Wikipedia talk:Pending changes/Straw poll/Archive 1

Archive 1

Archive 2

Archive 3

Tally, Consensus %, and Other Stuff

This header was added due to the table of contents appearing and was meant to be a summary of the items underneath.

Previous discussion here. —fetch·comms 00:06, 22 August 2010 (UTC)

• Tally added. Ғяіᴅaз'§Đоом | Spare your time? 00:43, 22 August 2010 (UTC)

The page's self-appointed guardian has removed the tally, without any discussion beforehand, and os edit warring to keep it removed. Not good. - BilCat (talk) 00:54, 22 August 2010 (UTC)

Don't you think it is a bit soon to bother with a tally, I will replace it if you feel it is valuable, I don't but as you do I will replace it. Off2riorob (talk) 01:00, 22 August 2010 (UTC)

It helps sure it's not needed now but better to get over and done with. Ғяіᴅaз'§Đоом | Spare your time? 01:02, 22 August 2010 (UTC)

OK, as I have been busy with this I am likely a little attached and am hopeful it will give a good feedback for the community. Thanks for joining in. Off2riorob (talk) 01:08, 22 August 2010 (UTC)

(edit conflict) You're welcome. Ok first thing's first the vote comment guidelines concern raised by Mkativerata should be dealt with, I think that we should bring in an admin or crat involved with PC, preferably User:RobLa since he is this extension's developer (I think). Ғяіᴅaз'§Đоом | Spare your time? 01:12, 22 August 2010 (UTC)

When did 60% become consensus? Isn't "consensus" usually considered to be 70% - 80%? --Yair rand (talk) 04:02, 22 August 2010 (UTC)

In many places seventy to eighty percent would be considered a landslide, we are only looking to either continue the use of a tool that didn't cause the wheels to drop of in the last two months and also won't make the wheels drop of in the foreseeable future either. Off2riorob (talk) 04:17, 22 August 2010 (UTC)

This is something that has the potential to affect the whole of Wikipedia, Off2riorob. 60% is too low a threshold; I think 70% would indeed be the minima for something that has the potential to affect the Wiki. (Also the wheels analogy can be seen as bad; bear in mind the US/Canadian school year starts during the vote, and with them comes a noted increase in vandalism.) —Jeremy ^{(v^_^v Carl Johnson)} 04:28, 22 August 2010 (UTC)

The usual threshold is 70%-80% for consensus, and those at or below 75% generally fail; I have no idea where the 6 v 4 comes from. Regardless, has anyone contacted the foundation on this? This is all for naught if they step in. Ryan Norton 05:29, 22 August 2010 (UTC)

Any kind of support is good. If sixty people stand over there and only forty stand over here then the guys over there have the weight to go forward, easy, its not a dangerous thing, it has been in use for two months and claiming it needs seventy people to support it for every thirty that object to is is excessive indeed. Claiming that this needs over seventy five people to support it if only 25 people do not support it is a joke. Off2riorob (talk) 05:42, 22 August 2010 (UTC)

It doesn't seem like a joke to me. Consensus has always been considered to be about 75%. --Yair rand (talk) 05:51, 22 August 2010 (UTC)

It's no joke, it is the way things have been done here for at least the almost 6 years I've been "hanging around here" with a registered account. I don't see why it is such a heated debate, this has been common practice since WP's founding for the most part... Ryan Norton 06:02, 22 August 2010 (UTC)

I'm going to have to go with Yair and Ryan on this one. Even in the case of the initial pending changes proposal, which only had temporary implications, a supermajority of ≥67% was necessary for approval. And more serious votes have historically required even more precedent than that. WP:RfA, for example, states that "[consensus] is sometimes difficult to ascertain and is not a numerical measurement, but as a general descriptive rule of thumb most of those above ~80% approval pass; most of those below ~70% fail, and the area between is subject to bureaucratic discretion."   — C M B J   07:08, 22 August 2010 (UTC)

The decision to make arbcom elections secret was passed with only 57% support. Rami R 11:55, 22 August 2010 (UTC)

Less it be misunderstood from my last comment, I firmly oppose a straight vote. I have no problem with a vote-like procedure, but at its conclusion an uninvolved admin judges the consensus without reference percentages, like in normal policy RFCs. Rami R 14:43, 22 August 2010 (UTC)

Random guess; that would be entirely impractical here. I imagine there will be a massive lot of "vote commenting". Figuring out a consensus without considering percentages at all seems impractical. --Errant Tmorton166^(Talk) 16:32, 22 August 2010 (UTC)

Just so you know, the proposal to turn on the feature passed at 59%, which I opposed because a) Jimbo closed the same poll he voted in, and b) 60% is too low. I'd consider two-thirds to be reasonable for a rough consensus. Sceptre ^(talk) 20:03, 22 August 2010 (UTC)

• Thats right two thirds that is exactly where I set it,

6 - 4

six from four is sixty six point six percent. That imo was and still a fair level as a minimum support to accept the continuation of this tool. Off2riorob (talk) 20:09, 22 August 2010 (UTC)

No, chummer, your math is still incredibly flawed; that's 60%. Did you even read what OverlordQ and I said on my talk page?! —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:11, 22 August 2010 (UTC)

• - Six from 4, you don't add them together and get ten...Its six from four....not sixty fourty, six from four...Off2riorob (talk) 20:14, 22 August 2010 (UTC)

• The way it's writ it looks like 60-40, Off. Take OverlordQ's advice. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:17, 22 August 2010 (UTC)

• - its very simple mathematics, 6 - 4, to work your percentage out ... four divided by six multiplied by one hundred equals sixty six point six percent. Off2riorob (talk) 20:20, 22 August 2010 (UTC)

• Off2riorob, do you deliberately not listen? "6-4" is not the appropriate way to express a damn percentage! Me and OverlordQ both told you this, and you insist on this confusing bullshit that any normal human being will construe as 60%? —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:23, 22 August 2010 (UTC)

I hate to ask a simple question in the middle of all this fighting, but where does it say 6-4 or 60% or anything else? I can't find it.--Bbb23 (talk) 20:25, 22 August 2010 (UTC)

Off2riorob initially included it in the lede. It was removed (multiple times) later as everyone here construed it like any other human being would, as 60% - too low for something that has the potential to damage the wiki. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:26, 22 August 2010 (UTC)

Aaah, thanks, found the sentence: "Consensus will be esablished [sic] by an Administrator by counting the votes (as there are no policy decisions to be considered) on the basis that 6 - 4 is a minimum consensus agreement." Now that I've seen it, and FWIW, I agree with Jeske, although I might not express myself quite the way he does. If Off2riorob is trying to express a percentage minimum for consensus, the usual way to do that is with a percentage, not an arithmetic structure that is supposed to resolve itself into a percentage. Besides, the math doesn't even work because it's "6-4", which is 2, not 2/3, at least in my math. And even if it's supposed to be "4-6", that would be -2, also not 2/3. Now, "4/6" would be in fact 2/3, but I've already devoted far more space to this than it deserves. Happy fighting.--Bbb23 (talk) 20:38, 22 August 2010 (UTC)

I don't care about the way it has been presented, 6-4 or 6:4 or 60-40, whatever. But what I do care about is that the actual percentage be made clear and unambiguous. Is it going to be 60%, 75%, 80%, or something else completely? Please make this clear. Thanks, HumphreyW (talk) 20:36, 22 August 2010 (UTC)

Agreed. Maybe this is different in other English speaking countries but int he States we don't say "six from four" to represent 4/6 (which is 2/3 and why on earth would you not simplify it as such). "Six from four" would mean to subtract 6 from 4 (which is -2). Just make it clear that it's a percentage and what percentage that is. Honestly, the more this strange obfuscation goes on, the more I want to just vote to kill this beast. It's starting to reek of hidden agenda. Millahnna (mouse)talk 20:46, 22 August 2010 (UTC)

• - The original intention was for two thirds, sixty six point six percent as a minimum to keep the tool in operation. Off2riorob (talk) 21:35, 22 August 2010 (UTC)

Concern over voting methodology

Concern has been expressed over the voting methodology. Pending Changes will either be discontinued, or continue in one of three forms. Will the outcome of the vote reflect community consensus regarding what happens next with Pending Changes? PL290 (talk) 16:18, 22 August 2010 (UTC)

I hate to sound like a whiner, but the way the vote is set up at present it looks like a show trial. —Jeremy ^{(v^_^v Carl Johnson)} 04:07, 22 August 2010 (UTC)

If sixty users support it and only forty users object that is a clear consensus for the continued use of this tool. Off2riorob (talk) 04:13, 22 August 2010 (UTC)

The problem is the options. One option against three combined ones is nowhere near fair as far as voting goes, which is why I propose that, at least for this first option, we use a straight up-or-down voting method; then we switch to specific options regarding PC should it pass. —Jeremy ^{(v^_^v Carl Johnson)} 04:19, 22 August 2010 (UTC)

@Offtoriorob: It isn't, though. If 40% support Option 1 (no Pending Changes), and 30% support Option 4 (big expansion), and 30% support Option 2(don't end Pending Changes, but don't expand it, either), it's disingenuous to claim that the Option 2 voters have more in common with one side over the other, and claiming to find any consensus in such numbers would be nonsense. That's why it's important to have a higher, clearer threshold. — Gavia immer (talk) 04:25, 22 August 2010 (UTC)

The other problem is the math as noted above. If 60 users support keeping and 40 users oppose it and we are going by a 2/3 vote then you need to double check that figure. 60 users out of 100 is NOT 2/3 (or 66%) but merely 60%. Millahnna (mouse)talk 21:35, 22 August 2010 (UTC)

There are two basic options, keep or remove, Off2riorob (talk) 04:22, 22 August 2010 (UTC)

Not the way that you had originally set it up. The way you had set it up had given the appearance that there were four options, three of which would be tallied together. (FWIW, I prefer the change Cyclone just made.) —Jeremy ^{(v^_^v Carl Johnson)} 04:23, 22 August 2010 (UTC)

There are two basic options, keep or remove, its a simple tool, we can also get rid of it later or expand its use later, its not a massive issue, imo. Off2riorob (talk) 04:24, 22 August 2010 (UTC)

Is that a dodge, Off2riorob? The way you had originally set it up was flatly biased in the absence of CycloneGW's edit just now! —Jeremy ^{(v^_^v Carl Johnson)} 04:26, 22 August 2010 (UTC)

No dodging here, I want this to work. Off2riorob (talk) 04:28, 22 August 2010 (UTC)

You're dodging the bloody issue! Answer me, Off2riorob: Did you or did you not set it up in such a way that the polling was fundamentally biased against one option? —Jeremy ^{(v^_^v Carl Johnson)} 04:30, 22 August 2010 (UTC)

No. please don't shout and bold. thanks,Off2riorob (talk) 04:32, 22 August 2010 (UTC)

I do not believe a bias was intended. He simply was trying to kill two birds with one stone. The problem is that the oppose votes aren't considered in the second half of the voting. CycloneGU (talk) 04:39, 22 August 2010 (UTC)

Regardless of whether bias was intended, the bias remains. See, for instance, the comment by User:GW Simulations, who says,

I am voting to close. I would also like to state my opposition to the way that this poll is being conducted; since all !votes for options 2-4 will be counted together, it will provide a bias in favour of those results. If, for example, someone !votes for option 2, but they dislike options 3 and 4, then their !vote for option 2 would count towards all three when the "keep" and "close" votes are compared, and therefore it could help to cause implementation of the proposal to which they are opposed.

The results should be split into a separate vote for each of the four options. PL290 (talk) 11:42, 22 August 2010 (UTC)

Excuse me? Fine, let's assume it's just an up and down vote. 65% say yes. We then need a separate two week poll determining the best way to continue implementation. Those who initially opposed - while a few might chime in to leave it alone for another month or two, and try a vote again then (stalling the process) - will otherwise abstain. So this poll is also getting a consensus from the supporters of the measure as to how we want to have the trial continue, up to a sitewide BLP implementation. This poll takes care of both issues at once; just because 15% like option 2, 30% like option 3, and 20% like option 4 doesn't mean the 35% for option 1 should automatically defeat them when 65% of voters support the measure!

Disagree. Take a user whose choice is "1 or 4, but nothing in between". That user has voted in both places. In your scenario, should "options 2 to 4" be most popular, and within that, option 3 have the most votes, you have artificially used that user's vote—and not only have you failed to give it the proper weight towards option 1, but you have actually used it to give undue weight to option 3, and bring about option 3 despite that user's explicit opposition to option 3. The results should be split into four sections, one per option—and as soon as possible, because the sooner it's done, the more chance there is of getting an unbiased result. PL290 (talk) 14:05, 22 August 2010 (UTC)

I am not really bothered about this tool, I am more bothered just that the community is able to opine about it in a decent way. It looks to be all set up in a good way and this is only the first day of fourteen days still time to change our minds. Off2riorob (talk) 04:48, 22 August 2010 (UTC) Off2riorob (talk) 04:46, 22 August 2010 (UTC)

I'm sorry, even with the fix (thanks for splitting up yeas and nays at least) this is still confusing and doesn't do what it's intended to do -- gauge consensus. There's should be two questions: 1) simple yay or nay on continuing and 2) IF CONTINUED, what changes (if any) are needed. As it is presently constructed 3 of the 4 options keeps it, that's stacking towards keep. My suggestion allows people who vote no a way to express what changes they would see if the consensus was to keep it. Otherwise the nays have no way of conveying where they think improvements are needed and obviously they think there are a lot if they're voting no. I would strongly encourage a stop and do-over for this. --WGFinley (talk) 16:54, 22 August 2010 (UTC)

This issue is discussed in the section labeled "?" (in which my comments are similar to yours). —David Levy 17:18, 22 August 2010 (UTC)

To get the best community consensus, use Approval Voting. Each person votes for "Yes" or "No" for each of the available options. The option that gathers the most votes has the most community support. With Approval Voting every voter can express an opinion on every option so you tend to eliminate most bias. (Of course, according to Arrow's law every voting system is flawed to some degree and subject to bias effects. It just seems to me that Approval Voting, or perhaps Range Voting where you assign points to each option (i.e, rate 0 to 9), would likely work best for this situation.) Rockmhoward (talk) 22:03, 5 September 2010 (UTC)

Changing Style

I originally raised concerns over the style of the page. I'm now going to change it. So far, there seems to be edit warring over the tally methodology being needed; I'm making the tally unnecessary by making the page use a numbered style, like RfAs. CycloneGU (talk) 04:16, 22 August 2010 (UTC)

There. All done. The tallying is now being done for us, and I don't think it's too confusing; all three keep options are together, the close (oppose) is by itself. If this were to switch to straight up and down, just remove the numbers from the comments and simplify the header. Thoughts? CycloneGU (talk) 04:21, 22 August 2010 (UTC)

Much better. —Jeremy ^{(v^_^v Carl Johnson)} 04:24, 22 August 2010 (UTC)

Yes, much better. thanks. Off2riorob (talk) 04:27, 22 August 2010 (UTC)

Sitenotice

Shouldn't there be a sitewide notice for this vote? —Jeremy ^{(v^_^v Carl Johnson)} 04:37, 22 August 2010 (UTC)

A vote with this serious of implications should warrant promotion to a site-wide notice. I have initiated a discussion at WP:AN accordingly.   — C M B J   07:11, 22 August 2010 (UTC)

?

According to the current setup of this vote, it seems that if there is consensus to keep PC, those against the use of Pending Changes would have no say in how it is used. Are "1" votes considered to be in favor of option 2 if the vote passes? --Yair rand (talk) 04:41, 22 August 2010 (UTC)

Fuck no for me. It's Option 1 or nothing. (I want no part of FlaggedRevs, as my user talk page will attest.) —Jeremy ^{(v^_^v Carl Johnson)} 04:43, 22 August 2010 (UTC)

Uh, I think you might have misunderstood my question. I am asking, if consensus exists for keeping pending changes, should it be assumed that all who voted against pending changes would prefer as minimal use of pending changes as possible rather than more frequent use of pending changes? --Yair rand (talk) 04:50, 22 August 2010 (UTC)

And again I say no. Let them vote in the "what to do" poll afterwards as they will. Do not automatically assume that they are all for FraggedRevs - particularly myself, as I view even the assumption I will support it as a personal attack. —Jeremy ^{(v^_^v Carl Johnson)} 05:00, 22 August 2010 (UTC)

I brought up this question, myself. I presumed they'd still have a chance to vote for the best way to try to implement it if the consensus is keep. Thing is, most would choose option 2 or otherwise wash the matter from their hands, so to speak. =) CycloneGU (talk) 04:57, 22 August 2010 (UTC)

The whole idea of this is to allow users an idea of understanding what they are supporting, users were moaning, well if I support what am I supporting and now we have it, again, this is not an terrible crime...it has been in operation for the last two months, there is nothing for us to worry about. If there is not consensus support for the tool in this discussion at least we have tried it wiki-live and we understand how it works a bit more and perhaps can use it in the future. Off2riorob (talk) 05:11, 22 August 2010 (UTC)

The layout I agree is clearer now a tally is not needed and then consensus between Keep and Oppose clearly established, I don't know why I engaged in that petty edit war in the first place, but really no one owns this vote comment thread. I apologise to anyone effected by the edit war Ғяіᴅaз'§Đоом | Spare your time? 05:47, 22 August 2010 (UTC)

In my opinion, the poll should have been configured with two separate questions to the effect of "Should the pending changes feature be retained?" and "If the pending changes feature is retained, what course should be taken?"

This would have been equally simple (the same number of sections) and easier to understand (no combining of votes or ambiguities), and it would have enabled those opposed to the feature's retention to explicitly voice their opinions on how to proceed if it isn't dropped.

Because it's too late to adopt such a format, I agree that we should assume that those selecting option 1 would prefer option 2 over options 3 and 4 (and count their votes accordingly if there is consensus to retain the feature). I understand your opposition to this, Jeremy, but you need to think of it as a vote against expansion. —David Levy 16:33, 22 August 2010 (UTC)

Absofuckinglutely not, David. I'm not supporting FraggedRevs in any way, even if it passes. As far as I'm concerned, my vote is for abolition, not nodding assent. —Jeremy ^{(v^_^v Carl Johnson)} 18:54, 22 August 2010 (UTC)

No one is proposing that your input be construed as any sort of support. The point is that your opposition shouldn't be disregarded.

Opposition to the feature's retention encompasses opposition to its expansion, so it should be counted as such. —David Levy 19:58, 22 August 2010 (UTC)

Absolutely not, David! I OUTRIGHT REFUSE to support *ANY* implementation of FraggedRevs, even so much as having it on just one article in Siberia! I DO NOT support FraggedRevs AT ALL! What the FUCK is so hard for you to understand about this?! —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:01, 22 August 2010 (UTC)

Again, no one seeks to construe your input as any sort of support for the feature. The idea is to respect (and take into account) your opposition to it.

The question is "Should the feature's implementation be expanded?", and I'm saying that your response should be counted as "no" (instead of being ignored). —David Levy 20:16, 22 August 2010 (UTC)

Then why are you contruing my "nay" vote as a vote supporting FlaggedRevs at all, David? I WANT NO PART OF FLAGGED REVS. PERIOD. THIS INCLUDES WHAT IS DECIDED ON IF OFF2RIOROB GETS HIS WAY AND THIS EXTREMELY BIASED STRAW POLL IS USED IN LIEU OF CONSENSUS. PERIOD. IF ACCEPTED MY VOTE SHOULD NOT COUNT AS FAR AS DETERMINING WHATEVER HAPPENS AFTERWARDS. PERIOD. If I wanted to vote for fucking FraggedRevs I would have voted for fucking FraggedRevs. Do not assume my opposition is conditional.—Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:20, 22 August 2010 (UTC)

I've plainly stated several times that I'm not construing your response as a vote supporting FlaggedRevs at all. I honestly don't know how to make myself any clearer on this point. —David Levy 20:41, 22 August 2010 (UTC)

Yet at the same time you say that if it is accepted it should default to a 2. That's bullshit, Levy. If I wanted FraggedRevs... I WOULD HAVE VOTED FOR FUCKING FLAGGEDREVS IN THE FIRST PLACE! Do not construe my "nay" vote as a fucking "support" one, dammit! —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:44, 22 August 2010 (UTC)

That isn't what I'm doing. I'm sorry that we're unable to see eye-to-eye. —David Levy 21:11, 22 August 2010 (UTC)

Let me make an attempt here. Jeremy, I think he means that if there is equivalent support to the measure continuing, and you were FORCED to pick between three options, it would be assumed you would not want to expand it but leave it as it is. Does this make more sense? CycloneGU (talk) 21:46, 22 August 2010 (UTC)

Given the infoboxen on my talk page, no, because that's completely against my will. I want ZEROpart of FraggedRevs if it is implemented. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 21:48, 22 August 2010 (UTC)

Some opposed users probably would want input but clearly some would rather, if PC is kept, just walk away from it altogether. We CANNOT assume any intent on opposed (or any user's) part other than what they state in their "vote". Period. A poll/discussion dealing with the specifics of implementation would be required if the consensus were to keep. Millahnna (mouse)talk 21:52, 22 August 2010 (UTC)

I'm not saying it's a right assumption. I'm just trying to clarify what he means. I understand you want absolutely no say in how it continues, just to say halt it. CycloneGU (talk) 21:53, 22 August 2010 (UTC)

Though I understand the feeling that oppose users have no say in how the feature gets implemented, they also have a full say in voting against its implementation at all. If users could vote against it, and then vote for the most restricted option in a separate poll, wouldn't that be tantamount to voting twice, or having one's vote and eating it too. If you vote none, you shouldn't be able to then vote for 'less'. If you want to vote for 'less' (option 2), then just do that from the beginning. Ocaasi (talk) 19:28, 22 August 2010 (UTC)

The users in question don't want to vote for option 2, but if the feature is retained, this is closer to their preference than options 3 and 4 are.

Someone voting to discontinue the feature believes that it isn't beneficial. If it's retained, this concern is best addressed by working to improve it before engaging in substantial expansion.

Do we seek consensus, or is this a winner-take-all/be-on-the-winning-side-or-be-ignored battle? "They lost, so their opinions don't matter" is neither fair nor consistent with our guiding principles. —David Levy 19:58/20:05, 22 August 2010 (UTC)

Given how Off2riorob conducted the poll to begin with and his attempts to own it, I'm getting the impression he intended for this to be the be-all-end-all. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:08, 22 August 2010 (UTC)

"we should assume that those selecting option 1 would prefer option 2 over options 3 and 4 (and count their votes accordingly if there is consensus to retain the feature)" - uh, no. Users have already specified, for instance, "only 1 or 4, but nothing in between." For a fair process, we need to keep track of the number of votes for each of the four options, not use the numbers creatively after the fact. PL290 (talk) 20:16, 22 August 2010 (UTC)

The problem is that this poll has been set up to "use the numbers." I don't advocate a straight numerical tally (with votes for "1" blindly converted to "2" if the feature is to be retained). I'm saying that everyone's input should be considered when gauging consensus (instead of saying "well, your side lost, so your opinions go out the window").

But if we are to have a straight numerical tally (meaning that we're ignoring the comments anyway), the assumption that "1" = "2" (in the event that the feature is retained) would better reflect the community's will than simply ignoring these responses would. —David Levy 20:41, 22 August 2010 (UTC)

I disagree: that is completely unacceptable, as Spinningspark has now eloquently expressed below. PL290 (talk) 21:03, 22 August 2010 (UTC)

Please see my reply to Spinningspark, which hopefully clarifies a position that I've struggled to convey. —David Levy 21:11, 22 August 2010 (UTC)

I have similar concerns and had a hard time voting because of it. Unless specific feature improvements are implemented before the continuation, I am very much opposed to the feature. IF those changes are addressed then I am in support of using it on some of the low traffic articles and BLPs. I ended up deciding that options 3 and 4 under the "keep" options were closest to my view but I still feel that is not right. I tried to clarify with my vote comment as best I could. There are things about PC I see as being potentially beneficial but I share the concerns of all of those who voted to disallow it. In fact the more I type this the more I want to change my vote to get rid of it. Millahnna (mouse)talk 19:42, 22 August 2010 (UTC)

It is completely unacceptable to make any assumption about the way votes should be transferred if the voter has not specifically indicated this. There must be no assumption that voters for option 1 would prefer option 2 if there is no consensus for option 1. There must be no assumption that voters for option 3 would prefer option 2 or option 4 if option 3 has no consensus. If voters want to indicate more than one preference they are free to do so (and have been doing so), if they have not done this it should be assumed that they are against anything else. Four options were offered, consensus needs to be judged for all four individually against the total of those voting. If you want to find consensus for continuing PC in any form, a different question should have been asked, but you are now stuck with a four-way proposal - or else the poll needs to be closed and started over. SpinningSpark 20:47, 22 August 2010 (UTC)

As noted above, the ideal manner of gauging the outcome is to consider everything written and ascertain the resultant consensus (instead of blindly tallying the numbers).

