%?" would mean the Oppose? vote could (and surely would) prejudice the percentage votes, eg. people could vote (or change their existing vote) to 70% if the Opposes get 'dangerously' high. Right now I'm inclined to think that opposing CDA could actually be better than settling for 70% (as it is so high it poses all kinds of problems) - but I might have gone the 'switching allegiance' way myself, had I rather seen a 70% baseline than vote 60% and see CDA completely fail. If the eventual public poll in is two phases of course, we may as well do it properly and poll "60 or 70%?" in 'phase one', and then poll the winner in 'phase two' as a For and Oppose. (It was a mistake of my own to put an unnecessary extra 'oppose' vote in my Phase One, as of course it causes the problems I've detailed above). We need to ensure that the poll simply allows people to vote for what they want without prejudice - or it will be ripped into, and rightly so. The alternative for us is to poll a 'phase one' poll in here asap (without the 'oppose' option of course), which will provide us with a single percentage to poll to the public.

• This has to be over when it's over. If we are curtailing things in any way at this point, whether or not we have set the pending RFC date yet is simply irrelevant.

• There is a possibility that the results will be split, so I suggest polling a realistic range for people to chose from (50%, 60%, 66%, 70% and 75% for balance), and asking for first (and) second and third preferences (if people have them), to take into account if the winning percentage is too close to second-place for comfort.

• Personally, I rather we didn't have the 80% figure (essentially an 'auto-desysop', as Bureacrats I think would find it hard to support an admin who failed this "rule of thumb"). The CDA proposal we are working on (5.4) is principally about Bureaucrats finding consensus, in my eyes at least. Adminship is hardly a paid job after all (no comments please), and 80% would certainly mean the 'community trust' is gone, whether that is truly fair or not. But adminship is a fiercely-protected thing of course, and even 80% could be an issue to some alas. It could be wise to add to the poll "change the auto-desysop percentage from 80% to 90% (yes/no)?" (90% being more of a token percentage perhaps, but one that could allay fears) – I am suggesting adding that here. For me, an eventuality like 60-90% does seem like a more reasonable 'consensus-finding range' for Bureaucrats to 'fish out' the mistakes and unfair prejudices from, should they need to.

• I originally started this section at the bottom of the page to grab attention. Are people going to see it moved up here? I do feel there has been something of a 'dimmed light' and a near-closed door (if not an entirely locked one) since 4th Jan. I think I've discussed things very reasonably, but it's now time for some real light to be cast on this matter again, in the time-honoured way of getting some decisive input...

CDA 'Finalisation Poll'

(proposal for polling asap)

The CDA proposal which had the most support was 5.4 (essentially, that Bureaucrats look for consensus within a 70% to 80% "rule of thumb" margin). Amongst those who support proposing a CDA proposal to the community, there is a consensus is that we utilise 5.4. (The finalised CDA will be proposed at an eventual RfC). However, the full support for 5.4 is not clear, especially on using 70% as the 'baseline' percentage. Percentage issues were posed by people who supported and opposed the principle of 5.4, and the most-suggested ammendment (which came from the those in support of it) was to change the baseline to 60%. Also, since Jan 4th, it has been suggested that another 'phase' is needed at this 'Draft RfC' stage, to gather more information from people, and to facilitate the finalisation of the proposal.

This poll (which hopefully will constitute the final phase) is intended to complete the CDA proposal. This poll will run for 7 days, after which the proposal will be finalised, and put to the community at RfC.

Please note, this is not the place to oppose the idea of CDA - that can be done at the final RfC. Please leave comments in the "Comments" section below, and not in the poll itself.

Vote 1:

Do you prefer a 'baseline' percentage of; 50%, 60%, 66%, 70% or 75%?

As a "rule of thumb", the Bureaucrats will need to see this 'baseline' percentage reached before they consider whether the Administrator standing under CDA should be de-sysoped.

Please give a single-value second preference if you can, in the format; "75 (70) - signed", or "66 (60 or 50) - signed" etc.

* 50 (60) – ConcernedEditor*

* 60 (50 or 66) - LiberalWikipedian*

* 75 – ConcernedAdmin*

Vote 2:

Do you prefer a 'desysop threshold' of 80% or 90%?

As a "rule of thumb", the Bureaucrats will automatically de-sysop the Administrator standing under CDA if the percentage reaches this 'threshold'.

* 80 – ConcernedEditor*

* 90 - LiberalWikipedian*

* 90 – ConcernedAdmin*

(* these are examples of how it could pan out, not intended as examples to provide in the actual poll)

• What do you think? Please bear in mind that I feel so strongly that a poll is needed, that I will be making one of some kind or other. If people are to be aroused from their slumbers, it may as well be a decent and enlightening one. If this is to be the last opportunity before the final jump, we need to get it right. Matt Lewis (talk) 00:10, 15 January 2010 (UTC)

Looks like I missed my promised "tomorrow" by 10 minutes (and the time to discuss it) - sorry! Matt Lewis (talk) 00:16, 15 January 2010 (UTC)

About doing it or not, I am kind of neutral, but do not have any objection. But I do feel strongly that no decision should be made without first hearing back from MacDui. If we do it, I'd like to suggest the following polishes to the wording of the poll questions. First, we need to be sure to provide a link to the archived discussion of the options that have already been discussed, for people to read if they are unfamiliar with it. Then, I think "baseline percentage" is confusing to people who haven't followed all the intricacies of this debate. I also think giving the option of non-factor-of-10 numbers makes things more complicated than they need to be. And I think we can assume that there will be a 10-point spread between the lower and the higher number. So, I would go with a single question. I would list the following choices:

• 50%-60%
• 60%-70%
• 70%-80%
• 80%-90%

and for each, let people indicate "first choice", "second choice", "third choice", or "fourth choice", with brief comments if they wish. --Tryptofish (talk) 00:43, 15 January 2010 (UTC)

Thanks for replying. I'm shattered right now so I'll address this tomorrow, other than to say you and MacDui can polish things as much as you like. MacDui seems to be neutral like yourself, but I won't go ahead until he comments for sure (obviously things are best if we are all in tune - consensus at this minute is a small crowd indeed!). Re 66% I used it as it is a two-thirds margin, and I imagined some awkward blighter demanding it! Who knows, it may even be the one to get through. The 'balancing' 75 is three quarters, and I couldn't use 80 obviously. Can't think what you mean by making them ranges at the moment. The choice thing I didn't think of - I'll have to think about that tomo. Matt Lewis (talk)

And thank you for working on it. Yeah, I think we both should get some sleep! I think MacDui said he has strong reservations, so do let's wait for him. --Tryptofish (talk) 01:13, 15 January 2010 (UTC)

I've made some revisions, mainly cursory. Hopefully it is close to something we can poll soon. I suggest informing all previously-interested parties in the way the Motion to Close was. I actually didn't get that notice (I came in too late and found all this by chance), so obviously we need to maximise all the allowable routes. As this isn't a FOR/AGAINST poll as such, and as the CDA in itself is now (essentially) one single proposal that can be either voted 'in' or 'out' at the RfC, I cannot personally see how WP:CANVAS laws can come into play now. Everyone should know about this in my opinion. I'd even put the final RfC in the watchlist 'notice space' if I could. I think there should be a specific place on Wikipedia just for major public votes (ie not RfC - which is clearly both over-used and underplayed), and polls in that new place should be able to make use of the top-page noticespace. Matt Lewis (talk) 13:39, 15 January 2010 (UTC)

Those following this page closely know my feelings; I will make it clear again for anyone else. The percentage to de-admin should be the same as it takes to make an admin... that roughly between 70 and 80 percent is at bureaucrat discretion, below 65 is no consensus (but could be taken up by ArbCom) and over 80% !vote to de-admin is clearly take-the-mop time. Keep it simple, please. My view: Matt is well-meaning but is needlessly complicating this. Time to start the RfC should be in the next 1-2 weeks max. Let's go, please. Jusdafax 17:08, 15 January 2010 (UTC)

If you like 65% then vote for 66%! We can't keep it simple, because people aren't 'stupid'. Sorry - I just cannot see how CDA is comparable to RfA at all - they both do completely different things, and CDA is nothing like a reverse process of RfA. I've written more on that to your comment on it below. I'm doing this for 4 reasons: 1) 70% was not a clear consensus, 2) that would make the final poll easily ciritcised, 3) I don't personally think 70% will win, 4) I do think that 70% could be gamable, lead to a false sense of accountability, and thus damage WIkiedia. This can be polled today, and the RfC can still be in 2 weeks max.Matt Lewis (talk) 18:29, 15 January 2010 (UTC)

Matt, I appreciate that you really mean well, but I have to say a few things here. First, I don't think you really took in what I suggested about re-wording the poll, assuming we do a poll at all; it's more like you are just re-tweaking your own wording. More importantly, I agree with something Jehochman says on this page: that we need to be careful about not just talking amongst a few of ourselves in an echo-chamber. I've been trying to keep an open mind and a neutral stance about whether or not to poll. But, reading all of the comments that have been coming in to this talk, I really do not see anyone besides Matt advocating the poll, and quite a few people advocating against it. Speaking as someone who has been relatively receptive to polling, I have to say that when I read consensus here, the consensus is to go forward with 70-80, and ask the community to respond to that. As Jusdafax just said, keep it simple. At this point, I do not see a compelling reason to poll, and I think continuing to push for a poll may be unhelpful, however well-meaning it is. --Tryptofish (talk) 18:35, 15 January 2010 (UTC)

I didn't see any suggested re-wording, so obviously I tweaked it myself.

I just don't understand the 'keep it simple' philosophy here - why? The poll isn't complex at all, but it is clearly needed to prove that we have a genuine concensus and are not simply a closed-shop 'echo chamber'. We already have had criticism below that too little collaboration has been seen in here - the only way around that is to cover all the angles. Nobody can complain then. We are simply laying ourselves up for masses of criticism if we don't - and it would be fair criticism too. In the light of Jehochman's comment here, as soon as MacDui responds, I'll be polling something. I've had support off this page for it. For some reason people can't get to grips with this page. But it IS a long and difficul page. Matt Lewis (talk) 19:24, 15 January 2010 (UTC)

There are two issues at hand.

First of all the general principle - if there community support for WP:CDA to be enacted as described above?

Secondly, there is the less clear-cut issue of the way the outcome shall be determined, expressed as a guideline in percentage terms. Four separate ideas were discussed at Wikipedia talk:Community de-adminship/Draft RfC of which only one received consensus, namely that for an Administrator to be de-sysopped a minimum of 50 editors should take part and that as "a general descriptive rule of thumb" more than 70% in favour of de-sysopping would be required. The wording is deliberately phrased to allow the closing Bureaucrat leeway, just as at an RfA. However, some participants expressed an opinion in favour of this guideline being more than 60% in favour of de-sysopping (an option that was not formally discussed). Choosing between these two options is then the second question to be addressed. (Note that both 30% and 50% were debated and rejected. An option for 80% is, in practical terms, more-or-less opposition to the process as a whole. 70% is the default position and consensus is required to change this.) Ben MacDui 20:24, 11 January 2010 (UTC)