According to the description, options 2, 3 and 4 are to be lumped together as "keep" votes (and then broken down if needed). I'm saying that if the first part of this process results in the determination of a "keep" outcome, everyone's responses should be considered in the following phase (with the sentiment that the feature should be discontinued being factored into the decision regarding the extent to which it should be expanded) instead of declaring that proponents of option 1 have "lost" (and their opinions no longer matter).

As for closing the poll and starting over, that might be the best idea I've seen yet. Had it not been rushed, it would have been very easy to format it in an unambiguous manner (rendering all of the above moot). —David Levy 21:11, 22 August 2010 (UTC)

First off, the instructions have now been changed at least three times and quite different to how they started out. You cannot say that options 2,3 and 4 are going to be lumped together when the hundreds of people have now voted without that being in the instructions. I am going to revert that. Secondly you cannot make assumptions about what the "option 1" voters would want and call that consensus if they have not indicated that is what they want. That is a huge abuse of process. SpinningSpark 21:32, 22 August 2010 (UTC)

Firstly, the explanation that you removed was present before the voting began. To be clear, I haven't edited that page.

Secondly, I don't advocate that assumptions be made; while I've cited such a possibility as less harmful than an alternative, I believe that the best option is to take exactly what is written (nothing more, nothing less) into account when gauging the consensus. —David Levy 22:22, 22 August 2010 (UTC)

I think you will find it was not present before voting began, it was not there when I voted. It was inserted in here at which point there where 154 votes and comments listed on the page, a large percentage of the current total. My apologies though, for implying that it was your edit, which it was not. SpinningSpark 23:37, 22 August 2010 (UTC)

I think you mean that it was reinserted at the point you mention. For example, the same kind of explanation was present much earlier here. My assumption is it bounced around, and you'd have to look at every revision to figure it out.--Bbb23 (talk) 23:52, 22 August 2010 (UTC)

It was removed by the same user in the previous edit. As indicated in the summary accompanying the edit to which Spinningspark linked, this was a two-step relocation from one section to another. —David Levy 00:02, 23 August 2010 (UTC)

I linked above to a pre-vote revision containing the explanation, which also was present when you voted.

As indicated in the edit summary, Ocaasi merely moved the text from one section to another. This was performed in two steps, the first of which (the text's removal) immediately preceded the edit to which you linked (its reinsertion in a different location).

For the record, I'm not endorsing such a tabulation method. I'm noting that I based my statements upon this pre-existing framework (which I had no role in devising). I agree with you that the poll should be redone and restarted. —David Levy 00:02, 23 August 2010 (UTC)

I don't think there can be too strong a complaint that 1 and 2/3/4 are collected together - only a few people have suggested an extreme (e.g. "1 or 4") preference and fundamentally the idea of comparing those who want to stop the trial, and those who want a continuation, is fair. Obviously very few people want the trial to continue "as is"; the 2/3/4 split is concerned with how many articles the trial is rolled out to, but it could equally have been split into "what technical changes need to be made as the trial goes forward" or "what should be the rules about the level of verification before content is approved if the trial continues". But then attempting to discriminate between the 2s, 3s and 4s in the event of the "keep" supermajority, in order to decide what to do next, would be a big mistake, as nobody opposed to continuing the trial would have an input on the form it was kept as. They might largely prefer option 2 to 3 or 4 if forced to choose, but that should be a choice they would make for themselves. My guess is 2 would predominate disproportionately compared to the distribution of the keep !voters, and the fact there would likely be a difference in distribution is illustrative that their voice needs to be heard. I'd prefer a massively expanded roll-out of flagged revs, but I want those people who disagree with me to have their say too - not for them to be told "you've lost the vote on whether we continue, now we, the victors, will choose how to proceed." TheGrappler (talk) 22:01, 22 August 2010 (UTC)

ten days

Theres no hurry but I think ten days on this will be long enough. Off2riorob (talk) 06:28, 22 August 2010 (UTC)

Seeing as consensus supported keeping the last proposal open for a full two weeks, the same should probably be done here. Also, the first hundred or so participants in this poll were informed that the vote would last two weeks, so we should not arbitrarily change the date without valid reason.   — C M B J   07:00, 22 August 2010 (UTC)

The number of votes will drop off pretty substantially around day 5, but the longer you leave it open, the less someone can complain that they were ignored. 14 days is fine, we'll all still be here.--CastAStone/^/(talk) 12:54, 22 August 2010 (UTC)

Leave it at 14 days. You rushed the conclusion of the discussion, but granted, it was already quite lengthy. Don't rush the poll. CycloneGU (talk) 13:06, 22 August 2010 (UTC)

Its not an election its not even a big issue, this is a simple tool, nothing more, ten days would be plenty imo. Off2riorob (talk) 14:11, 22 August 2010 (UTC)

I have serious concerns that you are trying to rush the poll in order to establish a consensus for keeping the tool as quickly as possible. Please note that this isn't a "simple tool", as it includes a new usergroup, a new form of protection, and will eventually affect over 10,000 articles. Also, according to some dedicated opponents, this will have a negative impact. So yes, leaving the poll open for as long as possible (at least until the voting begins to die down) is a fair option. _EricLeb⁰¹ (Page | Talk) 16:11, 22 August 2010 (UTC)

I got that distinct feeling myself and still have it, and I agree with Ericleb. Off2riorob, your construction of this poll is patently biased, and while Cyclone's change helps it doesn't nullify the fact that you seem to be giving us a Hobson's choice here. —Jeremy ^{(v^_^v Carl Johnson)} 18:59, 22 August 2010 (UTC)

Agree that any reasonable effort to remove the appearance of a rushed or unfair poll is worth making. Keep for 14 days. If it won't make a difference, it's no big deal, and if it will make a difference then it's imperative. Ocaasi (talk) 19:33, 22 August 2010 (UTC)

I'm seriously tempted to close the poll, at least temporarily.

So far, I've seen at least one user whom I know to have good judgement voting 4 while expressing reservations about the current implementation, which presumably means they see this as a final decision. We've got established users creating new sections, in both cases so far because they cannot make head nor tail of what is on the table, and we've got many contributors to this very talk page expressing serious reservations about the implementation. This poll was made in good faith at an appropriate time, but the explanation was flawed, and the original counting system, intentionally or unintentionally, appears heavily weighted towards an eventual outcome of option 3. Am I alone in thinking that we need to take a brief pause, structure this page a bit more robustly, and re-start the poll? --WFC-- 09:04, 22 August 2010 (UTC)

I had the same thoughts myself, so I'm glad to know that I'm not alone.   — C M B J   09:34, 22 August 2010 (UTC)

You will never be able to escape the fact that some people just don't get it. I've trimmed the recently added additional options into a 'Other responses' section, which should help prevent new sections being created. Stopping, adjusting, and then restarting polls is often far more disruptive than trying to improve and make sense of what has gone before. -- zzuuzz ^(talk) 09:51, 22 August 2010 (UTC)

Stopping and restarting a poll can also be disastrous for the results. Some may not realize another poll has been started, and when their opinion is not counted from this poll, they may themselves become upset that their vote was basically fed to seagulls. I put in the numbered system in an attempt to save this poll, and we're going to save this poll or otherwise ensure that existing votes stand in any new poll. CycloneGU (talk) 13:04, 22 August 2010 (UTC)

Stopping and starting is a very tricky proposition, although if it must be done, better to just scratch it now. That said, there would have to be a ready alternative, not just more hand-wringing, which will devolve into a poll about how to poll, and is not likely to resolve easily. On the issue of appearance of bias for option 3, I understand how the poll could leave the impression of being tilted, however I read the entire Pending Changes/closure discussion, took a tally of each user, and the overwhelming support by about 2:1 was for keeping pending changes, but focusing it less on high vandalism/high traffic articles and more on low-traffic/low watchlisted articles and blps. In other words the appearance of bias may just be the confirmation of consensus. Ocaasi (talk) 19:37, 22 August 2010 (UTC)

The bias isn't for any one specific option, Ocaasi; rather the way the poll is set up it's a three-versus-one proposition, hardly fair and inherently biased towards the side that has three options. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 19:40, 22 August 2010 (UTC)

Quite. And users are specifying specific, exclusive keep options. If a user says "keep, but only if it's option 4", what fair process can count that as "keep rather than not keep"? PL290 (talk) 19:53, 22 August 2010 (UTC)

@ Jeremy. I updated the poll instructions. I think they're clearer. If users were allowed to vote close and also to vote for option 2, it'd be like voting for Republican, but specifying which Democrat you'd like if they win. This method isn't perfect, but the alternative is a kind of double-voting, which is also problematic. @PL@290, I think, in fairness to the somewhat haphazard setup of the poll, if someone votes specifically and exclusively, then their vote should not count towards the general keep v. close discussion but only the secondary round between 2,3, and 4. Ocaasi (talk) 20:04, 22 August 2010 (UTC)

I strongly disagree with the premise of excluding those who voted 1 from having any say in the future of 2, 3 or 4. I'm tempted to go into the reasons why, but I can't believe we're even having this discussion.   — C M B J   21:37, 22 August 2010 (UTC)

I too am of the opinion that if FraggedRevs is accepted, then those who voted to oppose should have a say in how it's implemented iff they want to. It's crap like David Levy's suggestion that is rather insulting. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 22:07, 22 August 2010 (UTC)

1. It wasn't my suggestion.

2. As I've explained several times, I expressed agreement that it would be the least harmful method of tabulating the available numbers to arrive at an immediate outcome (with the alternative being that none of the feature's opponents would have any say). I don't, however, advocate such a process. —David Levy 22:22, 22 August 2010 (UTC)

There shouldn't be any need to make your reasons clear, I think most editors will find it obvious. I don't think it is necessary to stop the poll, simply because I can't believe anyone will fail to notice, when the poll is finally closed, that it would be incredibly disingenuous to only then consult the "keep the trial" comments on how to proceed next. There is some value in knowing how the 2/3/4s are being distributed - it's indicative of the type of discussion we need to have next, but it's certainly no replacement for the discussion we need to have next! TheGrappler (talk) 22:14, 22 August 2010 (UTC)

Confusion

I do not understand the meaning of 3 or 4. Us441(talk) (contribs) 11:56, 22 August 2010 (UTC)

3 is a kind of small expansion and 4 is a major expansion. Off2riorob (talk) 14:34, 22 August 2010 (UTC)

"Vote?"

When did we start taking votes instead of straw polls? I thought this was a consensus, not a democracy? --CastAStone/^/(talk) 12:52, 22 August 2010 (UTC)

It is. If more people choose to close the system, that's the consensus. Am I misunderstanding you? CycloneGU (talk) 13:05, 22 August 2010 (UTC)

We've always called these things Straw Polls. We don't use votes to determine consensus, we use discussions. That's the way we've done this for years. That's why VfD was renamed AfD so many years ago - the word "vote" has no place in consensus building. A consensus is not 50%+1. A consensus is when the clear feeling of the group as a whole is to take an action. in most cases, 51%, 55%, even 60% does not indicate a consensus, it indicates massive continued opposition. Admins don't get in on 50%+1 because that means that there's a ton of opposition out there; i.e. a lack of consensus. The same goes here.--CastAStone/^/(talk) 13:17, 22 August 2010 (UTC)

I'm afraid with Pending Changes, there always will be opposition; some are set in their ways and do not want the additional help of experienced non-admin editors to help fight spam using a tool that has, in my experience, helped to curb that very spam. It's a tool that would give experienced editors the chance to do some good for the encyclopedia aside from article writing. It eases the workload as anons. can put their additions right into the article instead of on a SP article's talk page. I will surely spark backlash over this, but some editors are just dead set against change, pure and simple. I am always for something that has potential to improve Wikipedia. CycloneGU (talk) 13:46, 22 August 2010 (UTC)

That's kind of the point. I'm for this too, but you've got to understand that doing things the right way is more important than doing the best thing. Those people have legitimate reasons for opposing this. Now I do think that this will ultimately gain consensus, but to call the consensus building a "Vote" is a huge mistake and very out-of-process.--CastAStone/^/(talk) 13:50, 22 August 2010 (UTC)

Then we need to find a way to convince those against it that this is worth keeping. But again, some are dead set against it and no matter what, they will always oppose. Heck, a few may even support now and oppose later, who knows? I'm just trying to save that which is already started here, and I'll let admins. decide if this stands. Either way, it does show some insight on which way people are leaning. CycloneGU (talk) 13:52, 22 August 2010 (UTC)

A poll is a vote. The difference is not in the title, not in the interpretation of the results, but in what happens after the vote count. East of Borschov 08:48, 24 August 2010 (UTC)

Can we please change the name of this to Straw Poll

And rename the sections accordingly? Per WP:Consensus, particularly WP:PRACTICAL:

The quality of an argument is more important than whether it comes from a minority or a majority. Editors decide outcomes during discussion, polls are regarded as structured discussions rather than votes.

--CastAStone/^/(talk) 13:55, 22 August 2010 (UTC)

• I concur but fear it's a little late now. TheGrappler (talk) 17:20, 22 August 2010 (UTC)

One day into a 14 day process? No one will notice/care and its the right thing to do.--CastAStone/^/(talk) 18:22, 22 August 2010 (UTC)

That was an expression of my cowardice more than anything ;-) It will take a bolder editor than me to do it! But I think you've got the moral right to, so as far as I'm concerned, go ahead! TheGrappler (talk) 18:35, 22 August 2010 (UTC)

Done. Lets see what happens.--CastAStone/^/(talk) 19:49, 22 August 2010 (UTC)

Since we've made this change, do we also want to tone down the language which discourages people from using this as a structured discussion? "Please respond to the questions below with an option number. ... There are two basic options: close or keep. ... Please keep any discussion at the talkpage, thanks." It seems to be forcing a vote. Andrew Gradman ^talk/_WP:Hornbook 20:10, 22 August 2010 (UTC)

I think we should encourage brief commentary from individual voters, but not responses, which will make the page unruly. The page introduction more clearly links to the PC/Closure discussion, which might also help. Ocaasi (talk) 20:24, 22 August 2010 (UTC)

Proposed instructions for "specific, exclusive votes"

The point was raised that if a user votes for support 4 but nothing else that their vote shouldn't be counted as part of the general 'keep' consensus. I think that's fair and should be listed in the voting instructions. Something like:

Users who support keeping pending changes but "only" under one specific configuration will have their vote counted towards the deciding between keep options (2 v. 3. v. 4), but not towards the general close v. keep consensus.

Ocaasi (talk) 21:15, 22 August 2010 (UTC)

That would, in effect, disenfranchise them. Against. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 21:16, 22 August 2010 (UTC)

Sorry, but that seems like too little, too late, too complicated. It should be either a straight choice of four options, or an initial keep/close vote that makes clear that "keep" means "keep regardless". The current hybrid risks fooling voters into thinking their expressed, exclusive choices are being honoured, when those votes will instead be bundled together with conflicting votes and used to bring about a result they didn't vote for. PL290 (talk) 21:32, 22 August 2010 (UTC)

Keep regardless, people have differing ideas and really I was hoping to clarify what the consensus was amongst the kkep voters wanted to do with the tool. You are welcome, to start here a collection of names that object to the format and see if there is a decent number that want to close it. Off2riorob (talk) 21:42, 22 August 2010 (UTC)

So far I count myself, PL290, Millahnna, and SpinningSpark who want to start over. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 21:45, 22 August 2010 (UTC)

Well so far that five out of about 165, thats 3 percent of the commenters.Off2riorob (talk) 21:48, 22 August 2010 (UTC)

No offense, but what will starting over do? There are still 80 yeas and 30 nays based onmy results from this morning. 80 people wanted it to continue in some way when I left this morning (I haven't checked yet this afternoon). CycloneGU (talk) 21:51, 22 August 2010 (UTC)

Starting over will allow the poll to be recreated to eliminate the bias that is currently present in this one, and allow us to not trot out a rushed poll that sweats bias out of every pore. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 21:53, 22 August 2010 (UTC)

I did not see a bias. If options 3 and 4 were gone and option 2 was simply "keep", we'd still have 120 yeas (keeps). CycloneGU (talk) 21:55, 22 August 2010 (UTC)

I suggest you look within the 2, 3, and 4 section. There are several votes there for one specific implementation only and nothing else. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 21:58, 22 August 2010 (UTC)

You are assuming that those who voted keep under the 3 and 4 options would still vote keep. Many may have voted conditional keeps or opposes. Millahnna (mouse)talk 21:58, 22 August 2010 (UTC)

I just counted the poll. There are 45 users who vote to end the tool. There are 121 users who vote to keep the tool. Of those 121 users, there are about 15 (may be off slightly because of my eyes having trouble going down the page), who voted ONLY Support 2 option. I'm not counting any users who voted for 3 or 4. Nor am I counting any who voted for 2 but also for something else, only unambiguous 2. That means there are more people voting to end the tool than an unambiguous keep.--Bbb23 (talk) 22:08, 22 August 2010 (UTC)

That's true but pretty selective. There are 45 close votes, and 90 keep option 3's. That's a 2:1 margin even if you counted 2,3, and 4 separately and didn't group them. Ocaasi (talk) 22:33, 22 August 2010 (UTC)

I thought the idea was to restrict the keeps to option 2 in the interest of only having one option to keep rather than 3 (not counting combos). Nonetheless, statistics are inherently malleable.--Bbb23 (talk) 22:47, 22 August 2010 (UTC)

I think this is the basic problem and the reason I attempted to find a consensus from the keep voters as to what the consensus is from them that they would like to do with the tool going forward. IMO from the quite lengthy previous discussion, users that want to keep the tool have differing opinions as to what they want to see it used for, personally I see it that way mostly and not that users voting keep-4 want only to keep it if their favored usage is implemented. Basically you can't ask people to vote keep without them knowing what they are voting for, so this was designed to solve that problem at the same time. Off2riorob (talk) 22:03, 22 August 2010 (UTC)

Restart the poll: yes or no?

As per the numerous concerns raised by independent users throughout this discussion page and even the poll itself, I propose that we have a frank and final discussion about whether or not this poll should be restarted. Please concisely indicate your position below, along with any thoughtful consensus-building commentary. Also note that, if the poll is restarted, all existing participants will be contacted accordingly.   — C M B J   21:57, 22 August 2010 (UTC)

As if to prove that we're not even capable of running a vote on whether or not to run a vote, there was a parallel vote running on the same question on the main poll page. Here is its content. – Smyth\talk 13:06, 24 August 2010 (UTC)

Oppose poll per Cybercobra's comment #39 under Keep, but if and only if certain issues are addressed. I feel very strongly that the feature should be monitored carefully for the problems related to discouraging anons. I also concur with those who think that the interface and guidelines need work. Particularly, the "don't accept", "stop reviewing", and "accepting of incorrect changes just because they aren't vandalism" issues need to be addressed. The latter is my strongest concern. Entire poll is poorly thought out at best as discussion page shows. Millahnna (mouse)talk 19:33, 22 August 2010 (UTC) Oppose this biased three-options-against-one poll. Biased towards the "yeas" and likely ending with votes being counted towards an option the voter had not supported. SpinningSpark 21:05, 22 August 2010 (UTC) I think the best way of doing polls with more than two options is multiple rounds in which the option with the fewest votes is dropped after each round, but that would require people to come back here a few times. Hopefully we can get a feel for the real consensus after this vote. —Arctic Gnome (talk • contribs) 21:10, 22 August 2010 (UTC) That's not the point. It should be either a straight choice of four options, or an initial keep/close vote that makes clear that "keep" means "keep regardless". The current hybrid risks fooling voters into thinking their expressed, exclusive choices are being honoured, when those votes will instead be bundled together with conflicting votes and used to bring about a result they didn't vote for. PL290 (talk) 21:25, 22 August 2010 (UTC) I dislike the format of this poll especially in its original "vote" incarnation. I don't think this should be used as a "3 vs 1" poll but it can legitimately be used as a "keep vs stop" poll (and that would be less chaotic than running pro/con votes on all four options simultaneously). There are a few people with a "one extreme or the other split" (option 1 or 4) but on the whole, this straw poll is addressing the basic question "Do we want to stop the trial or not?" After the poll has been run, if (as seems likely) the trial is to be continued, then there needs to be a new debate about in what format it is kept. It would be unfair just to use the 2/3/4 results of the "keep" !votes to determine what to do next! TheGrappler (talk) 21:33, 22 August 2010 (UTC) I disagree—see related discussion in numerous sections on the talk page. PL290 (talk) 21:38, 22 August 2010 (UTC) In my opinion, Spinningspark got it 100% right in this diff. That set of instructions would have ensured that, had general polling favored keeping the trial, nobody opposed to continuing the trial would have an input on the form it was kept as. That's potentially very significant, since those people supporting an end to the trial would clearly (judging from their comments) prefer the minimal possible continuation in such circumstances, and oppose a large scale roll-out. I would like a massive expansion of flagged revs, but I want those people who disagree with me to have their say too - not just for them to be told you've lost the vote on whether we continue, now we, the victors, will choose how to proceed. That would be ludicrous. TheGrappler (talk) 21:47, 22 August 2010 (UTC) Oppose the poll; poll was created with a bias that cannot be addressed short of redoing the whole thing. —Jeremy (v^_^v PC/VC is a show-trial!) 21:15, 22 August 2010 (UTC) Oppose the poll per Spinningspark and Jeremy plus related discussion in numerous sections on the talk page. PL290 (talk) 21:38, 22 August 2010 (UTC) Scrap this poll and notify all existing participants of the new forum. Far too many grave concerns have been raised by too many independent users.   — C M B J   21:45, 22 August 2010 (UTC) Scrap this poll The poll was constructed with an extremely clear steer that the outcome is either going to be scrapping or expanding. A significant proportion of people want the thing scrapped. The fact that they are a (large) does not mean they should be entirely ignored. --WFC-- 22:04, 22 August 2010 (UTC) Defective polling instrument - Since in essence there is only a boolean choice here (keep it on or off), it's inaccurate to present the choice as four distinct options where three have the same net effect. If this is the path to closure, there should be two or three separate polls: (1) keep PC we know today in force or not; (2) continue to pursue PC; (3) best way to improve PC (assuming #2 is "yes"). //Blaxthos ( t / c ) 22:05, 22 August 2010 (UTC) I don't like the poll in its current form, but there is some worth in seeing how the 2/3/4s are distributed among the "keep" voters (and indeed, what "second preferences" would be among those who'd rather end the trial now) because after this binary choice, other choices are far more open. Seeing the strength of feeling about expansion of flagged revs can help inform what kind of discussion we have next, but obviously it is not a replacement for the next discussion. For instance, if the consensus is to keep, and the vast majority of "keeps" are mostly strong in favor of option 4, then we need a subsequent discussion about mass roll-out to BLPs; if consensus is to keep, but there is a more equivocal mixture of 2s and 3s with lots of comments about how the system needs to improved before it's rolled out, then we probably need to be having a very different discussion afterwards. TheGrappler (talk) 22:22, 22 August 2010 (UTC) No matter what option could be selected, it will require continued discussion and improvement. None of the options clarifies reviewer policy, or fixed the change/accept issue, or sets down page-use guidelines. All of those things need to be figured out no matter which option gets the most support. Still, if the 2:1 ratio of option 3 (keep with minimal expansion) : option 1 (close) remains, I don't see how we could justify ignoring that. To use the voting lingo, you only have a run-off if there's not a clear winner. But if overwhelming support is for option 3, then what is left to formally discuss besides how best to implement it? Ocaasi (talk) 23:38, 22 August 2010 (UTC) It is not clear whether or not option 3 would have overwhelming support in a run-off discussion. Thirty percent of the participants in this poll have essentially been deprived of their opportunity to weigh in on how to move forward if consensus is in favor of continuing the trial.   — C M B J   02:52, 23 August 2010 (UTC) Yeah, I've done that math, and it is worth considering. I don't quite accept the premise that the oppose voters have been deprived of anything, since they have full opportunity to pick any of the keep options. Also, consider that in a run-off, 2's and 4's may shift towards 3 in anticipation of what you just described. Last, don't forget that the difference between 2 and 3 is at max 8k low-traffic articles. That's not insignificant, but it is also not likely to be more problematic than what's already going on. It might even be necessary to test the scalability of the feature. That's not a comment on the voting procedure, just the possible outcome. Ocaasi (talk) 03:08, 23 August 2010 (UTC) "I don't quite accept the premise that the oppose voters have been deprived of anything, since they have full opportunity to pick any of the keep options." Huh? I truly don't understand the logic behind that at all. Those completely opposed to keeping PC are not deprived of anything because they can always vote to keep PC? WTF? The perception is, rightly or wrongly, that those voting opposed (of whom I was not originally one of) will have no say in how the implementation is carried out (as implied by options 2-4) because they are voting against (which has no such implications). If there is a follow up poll planned discussing that very issue then it has not been made clear. And based on how abruptly the poll went up after conversation tapered off, I see plenty of reason why many may have jumped to the conclusion that this poll had more finality in that regard than many seem to feel that it had. Millahnna (mouse)talk 03:32, 23 August 2010 (UTC) I agree the poll wasn't implemented as carefully as it could have been, but I'm not convinced the current structure is fundamentally flawed. You get to vote on: close, keep as is, keep with limited expansion, or keep with expansion to BLPs. Those are four options; you get to select the one you prefer. All options assume improvement can and must continue. Waiting for improvement before continuing is not technically possible because the developers can't just turn the feature on and off. It is reasonable to not just lump together all keeps vs. all opposes, since some keeps are only for one type, but most support for option 2, 3, or 4 is not exclusive of a general keep sentiment. I think it's reasonable not to expand until improvements happen. That is already somewhat expressed in the "gradual" expansion, and could be specified further. Re: keeping out close votes, consider this logic. If I vote for a Republican candidate, should I also get to select which Democrat I prefer if the Republican loses? Ocaasi (talk) 09:08, 23 August 2010 (UTC) Oppose the poll. This should be redone as two separate polls. I voted for option 1, but I would also like to express an opinion on options 2-4. Right now, the only way to do this is to vote for both and thus cancel out my initial vote. Kaldari (talk) 03:29, 23 August 2010 (UTC) Oppose the poll as biased. Should offer only two options, as we do in any RFC: either remove it or improve it. Don't just call for expansion or keeping it wholesale if there are other options. Shooterwalker (talk) 03:50, 23 August 2010 (UTC) Oppose poll This poll is terribly designed. "Keep-as-is" and "keep with changes" are separate things and shouldn't be lumped together. The current poll, or more accurately, the way it would be tallied, is almost assured to produce a "keep" result.oknazevad (talk) 03:50, 23 August 2010 (UTC) Agreed. This poll is seriously flawed as it stands, and it almost appears like there was an attempt, conscious or not, to stack it a certain way. SchuminWeb (Talk) 05:00, 23 August 2010 (UTC) What is with these opinions? This logic is flawed. At the end of the day, the poll's trying to determine how many people want to keep it, and how many want to get rid of it. Keep is a very broad stance, so asking the "keeps" to what extent they want it kept is completely reasonable. However, those who want it to cover all BLPs and those who want it expanded no more than it already is both want it kept - and that's what's important. SwarmTalk 05:35, 23 August 2010 (UTC) Oppose this poll I have gone for option 1 but only really because of the poor format of this poll that does not allow my real opinion to be presented. Davewild (talk) 07:10, 23 August 2010 (UTC) Oppose the poll although I have already voted for close. But I am not opposing because of perceived bias. I am opposed because the premise is wrong. The poll is asking the wrong question. Rather than ask about how much to expand the PC, it should be asking: 1) whether to keep PC the way it is, 2) change PC to improve the implementation workings, or 3) completely close it. Voting on how much expansion we desire distorts the results because the question is wrong. If the PC implementation were improved then my current vote of "close" could be different simply because if it works better then I would have less resistance to it being kept. HumphreyW (talk) 07:13, 23 August 2010 (UTC) Comment: Yes, the poll is not very functional. We should start the poll over, and vote simply yes vs. no. If yes prevails, then we vote on the other options to narrow it down. And obviously if no prevails we scrap the system. I strongly suggest a restart. |Finalius|T|S|C 12:59, 23 August 2010 (UTC) Oppose the poll per reasons stated by Kaldari. If the "Scrap it entirely" option lost, those voting on it should have the opportunity to vote on the next steps to be taken. Should be done as two polls. Keep/Scrap Entirely and (assuming Keep won), Large Expansion/Change Implementation/Limited Expansion. --Resplendent (talk) 14:15, 23 August 2010 (UTC) I don't see a strong reason to restart the poll, but once finished we should _then_ discuss 2 vs. 3 vs. 4. It may well be that some people prefer 1 but would pick 3 or 4 over 2. Hobit (talk) 16:05, 23 August 2010 (UTC) I agree, I think almost everyone voting 1 would change to option 2, rather than see option 3 or 4 implemented. Lets not take this poll as a final decision. If the decision is to keep, then lets vote again as to what extent it will be kept. We are asking two questions, but only allowing for one answer as I see it. —Charles Edward (Talk | Contribs) 17:46, 23 August 2010 (UTC) Speak for yourself and stop assuming. —Jeremy (v^_^v PC/SP is a show-trial!) 00:09, 24 August 2010 (UTC) Re-orgonize votes options 2,3,4 should be separates into thier own sections. If they are not separated the vote will be setup so that all versions of keep will be mashed together into one big keep, unfairly outnumbering don't keep.Sumsum2010 · Talk · Contributions 20:59, 23 August 2010 (UTC) Does that mean that I get three !votes, since I'll accept any "yes" option? Or is the goal to divide-and-conquer the "yes" crowd, so that a mere one-third of respondents (=editors voting no) can prevent two-thirds of respondents from getting what they prefer (=editors voting yes)? (Well, it's how the California leg works...) WhatamIdoing (talk) 21:27, 23 August 2010 (UTC) I see your point. This should be done as "close" or "keep" then if keep wins THEN have 4 options one of which being "any of the three".Sumsum2010 · Talk · Contributions 21:33, 23 August 2010 (UTC) I don't really see the problem here to be honest - You can either vote "Keep" or "Close", with the keep side having the ability to add some nuances to their vote. If we boiled the entire vote down to "Keep" or "Don't keep" we would technically pit option 1 against option 2, while discussion option 3 and 4 once we were certain we had consensus to keep. Think about it; Option 3 and 4 actually call for an extension of pending changes - gives only two options, would it be likely that people voting 3 and 4 would switch their vote to option 1? Excirial (Contact me,Contribs) 22:07, 23 August 2010 (UTC) If you're voting to expand it then you obviously want to keep it, albeit with caveats. —Jeremy (v^_^v PC/SP is a show-trial!) 00:07, 24 August 2010 (UTC) Restart the poll with a preferential voting system (see talk). Sceptre (talk) 22:38, 23 August 2010 (UTC) WTF??? (Okay, I admit it's not the most creative way to start a response, but it captures my feelings here on many levels.) First, I'm coming to this straw poll/discussion/whatever as an uninvolved outsider; I don't have a strong opinion on this matter. If this actually improves Wikipedia, I not only can live with this, I may even come to support it. Second, reading the relevant pages here -- Wikipedia:Pending changes/Metrics & Wikipedia:Pending changes/Closure -- I have no clue how to vote here. Nowhere does anyone state if this has improved any specific articles, made no difference on any articles, or hurt any articles, which is what I would like to know. Thirdly, there is no clear rationale for the four options: are we deciding whether "Pending changes" is a clear success (which I guess justifies option 4). or we need more information, requiring either a longer period of time or a wider sample to sample (justifying options 2 or 3), or it is destructive (justifying option 1). And if I'm baffled, how are the rest of Wikipedia's membership supposed to react to the result here? Jimmy Wales has been pushing this for years, claiming that "Pending changes" will be the final solution to all of the BLP problems -- so approving this will appear like the fix is in. Then if this straw poll fails, there will be a number of people who will believe "the inmates now run the asylum". And lastly, this whole matter still looks like a case of a solution looking for a problem to solve: since at no time has anyone explained just what this is supposed to do & provided a metric to see just how well it does it. -- llywrch (talk) 04:46, 24 August 2010 (UTC) Close poll, then have a community discussion on how to implement this - per my comments on the discussion page. This is a farce of a 'poll', and should be halted now. Jusdafax 06:43, 24 August 2010 (UTC) Keep the poll 1. Instructions have been consistent for the majority of participants. 2. All keep options include at least keep as is. Options 2 and 3 add additional support, but are keep-inclusive. 3. Votes to only support one option are few 4. Oppose voters had/have the option to vote 'keep as is (with improvements)'. Since that is the minimal keep scenario, anything less is naturally a close vote. Voting 'close', and then wanting to express support for a version of 'keep' allows two contradictory positions, essentially two votes on the same poll. 5. Proposals for a conditional oppose or conditional supports are not currently possible, as the feature will be turned off in 1 month, and technical developments are in the works. It is assumed that improvements will continue, particularly to the interface. Their exact date is not known but is a high priority. The option 3 limited expansion is minimal and would likely be rolled out over time, as improvements happened, reducing the need for a conditional expansion option. 6. The whole feature is still under evaluation and development. A keep vote does not mean keep this way forever. Even support for keeping the feature is conditional on its improvement. 7. Consensus still needs to be determined. A vote is not binding, though it is important. 8. Discussion is not finished. There is a massive list of problems and recommendations at Wikipedia:Pending_changes/Closure which need it. Ocaasi (talk) 07:33, 24 August 2010 (UTC)

• This poll should be Redone from the ground up and restarted. It was initially created by Off2riorob shortly after discussion petered to a trickle on /Closure, and set up in such a way that there were four options, three of which counted together. In addition, when users altered the page to reflect concerns here, he edit-warred with them to keep his original intent. I cannot trust this poll if its creator seems dead-set on it being his sacred cow. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 22:01, 22 August 2010 (UTC)
• Redo and restart per Jéské. Not enough discussion or thought went into a polling design. Yworo (talk) 22:12, 22 August 2010 (UTC)
• Redo and restart per Jéské & the discussion on this talkpage. Jarkeld (talk) 22:13, 22 August 2010 (UTC)
• Redo and restart. Details should be determined by consensus, by IMO this should run in two phases. The first should determine whether we continue or scrap. The second (on the assumption that there is consensus to continue), should determine whether its ready for expansion, or should be kept at current numbers until improvements can be made. --WFC-- 22:15, 22 August 2010 (UTC)

• Restart the poll. As someone else noted, the poll instructions have changed so many times at this point that folks who voted hours ago may not, were they check back on the page now, feel their original vote reflected their opinion accurately at all. Additionally, questionable logic has plagued the polling process: extremely bad math (or at least badly represented math), assumptions about what votes mean that aren't explicitly stated by voters, etc. This combination of problems has led to feelings of disenfranchisement and a sense of lurking cabals, even though I personally know that several of the keep voters are not part of any such activity. I would tentatively put forth that we need a poll with either 2 or 4 options: support keeping PC, oppose keeping it, and (possibly) conditional support or opposes. IF conditional options were allowed, a voter could briefly specify their stipulations (oppose but if kept then, support but only, etc.) but we would need much clearer guidelines so that voters were able to accurately represent their opinion. The other option I see as potentially viable is to set up a range along the lines of:

1. Opposed to keeping PC entirely
2. Opposed to keeping PC but if it is kept were prefer minimal implementation
3. Support keeping PC but only if it is implemented minimally
4. Support keeping PC and seeing considerable expansion

Though I can't think of any middle ground/neutral options at the moment, something could certainly be added to reflect those of us who feel very on the fence about the whole process. Perhaps something along the lines of, "Until certain issues with implementation and/or interface are addressed I am neither opposed to or in support of." That's all I've got for now anyway. Millahnna (mouse)talk 22:19, 22 August 2010 (UTC)

• Redo and restart. Per my comments above (and those of others), this poll was hastily formatted in a manner that encourages polarized ballot-casting and hinders efforts to gauge consensus. —David Levy 22:22, 22 August 2010 (UTC)
• Don't restart. It's unnecessary. The instructions have changed 'slightly' but the opinions themselves haven't. The only question is how to weigh the individual comments. So, worst comes to worst, we might have to actually read and think rather than just count. But that's what consensus is about anyway. Also, it won't be any of us 'closing' the poll, but an outside admin or two who can try and gauge the straw poll for whatever perspective predominates. More importantly, there are 45 close votes, and 90 keep option 3's. That's a 2:1 margin even if you only counted that selection. Ocaasi (talk) 22:34, 22 August 2010 (UTC)Ocaasi (talk) 22:25, 22 August 2010 (UTC)

As an illustration of my point, originally I voted keep 3 or 4. Were I to vote with my hypothetical options I would vote opposed but minimal if kept or neutral on the gruonds that too many issues need to be addressed before I can be swayed in either direction. Millahnna (mouse)talk 22:38, 22 August 2010 (UTC)

I think I follow that, but doesn't a vote involve some decision? You want it closed, or you want it minimally kept; you chose one of them. What about the current poll prevents you from selecting either of those outcomes? Ocaasi (talk) 23:12, 22 August 2010 (UTC)

Although not banned, it's undeniable that the poll is structured to discourage that. There is little way of telling whether people want to see the thing scrapped because it's been a nightmare, or whether they're irrevocably opposed to the concept. There is little way for people who are skeptical based on the experience up until now to express that, and there are grounds to believe that this would form a large proportion of editors. Equally, those voting for option three have no way of expressing that they want to see this expanded, but are concerned about doing so now; the poll gives no time for the next decision to be made, and therefore would have led several to conclude that if this isn't expanded now, it could be ages before it is, if ever. I think a re-start along the lines of Millahnna's suggestion is a better way to go, with an explicit timeframe on when the next decision will be taken. --WFC-- 23:29, 22 August 2010 (UTC)

• Redo and restart. per my comments above. SpinningSpark 23:25, 22 August 2010 (UTC)
• Redo and restart and let people critique the proposed format before we even bother restarting. The present poll is a mess, and it will be more of a mess the more that it changes while underway. — Gavia immer (talk) 23:31, 22 August 2010 (UTC)
• Redo and restart. But, I think the current poll focusses completely on the wrong aspect. It currently focusses upon the size of the implementation, but that is wrong. It should instead focus on the way it is implemented. The current implementation is terrible, confusing and unclear.

More sensible options would be:

1. Close
2. Keep only if implemented properly (no confusing greyed out buttons and the like)
3. Keep as is

The size of the implementation can be changed later as an when the need arises. It is too early to decide upon how pervasive it should be until the way in which it works is decided first. HumphreyW (talk) 23:50, 22 August 2010 (UTC)

• One of the few sensible suggestions to date. I endorse that. SpinningSpark 00:00, 23 August 2010 (UTC)

The second option is reasonable but technically kind of impossible. The interface issues may not be finished within the next two weeks, though they are currently being addressed. The feature is still ongoing and can't technically be turned on and off depending on whether or not an update is ready yet. The next best is just to start focusing on the issues immediately, and (if option 3 predominates) to expand slowly in step with with interface and policy improvements. Ocaasi (talk) 00:26, 23 August 2010 (UTC)

So how about this instead:

1. Close
2. Allow trial to continue a while longer and have a new poll later when the interface is updated and people have had a chance to get a feel for it
3. Keep as is

But whichever options for a new poll are selected the separate issue of usage policies for PC will affect the size of the implementation and should not be voted upon here just yet. HumphreyW (talk) 00:38, 23 August 2010 (UTC)

• No opinion (changed to Don't Restart CycloneGU (talk) 17:55, 23 August 2010 (UTC)) - Basically, the results will still generally be the same, the only difference is a change in how the votes themselves are cast. Waste of the first day IMO. But consensus is leading towards restart because of a perceived bias that, frankly, wasn't there in the first place. CycloneGU (talk) 23:56, 22 August 2010 (UTC)

  I agree that the format was not perfect, and that results would generally be the same. I disagree that the 10 or so editors who vehemently oppose the poll as is represent a broad consensus to restart. Ocaasi (talk) 00:04, 23 August 2010 (UTC)

  I base my opinion not on the view that the poll is biased, but on the view that it's a confusing mess that fails to solicit the information needed to properly gauge consensus on how to proceed. —David Levy 00:09, 23 August 2010 (UTC)

  Has anyone thought about the following three issues: (1) how many people who voted the first time will be discouraged from voting again; (2) how long will it take to reach a consensus on the proper way to conduct a restarted poll and (3) who will decide whether a consensus has been reached as to the appropriate way to reconduct the poll?--Bbb23 (talk) 00:18, 23 August 2010 (UTC)

  I have thought about people being unwilling to vote again and anticipated it coming up as a negative to restarting the poll. While it is a downside, it is entirely the consequence of the poll being put forth so haphazardly. You second and third questions are very good ones. I have no answer for them, myself. Millahnna (mouse)talk 00:23, 23 August 2010 (UTC)

  1: Any number would be unfortunate, but allowing a highly flawed poll to run its course would be worse. The longer we wait, the worse the situation will become.
  2: It would take however long it takes. Again, the alternative is to proceed with a highly flawed poll.
  3: That's something to ponder. I would hope that widespread agreement would be clear. I honestly believe that it would be fairly easy to devise a good format. The problem is that this poll was hastily initiated without discussion. —David Levy 00:26, 23 August 2010 (UTC)

  I sincerely hope that your optimism turns out to have been justified. The process will certainly go a lot more efficiently if the accusations and offensive rhetoric present in the current discussion are eliminated in any future discussion.--Bbb23 (talk) 00:34, 23 August 2010 (UTC)

  Agreed. Of course, much of the unpleasantness reflects confusion/misunderstanding stemming from editors' mutual struggle to make sense of the current poll. —David Levy 00:38, 23 August 2010 (UTC)

• Restart. Also, restart the vote on restarting, I sense some bias (joke). But honestly, this discussion is predicated on the idea that the results of this straw poll are binding in some way; they aren't. The straw poll is useful in that it can reveal an existing consensus; however this poll started as a vote to determine a consensus. This isn't the ideal way way to hash out consensus: it was extremely unclear from the beginning what the objectives were of this poll/vote (that being to get a consensus).--CastAStone/^/(talk) 00:41, 23 August 2010 (UTC)

I recommend that we start the poll with just one voting area, and the option to lay out your preference as well as other acceptable options (if they exist). With everything put together in one area, we can take arguments at their value instead as up or down votes, because only discussion can reveal an existing consensus. Also, message everyone whose laid out their opinion on this page already to tell them that this is happening so they can do it again.--CastAStone/^/(talk) 00:41, 23 August 2010 (UTC)

• There has been fairly extensive discussion at Wikipedia:Pending_changes/Closure. If you haven't read through it, please do before deciding that there hasn't been a lot of groundwork towards consensus.
• If the vote is "not binding" then why aren't the opinions expressed on this page sufficient?
• If you're serious about this, you need to make sure there are more than 10 people who want a new poll. I'm not convinced that the views in this section aren't a minority. That may sound hypocritical in light of the way the poll started, but it's still worth figuring out. Ocaasi (talk) 00:59, 23 August 2010 (UTC)

• Don't restart. I don't think we'd get a different result if we changed the polling structure to "keep / don't keep", even if we followed it up with another poll on "how much to expand". I think the reason the "don't-keep" people are frustrated is because they assume that the !"winner" of this poll will be the group which got the plurality of !votes (which, so far, is 3). But if we instead treat each !vote as a weighted preference, and take the average, then we might find that the 2's are actually out in the lead. (By the way, I think this was a finely constructed poll, and I don't understand why folks are throwing accusations of bias at Off2riorob.) Andrew Gradman ^talk/_WP:Hornbook 00:46, 23 August 2010 (UTC)
  • If it were made clear that this is a short term decision, people would be more inclined to vote 2, on the grounds that when the developers sort the issues out we would look at it pretty quickly. The poll is completely ambiguous as to the timescale, and the result is lot of people voting for 3 while saying "I want to see this expanded, but there are still some serious issues right now that need to be fixed", in all probability because they think this is long term. Assuming we're not simply going for the most popular of the four options, this makes a big difference. I haven't bothered to count, but at a glance it would take the balance of the votes from a weak preference for expansion, to quite firmly saying "we want to expand, but not until some of the issues are mitigated or resolved by the developers." --WFC-- 03:54, 23 August 2010 (UTC)
    • Indeed, I was under the mistaken impression that we were being asked to determine the feature's long-term implementation.
      But the original instructions contained the statement that "consensus will be judged by an Administrator on the basis of a simple head count," which likely discouraged many users from commenting beyond their ballots (and might even have dissuaded some from participating at all). So if the poll is allowed to continue, we must either evaluate it on the basis of "a simple head count" (thereby ignoring feedback essential for determining the community's true consensus) or discriminate against editors whose simple ballots were cast in good faith (on the basis that any further comments would be ignored). And in either scenario, we're left to divine the intent of users whose input might have been based upon a misunderstanding of what was being asked (as mine would have been). —David Levy 07:43, 23 August 2010 (UTC)
• Redo and restart. I think a number of errors were made up front that would make the current poll of dubious value, the only way to correct that is to start over. --WGFinley (talk) 04:59, 23 August 2010 (UTC)
• Redo and restart This was set up very badly, with little outside input, and to those who say it would make no difference, my vote would almost certainly be different if it was not a forced choice on the options presented, as my original opinion was judged as not being allowed. Davewild (talk) 07:09, 23 August 2010 (UTC)
• Don't restart: It is very unlikely that restarting the straw poll would lead to a different outcome. --Forty two^{the answer?} 07:11, 23 August 2010 (UTC)

Doesn't it depend upon which question is asked as to how the outcome might be affected? Ask a different question and I expect we would get a different result. HumphreyW (talk) 07:17, 23 August 2010 (UTC)

I !voted in favour of 1 and would vote so even if there is a second poll, irrespective of the questions asked. I believe people who are voting in support of keep aren't so stupid that rewording the question would elicit a different reply from them. Those who are in favour of 2, 3 and 4 would !vote in support of "keep" should there be a second poll.--Forty two^{the answer?} 14:26, 23 August 2010 (UTC)

• Add more options and restart PC is a good idea but as currently implemented has significant shortcomings. It should not be discarded or expanded, but instead the way it works should be changed. Also, the stuff about vote percentages is misguided; consensus is in principle supposed to be about actual reasoning presented and other subjective factors, rather than voting. In practice that tends to mean X% +/- some margin left open to the closer's discretion, where X depends on the type of question, but anyway never some precise count. 67.122.209.167 (talk) 07:41, 23 August 2010 (UTC)
• Redo and restart - (1) decide single or multiple polls, (2) decide unambiguous, measurable question/answer mechanism (no narrative, other than to demonstrate reason for choice—we've had the discussions), (3) clearly state that mechanism in the header so voters understand the effect of their votes, (4) do not change the process once the poll goes live (why does the importance of this need to be pointed out! The process has changed several times since this poll went live, deeming existing votes to apply just the same to the changed process even though those voters may be unaware and disagree, and despite objections to this grave abuse of process, further suggestions to change the process of an ongoing poll keep coming!) PL290 (talk) 09:25, 23 August 2010 (UTC)
• Redo and restart there are 2 issues I see. The first one is that this poll is confusing and slightly weighted towards keep (disclaimer; I !voted to oppose the tool). It started off as one thing contrary to normal policy and then morphed into this slightly better version. We need to redo it as a keep/oppose poll AND a option "2/3/4" poll. The second issue is that this is a clearly contentious straw poll, it did feel somewhat hurried (despite being done in the best intentions) and I worry that any close will be controversial and cause fall out in the future. Call this a wash and restart with the constructive suggestions already made. --Errant Tmorton166^(Talk) 10:00, 23 August 2010 (UTC)
• Close, redo and restart while I believe the drafters had good intentions, and I understand PC is supposed to end in a few weeks so we sorta need to make a decision soon, I feel and have felt for a while that there was insufficient discussion in drafting it. (It reminds me of the Ireland naming vote to be honest.) Who's fault this was is IMHO irrelevant. It should be closed ASAP to stop people wasting time and then needs to be well discussed and then re-done. Once this has happened, it can be restarted. (Of course it goes without saying that once a new poll has been designed, anyone who was part of the design/discussion but failed to raise the issue during the design should generally be ignored if they later complain about the poll being flawed, unless there's a good reason they didn't notice earlier.) As I've remarked before, it would be silly to end PC only to restart it in a few days so if absolutely necessary, the trial should be extended until the outcome of the poll is determined. (However this is no excuse to spend many weeks designing the new poll.) We will undoutedly lose people due to fatigue with the whole process but that can't be helped. As an example, some people, including me, were under the impression there was no need to explain our rationale. Regardless of whether this is a good way to determine consensus, people who in good faith simply left a vote are seemingly now going to be ignored. This is clearly a bad thing. Some people have some other issues, which while I don't find that bad myself (well the biggest problem I see is that it wasn't apparently clear to some that we are only discussing a continuation of the trial), it seems their feelings are strong enough that they are not going to be happy whatever the outcome which is also a bad thing. The alternative is to continue but only count keeps (2,3,4) and abandon, and a new poll for deciding what to do if we keep. This would be possible if we 1) continue in the poll as originally designed (i.e. no disenfranchising those who didn't explain why they chose whatever option) and 2) those who feel the current poll is flawed agree that this is an acceptable way forward (I'm far from sure that this is the case hence it isn't my primary recommendation) Nil Einne (talk) 15:35, 23 August 2010 (UTC)
• Why is this poll still running when there is a clear consensus to close it and start over? The longer it is left open the more damaging it is going to be when you do close. And if you don't close it at all you are giving ammunition to those who are saying it was rigged from the start. SpinningSpark 15:58, 23 August 2010 (UTC)

  ...because people are still !voting, oblivious to this discussion here? I've not !voted here, because I didn't consider this thread to have nearly as much merit as a straw poll advertised site-wide. I'll rectify that now. TFOWR 16:06, 23 August 2010 (UTC)

  That's no reason to carry on with a flawed poll, it's more an argument to stop it before too much damage is done. Besides, there is a similar response on the poll page itself. SpinningSpark 16:20, 23 August 2010 (UTC)

  Because no one with the authority to stop the current poll (whoever that is) has done so. Frankly, I don't understand how these things work in the anarchistic/structured world of Wikipedia, but who gave authority to Off2riorob to start the poll? If such a person exists, then that person perhaps could also stop the poll. If it was one of those consensus decisions, then perhaps Rob (is that his name?) could weigh in on stopping the poll. He originally invited people to express their opinions on stopping the poll (see above), but quickly said that three people in favor of stopping didn't establish a consensus. If Rob agreed that there is now a consensus for stopping the poll, that would help.--Bbb23 (talk) 16:39, 23 August 2010 (UTC)

  The usual way to close debates if for an uninvolved (ie someone who hasn't voted) to gauge consensus and take the appropriate action. An administrator is needed, since if it is closed the site wide banner needs to be removed. The attention of an administrator is attracted with a template like this: SpinningSpark 17:03, 23 August 2010 (UTC)

{{adminhelp}} The template was placed here to request that an uninvolved admin assess whether or not there is consensus in the debate to close the poll and start over and take the appropriate action if ther is. If you have good grounds for removing the template then say what they are on the page, don't just remove it. SpinningSpark 05:58, 24 August 2010 (UTC)

Thanks, SpinningSpark. For some reason, David, who is usually very responsive didn't respond to my plea below to put in the question. Just for your information, though, the admins have been posting responses to the request in the edit summaries when they remove the template. Hopefully, now that you've actually posed a question, they'll do something more substantive.--Bbb23 (talk) 15:09, 24 August 2010 (UTC)

My apologies for overlooking the below message. If I fail to respond in the future, please feel free to leave a note on my talk page. Sorry again. —David Levy 15:18, 24 August 2010 (UTC)

It's nice of you to apologize, but there's really no need. I figured my post just got lost in the chaos. You've been besieged in all of this and are doing an admirable job (at least in my view).--Bbb23 (talk) 15:28, 24 August 2010 (UTC)

I have removed the {adminhelp} template, it clogs up the helpers IRC channel, and hasn't been addressed for over 20 hours. I suggest a request at the Admin's noticeboard.--Commander Keane (talk) 03:10, 25 August 2010 (UTC)

• David, you know better than I how this is supposed to work, but shouldn't the question to the admin be included below the adminhelp template? Granted, it might seem obvious based on the context of the discussion, but it's what the template documentation says.--Bbb23 (talk) 20:11, 23 August 2010 (UTC)

  I just counted 1 "oppose poll" amongst the "support 2,3,4" !votes (none, surprisingly, amongst the "support 1"s) and 15 "oppose polls" in the "Other responses" section. I've not considered duplicates in the poll and in this thread. I consider the poll far more legitimate than this thread, but even adding in the !votes here it's still a tiny fraction of the total !votes in the poll. TFOWR 16:29, 23 August 2010 (UTC)

• Don't close. The closing 'crat admin can assess whether options 2,3,4 taken together can be considered "keep" !votes. I'm unconvinced that anyone would !vote for options 2,3,4 when they would have !voted "don't keep" had the choice been between option 1 and option 2,3,4 combined. The straw poll may have been less than perfect, but it is by no means fatally flawed, and binning it now would be like throwing out the baby with the bath water. TFOWR 16:06, 23 August 2010 (UTC)

  Who said anything about a "closing 'crat"? The original instructions indicated that "consensus will be judged by an Administrator on the basis of a simple head count" and that "all the keep vote comments for option 2,3 and 4 will be first counted together." So if the poll continues, we must either abide by those terms or effectively punish those who based their responses upon them.
  On top of that, there appears to be widespread misunderstanding regarding what's being asked (so we can't even be sure that the responses are applicable). —David Levy 16:35, 23 August 2010 (UTC)

  'crat struck, admin inserted. We abide by those terms. Closing a poll that's run for several days, and in which nearly 200 editors have !voted in good faith, on the basis of a non-advertised talkpage thread with around 20 !votes is just not on. I'm not dead-set on the poll - I'm not big on voting anyway - but the time to close the poll has passed. TFOWR 16:40, 23 August 2010 (UTC)

  So we're to gauge consensus via a "simple head count" (with all elaboration ignored)? —David Levy 16:56, 23 August 2010 (UTC)

  Well, as I said, my preference would have been for a discussion rather than a straw poll. But at this stage - really, what's wrong with a simple go/no-go decision? We either keep PC or we don't. if we decide to keep it, that's the time (IMO) to decide how it develops. The features available to us are always subject to community-driven change - PC is (or should be) no exception. TFOWR 17:11, 23 August 2010 (UTC)

  I don't object to a simple go/no-go decision. I object to said decision being based upon a head count (instead of the weight of the arguments presented).
  But even if we assume that a head count is an appropriate method, that stipulation was removed long after the polling began. So some users based their responses upon such an instruction (which encouraged them to cast ballots without elaboration) and others didn't. Therefore, no matter how consensus is gauged, it won't be fair.
  I also have no objection to the idea of first deciding whether to retain the feature and then (if it's kept) deciding how it develops. But the poll purports to accomplish both steps now, with some users (myself included) misinterpreting it as a means of determining our long-term course of action (which has led to responses conveying opinions contrary to those of their authors). —David Levy 17:30, 23 August 2010 (UTC)

  Well, I'm (obviously) not keen on a simple head-count either. But I'm less keen on throwing away what we have, and using what we do have to assess consensus does seem to me to be reasonable, whether !voters elaborated or not. Certainly not ideal, but more attractive than going back to the community and asking that they, effectively, "!vote again but !vote right this time". Basically, I'm not happy with the idea of letting a poll run for several days, then pulling the plug mid-way. Deciding, mid-poll, that the poll is a go/no-go poll but that a second poll will be run if a go-decision is taken, that seems to me to be far preferable to re-running the poll. TFOWR 17:40, 23 August 2010 (UTC)

  You and I are in agreement that neither scenario is ideal. We disagree on which is the lesser of two evils. —David Levy 17:51, 23 August 2010 (UTC)

  Well, some agreement is better than none ;-) It does occur to me that we're maybe over-thinking this right now. One solution is to treat the current poll as partly flawed, and take its result as continue/discontinue the trial. We'd then re-run a new and improved poll in a month or two. I'd suggest if we were to do that we should reach consensus on the nature of the closing poll before the extended trial ends... TFOWR 17:55, 23 August 2010 (UTC)

  Indeed, given some of the unpleasantness on this page, I'm happy with any agreement (even agreement to disagree). (-:
  If the poll is allowed to run its course, your suggested approach probably would be the best one to take. —David Levy 18:06, 23 August 2010 (UTC)

  If the main objection to continuing with the poll is changes in the instructions, i.e., people voted with one understanding and that subsequently changed, what about continuing the current poll with a notice to all previous voters that they can change their vote if they wish based on the current (and hopefully not-to-be-changed) instructions? Just a thought.--Bbb23 (talk) 18:11, 23 August 2010 (UTC)

  I don't know whether that's the main objection. It's one objection, and your suggestion could be combined with Ocaasi's (see the section Two Polls Are Not Needed - One With Two Questions Is) if the poll is not halted. —David Levy 18:18, 23 August 2010 (UTC)

• Don't restart: Poll questions were designed and stated very carefully and i.m.o. correctly. This was a neat and effective way to combine two polls in one: first poll asked whether we should stop or continue, whereas second poll asked how we continue if first poll dictated continuation. DVdm (talk) 16:07, 23 August 2010 (UTC)

  ...with one response to the first poll resulting in exclusion from the second. —David Levy 16:35, 23 August 2010 (UTC)

  Yes, there are no second choises here. We also cannot choose continue_2 or, it that doesn't work, continue_4. Some have explicitly voted that way, but it's no use: that vote will be counted as a continue_2. DVdm (talk) 17:33, 23 August 2010 (UTC)

  How is acceptable that users advocating the feature's discontinuation have no say in its implementation if it's retained? You just acknowledged that these are two separate questions, so this isn't a matter of denying respondents a second choice; we're basically presenting this as a competition whose "losers" suddenly have their opinions deemed meaningless and play no role in shaping consensus from that point forward. —David Levy 17:51, 23 August 2010 (UTC)

• I have changed my vote above. We have now had over 250 votes, and when I implemented my changes to the poll layout (which at the time was muddled and mixed up), there were exactly 31 votes (including me at position 2 in the yes votes). Thus, the poll results are now based on a simple yes or no option, and the tertiary option inside the yes option is its own poll. Might not be the cleanest, but gives the quickest feedback. CycloneGU (talk) 17:55, 23 August 2010 (UTC)

  Thus, the poll results are now based on a simple yes or no option, and the tertiary option inside the yes option is its own poll.
  ...which is off-limits to those selecting the no option, meaning that if such respondents "lose," their opinions are deemed meaningless and they play no role in shaping consensus from that point forward. (Of course, this flaw predates your reformatting.) —David Levy 18:06/18:12, 23 August 2010 (UTC)

  I am thinking that there would still be a need for a second poll to clarify how to proceed. Right now, the yes or no option is the main one. This suffices, but perhaps send notices to all voters that the voting text has changed? Restart with 250 votes seems infeasible. CycloneGU (talk) 20:11, 23 August 2010 (UTC)

• Close now, then discuss polling - This is an incredible mess of a 'poll'. I say close this farce down at once, then have a community discussion on how to make a poll like this a meaningful outreach instead of something that a sizable number of editors have major issues with. Poll? More like troll, as I see it. Jusdafax 06:37, 24 August 2010 (UTC)
• Keep the poll, on the basis of the following compromise: It goes without saying that if forced to choose between options 2, 3 and 4, those people the vast majority of those people who hate pending changes would all vote for 2, the most limited deployment of the feature. So when (as is probably going to happen) we have to choose between 2, 3 and 4, it should be done by first including votes for 1 in the total for 2. As 2 is the status quo, it will be selected by default unless its total level of support is significantly exceeded by 3+4 (the meaning of "significantly" being at the discretion of the developers). – Smyth\^talk 12:28, 24 August 2010 (UTC)
  Changed my mind, see below for my new comment with the following timestamp – Smyth\^talk 13:14, 25 August 2010 (UTC)

  At the time of your comment, there are roughly 30 option 2's (although some include 2 or 3, or 2, 3, or 4). With a little over 100 closes, and 190 total keeps, that would leave 130 support 2 vs. 160 support 3 or 4. I think that is the maximum way to phrase the situation in support of option 2. Other rationales make the case more strongly for option 3, for example, the fact that anyone who prefers the status quo over a close can currently just vote for it. Ocaasi (talk) 12:51, 24 August 2010 (UTC)

  Again?! How many times am I going to have to say this? I DO NOT WANT MY* SUPPORT 1 VOTE CONVERTED TO SUPPORT 2, PERIOD. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:29, 24 August 2010 (UTC)

  Then place an obvious notice next to your vote saying so, and calm the hell down. – Smyth\^talk 20:59, 24 August 2010 (UTC)

  That's the thing, I shouldn't have to be forced to do that to make sure my wishes are respected, Smyth! —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 21:15, 24 August 2010 (UTC)

  Smyth: You're better off ignoring Jeremy. Nothing that anyone writes will convince him that no one seeks to construe anyone's opposition as support. I spent a great deal of time responding to his tirades with detailed explanations, to which he invariably replied with the same vitriolic insults and accusations. —David Levy 22:16, 24 August 2010 (UTC)

• Close poll and redo As it stands now the poll has attracted serious concerns from many users about not just its result but its very legitimacy as an instrument, because they view the poll as flawed from its inception. In a consensus driven environment legitimacy of discussion or polling is paramount. Widespread concern about legitimacy is reason enough to scrap the current poll, whether or no one agrees with those concerns. Making a determination (or even appearing to do so) about such a far-reaching and controversial feature by means of a discussion that has lost legitimacy (or never had it) would be disastrous for Wikipedia, and whatever the consequences of closing the poll, they are preferable to the other potential outcome. Revcasy (talk) 15:00, 24 August 2010 (UTC)
• Redo and restart. Exploding Boy (talk) 15:10, 24 August 2010 (UTC)
• Restart poll about restarting the poll. What? No, seriously, I think that the poll (the actual poll) should not be restarted, far too much confusion would result from it. People would assume they've already voted, etc. and a massive amount of time would be wasted. The question of how much to used PC will likely prove to be irrelevant, as it doesn't look like there's going to be consensus for continuing the trial in the first place. IMO, just fix up the instructions to make it pretty much a yes/no poll, with people commenting about how they would like it done if necessary. --Yair rand (talk) 19:40, 24 August 2010 (UTC)
• No need to restart poll. Yes it's biased, but it's biased towards keeping flagged revisions, which is the right decision for the project. It's in your best interests to keep the bias to prevent a good feature from being voted down by editors voting for the wrong side. —Preceding unsigned comment added by 67.208.6.108 (talk) 19:52, 24 August 2010 (UTC)

  That's very much debatable, 67. Read the lists at the top of Wikipedia:Pending changes/Closure and you'll see that PC raises just as much issues as it attempts to solve. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 20:06, 24 August 2010 (UTC)

  Most people who want to get rid of it are saying that anonymous people won't be able to edit as much, but it actually lets us edit MORE than we did before because now we can edit articles that use to be blocked and are now reviewed. If you ignore the wrongly-informed arguments there are only a small number people who are actually opposed for a good reason and a very large majority wants to keep it. —Preceding unsigned comment added by 67.208.6.109 (talk) 21:17, 24 August 2010 (UTC)

  That's not the only reason they're opposing, 67. They're also opposing because of speed issues, ineffectiveness against long-term abuse or spam, false impression of authority, and extremely opaque documentation and interface. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 00:14, 25 August 2010 (UTC)

• Very cute, Smyth, but moving all of these comments that the process is broken still doesn't address the problem that this straw poll is broken. If or when this poll is closed, no one will be satisfied with the solution, & the folks who are disappointed with the result will complain long & hard. Sheesh, this is becoming the classic definition of "cluster fuck". -- llywrch (talk) 22:27, 24 August 2010 (UTC)

  I agree, it's a farce. I still don't think that redesigning and restarting the poll would give a substantially different result, but I'm leaning towards doing it anyway just so that the result will be accepted as legitimate. But administrators are (quite rightly) declining to impose their authority on this situation, so what can we do? If someone tries to shut down the poll at this stage, they'll just start an edit war. – Smyth\^talk 12:32, 25 August 2010 (UTC)

  I think we have established a consensus to close the poll and redo it. If it turns into an edit war then the whole thing will be forced into the normal dispute resolution process. WP:Bold. Someone should archive the entire poll page and make an announcement. Revcasy (talk) 12:43, 25 August 2010 (UTC)

  Well, I'm not willing to do that, I think the fall-out would just waste even more of everyone's time. So I've cancelled my suggestion above, and here's a new one:

  Keep the poll (with its current form and timescale) but only use it for the purpose of answering the yes-or-no question. Consensus on that question will be evaluated by the developers at their discretion. If they decide to keep the feature enabled, then its usage limits will remain at what they were during the trial, and consensus for any future changes will be established by a separate, properly-thought-through process. – Smyth\^talk 13:14, 25 August 2010 (UTC)

  I still think that consensus has been established by the above discussion and the "other comments" section on the main poll page. It is actually a surprisingly clear consensus. Furthermore, I am not comfortable with continuing to ignore what we have decided through discussion, particularly since we are operating under a time constraint and also because the longer this poll is allowed to continue the more damage it does. The one and only reason I do not carry out the closure myself is that I lack the expertise to set up another straw poll or to notify all of the people who have already voted about the change, both of which should be done in conjunction with the closing of this poll. Revcasy (talk) 13:35, 25 August 2010 (UTC)

Seems to me like there is a clear consensus to just add your fovoured position to the poll and wait to see how it is interpreted, as in the wheels are not dropping of, there is a vocal minority of users complaining but that is not a reason to close a poll that three hundred and seventy users have commented in the vast majority of which have just offered their position and moved on. IMO the most destructive thing would be to close the poll at this point. Off2riorob (talk) 14:07, 25 August 2010 (UTC)

Off2riorob, you're part of the reason why people want to redo this poll. You initially created it with a strong bias against one side, edit-warred with users when they tried to simplify or clarify the instructions, and attempted to use a mathematical formula to express a percentage! Given all this, any sane person would declare the poll useless or invalid and start over. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 20:13, 25 August 2010 (UTC)

I am part of the reason this poll exists through my contributions. I did my best in a good faith manner as best as I could contribute. Almost 400 users have simply commented and moved on and we are more knowledgeable than before the poll started. Off2riorob (talk) 20:24, 25 August 2010 (UTC)

Non-restart options that might help

• limit the 'keep' period to (6 months)
• emphasize that 'keep' is still a trial, for testing and improving--not for ingraining and expanding
• focus first on improving the interface and policy guidelines
• emphasize that if PC isn't working on a page, don't use it
• per option 3, only spread to new pages gradually, and in step with other improvements
• emphasize the need for continued discussion, statistics, troubleshooting, and specific RfCs on issues
• use the working summary Wikipedia:Pending_changes/Closure to prioritize and delegate specific issues

Ocaasi (talk) 02:38, 23 August 2010 (UTC)

Yup, this is all part of the general procedure. Continued discussion, etc. is part of the original trial as well as spreading gradually (though requests were taken for new pages to add for testing). CycloneGU (talk) 03:37, 23 August 2010 (UTC)

Wikipedia:Pending changes/Closure and Wikipedia:Pending changes/Straw poll: an analytical perspective

For anyone interested, I have processed the content of both the discussion and the straw poll:

Wikipedia:Pending changes/Closure Position Pages Characters Words Support 10 36615 6393 Oppose 9 30362 5226 Other 7 23640 4175	A graphical version of the data
Wikipedia:Pending changes/Straw poll Position Characters Words Option 1 12720 (41.87%) 2108 (40.76%) Option 2, 3, and 4 17661 (58.13%) 3064 (59.24%)	A graphical version of the data

While quantity of content alone does not prove or disprove any specific consensus, it is a notable factor in this particular instance. Many of the users that voted in this poll did so without contributing a meaningful rationale to support their position. And as WP:DEMOCRACY and WP:VOTE both firmly tell us, consensus is always established by the relative strength of relevant arguments.   — C M B J   06:01, 23 August 2010 (UTC)

Please keep in mind that the original instructions contained the statement that "consensus will be judged by an Administrator on the basis of a simple head count." This likely affected many editors' responses. —David Levy 07:43, 23 August 2010 (UTC)

That's a good observation -- the instructions could have indeed played a role in causing this. And if that is true, it is further evidence of why this poll is fundamentally flawed to such an extent that it is not reliable as a source of consensus.   — C M B J   09:52, 23 August 2010 (UTC)

Counting words is spectacularly more stupid than counting votes. Just because someone has used a lot of words does not mean that they have a reasoned position; the implication is that one could influence the outcome simply by being garrulous. If I had known this analysis was going to be done I would have written a lot more, but I understood that discussion was to go on the talk page, not in the poll itself. Besides, there could be a lot of words because the voter's position is being challenged by the opposite view. Consensus has to be assessed by an intelligent human for goodness sake, you can't do it robotically. SpinningSpark 09:26, 23 August 2010 (UTC)

Also, even if number of words did reflect strength of argument, it would be the number of words per contributor that would be significant, not the total weight of verbage. SpinningSpark 09:28, 23 August 2010 (UTC)

I thought that I was unequivocally clear in stating that the data alone does not override human-interpreted consensus. And I don't really understand why you're assailing me for simply making data available. A methodology similar to what I used here has been found to be accurate in some other areas of statistics, though I did not have this in mind when preparing the data here.   — C M B J   Edited 12:15, 23 August 2010 (UTC)

I can't see how it's of any use whatsoever. --Pontificalibus (talk) 10:01, 23 August 2010 (UTC)

The straw poll Position Tally Option 1 76/227 (33.48%) Options 2+3+4 151/227 (66.52%) Versus My data analysis Position Characters Words Option 1 12720 (41.87%) 2108 (40.76%) Options 2+3+4 17661 (58.13%) 3064 (59.24%) Versus The straw poll, minus strict !votes Position Tally Option 1 71/170 (41.76%) Options 2+3+4 99/170 (58.24%)

Here is an alternate view that might help clarify its usefulness. At a glance, the data illustrates an error in how this straw poll was being interpreted by almost everyone here, including myself. Conventional polling wisdom suggests that the current tally is 76/227 (33.48%) versus 151/227 (66.52%). However, when we subtract the number of people from the argument who contributed absolutely nothing at all, the tally (currently) comes out to 71/170 close versus 99/170 keep, or in percentages, 41.76% versus 58.24%, which is almost 100 percent identical to the quantity of content as per my research.   — C M B J   10:53, 23 August 2010 (UTC)

Temporary note: click the image, and then click the image again (on the image page) to see it. There's something going on with the image server right now.   — C M B J   11:30, 23 August 2010 (UTC)

T26834 (and possibly others). TFOWR 11:33, 23 August 2010 (UTC)

Comment

I have already commented extensively in the previous RFC. I was under the impression there was no need to say anything here let alone repeat what I've already said in this straw poll. Why is the fact I haven't chosen to repeat what I've already said suddenly being taken as a sign I have nothing to say? I should warn you it's a well known fact I'm very good at long winded responses, so if you want to count words, mine may outweight 4 or so other responses, no matter the quality of responses. Nil Einne (talk) 15:00, 23 August 2010 (UTC)

This is just an analysis. It is not binding. It's open for interpretation. It's never been used to determine consensus in a vote. It is interesting in theory but not applicable here. Don't sweat it (or write an essay). Ocaasi (talk) 15:06, 23 August 2010 (UTC)

But as others have said as well, it's a dumb analysis which completely fails to take into account that people have been told they don't need to explain their reasons, and that plenty of people have already explained their reasons, and have not been asked to repeat or link to these. At the very least, if you want to do a marginally useful analysis you should take into account comments in the other RFC although I question whether that's enough since people have already been told there's no need so even if they didn't comment in the RFC but had plenty to say here, they may not have bothered since they saw no reason to waste their time on something they were told they didn't need to do. You can debate whether or not a head count is a good way to determine consensus, but at least people weren't told 'you don't have to vote if you previously commented in the RFC, we will analyse your comments and determine which way you voted' but then we don't do that. Nil Einne (talk) 15:09, 23 August 2010 (UTC)

Wikipedia:Pending changes/Closure, by crude content Position Characters Words Keep 36615 (54.67%)* 6393 (55.02%)* Close 30362 (45.33%)* 5226 (44.98%)* Other 23640 4175 Versus The straw poll, by refactored content Position Characters Words Keep 17661 (58.13%) 3064 (59.24%) Close 12720 (41.87%) 2108 (40.76%)

I did take into account the other RfC--that's what the first chart was based on. In fact, I'm still quite frustrated that my more lengthy comments there have been swept under the rug by an arbitrarily created poll. No one was even given a chance to see or respond to what I wrote. Still, the data is not stupid or inapplicable by any means -- and it is not authoritative in the matter -- but it is an aide that may help illustrate errors in our thinking and in the poll's methodology.   — C M B J   22:58, 23 August 2010 (UTC)

Consider the difference between:

• the article deletion process, which has a well-established set of rules, and where the result is determined by arguments in terms of those rules and not by weight of numbers
• the administrator promotion process, which is all about how many people like and trust the candidate, and where the result is usually determined entirely by headcount

I think there's a distinction between discussions about how a feature or policy should be applied to a particular situation, versus what the policy or features should be. In the first type of discussion, people who vote without explaining themselves can be ignored. But in the second type (which is what we are involved in just now), it's perfectly valid to vote solely on the basis of "I do/don't like this", because there are no other external criteria to work from. In those circumstances, Wikipedia is a democracy (within the limits of what would be acceptable to the Foundation and the developers). – Smyth\^talk 13:26, 24 August 2010 (UTC)

CMBJ, where's the chart that compares individuals' !votes to their desire to close the poll? My brief scan suggests that a lot of people who oppose pending changes -- and who are now aware that they're "losing" -- are pushing to close the straw poll. This looks like an effort to WP:GAME the system from where I'm sitting. WhatamIdoing (talk) 15:51, 24 August 2010 (UTC)

May I suggest, then, that you look at the earliest revision of this page? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:38, 24 August 2010 (UTC)

I have not been following the discussion to continue/restart, but as of the last time I was, it looked like the issue was pretty much universally divisive. Almost everyone who supported restarting had voted close, and almost everyone who supported continuing the current poll supported keeping the proposal. That is actually why I requested that an uninvolved party interpret the discussion. As for "losing" this poll, or any poll for that matter, WP:VOTE and WP:DEMOCRACY make it unequivocally clear that poll results are non-binding. There is no "win" or "lose" based on vote or even content ratios, because we make our decisions based on consensus as determined by the relative strength of ideas.   — C M B J   23:06, 24 August 2010 (UTC)

Update

As of present, the data has changed as follows:

• Option 1 is supported by 124 total users. Of these 124, an estimated 119 (95.9%) provided input, and 5 (4.03%) provided no input. The adjusted percentage of supporting votes relative to all votes is 42.3%.
• Options 2+3+4 combined are supported by 232 total users. Of these 232, an estimated 162 (69.8%) provided input, and 77 (33.19%) provided no input. The adjusted percentage of supporting votes relative to all votes is 57.6%.

  — C M B J   06:45, 25 August 2010 (UTC)

It's a Straw poll!

"A straw poll or straw vote is a vote with nonbinding results. Straw polls provide important interactive dialogue among movements within large groups, reflecting trends like organization and motivation." I responded on that understanding, to provide nothing more than dialogue aimed at having a quick look at how people feel - and I expect many others did too. Absolutely no decision about PC should be made as a result of this poll. -- Boing! said Zebedee (talk) 17:53, 23 August 2010 (UTC)

The way it was set up, Off2riorob intended it to be a be-all-end-all. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:21, 23 August 2010 (UTC)

Well, whether he meant that or not, it simply can not be - one straw pollster does not a community make. I !voted "Support 3", but only on the grounds that this was a Straw poll and was not in any way binding - there's a very big difference between me saying "Here's how I feel at the moment, at this transient stage" and "Here's my final decision". Had I been faced with !voting in a final decision, I would have spent a lot more time considering the various points raised in the discussion (which is a very good discussion, and raised issues I hadn't considered), and I might have !voted differently - in fact, I would probably have refused to decide, because I think it is way too early for that and we have not yet adequately evaluated the results of the first trial. A straw poll is an opportunity for brainstorming and non-binding discussion only - any actual decision based on one would be dishonest. -- Boing! said Zebedee (talk) 19:44, 23 August 2010 (UTC)

Yes, it is a straw poll, but it is a straw poll that asks the wrong question. I stated above the same but it is worth repeating here in case someone missed it. It is too early to be asking about how far to extend the usage of PC (which is what is currently being asked) because we have not yet decided if PC is even working as intended, or even if the way it operates is a good way. We need to decide whether or not the PC trial should be extended with a new interface paradigm (that is apparently being worked upon now), and only after that should the decision about how is is used, and what pages it applies to, be made.

Right now it appears as though someone has already decided that it should be used for BLPs and pretty much nothing else. But to the best of my knowledge this has not been voted on previously, it just seems to have been pushed into the poll with the expectation that it is what people want.

I say this: Vote on whether to get the implementation cleaned up, made more presentable, less confusing and extend the trial a little longer with the new setup - and then vote on discuss where to use it and how to apply the policies etc. HumphreyW (talk) 19:55, 23 August 2010 (UTC)

The BLP stuff comes from the fact that several comments on /Closure expressed an interest in using it solely for BLPs and a select few other articles. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 20:04, 23 August 2010 (UTC)

As a straw poll, there's nothing wrong with asking any old question at all, because a straw poll is explicitly non-binding and no actions should be taken as a result of one. Using a real poll to decide future direction would a very different issue, and that would need to progress slowly and carefully, with all options carefully introduced and discussed first. And that's really my only point here - this is explicitly described as a straw poll, and no decisions can be taken on its outcome - but they can be used to frame a real poll that should probably come later. -- Boing! said Zebedee (talk) 20:19, 23 August 2010 (UTC)

The poll began in this form, under the title "Wikipedia:Pending changes/Vote comment." It subsequently was renamed and reworded, but the concern that it might be treated as a binding vote (as appears to have been the intent) is valid. —David Levy 20:27, 23 August 2010 (UTC)

Hmm, so it changed part-way through too - that just makes things worse. -- Boing! said Zebedee (talk) 07:23, 24 August 2010 (UTC)

The wording changed, which is poor form, but I fail to see how the slightly different instructions would have changed anyone's preferences or vote choice. Ocaasi (talk) 10:30, 24 August 2010 (UTC)

Well, there's the issue noted above. (Some people would dedicate greater time and care when considering a binding decision than they would when engaging in a non-binding brainstorming session.)

Additionally, the earlier respondents were advised that the results would be evaluated on the basis of a "simple head count" (meaning that any elaboration beyond a letter choice would be ignored), so they might not have explained their rationales in as much detail as they otherwise would have.

Respondents arriving at the page after the aforementioned advice was removed may have followed the example set by previous editors. And reading arguments of greater depth might have provided insight that altered their opinions (and so on, down the line).

So we're left to somehow gauge consensus, and any method used will contradict that which reasonably would have been expected by someone.

Perhaps of greater concern, however, is the ambiguity that apparently has caused some users (myself included) to misunderstand what was being asked (leading to responses inconsistent with actual opinions). —David Levy 11:04, 24 August 2010 (UTC)

Good points. However: any lack of elaboration would plagued both close votes and keep votes equally. Most straw polls combine quantitative counts as well as qualitative interpretation. Though not all votes have more elaborate comments, many do, and as many from the closes as the keeps. Almost all votes have a brief qualitative explanation anyway. And, I'm still not sure about your last point regarding what was being asked. Comparing the original version of the poll to the final version, I still don't see how anyone would change their choice between 1, 2, 3, and 4. Is the issue simply that 2/3/4 were originally going to be counted together and that would make a 1 turn into a 2, as a form of strategic voting? If so, that would mean that the current vote has more 2 votes than it should, but that is both against the closers' benefit. Most of the first 50 keep votes were 3's anyway (actually most of all the keep votes are 3's). So again, aside from the theoretical nebulousness of it all, do you have an idea what the actual effects would have been? Ocaasi (talk) 13:50, 24 August 2010 (UTC)

1. I don't assert that anyone's viewpoint is at a disadvantage. The concern is that we aren't soliciting the proper information on which to base a determination of consensus. (Also contributing to this is the sentiment that we aren't asking the right questions.) That the various opinions are being conveyed equally poorly does not mitigate the problem. Our goal is to collaboratively arrive at the solution of greatest benefit to the community, not merely to provide all sides a fair chance of "winning" the debate.

2. As noted elsewhere on this talk page, the description's ambiguity has led users (myself included) to mistakenly believe that we're being asked to decide upon a long-term course of action. (I'm told that the poll actually refers to a short-term strategy.) Evidently, this has prompted some to express support for expansion (primarily in the form of votes for option 3) that they regard as premature (either unknowingly or because they don't realize that they needn't lock in a permanent implementation now). —David Levy 14:25, 24 August 2010 (UTC)

That makes much more sense: If this is the last vote forever on pending changes, then it makes sense to add a little more slack but not too much (option 4 is pretty premature no matter what you think of PC). So... you think that 3's are inflated because voters felt locked in. Do comments reflect that? (Could they?) I think the description of a 'gradual, limited expansion tempered the false enthusiasm of the 3 voters, since it basically described the alternative I assume you think is lacking, just in a non-specific way. I think an easy solution to this situation is just to count the vote as a win for keep, and then focus the majority of efforts on improving the actual interface and policy. Even if three 'wins' under possibly misinterpreted pretenses, it's still critical to make the system better. And once the system is improved, all of the current 3 votes indicate that there is enthusiasm for at least a small expansion. Can't we basically effect that in the implementation?

Also, I always assumed that this poll would only be binding for 6 months to a year at which point the feature would be re-evaluated based on the improvements and performance. Ocaasi (talk) 15:41, 24 August 2010 (UTC)

The issue was brought to light by WFCforLife in this message. Until I read it, I was among those under the mistaken impression that we were being asked to determine the feature's long-term implementation.

Regarding your proposed outcome, TFOWR made a similar suggestion (stressing the importance of additional discussion), and I agree that this is the best course of action if the poll runs to completion. —David Levy 15:48, 24 August 2010 (UTC)

(ec) I went for 3 partly because it was only a Straw poll and was nowhere near a final decision, thinking it was the way I'd like to see the next phase of a trial going. And after some tweaks, more trial, and more feedback, I'd be considering whether a move to 4 would be a good idea - but still with the possibility of 2, or even 1, if feedback further highlights the existing problems. So mine was a pretty non-committal opinion, which is fine if we're just brainstorming in a Straw poll. -- Boing! said Zebedee (talk) 15:54, 24 August 2010 (UTC)

Still, I assume that once the majority of interface and policy changes happen, that the support for a minimal expansion is already there from the 3's. That is a little presumptuous as to exactly which changes are 'the critical' ones, but having read through the entire pc/closure discussion and helped compile the working summary of issues and recommendations, I have a pretty good idea what bothered whom. I'm not too concerned about the outcome of this poll, as long as people don't feel shafted and pending changes keeps getting worked on. The only tradeoff I see is between minimally expanding before changes are ready --or-- not getting the added info from a slightly expanded trial. Neither is make or break in my mind. What about splitting the difference, authorizing up to 5k in an expansion, with the emphasis on fixing issues first, and a re-evaluation after 6months to 1 year? Ocaasi (talk) 15:58, 24 August 2010 (UTC)

That seems reasonable to me, assuming that it ends up accurately reflecting a manual analysis (i.e. not a "simple head count") of whatever discussion occurs (including the poll, taking into account the various flaws). —David Levy 16:22, 24 August 2010 (UTC)

Yep, that would work for me too. -- Boing! said Zebedee (talk) 16:46, 24 August 2010 (UTC)

An analogue

This situation where we have several preferences, with users supporting from one to four of them, reminds me a lot of the Ireland poll last year, where we had seven different proposals with varying levels of support. Clearly, like the Ireland poll, this is something that preferential voting was devised for. Thus, if the poll is reset, I propose the following:

• The poll be closed while we accumulate options for a poll. At least the current four options remain.
• If possible and uncomplicated, to gauge support for the tool, we subdivide the options into "oppose", "conditional oppose", "conditional support", "support".
• The poll be re-opened quickly (say, the end of the week), with the old votes removed, and the new votes using a preferential list (e.g. {{stv-ballot}}).
• All voters in the current straw poll are notified of the new poll.
• The poll lasts for, say, two weeks.
• After the poll, we calculate the winner after preferences. I'm not sure what the threshold for [2+3+4] support is under preferential voting, but I'd reckon it'd be around 66%.
• Comments are in a different section to the poll. In the case of consensus close to the threshold of support (threshold>50%), they'd probably hold weight in determining consensus.

Thoughts? Sceptre ^(talk) 22:34, 23 August 2010 (UTC)

I don't think that keeping the current four options is an acceptable choice, but I agree with mostly everything else. I support the binary YES or NO question, because otherwise there's going to be a whole lot of voices that go unheard if we move forward with implementation. If, however, we must have more than two options, then it should be limited to the "oppose", "conditional oppose", "conditional support", "support" format. In the event of the latter, guidelines for tabulation need to be made unequivocally clear in advance.   — C M B J   23:23, 23 August 2010 (UTC)

Maybe "abolish", "scale back", "retain levels", "increase levels", and "full BLP rollout"? The problem is, people may just vote for 5 without voting for 2, 3, or 4 (although I expect most preferences for 5 to flow to retaining the tool). Maybe we need to explicitly say "you have as many preferences as options; rank your preferences with "1" being your first preference, and so on"? Sceptre ^(talk) 23:32, 23 August 2010 (UTC)

I guess we might be able to learn about the pros and cons of that system somewhere around here. I personally would fall into either the "oppose" or "conditional oppose" category, but I do not think that I fit into a "scale back", "retain levels", "increase levels", or "full BLP rollout" category, because my objections are not based solely on scale. With regard to our particular dilemma here, though, the problem with making too much out of the !voting aspect is that head counts are non-binding, and for good reason.   — C M B J   23:55, 23 August 2010 (UTC)

Its EXACTLY like the Ireland thing! I remember that - after years of fighting about what to change it too, we decided...to keep it exactly the same. Good memory though. I'd be okay with that voting method, except that this is not a vote, its a straw poll/discussion intended to reveal consensus, not be a binding democratic decision. So we'd have to figure out how to make that all work together. --CastAStone/^/(talk) 02:48, 24 August 2010 (UTC)

About improvements

Improvements to pending changes will continue regardless of the outcome on en as de (German) uses it. In other words, the wording is really misleading, and assumes en exists in some kind of vacuum. Unless I am missing something, because I didn't notice it despite the massive amount of discussion here. Ryan Norton 12:09, 24 August 2010 (UTC)

There is a lot of concern about approving pending changes for continued use on en.wiki because of the fairly obvious user interface problems. The point of mentioning the improvements is to remind users that supporting the feature doesn't mean that it's perfect, but that it is going to still be in use as improvements happen. I don't know if de and en are using the exact same version of pending changes, but it's pretty much irrelevant as long as users here are comfortable with whatever version we have. Ocaasi (talk) 12:36, 24 August 2010 (UTC)

AFAIK, they are not. en's using a watered-down version; de's using full-blown Flagged Revisions. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:23, 24 August 2010 (UTC)

de uses it on all pages IIRC... not sure how different the implementation is. Actually would be really nice to know what parts worked and haven't worked for them, but then again en's documentation on its trail is.... almost non-existant besides what was said at Closure. Ryan Norton 23:43, 24 August 2010 (UTC)

Attempt to find a fix

Major problems:

• Poll changed
• Was initially presented as a straight vote
• Didn't have a timeline, which might have led people to be more ambitious
• Possibly excludes close-voters from expressing a preference

Preliminary results:

• 2:1 support for keep; 170 votes for 3 or 4, 30 votes for 2, 110 votes for 1
• If every 1 vote supported the 2 option, it would be 170:140 — option 3:option 1
• If some 3 votes would have been 2 votes due to the poll not having a timeline, this ratio might be closer or even reversed

Fix:

• Since the majority of support is for keep, we continue using the feature as a trial
• Since the poll was presented ambiguously, we emphasize correcting interface and policy issues first, no matter what
• Since the feature is still significantly under development, we agree to a full reevaluation after 3-6 months
• Since the 3 votes expressed confidence in a minimal expansion, even if somewhat premature, we split the difference, allowing a maximum expansion up to 5k articles (instead of 10k—currently 2k is authorized and ~1.4 k in use)

Ocaasi (talk) 16:50, 24 August 2010 (UTC)

Your statement that "Since the majority of support is for keep, we continue using the feature" is wrong. In order for Pending changes to continue there needs to be consensus in favour, not just a majority. I have commented above that this is a bad poll and certainly don't think such a major decision should be made on the basis of this poll, but if Pending Changes is to continue there does need to be a consensus in favour of whatever format it does continue in, when the decision itself is made. Davewild (talk) 17:39, 24 August 2010 (UTC)

The numerical majority of support is for keep, and seeking consensus doesn't mean denying numbers. Those votes generally express the sentiment that pending changes serves a useful purpose and that it is worth exploring further with a limited expansion assuming improvements to the interface and policy continue. That sentiment is not lessened or less relevant because there are twice as many people expressing it. Virtually no poll on wikipedia reaches unanimity and many ultimately resolve to an intelligent, opinion-weighted, head-count. Otherwise, how would you prefer to ascertain what 300+ people, 1/3 of whom fundamentally disagree, agree on? Also, the timeframe to figure this out was 1 month, with no option to turn the feature off temporarily. Extensive preliminary discussion has taken place at WP:Pending_changes/Closure. It does not appear to have an easy resolution which could be found within that timeframe, which is why my recommendations emphasized continued improvement, reevaluation, and minimal expansion in balance with changes. In the case of a new poll, what else do you think would be revealed that would more likely achieve consensus? Ocaasi (talk) 18:00, 24 August 2010 (UTC)

If there was over two thirds support for Pending Changes in a well designed format for making a decision then that could be ok. That is not the case here, firstly support even when added between options 2, 3 and 4 (and there are people who say they would only support one option (see support 20 or 51 for instance) is less than two-thirds so is certainly not a consensus. This could change but secondly this straw poll should not used for making a decision on, only as a general indication of where people stand. The discussions on this talk page show why this poll was not done well and indeed if I knew the right implementation was going to be used then I would very likely be switching to supporting. Davewild (talk) 18:10, 24 August 2010 (UTC)

Yes, I'm aware of the problems with the poll and the weighting of opinions. But what about the compromise I recommended. Do you think it is fair and generally in line with sentiments expressed among comments? Your point that there needs to be consensus on a format to continue seems to overlook the numbers, that in the most conservative case there would be a 50:50 split between option 2 and 3, hence, the straight compromise between them offered above. As it stands now, there are 112closes : 203keeps, which is 64 % It's pretty close to 2/3, although some of the keeps are for a specific option. Option 4 represented a more expansive approach, but the compromise is far more similar to option 2, which is just a continuation of the trial as it was. The only difference between that and what I suggested is a possible expansion of 3k articles, in light of the significant support for a gradual, limited rollout. Does this still seem excessive to you? Ocaasi (talk) 18:20, 24 August 2010 (UTC)

Two points, first two-thirds support is an absolute minimum for implementing a major policy such as this (really 70% would be good). If it falls below that then we cannot implement it unless changes are made (such as your proposal) and a new discussion produces at least two-thirds support.

Secondly for me personally the number of articles is far less important than the discussion about changing what edits should be accepted or rejected, the discussion about which articles should to be included in the trial and under what circumstances should semi-protection be used instead of pending changes. For me these are the important issues and I will not support a permament implentation until I know what the results of these are going to be. Davewild (talk) 18:34, 24 August 2010 (UTC)

I added as a trial and changed 1 year to 6 months. Actually trying to find an agreeable compromise, otherwise an admin is likely to just do it themselves. This way we can start framing actual issues rather than going back and forth about this poll.

This is not a major policy, but a continuation of an approved trial. That is how I always saw it.

The major issues you raised are on the Pending Changes/Closure summary and absolutely need to be resolved. This poll was mainly about not having pending changes expire after the 1 month deadline that was set. If you read through those discussions, there was a general sentiment that PC not be used on pages with high amounts of vandalism relative to good i.p. edits, that it wasn't ideal for content disputes or highly technical articles, and that the page protection log (listing why PC was used on a page) is something every reviewer should refer to in order to properly review a page. I agree there's a lot to figure out, which is why this compromise is just a continuation of the trial plus the possibility of a limited expansion to (less problematic) lower traffic articles, to get another small dataset for comparison. In total, I'm not sure why you're still referring to this as a permanent decision when the way I phrased this section clearly qualifies that. Ocaasi (talk) 18:43, 24 August 2010 (UTC)

The original trial got 259 supporting to only 61 opposes - that was a consensus and was reached on a particular implentation on which most of the broad details had been agreed upon. I also note at the time of that poll the level set for that trial to go ahead was two-thirds not 64% ,60% or any other level. Why should a further longer trial need less support?

A compromise needs to be put to the community to let them decide if it is ok, you can't just implement something for which you (or a few editors on a talk page or any one admin/arbitrator/bureaucrat) decide is what the community said, when hundreds of editors commented on different options. 6 months or a year is a long time on wikipedia, any further trial (which I would welcome if designed well) should not in my opinion be for that long (3 months is more appropriate). Also the details on those issues I raised such as when to accept an edit need to be shown to the community first before they make a decision, not afterwards when editors could well decide they would have "voted" the other way in view of those decisions.

I am not just trying to be negative but a better way forward in my opinion would be to have separate discussions (at the same time) on each change in the implementation that is proposed. Once each of those discussions reaches a decision then the whole package can be put to the community to make a decision by consensus. Davewild (talk) 19:22, 24 August 2010 (UTC)

I realize we can't resolve this ourselves, but it's possible that a compromise could be worked out that would be agreeable to many of the disputing parties. I think this somewhat untenable position is the result of only having 1 month to figure this out. That basically left 2 weeks to discuss all of the issues. Some of the issues you mentioned can't be voted upon, because it's assumed that they would be worked out over the intervening months. Out of curiosity, would you consider a 3 month extension of the current trial for more details to be worked out to be reasonable? Otherwise I can't see how this resolves besides either shutting down the feature due to failure to meet 2/3 by a few percent or an admin who goes with the page's weighted opinions and extends the trial via option 3. I can see how either would be undesirable. What do you think is actually going to happen given the time limits, the state of this poll, the available options, and the complexity of resolving some of the underlying issues? Ocaasi (talk) 19:33, 24 August 2010 (UTC)

Well firstly what I would do is scrap this poll straight away as I and many others have commented further up this talk page. Then I would have about 3 weeks discussing the implementation details and what we want to change (or keep the same). The conclusions reached would then be put to the community to decide if this discussed implementation should go forward to a further trial for another 3 months. Any areas where agreement has not been reached would be clearly stated to the community when they come to make that decision with the understanding that they will continue to be worked on during the extended trial (with the community making the decision on these areas once a proposal has been agreed on these areas). Extending the 1 month deadline by a couple of weeks should not be a major problem to allow this process to be done.

My expectation however, is that nothing like this will be done and due to the way this has been handled, whatever happens will alienate a substantial number of editors, who will feel that any decision made was something which they were not given a proper say in. Davewild (talk) 20:06, 24 August 2010 (UTC)

I think stopping the poll will also leave many feeling like something unseemly went down. Also, who said we have the ability to extend this for that 3 weeks to discuss it? The note at the top of the PC/closure page says we have 1 month from the end date of the trial to reach consensus or the developers turn it off. I doubt many people want this shut down just because we ran out of time to discuss it. All of which is why I'm offering... the above compromise. If a few editors can agree to make it reasonable, we can try and spread some support for it among the disputing parties, and then when admins come around, they can take that into account as an option that has received some approval from those involved with the poll and its issues. I frankly don't think there's any way the developers are going to let all of their work just vanish over a poll which is 5-10% short of consensus. So I think the best that's likely to happen is we can try and move opposition towards a reasonable alternative and start discussing the substantive issues. Ocaasi (talk) 20:48, 24 August 2010 (UTC)

Many people already feel that something unseemly is going down. The results of this poll will lack legitimacy in the minds of many editors. I appreciate your attempts at compromise, but I think at minimum this poll, as a failed instrument, should be abandoned. Perhaps an extension could be requested. I am open to that, as opposed to trying to kludge together some sort of result from this mess. Revcasy (talk) 21:04, 24 August 2010 (UTC)

Is 3 weeks even close to enough to resolve the 40 or so issues, recommendations and questions on the PC/Closure page? I tend to doubt that there will be much more consensus on anything until the developers whip out the next update. 3 weeks would give time to reframe a poll, but not to likely fix the issues themselves. It's probably worthy trying, but I'm worried that after 2 weeks, with the poll approaching we'll be in the same place. My next suggestion then would be to try and extend this for 'longer', like 2 months, but given the very short time of the original trial, that is close to success through delay (at least for pc supporters, among whom I count myself). Developers said they wouldn't 'turn off' the feature only to turn it back on; what about the option of just not approving any new pages for PC (but letting existing pages choose to keep it) until more issues are resolved? Ocaasi (talk) 21:22, 24 August 2010 (UTC)

That got kicked in the head. The devs recommended exactly that, but people are still putting pages on PC. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 21:47, 24 August 2010 (UTC)

• I will give this meaningless pol1 somewhere between 24 hours and a week. If it is still running when I next see it, I will take the originator of this mess to Arbcom. Septentrionalis PMAnderson 01:12, 27 August 2010 (UTC)

Results of trial

Wondering if there are any results from the trial. Were there many cases of BLP violations which were caught by the pending changes. Slp1 provided me one about the Wisconsin Hoofers on my talk page. [1] Are there any more? Are there any other statistics or observations about what has happen through the trial. --Salix (talk): 23:03, 24 August 2010 (UTC)

• You can find observations here. But, yes it would be good to have some numeric measure of the trial. As anecdotal evidence I could add Antoni Gaudi, an article with 50K+ views per month and constant weekly vandalism, temporarily semi-protected in the past on multiple occasions. PC has been implemented in Week 2 of the trial, 25 June. In the two months since, 10 pending edits by IPs were reverted and therefore not visible to readers, while only one minor constructive edit by an IP has been made and accepted though the PC process. --Elekhh (talk) 02:15, 25 August 2010 (UTC)
  • Those two webpages you link to aren't helpful. Though I have no doubt that they were created in good faith, they don't answer the question which anyone interested in this experiment needs answered: did it fix any problems? Discussing how it affected the page you mention, however, does provide part of the information needed here. And the point I tried to make on this matter -- which some here still appear to fail to understand -- touches upon that: give us the tangible results. Declaring a consensus & making a decision without giving us that information will only make one of two schools of thought on this issue feel that the decision was shoved down their throat. -- llywrch (talk) 20:24, 25 August 2010 (UTC)
• There is also a link to a Pending Changes Feedback page, which seems to show responses concerning the trial, mainly from while it was taking place - mostly firsthand accounts, I think. It's very long, though, and I can't say that I've given it more than a cursory glance. But it looks as if it would be rather helpful to get a general idea of how it worked with some pages or for some reviewers. Layona1 (talk) 03:42, 25 August 2010 (UTC)

This page on Wikipedia:Pending changes/Metrics has links to some stats regarding the various articles in the program. For some articles less than 10% of the anon-edits were accepted. --Rocksanddirt (talk) 16:22, 25 August 2010 (UTC)

Ideas for better polls

Two Polls Are Not Needed - One With Two Questions Is

There's a clear consensus that the original poll and its revisions are not constructed in a sound manner. However some arguing against redoing the poll have said we don't need to drag this out with two polls, and they are absolutely right, we don't need two polls, we need one with two questions.

1. Should the Pending Changes process be kept?
2. If there is consensus to keep, what changes (if any) would you like to see to the process? Comments are encouraged: Choice A, B, C and D

Simple, effective, easy to count, easy to gauge consensus which should be the intent. --WGFinley (talk) 16:16, 23 August 2010 (UTC)

There is not a clear consensus but a vocal opposition with some good points.

At the very most, all that we are missing is the preferences of the oppose voters. A simple, minimal solution would be to let the oppose voters add on a preference between 2, 3, and 4. That way, an entirely new poll wouldn't be necessary, just an effort to contact approximately 1/3 of the voters. Ocaasi (talk) 16:43, 23 August 2010 (UTC)

Without conceding the first point I agree that something along the lines of your second point would be a good alternative Allow Oppose voters to go back and tag a preference along with their oppose vote. --WGFinley (talk) 16:47, 23 August 2010 (UTC)

This addresses only one concern. Other concerns are that the poll's explanation is unclear (with some editors—myself included—mistakenly interpreting it as a means of determining our long-term course of action) and that it originally contained problematic terms (removed long after the polling was underway) that affected the manner in which users responded (including encouraging them to cast ballots without elaboration), leaving us to discriminate against one group or the other (depending on whether we abide by these terms). —David Levy 16:56, 23 August 2010 (UTC)

Disagree. We just need one question: Which of these four options do you support? It allows people who would only support if... to have their say. --CastAStone/^/(talk) 17:54, 23 August 2010 (UTC)

Not if their responses are initially combined and treated as unconditional "keep" votes (as indicated in the original instructions). And as noted elsewhere, this format prevents users who wish to discontinue the feature from having any say in its implementation if it's retained. —David Levy 18:22, 23 August 2010 (UTC)

Isn't it significant though that the only difference between 2, 3, and 4 has to do with how to expand the feature? In other words all 2, 3, and 4 votes are keeps, just 3 and 4 are keep plus. Also, discussion about the future implementation is going to be ongoing no matter what. Please see WP:Pending_changes/Closure, where the discussion is detailed and just getting started. Last, part of the rush for this poll was the 1 month limit which was set before the feature would permanently turn off. Extensive discussion about details didn't seem possible within those time limits, so an integrated 2-question poll was sought to make the poll more efficient. Ocaasi (talk) 14:07, 24 August 2010 (UTC)

I don't dispute the logic of initially counting options 2, 3 and 4 collectively. I'm pointing out that this precludes the purported benefit cited by CastAStone. —David Levy 14:25, 24 August 2010 (UTC)

Ask each question separately

Regardless of what happens with the current poll, it's clearly not going to settle everything it was intended to. I suggest we develop a set of independent questions in a wiki-fashion, and whatever questions aren't concluded by the current poll can be used to construct a follow-up. – Smyth\^talk 14:04, 25 August 2010 (UTC)

Workshop (feel free to edit)

• Should the pending changes feature be enabled? Possible answers:
  • Yes
  • No
  • Only if... (specify functional, performance, usability or policy issues that must be fixed)

• If enabled, should there be a numerical limit on the number of changes using the feature? (Note: the trial limit was 2000.) Possible answers:
  • No limit
  • (any number)

• If enabled, should certain classes of pages have the feature automatically applied? Possible answers:
  • No
  • (any class of page)

Discussion

I approve of this approach; I would prefer my own wording, far below; but I would have no objection to the poll outlined above. Septentrionalis PMAnderson 19:38, 27 August 2010 (UTC)

Actually I think that the current straw poll actually does fit into this schema, or can be fitted into it. Q2 is contingent on Q1 and Q3 depends on both. Q2 as phrased here may be hard to find consensus on because of the difficulty picking the "any number" first - is it more sensible to ask "Should the number of articles covered be extended?" and only try to work out how far if if there actually is consensus to extend at all? Q3 will lead to subsequent questions about e.g. featured content and BLPs, so certainly not the end of the debate. Personally I don't see the point of asking a whole bunch of contingent questions at once; it makes much more sense (and will save unnecessary debate) to do one thing at a time. For instance, if at the close of this straw poll, we see 85% for Option 1 and 5% each for Options 2, 3 and 4 then debate of subsequent questions is unnecessary. If at the close of this straw poll, we see 30% Option 1, 60% Option 2, 5% Option 3 and 5% Option 4 then there would be clear consensus to keep going with PC and also clear consensus against extending the number of articles covered, again saving on debate for now. If the straw poll ended up 1% each for Options 1, 2 and 3 and 97% for Option 4, the community consensus would be pretty clear (fat chance! but had it been so, a lot of debate would be saved). If the straw poll finishes at 30% Option 1, 20% Option 2, 20% Option 3, 30% Option 4, then that shows more of a divide between the "stop", "keep as is" and "radical extension" camps and would need more debate about what to do next. This is only a straw poll, it can't be the be-all-and-end-all, but the ability of "keep" voters to state preference for Options 2, 3 and 4 (and the fact that some "stop" voters have stated alternative preferences for if the trial continues), will actually help inform us what question to ask next. So while I agree with the idea that we need to ask a series of questions, I think it's (a) fine to ask one question at a time, since asking contingent questions just makes things more complicated, (b) it's fine to allow people to state how strongly they feel and what they think the next step should be, and (c) it's actually a good idea to fine-tune the subsequent questions based on those results. It seems pretty likely from the way the poll is going at the moment, that the next debate is mostly going to be about how to move PC forwards with a small expansion, or whether to keep the number of articles frozen - there isn't a critical mass of support for Option 4 to make it viable for consensus to crystallize around for now (subsequent discussion may change things of course!). TheGrappler (talk) 19:56, 27 August 2010 (UTC)

I agree, as I said far below, that the right second question is whether to expand. But I cannot agree that 70% is consensus to keep PC - and will contest any decision that it is; even though I would like to continue the present experiment. Septentrionalis PMAnderson 20:16, 27 August 2010 (UTC)

Point taken. The point of the numbers I suggested was simply to illustrate that the distribution of 1/2/3/4s could help inform the next step. I appreciate that some people will dislike the use of 70% as "consensus". I don't think it's a good idea to do things by an arbitrary cut-off point, but my personal fear is that most large-scale discussions will fail to result in consensus if a rate higher than 70% is considered necessary. (Wikipedia has given me endless respect for the Quakers, goodness knows how they come to a consensus about everything.) TheGrappler (talk) 20:37, 27 August 2010 (UTC)

That's why we have defaults for what happens if there is no consensus; this is unusually clear about what is the status quo ante. Septentrionalis PMAnderson 20:43, 27 August 2010 (UTC)

My fear is precisely that if the required agreement rate is kept so high that almost all large-scale discussions will fail to meet that definition of consensus, arguments will simply degenerate into "what is the default if there is no consensus?" At least in this case the status quo ante is clear, that's true, but I was thinking about the general principle. I find it very troubling that the consensus-forming system hasn't scaled well. Admittedly, poorly designed polls don't help form consensus either! But there's a community mechanism issue here, not just a question of poll design. TheGrappler (talk) 20:49, 27 August 2010 (UTC)

I agree. We should discuss this more; the search for compromise is disappearing - in content disputes as well as these large-scale polls. Septentrionalis PMAnderson 18:22, 28 August 2010 (UTC)

Preliminary Results

2010-08-25

I have analysed the votes recieved prior to 02:46, 25 August 2010 (UTC) and have compiled the following preliminary results.

The votes are tallied as follows.

• Where a user has selected one option, that option is given 4 points.
• Where a user has selected multiple options:

• The first option will recieve 4 points
• The second option will recieve 3 points
• The third option (if any) will recieve 2 points
• The fourth option (if any) will recieve 1 point
• These rankings are detemined by the order in which the user lists the options.

• Strong, weak, etc. have no bearing on the results.

The following votes were discounted. A reason has been given beside each.

Close No votes discounted

Keep

• 73. Voting unclear
• 167. Voting unclear
• 184. Voting unclear
• 187. Voting unclear
• 218. Confusing vote

Other Responses All voting made in this section has been discounted from these results.

Option	Points	%
1	496	31.21
2	187	11.77
3	649	40.84
4	257	16.17
Total	1589	100

Conclusion... there is no clear consensus at the present time. Option 3 is the most popular, with 144 users making this their first choice but this is not enough to win against Option 1 which 124 users said would be their first option. --tb240904 ^{Talk Contribs} 03:34, 25 August 2010 (UTC)

Alternative conclusion: These statistics are meaningless. Users who opposed were able to allocate 4 points, users who do not were able to allocate up to 9 points.  Chzz  ►  04:21, 25 August 2010 (UTC)

Alternative conclusion: Good Point!--Gniniv (talk) 09:44, 25 August 2010 (UTC)

Alternative conclusion(s): 125 closes and 235 keeps is 65% keep support (the minimal keep as is case is implicit in both 3 and 4). Also, though subject to Chzz's 4 v 9 fallacy, the total 2+3+4 points are 68% of the total points. These stats are fuzzy though. Ocaasi (talk) 11:14, 25 August 2010 (UTC)

---

Based on the 2010-08-25t10:58:17z version, looking just at signatures and not actual edits, counting just the # generated numbers:

option	# of votes	percentage
close (1)	125	33.24%
keep (2-4)	240	63.83%
other	11	2.92%
Total	376	100%

-- Jeandré, 2010-08-25t11:42z, -- Jeandré, 2010-08-25t11:54z

I hardly see how anyone can be called "naughty" for first voting in a poll, and then voting in what appears to be another poll about whether the first poll should continue. – Smyth\^talk 12:20, 25 August 2010 (UTC)

section 1 Straw poll instructions

section 2 Straw poll

subsection 2.1 Close: option 1

subsection 2.2 Keep: options 2, 3, or 4

subsection 2.3 Other responses

Were the levels of sectioning different previously?

If the double votes are counted, for a total of 391 votes, then the percentages for close are 1.2% less, for keep 2.4% less, and for other 3.8% more. -- Jeandré, 2010-08-25t18:09z

The double votes that explicitly pertain to whether or not the poll should be restarted or continued should not be tallied as part of the straw poll. In effect, there are two distinct polls running in parallel on the same page.   — C M B J   20:52, 25 August 2010 (UTC)

2010-08-28

Based on the 2010-08-28t11:35:30z version, looking just at signatures and not actual editors, counting just the # generated numbers:

option	# of votes	percentage
close (1)	160	33.26%
keep (2-4)	307	63.82%
other	14	2.91%
Total	481	100%

-- Jeandré, 2010-08-28t15:56z

Seriously, can you ditch the "naughty" description? It's profoundly silly and, frankly, more than a little insulting. Millahnna (talk) 22:18, 28 August 2010 (UTC)

Yes please do.Sumsum2010 · Talk · Contributions 00:53, 29 August 2010 (UTC)

My name appears in this list four times for some reason.   — C M B J   05:59, 29 August 2010 (UTC)

Proposal

This poll is hopelessly flawed. It is also not a majority vote; we don't operate that way.

There may still be some chance to salvage a continued test, option 2, from the misguided enthusiam of its friends. by posing the question whether the test should be continued at all; some people, like Casliber, will oppose that as inadequate - but there aren't very many of them, and his first choice is option 1 anyway. I don't think there is hope of consensus to expand; but that should be asked in a separate poll at a separate time. Let's close this down and start a straw poll with two choices: continue as is, or abandon. If you don't, the result will be no consensus, which, given the way the trial was set up, is equal to abandon. Septentrionalis PMAnderson 02:18, 27 August 2010 (UTC)

The vote is balanced as that. the two first votes are YES and NO. Then, If yes, the option with the most votes is implemented. this was the purpose of the vote. It's efficient and quick. "kill's two birds with one stone." How ever you want to put it. It is a Democrocy, and this is not life and death

The purpose of the vote is to see which way the community wants to go. intelati 02:30, 27 August 2010 (UTC)

That is contrary to policy. Wikipedia is not a democracy; it does not run by majority vote. The way for a majority to behave here is to offer concessions to the minority to see if they can be won over; I can be, although many comments suggest they cannot. Septentrionalis PMAnderson 02:35, 27 August 2010 (UTC)

But, in the case of Wikipedia, the community wants to voice opinion on what they want to do. In the case of this vote it is to find the "community consensus is required for its continued use." If the Majority "wins" then that is the "law," exactly like the USA Impeachment process. Only if there isn't a Majority vote, It will stop no matter if 55% of the people want it to continue--intelati 04:44, 27 August 2010 (UTC)

And as democracy, this bears a certain resemblance to Piet Hein:

His party was the Brotherhood of Brothers,

That is, they constituted that minority

which formed the greater part of the majority.

Within the party, he was of the faction

that was supported by the greater fraction.

And in each group, within each group, he sought

the group that could command the most support.

The final group had finally elected

a triumvirate whom they all respected.

Now, of these three, two had final word,

because the two could overrule the third.

One of these two was relatively weak,

so one alone stood at the final peak.

He was: THE GREATER NUMBER of the pair

which formed the most part of the three that were

elected by the most of those whose boast

it was to represent the most of the most

of most of most of the entire state --

or of the most of it at any rate.

He never gave himself a moment's slumber

but sought the welfare of the greater number.

And all people, everywhere they went,

knew to their cost exactly what it meant

to be dictated to by the majority.

But that meant nothing, - they were the minority.

Septentrionalis PMAnderson 13:44, 27 August 2010 (UTC)

Alternative proposal

Two questions:

• Should PC be kept?
• If kept, should it be expanded?

A yes result on either would require consensus; and everybody could vote on both. (There is a possible third question, how far it should be expanded? - but that can certainly be asked later.)

At the moment, it looks as if it is unlikely, but possible, that the first question would get consensus. The second question may have a bare majority - it has no hope of consensus. But it may be that the numbers would change if this poll had actually been conducted in a straightforward fashion. Septentrionalis PMAnderson 14:13, 27 August 2010 (UTC)

See #Ask_each_question_separately above. – Smyth\^talk 18:10, 27 August 2010 (UTC)

Where's the option to keep, but use it across any form of article

Pending changes isn't 'Biography of Living Persons protection', so why is everything centred around BLP's? Are non-BLP articles not important enough to be worthy of pending changes? I'm finding it invaluable in protecting any sort of mid-level vandalism. - ʄɭoʏɗiaɲ ^τ _¢ 12:26, 28 August 2010 (UTC)

BLPs are a particularly urgent problem because of the privacy/reputation/liability issues. The current version, option 2, can be applied to any individual article, as can option 3. The original trial proposal emphasized PC as an i.p.-friendly alternative to semi-protection, and feedback from the trial indicated that use on very high-vandalism articles which had previously been semi-protected produced more work than was worth the benefit. I believe expanding the feature to all articles was seen as a non-starter and excluded from the options because it wouldn't have enough support. If the relatively minimal enthusiasm for option 4 is any indication, that hunch was pretty accurate. I don't think PC is robust enough yet to even be considered for use on all articles, not even considering the massive opposition to that large an implementation. Ocaasi (talk) 18:35, 28 August 2010 (UTC)

PC will never be robust enough for use on all articles because of the fact that it would be so severely understaffed that there is a good possibility that articles with pending revisions on them will remain unexamined for large periods of time. This is something most everyone understands. There's just not enough willing manpower to man PC feeds. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:24, 28 August 2010 (UTC)

Conditional on not having a backlog, I agree. The obvious counterpoint is German Wikipedia, which uses the feature on all articles, despite having a considerable backlog. That test-case has already been handled for us, and it seems the results wouldn't be acceptable for this Wiki or its community. I don't think it's impossible, but barring massive improvements to the interface, full integration with edit-filters and anti-vandal tools, and a whole lot of new editing talent, I think it's unlikely to ever happen. Ocaasi (talk) 19:41, 28 August 2010 (UTC)

Include abbreviated list from Closure page?

Would it help anyone coming to this page if there were a list of all the pros and an abbreviated list of all the cons, as there is at Wikipedia:Pending changes/Closure? —Jeremy ^{(v^_^v Carl Johnson)} 19:27, 22 August 2010 (UTC)

I think that's a good idea, although it is somewhat long and might make navigating the page more confusing. I added a small note in the introduction that a summary is on the discussion page, but it's not that prominent. Do you know how to transclude that section, if we decided to? Ocaasi (talk) 19:46, 22 August 2010 (UTC)

I would think {{Wikipedia:Pending changes/Closure#Pros (What worked)}} and {{subst:Wikipedia:Pending changes/Closure#Cons (what didn't work)}} would transclude/subst them. (Note that I said an "abbreviated" list of negatives, so we'd have to subst the cons list.) —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 19:50, 22 August 2010 (UTC)

FYI, I don't think transclusion works for sections. (I've always found that so frustrating.) Andrew Gradman ^talk/_WP:Hornbook 20:13, 22 August 2010 (UTC)

I was under the impression that it does, since it's been used in AN/I threads I've been in to allow a blocked user to say his piece. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:15, 22 August 2010 (UTC)

Can you read over the poll instructions? I edited them to reflect concerns in this discussion. Ocaasi (talk) 20:23, 22 August 2010 (UTC)

At this point I'm too pissed because Off2riorob has selective reading and Officer Levy above thinks I need to be indoctrinated into CRASH. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:27, 22 August 2010 (UTC)

No problem. Whenever you get a chance. I think the instructions are better, the vote has only just begun, and the current poll appears to fairly represent prior levels of support and opposition. In other words, little to no harm done. Plus, this is still a poll for consensus, and if a vast majority of voters express an opinion, it can and should be incorporated into future policy/implementation. Ocaasi (talk) 20:36, 22 August 2010 (UTC)

I don't appreciate your incivility. You're entitled to disagree with me, but there's no need for name-calling. I'm only trying to help. —David Levy 20:41, 22 August 2010 (UTC)

Your "help" is insulting me, Levy. Go find someone else to offend. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:42, 22 August 2010 (UTC)

Most of your responses are quite insulting (and clearly intended as such), but I refuse to be driven away by them. —David Levy 20:47, 22 August 2010 (UTC)

That's because you're not getting the fucking point! I don't want my "nay" vote to count as any form of "support" vote! What you're suggesting would change it to that! I want no part in how FraggedRevs survives on en.wiki, and you fucking well know that! —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 20:49, 22 August 2010 (UTC)

Jeremy, please listen for just a minute. The only reason anyone suggested counting oppose votes as support 2 votes, was to give you another say at limiting the feature. It'd be like, "Hey, want to get shot or not? (Oh, and if you end up getting shot, do you want it in the foot, the belly, or the head)?" It's NOT against you to say that if the vote winds up on you getting shot, that your oppose vote will count for getting shot in the foot: after all, if you must be shot, better to get it in the foot, right? Ocaasi (talk) 21:10, 22 August 2010 (UTC)

As I've stated several times, Jeremy, I don't propose that your response be construed as any form of support for the feature. I'm only arguing that under no circumstance should your opposition (and that of others) be ignored. —David Levy 21:18, 22 August 2010 (UTC)

I have made it explicit in several places, including my talk page, Ocaasi and Levy, that I am vehemently against FlaggedRevisions of any sort and would NEVER support them. The vote change should PendingChanges be accepted thus goes against my will, which I construe as a personal attack. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 21:20, 22 August 2010 (UTC)

You seem to be misunderstanding the concept, which has nothing to do with converting your input to any sort of support. Let's try this another way.

If you were answering the question "Should the implementation of pending revisions be expanded?", what would your response be? —David Levy 21:28, 22 August 2010 (UTC)

Silence. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 21:29, 22 August 2010 (UTC)

If that were this poll's question, you wouldn't have responded "no"? —David Levy 22:22, 22 August 2010 (UTC)

I wouldn't have responded. Why the hell have you failed to figure this out? —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 22:42, 22 August 2010 (UTC)

Everyone take a deep breath and focus on the issue. Please?--intelati 22:48, 22 August 2010 (UTC)

I'm trying, despite Jeremy's continual incivility. —David Levy 23:06, 22 August 2010 (UTC)

Perhaps because you've continually shouted profane insults and accusations instead of calmly conveying this.

However, that doesn't affect my position (which, as Millahnna has explained to me, you might have misinterpreted). 23:06, 22 August 2010 (UTC)

Am I getting this right? David what you are saying is that if the consensus is to keep PC, then people who voted oppose should have a say in how it is implemented. Meanwhile what Jéské is saying is that if the beast is kept, he wants no part of it, not even determining how it's used. Is that correct? Millahnna (mouse)talk 21:31, 22 August 2010 (UTC)

That is correct, Millahnna. —Jeremy ^{(v^_^v PC/VC is a show-trial!)} 21:35, 22 August 2010 (UTC)

And that's why it was necessary to shout profanities at me and engage in name-calling? —David Levy 22:22, 22 August 2010 (UTC)

As I stated above, I have stated explicitly both on this page and in the userboxen on the top of my talk page that I want no part of FlaggedRevs at all. Your repeated insistence that in the event that FR/PC passes that my vote should be converted into one that is obviously counter this viewpoint has only served to repeatedly rile me, especially when I indicated that I viewed such comments as tantamount to personal attacks. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:29, 23 August 2010 (UTC)

1. Referring to me as "Officer Levy" was a personal attack. Non-personal statements with which one disagrees are not.

2. You made it very clear that you objected to the suggestion with which I expressed agreement, but you consistently described it as a means of converting your opposition to support (a misunderstanding of what was under discussion). I attempted to explain that the intent was to ensure that your opposition (and that of others) not be ignored, and you continually replied with profanity-laced tirades reiterating your mistaken impression. —David Levy 19:59, 23 August 2010 (UTC)

If by "not be ignored" you mean "twisted into the opposite meaning" then you'd be right. If you meant otherwise, your words defied your meaning. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 20:10, 23 August 2010 (UTC)

I'm sorry that you still don't understand the intent. —David Levy 20:12, 23 August 2010 (UTC)

Why Can't you Be Friends?--intelati 20:22, 23 August 2010 (UTC)

The only "intent" I'm getting out of your statements is that IF PC passes THEN my vote nay gets converted into a vote for it! —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 00:11, 24 August 2010 (UTC)

And you're mistaken. —David Levy 01:08, 24 August 2010 (UTC)

How am I mistaken? Haven't you been saying something like that this whole time? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 01:10, 24 August 2010 (UTC)

No. This whole time, I've repeatedly stated that in no way would the responses in question be construed as support for the feature. In my very first edit to this page, I noted that "you need to think of it as a vote against expansion," and I've since stressed over and over again that the idea is to assign appropriate weight to the opposition (instead of ignoring it). But you refuse to believe me, so I can only reiterate that you're mistaken and express my regret that I'm unable to successfully convey my position to you. —David Levy 01:26, 24 August 2010 (UTC)

But it's a vote FOR FR, something you seem unable to understand! What part of "I will NEVER vote for any iteration of FlaggedRevs" are you ignoring? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 01:40, 24 August 2010 (UTC)

I'm not ignoring you. You're ignoring me (or are "unable to understand") when I state over and over and over again that I do not seek to construe anyone's opposition as "a vote for FR." —David Levy 01:59, 24 August 2010 (UTC)

Is that in between every instance where you say that a "support 1" option should be converted to "Support 2" if PC passes? Because if so, I'm not seeing it. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 02:10, 24 August 2010 (UTC)

Beginning with my first edit to this page, I've stressed that the idea is not to construe such responses as any sort of support for the feature, but to carry over the opposition.

If the poll isn't scrapped and restarted, I hope that we don't end up evaluating its results via the system originally described (the basis for the suggestion). If we do, there will be two phases (the latter contingent upon the former's outcome).

In the first phase, a "simple head count" will compare votes for option 1 against votes for options 2, 3 and 4 combined. If the votes for options 2, 3 and 4 combined compose a 2/3 or greater majority, that will be deemed consensus to retain the feature.

This triggers the second phase, in which votes for options 2, 3 and 4 are counted separately to determine the feature's upcoming implementation. At this point, the issue of whether to retain the feature has been settled and no longer is under consideration. The question becomes "Now that we've decided to retain the feature, should we expand its use (and if so, to what extent)?"

If votes for option 1 are simply set aside, this means that the opinions of everyone behind them are completely ignored when deciding how to proceed. The idea is not to pretend that these individuals voted for option 2 (or supported the feature's retention in any way), but to consider their opposition when evaluating whether there is consensus to expand the feature's use (and if so, to what extent). The suggested assignment of the numeral "2" to these responses is purely for the purposes of mathematical tabulation (in a context it which this label represents opposition to the feature's expansion, not support for its existence). It's an arbitrary designation, and we could just as easily relabel "option 2" votes as "option 1" or relabel both as "option x." No matter what term is used, in this context, it represents opposition to the feature's expansion and nothing more.

I hope that the poll is scrapped and restarted (and if not, that its results aren't gauged in this manner), rendering all of the above moot. —David Levy 03:26, 24 August 2010 (UTC)

You still aren't fucking getting it! Even after I've spelled it out several fucking hundred times and Millhanna confirmed it! I do not want my vote converted into one that obviously supports FraggedRevs, Levy! This includes fucking Option 2! —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:33, 24 August 2010 (UTC)

Wow. —David Levy 22:16, 24 August 2010 (UTC)

I think it's about time for User:Jéské Couriano to step away from the issue. BigK HeX (talk) 15:15, 29 August 2010 (UTC)

Not happening, BigK, as long as someone assumes that my "nay" should be twisted into a "support" of any form. If I wanted to support any of them, I would have voted to do so. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:48, 30 August 2010 (UTC)

David, for the record, while I appreciate that you want to find a way to make sure those opposed are still heard if PC is kept, a new and separate poll seems a better solution than assuming anything from their oppose votes unless they explicitly state it (eg "Opposed to keeping PC but if it is kept would prefer option 3). To assume that you can count all opposes as one of the options of keep if the consensus goes that way is more disenfranchising than you seem to realize. Millahnna (mouse)talk 21:31, 22 August 2010 (UTC)

Perhaps so. It's my opinion that a follow-up poll would be ideal, but I'm under the [mistaken?] impression that none is planned.

As noted above, I believe that everything written should be considered when gauging the consensus. I've expressed opinions on how a straight vote count could be carried out in the least harmful manner, but I do not advocate such a method. —David Levy 22:22, 22 August 2010 (UTC)

OK your view is starting to make more sense to me now. I hope your impression is mistaken because were it not, that would be detrimental to the whole WIKI concept of consensus and community writing. Which, I think, is essentially why Jeremy felt so strongly. Unless I'm misreading him, he felt that your attempts at finding a "least bad" tabulation method were a little more along the lines of endorsing a lack of a secondary poll than you actually intended them to be. Again, just my interpretation; I know I was certainly getting a similar impression myself. Millahnna (mouse)talk 22:34, 22 August 2010 (UTC)

I didn't realize that I appeared to be endorsing the idea of halting discussion after this poll. I only meant that if this poll is to be the final word on the subject, opponents' input shouldn't be ignored.

My [hopefully mistaken] impression that no follow-up poll is planned stems from the fact that this poll seeks to determine the feature's future implementation in the event that it's retained. (If a separate poll were forthcoming, why would we be addressing this now?) If a follow-up poll has been mentioned somewhere, please accept my apology for overlooking this. —David Levy 23:06, 22 August 2010 (UTC)

Based on Jeremy's confirmation that I guessed his intent correctly, that should clue you in David to a big part of the problem in what you are suggesting. Some users who are opposed will want no part AT ALL in the implementation if PC is kept. To assume that their votes mean anything about how to implement PC could create false priorities in the consensus at that point. As I said, best to just have a separate poll/discussion about the "how" if keep is the outcome. That way any opposed voters who ARE interested in contributing to implementation can contribute and those who want no part of it can walk away without any concerns that their original "oppose" vote will be translated to an assumptive vote for a specific variety of implementation. Millahnna (mouse)talk 21:43, 22 August 2010 (UTC)

To be clear, this wasn't my idea. I've expressed agreement that it would be the least bad method of arriving at a numerical tabulation (which I oppose). As indicated above, I agree that a follow-up poll would be best. —David Levy 22:22, 22 August 2010 (UTC)

Fait accompli

I'm very disappointed, and disillusioned, with the way this poll and discussion are being conducted.

I feel that we're being railroaded into acceptance.

I approached this idea with an open mind, and I - like many others - agreed to a specific 2 month trial. That trial is now over, yet PC remains active — without clear consensus.

Note, The trial will last for two months and then a community discussion will decide the future of the implementation, the default being deactivation.

When I first considered the trial, one of my greatest concerns was that PC might be 'introduced by stealth', and could run away, resulting in excessive protection across the board. I supported the specific fixed trial on a small number of articles, to see how it would work. The trial has produced some results, but there are many concerns. My personal vote, were it possible to do so, would be to work out a schedule for a further trial, after some of the issues have been addressed.

I appreciate that others might want to simply implement a live system, in some shape or form...but this 'straw poll' makes a mockery of the process of consensus.

I suggest that the only way to clear the air here is to remove PC, for now, and then we can discuss a next move.

And if that is difficult...well, that should have been considered when suggesting a 2-month trial with a default of deactivation. Otherwise, there was an assumption that we'd just play along and accept it as a fait accompli.

I am still trying hard to keep an open mind, and to believe in the system of consensus. Therefore, I hope you will withdraw the system ASAP, because there is no clear consensus to implement it, and then start discussions on how to move forward.

I, and I suspect many others, may well be prepared to consider implementing some form of Pending Changes on a permanent basis...but I feel that the methodology of this discussion is being conducted in a very unfair manner, and may actually backfire, causing people like myself who want to help introduce this very important feature with due consideration and care. Best,  Chzz  ►  01:39, 25 August 2010 (UTC)

I agree, though my timing concern is more for the timing with which the trial was conducted (June-August, when vandalism is low). Someone needs to talk to the higher-ups about this, and soon - this is essentially kicking consensus in the head. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 03:30, 25 August 2010 (UTC)

No one is railroading anything. The developers said that we couldn't turn the feature off and then back on. That meant we had to figure these issues with the features still running. That's option 2 in this poll. Option 3 allowed a limited expansion up to 10k articles for the purpose of seeing if applying PC to lower traffic articles is better suited to its use. (Option 4 has a few supporters, but not many of them, and it's generally considered premature to roll out to that many articles, even though BLPs might be the best candidates in the long run). All said, your preference "to work out a schedule for a further trial, after some of the issues have been addressed", I think is implicit in both option 2 and 3, since neither would rush towards a massive change until there were significant improvements. Believe me, many of the PC supporters see the numerous flaws in its interface and are looking forward to both developer updates and policy clarification. All of the votes for option 3 are not content with the feature, just reasonably comfortable enough to keep developing it. This poll didn't spell it out, but the notion of "trial" was still part of it. Ocaasi (talk) 08:51, 25 August 2010 (UTC)

"The developers said that we couldn't turn the feature off and then back on" - Translation for people who aren't developers: Said developers actually have to do some work if they want to turn the feature back on again. They'll do it regardless, it might just take a month or two to re-implement. This is all assuming this is still their position, and it would be incompetence on thier part if the code is structured so badly that this has to happen, so they are trying to take the path of least resistance. Ryan Norton 13:16, 25 August 2010 (UTC)

Noooooooooo.....! Wikipedia:Do_NOT_bite_the_developers Ocaasi (talk) 15:35, 25 August 2010 (UTC)

I work in software development and have a fair understanding of how this wiki is designed. Things like this pending changes are just pluggins to the main platform. They can be added. They can be removed. Removing will cause a loss of data, I would expect, such as the revision information. But it can be turned off. (and was advertised as such, otherwise our supposed "trial" was no trial at all) I too agree with the above opinions. I personally supported keeping the PC, but I believe this poll is constructed in such a manner that it disenfranchises the opposers. The supporters are able to answer two questions, one - do you support keeping the program, and two, what will its future look like. Opposers have only been able to say they do not want it, and give no say on what its future would look like, should their position not carry the day. Good solutions to this problem have been brought up days ago now, but no action taken. It getting to the point where this poll has gone to far down the wrong path and cannot recover. I am beginning to favor restarting it in a better format. —Charles Edward ^{(Talk | Contribs)} 17:05, 25 August 2010 (UTC)

Disenfranchises the opposes, I asked them, the people that oppose the tool oppose it, that is if there is a consensus to keep the tool running they want it to run as minimum as possibe, makes sense really. Off2riorob (talk) 17:08, 25 August 2010 (UTC)

No, that is an assumption, Off2riorob. Some of us do not want it running period and will not support any other option. Read my shouting matches with David Levy. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 22:36, 25 August 2010 (UTC)

Yup, I agree with that. But I feel like once this poll is done, their votes will not be counted in that manner. Maybe I a mistaken though? —Charles Edward ^{(Talk | Contribs)} 17:23, 25 August 2010 (UTC)

This has been a point of contention. —David Levy 17:40, 25 August 2010 (UTC)

I am sure worst fears will not materialize. Clearly all positions will be considered in the closure and in the decision as to how we progress after that, might be a period of discussion and then another poll, lets not dis-rail the ongoing discussion and the information that is being collected in this poll. Off2riorob (talk) 17:49, 25 August 2010 (UTC)

I don't think it would be a problem to turn the extension off, to delete all the data and protection settings, and then to turn in on again later with empty tables. It would be more difficult to do something useful with all the stale data. There's an issue of referential integrity, for example if pages are deleted, and we wouldn't really want to revert all the pending-protected pages to old "stable" versions after a delay of several months when no reviewing was being done. Ryan is correct that it is just a matter of extra development and sysadmin work, but there's no guarantee that we will "do it regardless". -- Tim Starling (talk) 04:33, 27 August 2010 (UTC)

That's the advantage of a small trial: this is the list of pending changes; there was one - and since it was obvious nonsense, I rolled it back. That's typical. If somebody is standing by to clean up the list, just before turnoff, this shouldn't be a problem. (How many of the pending pages have been deleted?) Septentrionalis PMAnderson 14:49, 27 August 2010 (UTC)

I think you misunderstood. The main issue with stale data occurs after the extension is turned back on, not after it's turned off. When the extension is off, the most recent version will be shown, unreviewed or not, and I agree that there wouldn't be many instances of vandalism exposed by doing this. When the extension is turned back on, the most recent stable version will be shown, and since reviews were disabled, that stable version will date from before the extension was turned off, so anonymous users will see a version of the article that is several months old. This will affect most PC-protected pages. As for deletion and other referential integrity issues: it's the sort of thing that can waste a lot of developer time regardless of how many pages it affects. So like I said, the simplest way to turn the extension back on is to wipe all the data first. -- Tim Starling (talk) 07:25, 29 August 2010 (UTC)

Wouldn't the sensible thing in that case be to remove Pending Changes from all pages before the feature is switched off - if necessary replacing it by semi-protectionIf a page is that unstable that wedon't have time to fix it then PC isn't working anyway.Nigel Ish (talk) 10:51, 29 August 2010 (UTC)

People are missing one point: it doesn't matter if the feature is available, we don't have to use it. Until we have consensus to use the feature after the trial, all articles under pending protection should be converted to another form of protection or unprotected.—Kww(talk) 17:44, 25 August 2010 (UTC)

Why is that, its working fine without problem and there is a clear support for the tool to continue in some way. Off2riorob (talk) 18:03, 25 August 2010 (UTC)

Because the agreement was to have a two month trial. Two months is over, and there is no consensus to continue.—Kww(talk) 18:32, 25 August 2010 (UTC)

I agree. --Yair rand (talk) 18:34, 25 August 2010 (UTC)

Yes there are some users that hold this position but there is a clear majority of users that support the continuation of the tool in some way, and while discussion is ongoing I can see no reason to switch off the tool at all, as there are no problems at all with the tool as it is operating on the limited article as it is now, we are not to a week or two extension while discussion continues. Off2riorob (talk) 19:48, 25 August 2010 (UTC)

"I'll only put it in a little bit" is an argument that only works with young girls. The trial is over. There is no consensus to continue it (on !votes, we normally measure consensus at boundaries arund 70%, which this has not met). That means that status quo (no pending changes) continues until there is consensus to change.—Kww(talk) 20:02, 25 August 2010 (UTC)

Could you edit that comment for our diverse audience? Ocaasi (talk) 20:15, 25 August 2010 (UTC)

Yes I can see your position but wikipedia is not so cut and dried as I have seen and I personally doubt that there is any good reason to do that, I also doubt there is some kind of countdown clock about to switch it off on the day a month is over.Off2riorob (talk) 20:19, 25 August 2010 (UTC)

Re Ocaasi: yes, that was probably a bit obscure for people with English as a second language. "I'll only put it in a little bit" is a reference to the kind of lie a young man uses to convince a girl to have sex. Idiomatically, I'm saying that it is deceitful to say "I'm only going to have a trial, and I will stop in two months", and then, two months later, say "well, I don't see that everyone wants me to stop, so I'll keep going". And yes, Off2riorob, it is that cut-and-dry: to say you will have a two-month trial and then not stop when the trial is over is deceitful.—Kww(talk) 21:27, 25 August 2010 (UTC)

Kww, I know the phrase. Can't say I practice the technique. I was hoping you'd "edit" it for all of the readers who might prefer a different analogy. Thanks, though, for your thorough explanation. Ocaasi (talk) 22:14, 25 August 2010 (UTC)

Wikipedia needs a way to partially protect FAs and GAs and this is an ideal tool

Recently I was reviewing some of WP:MED GAs and came across some vandalism that was a couple of months old. Section of the article had been removed by an IP [2]. Another IP had come along and replaced it with content of much worse quality.

The least we can do for people who put in the work to get an article to GA is properly protect it from IP vandalism. I wish to role out pending revisions to all nearly 100 GA in WP:MED.Doc James (talk · contribs · email) 20:56, 27 August 2010 (UTC)

No, we don't. If the FA star were worth the electrons it is written on - and I say nothing of GA - we might (although the appropriate solution for those who care about "their work" is to watchlist the article); but vandalism of ill-written, ill-sourced articles (but with all the quote marks in the "proper" places) is no more important that vandalism of other ill-written and ill-sourced articles. Septentrionalis PMAnderson 21:01, 27 August 2010 (UTC)

There's some good logic in trying out PC on medical articles, especially given how frequently they are consulted for serious information (WP:Disclaimer aside). FA and GA articles are hardly just formatted correctly; regardless, what is the point you're suggesting: that since all Wikipedia articles suck we shouldn't try to improve any of them? Ocaasi (talk) 21:05, 27 August 2010 (UTC)

I am attempting to attract my colleagues to edit Wikipedia. These are people who may be willing to bring articles to GA / FA but are probably not willing to protect them going forwards. If they see that the work they put in quickly vandalized an no one protects it they will not contribute. There are a lot of wikis out there to which people can contribute medical content List_of_online_encyclopedias#Medicine_and_surgery. I occasionally go on holidays and am not able to keep track of all the GAs in my list of a nearly a thousand articles I follow. As the above example shows Wikipedia does not have great mechanisms to protect its current high quality content. If we wish to attract academia we need way to improve our reliability.Doc James (talk · contribs · email) 21:15, 27 August 2010 (UTC)

Installing Pending Changes does not improve articles; like semi-protection, it may prevent them from getting worse when Recent Changes Patrol is not enough - that's the most we can hope for. If anything, it slows down the rate of improvement by keeping out anon improvements. Anybody who supports it in that hope is dreaming. Septentrionalis PMAnderson 18:27, 28 August 2010 (UTC)

Quite. This recurring insight that Pending Changes equates to "We can finally keep the rabble out!" is not a point in its favor, but the most fundamental reason to oppose any such feature. — Gavia immer (talk) 18:35, 28 August 2010 (UTC)

Something is off in both arguments. Pending changes lets anonymous ips edit. Semi-protection doesn't. PC gives more opportunity for good edits to come through while limiting the effect of vandalism, particularly on sensitive articles. It depends what its being compared to and what its being applied to.Ocaasi (talk) 12:58, 29 August 2010 (UTC)

Every time I see the argument that PC "goes against the WP spirit" or whatever I can't help but wonder how it could be said with a straight face, especially since they usually will supposrt semi-protection at the same time. I REALLY wish someone would explain to me why so many people say this, where it's so incredibly obvious to me that PC let's MORE people edit, not less. ♫ Melodia Chaconne ♫ (talk) 13:53, 29 August 2010 (UTC)

Totally agree with Melodia, as it's much more friendly to new and casual contributors. Flatterworld (talk) 14:06, 29 August 2010 (UTC)

But that's only a fair comparison if we were proposing to semiprotect wide swaths of articles permanently. We don't, because we lose by doing so. We would lose less by PC, but not that much less. Septentrionalis PMAnderson 23:57, 29 August 2010 (UTC)

We would actually lose more, because (see above responses) well-meaning contributors would be so much more willing to implement broad swathes of PC protection without thinking any further than "there's no downside for me, so there's no downside". — Gavia immer (talk) 00:22, 30 August 2010 (UTC)

Agreed in part. While I doubt any reasonable admin would implement PC on an article without first thinking long and hard about its appropriateness, many of the issues I have with the feature lie in the psyches of the community, not in any technical issues. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 21:00, 30 August 2010 (UTC)

Pending changes in fact lets the "rabble in". If people are serious about scholarship they will stick around. They will also be encouraged that their work here has at least slight oversight. The greatest benefit of PC is that it will improve the standing of Wikipedia in the eyes of our reader and new editors.Doc James (talk · contribs · email) 18:25, 1 September 2010 (UTC)

Pending changes keeps GAs and FAs peer reviewed

Both GAs and FAs are to be peer reviewed. Pending changes will keep them partial peer reviewed going forwards. Address another of the accusations from academia that Wikipedia is not peer reviewed ( when in fact it is to a limited extent ).Doc James (talk · contribs · email) 21:28, 27 August 2010 (UTC)

The requirements for GA are purely formal; all too many FA reviews do not consider content at all - because that is the hardest thing to review. Presenting them as "peer review" would border on fraud; to expect Pending Changes reviewers to do even as much as FA does is fantasy. Septentrionalis PMAnderson 18:44, 28 August 2010 (UTC)

Anderson's right. Doc, what you're describing is basically Citizendium Lite. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:22, 28 August 2010 (UTC)

I know little about other projects but believe me we look at content when we review GAs and FAs at WP:MED. Pending changes is the best of Wikipedia and Citizendium. It may get academia to take us more seriously. The easy stuff has been done ( ie the low fruit have been picked ). We now need more scholarly work to improve from here.Doc James (talk · contribs · email) 18:21, 1 September 2010 (UTC)

Listen to what PMAnderson said, Doc. Legitimizing GA and FA as "the best of the best" is crap because neither have subjective content accuracy standards. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 05:49, 2 September 2010 (UTC)

This poll is such a MESS!

So have we really ran out of anything meaningful and useful to do in Wikipedia?! well, this whole "pending changes", and even the poll made to decide upon it, is such a terrible mess; i can't believe it, EVERYONE is confused!! Just bring down this whole thing and move on fellows. This is the kind of stuff you should be doing perhaps in the sandbox, but not here! I'll be damned if any of this confusing play-with-the-toy thing you've been on about takes place in here. Non of this is needed, non of it is necessary. No pending changes no flagged revision, we don't need any of this. Maysara (talk) 12:55, 30 August 2010 (UTC)

Ah, indefinite semi-protection. So much less confusing to the non-autoconfirmed. -- zzuuzz ^(talk) 13:34, 30 August 2010 (UTC)

(I'm assuming you're being sarcastic here.) I don't see how a lock at the top of the article is confusing; it's pretty straightforward, if you ask me. Aside from that, there should be a notice on the edit page of a protected article asking non-autoconfirmed users to use {{editsemiprotected}} on the talk page to ask for a change. It allows more anonymous input, but it's not used often enough. _EricLeb⁰¹ (Page | Talk) 15:17, 30 August 2010 (UTC)

But it's not used because it's inconvenient. Pending changes is the same as edit semi-protected in effect, but the user experience is seamless. Ocaasi (talk) 18:22, 30 August 2010 (UTC)

It's slightly inconvenient, perhaps, but it's much more efficient to allow the article experts (who have the page on their watchlist) to decide whether or not something goes in, instead of leaving it to the community, who may not have a clue as to what is vandalism or what is not. Why fix something which isn't broken, which then leads to additional problems needing to be solved? How do we prevent a user to take advantage of his reviewer rights to insert a controversial bit into an article, even though others disagree? How do you prevent new users from being discouraged, seeing that their revision has not been accepted immediately after inserting it? How are we supposed to cut down a possible 10,000 article backlog (and even more if all BLPs are affected) voluntarily? How is this supposed to reduce the time spent fighting against vandalism when the process takes twice as long?

It's a very useless feature. Occam's razor says we should leave this as simple and as effective as possible. _EricLeb⁰¹ (Page | Talk) 18:56, 30 August 2010 (UTC)

Article experts aren't perfect either; they can be biased in a way reviewers are not. Besides, pending changes isn't optimal for situations with edit wars, more for situations with a high potential for harm from vandalism, and with the possibility of improvements from anonymous i.p.s. Watchlists are great, but part of the concern is about pages that don't have many watchers, especially BLP pages where vandalism can have such a serious impact. Also, edits accepted by reviewers on pages with many watchers can be easily reverted if incorrect. Reviewers are not 'the community', they are a subset of experienced members of the community. Just because something isn't broken doesn't mean it couldn't work better. Currently semi-protection excludes i.p.s and is not applicable to large numbers of articles. Reviewers can't be guaranteed not to accept controversial edits, just like any article; mistakes or bad edits from reviewers can be dealt with the way any bad edits are, with another edit. The lack of instant gratification for new users has to be balanced with their exclusion from protected articles and the risk of vandalism to BLPs. Good page documentation can help explain the process. Appropriate application of the feature can keep the reviewer backlog clearing quickly. We don't know how large a backlog would be; that's the point of trial. So far, pretty good. Most BLPs are lower traffic than the current sample, so a linear expansion of the backlog would be unlikely, though there would be some. Pending changes works with vandalism fighting--an edit can still be rolled back as normal--and better integration with anti-vandal tools would be likely if the feature had traction. Pending changes as a feature is just one option; it's neither finished being developed or settled what article it works best with. That's why a continued trial makes some sense. Ocaasi (talk) 19:58, 30 August 2010 (UTC)

Why can't we just have a vote with buttons to click, tally them up, and get it over with????--The Wing Dude, Musical Extraordinaire (talk) 19:13, 30 August 2010 (UTC)

See policy: WP:NOT#DEMOCRACY. BigK HeX (talk) 21:15, 30 August 2010 (UTC)

What we do currently is barely working. Recently I came across 2 month old vandalism ( large sections of text deleted ). Many edits had occurred afterward. If it was reverted when it occurred work would have been decreased. Yes Wikipedia is the encyclopedia that everyone can edit, pending changes in no way changes this.Doc James (talk · contribs · email) 18:36, 1 September 2010 (UTC)

Where do we go from here

As straw polls aren't a necessarily binding tool, how does the conversation proceed next? BigK HeX (talk) 15:11, 29 August 2010 (UTC)

For one, the poll is open for about another week, so we have to wait for everyone to weigh in. After that, some very cunning, diplomatic admins will try and ascertain consensus and close the poll. Ocaasi (talk) 20:36, 29 August 2010 (UTC)

Pending Changes will probably die, because while a super-majority is in favor (including me), IMPO consensus is lacking. Percentages on straw polls don't tend to change over time, so I feel comfortable predicting this now. Many (not all) of the arguments against are things that can be addressed and corrected in the future, so it is possible we will see a better version of this spring to life again someday. Another possibility is another trial period with some changes made, although I don't see the support there for that. Next time we pilot something, perhaps we should have a clear drop dead date, and make these decisions prior to that date, not after it.--CastAStone/^/(talk) 03:52, 30 August 2010 (UTC)

The poll as is presently is at 66 percent and is imo passing to proceed with the trial in some way. Off2riorob (talk) 13:48, 30 August 2010 (UTC)

I think that's right: there isn't consensus to discontinue PC, but there isn't consensus to continue either. What I'd like to see is the straw-poll being seen as a decision to keep the trial running for now. That would give us an opportunity to address the concerns raised about the closure discussion and the straw poll (and I'll state my position now: discussion is good, voting is evil ;-) TFOWR 13:56, 30 August 2010 (UTC)

I think we'd be taking a step backwards to remove pending changes, but without a lot more ground work, pending changes isn't entirely workable. Edit filter integration, and something resembling a full workflow along with the accompanying automated and semi-automated tools to manage a large backlog are needed to make pending changes really work. I do think we need to stick with pending changes while we make the improvements to make it work, because once we abandon it, it would be nigh-impossible to get another try even with improvements. Triona (talk) 07:19, 30 August 2010 (UTC)

I agree. I'm pro-PC, but I don't see how this straw poll could be massaged to mean "consensus to keep". The next step would be to seek consensus for a new trial in a few months, after a few fixes.

This makes me wonder: Were attempts made to make improvements to PC during the course of the trial? That may have been our mistake. We could have fixed some of the bugs that were identified early on, and blunted some of the votes that we're seeing now against retaining it. Andrew Gradman ^talk/_WP:Hornbook 08:40, 30 August 2010 (UTC)

Attempts were made to try and address the speed issues, but the devs seem to run on a variant of Valve time and as such the attempt came out, IIRC, about halfway thru the trial instead of right as it was pointed out. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:24, 30 August 2010 (UTC)

Support changes recommended by Triona. I am willing to keep it in the interim, but I still went for option 1 to avoid the acceptance implied by option 2. --UncleDouggie (talk) 09:57, 30 August 2010 (UTC)

I think the thing to do at this point is to cleanly discontinue the trial. I've been doing my small part by removing pending changes protection from every article I get a notification about, but I haven't been actively searching them out. Then people need to get a consensus as to what modifications are required to start a new trial. Just keeping an old trial running forever while you are trying to get the new one ready is wrong.—Kww(talk) 14:55, 30 August 2010 (UTC)

Finding the community's comfort level

I agree that this poll isn't going to give a clear consensus. However, I do think the poll has done it's job -- which is to highlight the contentious areas. Also, I think that there is clearly enough support to justify an extension of the trial. Later, we may be able to pose narrower questions to the community to see where any broad agreement may lie. I believe the demographic breakdowns that would be useful to identify are:

1. Oppose (per Page Protection being superior to Pending Changes in anti-vandalism efforts): People who do not support the concept of pending changes being used as an alternative to Page Protection in any fashion, and prefer continuing the status quo usage of Page Protection against vandalism, because Page Protection is a superior tool to Pending Changes.
2. Conditional Oppose (per fear that Pending Changes will be used pre-emptively). People who hold a fear that Pending Changes will be abused, despite guidelines being written that Pending Changes should be used only where Page Protection would previously have been applied.
3. Weak Oppose (per Pending Changes being unusable at present): People who -- at present -- do not support any usage of the current version of pending changes, but may change their minds if certain issues are improved with the pending changes system, including (but not limited to) programming issues (such as speed) or uncertainties in the guidelines which govern Pending Changes. The issue should be dropped for now, and possibly broached later (perhaps, 6 months from now). In the meantime, the status quo usage of Page Protection against vandalism is preferred.
4. Weak Support: People who do not yet support final acceptance, and would prefer yet another extension of the trial, possibly scaled up to gather more representative statistics on vandalism.
5. Weak Support: People who are wary of either the programming for pending changes in its current state or the mission/policies of pending changes, but support very limited and judicious use of Pending Changes on low traffic articles (less than 40k hits or so per month?) that would otherwise be Page Protected.
6. Conditional Support (per Pending Changes being a superior alternative to Page Protection): People who believe Pending Changes is an acceptable alternative to Page Protection (and preferable to the status quo) and believe it should be used reactively on vandalized articles, but only in groups that will be clearly defined in later discussion, and may be, for example, some combination of the following categories: low-traffic BLPs, Featured/Good articles in only the WikiProject Medicine, and/or heavily sockpuppet targeted articles (and perhaps some other small category). Support is conditional upon Pending Changes being used only where Page Protection would have been applied.
7. Conditional Support (per Pending Changes being a superior alternative to Page Protection): People who believe the Pending Changes system is an acceptable alternative to Page Protection to use on any article throughout Wikipedia, as needed, with the firm exception that it should almost *never* be used on high-traffic articles (greater than 150k hits or so per month?) unless there's some compelling community consensus for an individual high-traffic article. Support is conditional upon Pending Changes being used only where Page Protection would have been applied.
8. Support: People who believe Pending Changes is an acceptable alternative to Page Protection (and preferable to the status quo) and believe it should be used on vandalized articles as needed, and also should be used proactively. A bot should apply Pending Changes to articles that fall into groups that will be clearly defined in later discussion, but may be some combination of the following categories: all low-traffic BLPs, all low-traffic Featured/Good articles throughout Wikipedia, and/or heavily sockpuppet targeted articles. Though acknowledging that Pending Changes guidelines have been written such that it is only used where Page Protection would be applied previously, there is support to explore the idea of more liberal use.
9. Support: People who support liberal use of Pending Changes in broad areas where vandalism has been a problem (or may be a problem in the future). Though acknowledging that Pending Changes guidelines have been written such that it is only used where Page Protection would be applied previously, there is support to explore the idea of more liberal use.

I think this covers all of the issues raised in the straw poll. Getting responses from the community on something like this before final acceptance should help to identify just how comfortable editors are with the Pending Changes system. BigK HeX (talk) 15:32, 30 August 2010 (UTC)

Nice write up; I can't think of anything to add at the moment. I know what my biggest concerns are in how PC has been implemented (chief among them being poor guidelines as I just reverted blatant vandalism that someone accepted) and you seem to have encapsulated them all. The statistics are very important also. Many have asked for data on how well PC has done at it's potential task and all any of us seem to have is "anecdata". Millahnna (talk) 17:58, 30 August 2010 (UTC)

How about me? My feelings are that, if it were not for the tactics and arguments of the supporters, I would be willing to let PC continue at the present pace; but I cannot trust them. Septentrionalis PMAnderson 19:06, 2 September 2010 (UTC)

Will attract more editors to Wikipedia and increase the reputation of Wikipedia

One hears the argument that PC will keep people from editing. I believe this to be completely false. One of the reasons I hear from colleagues why they are not interested in contributing is that they fear that the work they put in to generate good or excellent content will be destroyed by anonymous vandalism. PC will alleviate some of these fears and thus will attract hopefully many productive editors.

One of the concerns from readers and teachers we hear is regarding Wikipedia's reliability. PC by increasing our reliability will thus increase the confidence of our readers. Hopefully increasing readership and thus potentially those willing to contribute financially or as editors to this encyclopedia.Doc James (talk · contribs · email) 04:34, 2 September 2010 (UTC)

Sanger tried that and thus far has failed. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:45, 2 September 2010 (UTC)

Ah no Sanger did not try "Pending Changes" as far as I am aware? Doc James (talk · contribs · email) 04:46, 2 September 2010 (UTC)

Pending Changes is very similar to Citizendium's article-approval process. The only difference is that PC includes the wildcard Sanger specifically omitted. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:47, 2 September 2010 (UTC)

Funny, I seem to recall hearing you argue the exact opposite point a few threads up. People stay away from WP because it's a wiki (anyone can edit, contributions might be vandalized, people doubt its reliability), so you think it should be less of a wiki, so you think we should implement PC, but you strangely seem to think that it will still be a wiki so implementing it isn't problematic. You're not making any sense. One major point you don't seem to understand: Wikipedia exists because it is a wiki. Edits become visible instantly. That is Wikipedia, that's all there is. --Yair rand (talk) 04:52, 2 September 2010 (UTC)

What we will have with pending changes is still completely different then what Sanger used. 95% of Wikipedia's pages will still be free for anyone to edit. No credentials will be verified before or after. etc. This comparison is a false argument. Pending changes still allows anyone to edit and more so than semi protection. I do not understand what is not clear.Doc James (talk · contribs · email) 04:54, 2 September 2010 (UTC)

If the comparison is a false argument, then what is the method used for approving edits between PC and Citizendium, Doc James? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:57, 2 September 2010 (UTC)

One can still edit Wikipedia easily while not logged in. Reviews are not required to be experts. Most page edits will not be reviewed as this will only be used on certain pages (similar to semi protection).Doc James (talk · contribs · email) 05:05, 2 September 2010 (UTC)

That's a dodge. Answer the question. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 05:07, 2 September 2010 (UTC)

If you go to Citizendium (which I just did) it does not let you change anything without logging in. If I go to a pending changes page on Wikipidia while not logged it you can make changes ( all that is required is approved ). Thus will I cannot contribute anonymously to one I CAN to the other. Doc James (talk · contribs · email) 05:11, 2 September 2010 (UTC)

Chummer, answer the question and stop evading it. I have a point in asking the question. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 05:14, 2 September 2010 (UTC)

Sorry I am unsure what "Chummer" refers too? but think I have answered the question clearly. What PC will create is clearly different than what Citezendium currently does.Doc James (talk · contribs · email) 05:19, 2 September 2010 (UTC)

You're still dodging it. I will ask again, assuming you lack selective reading: What is the method used for approving edits between PC and Citizendium? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 05:21, 2 September 2010 (UTC)

(undent) Sorry you lost me at Chummer. Do you mean "Each member employed a "chummer"—a young boy paid to bait the member’s hook with lobster tail, and cast chunks of lobster into the surf"? Some of you vernacular is unusual.Doc James (talk · contribs · email) 05:23, 2 September 2010 (UTC)

• facepaws* Jesus Christ, and I thought El Machete Guerrero had selective vision... —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 05:25, 2 September 2010 (UTC)

You seem to be asking a question for which you obviously already have the answer... I have however clearly explained how these two entities differ and no amount of incivility will improve your argument.Doc James (talk · contribs · email) 05:29, 2 September 2010 (UTC)

What is the method used for approving edits between PC and Citizendium? Focus on that question right there. You have not answered it. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 05:32, 2 September 2010 (UTC)

Neither have you answer my previous question above. Maybe we should just leave it at that.Doc James (talk · contribs · email) 05:35, 2 September 2010 (UTC)

People, Doc. Both PC and Citizendium article-review processes require the edits go through human beings, whom naturally are biased, spiteful, and looking to be the top of the heap. It is for this that PC will fail. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 05:40, 2 September 2010 (UTC)

Ah I guess you and I have a different take on the fundamental nature of ALL human being. I see both the best and worst on a daily basis. I see PC as a way to let the best shine through and decrease the impact of the worst. Wikipedia exists because of the best of people. Our methods of protecting the project exist because of the worst of people.Doc James (talk · contribs · email) 05:50, 2 September 2010 (UTC)

If I shared a vacuous optimism on the nature of reviewers, and unbounded fears on the nature of vandals, I might support this despite its advocates. But I do not see that angels have descended in the form of reviewers, and these arguments are not tenable without them. Septentrionalis PMAnderson 19:14, 2 September 2010 (UTC)

Wikipedia exists and that should at least generate some optimism about those of us who edit here and thus will be reviewers.Doc James (talk · contribs · email) 19:28, 2 September 2010 (UTC)

Speak for yourself, Doc. I have seen enough to know that anyone with Reviewer tools is in no way worthy of using them. All human beings have some form of bias, and many will fight to support said bias. It's as PMAnderson says: Those who believe all humans are universally good will ultimately end up destroyed by the "good" intentions of others. You're gambling on a psychological and sociological impossibility, unfortunately the same one FraggedRevisions depends on. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 22:35, 7 September 2010 (UTC)

I'd take your word for it, but you keep saying we can't trust eachother ;p. I'm sure you skepticism is rooted in experience, but like the Wiki project in general, PC is not "supposed" to work, but it does. Partly this is do to the open and transparent nature of things, but largely it's due to the generally good motivations of editors. Vandals are a minority. They need to be managed. Wily editors are a minority. They need to be educated or restricted. Aside from that, nothing you're so sure about is really anything new. Wikipedia has never been ruled by angels and yet it has persistently worked pretty great. Pending changes is no more vulnerable to gaming, subtle vandalism, vandal gangs, piranhas, or anything else than Wikipedia already is. The mutual oversight which makes Wikipedia function isn't less robust under PC; if anything it's more. You can dislike the feature for all kinds of legit reasons, but I don't think editors are inherently evil is one of them. Ocaasi 23:39, 7 September 2010 (UTC)

New proprosal - soft shutdown of PC protection

Because the developers have indicated that it would not be feasible to turn the feature off and turn it back on again, but there is a significant minority objecting to the feature, I propose an interim solution. Keep the feature enabled in MediaWiki, but end the current trial "softly" by removing PC protection from all pages within 30 days while we sort out where to go from here. The rationale for this is simple - while a majority wants to continue, the objections of the minority need to be addressed before going forward. This is likely to be the least disruptive way to proceed from here, and its the option that is least likely to "tie our hands" going forward. At same time this is going on, or shortly thereafter, we can discuss the objections and what would make this more palatable to those currently opposed. Triona (talk) 01:34, 8 September 2010 (UTC)

• Support. Makes sense to me. --Tryptofish (talk) 01:39, 8 September 2010 (UTC)
  • Having read the subsequent arguments, I now think that it should be much less than 30 days. --Tryptofish (talk) 15:26, 8 September 2010 (UTC)
    • I'd suggest removing the "30 days" and leaving it at "however long it takes us to remove PC from pages". As far as I know, every page that I applied PC to is now un-PC'd (and I loved PC - I used it in preference to semi on a whole load of articles), so I can't see it taking too long either. The "30 days" issue is just going to be another stick to beat PC with, so let's amend it to "until it's removed". That means that admins opposed to PC can hasten the process by un-applying PC. TFOWR 15:32, 8 September 2010 (UTC)
• Oppose. The devs said two months; we gave them three. It's their fault they didn't make it easier to shut down. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 01:41, 8 September 2010 (UTC)
• Conditional Support proposal as phrased, as the best we can do. Oppose any caveat as below. Septentrionalis PMAnderson 01:46, 8 September 2010 (UTC)
• Support with the caveat the continued use in small trials is permissible.Doc James (talk · contribs · email) 01:50, 8 September 2010 (UTC)

• What small trials? You've had three months, Doc, and the sample set was anywhere from 25-50% of what was intended (500-1000 articles, instead of the planned 2000). You've had your small trials. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 01:56, 8 September 2010 (UTC)

• Oppose, as pointless. Those with objections should be discussing what would make it more palatable now. For being in the (significant) minority, those who oppose PC are still demanding an awful lot, without any indication of giving much. BigK HeX (talk) 01:55, 8 September 2010 (UTC)

• Chummer, those who opposed (or even supported) with caveats told you the caveats directly on the straw poll or came to the talk page and explained themselves. It doesn't help that it appears the pro-PC side is trying to fuck over the anti-PC side. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 01:58, 8 September 2010 (UTC)

I have no idea what point you're trying to make about the "caveats" on the straw poll. Also, you don't have to repeat your uncompromising opposition at every opportunity. At this point, the discussion is going to involve a lot of compromise. *IF* you want your opinion to carry more weight, it might not a bad idea to decide where it would be least uncomfortable for you to give ground on the issue. But, again, that's only if you would like to help decide the compromise. BigK HeX (talk) 02:06, 8 September 2010 (UTC)

• • No compromise has been suggested by the majority; the minority "demand" the original terms of the trial, and decision making by general agreement. Majoritarianism can go too far; this is not the Social Democratic Labor Party, vintage 1903.Septentrionalis PMAnderson 02:12, 8 September 2010 (UTC)

Doc James has suggested an amazingly low-impact compromise that, as far as I can tell, seems to have been roundly rejected by opposers who have voiced an opinion. The GA's and FA's in his favored WikiProject covers what ... like a couple hundred articles, if that? It's very unclear where those who oppose PC and are willing to compromise would like to move forward ... at least to me. I guess I can just come out and ask you ... on what aspects of PC would you want to "meet in the middle"? BigK HeX (talk) 02:16, 8 September 2010 (UTC)

By considering, and probably supporting, a future proposal for a small trial, with assurances that it will be limited - and will be turned off, without this sort of disingenuous footdragging, if it fails to win general support. This is "we won't turn it off unless you agree to a compromise", for which there is an unkind - but accurate - name. Septentrionalis PMAnderson 02:23, 8 September 2010 (UTC)

You seem to think it already has "failed to win general support" as if this were not a subjective measure. You seem to not realize that many here DO see the roughly 2/3rds in favor as having won general support. BigK HeX (talk) 02:28, 8 September 2010 (UTC)

Anybody who thinks so does not understand how Wikipedia operates. The test of general agreement is that no significant voice is unwilling to tolerate the idea; that is not the case here. Septentrionalis PMAnderson 02:38, 8 September 2010 (UTC)

Ummm ... there's very significant voice opposed to a rejection of PC. Pretty obviously, on the binary question of whether PC will run or will not run, we're not going to have a true consensus where most everyone can leave the issue happy. Anyone who understands how Wikipedia works is well aware that we must accept rough consensus in such cases. So, I hope to find where the least objectionable points lay, so the project can gain as much acceptance as possible. BigK HeX (talk) 02:55, 8 September 2010 (UTC)

Consensus is arguments, chummer, not headcount. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 03:32, 8 September 2010 (UTC)

Again ... I have no clue what point you intend to make with that statement. Are you trying to suggest that the arguments in favor of keeping PC are somehow underwhelming??? Dunno.... BigK HeX (talk) 03:50, 8 September 2010 (UTC)

No, the arguments in favor of PC aren't enough to counter the legitimate arguments against, meaning there is no clear-cut consensus to retain the tool. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 03:54, 8 September 2010 (UTC)

We disagree on that point, I'm afraid. BigK HeX (talk) 04:02, 8 September 2010 (UTC)

Because you see numbers, not arguments. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:12, 8 September 2010 (UTC)

You presume too much about what I see, and, in fact, you presume inaccurately. BigK HeX (talk) 04:13, 8 September 2010 (UTC)

You, on the other hand, turn everything into an argument. – Smyth\^talk 08:53, 8 September 2010 (UTC)

How can I presume something which you yourself have brought up as an argument for retaining the tool, chummer? You've repeatedly said that 65% is good enough to keep the tool, and damn discussion. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:17, 8 September 2010 (UTC)

You are misstating my position. If nothing else, my repeated attempts at discussion should make it extremely evident that "damn[ing] discussion" is the pretty far from the position that I hold. BigK HeX (talk) 04:21, 8 September 2010 (UTC)

• Support, as it may be unclear as to how many people who voted support may want features changed, removed, or added before truly supporting the feature. ialsoagree (talk) 01:59, 8 September 2010 (UTC)
• Support, I supported PC with expansion, but you have to crawl before you can run.--intelati^(Call) 02:05, 8 September 2010 (UTC)
• Soft oppose. The remaining articles would likely be under 1000. They'll be useful to have continuing data on. 65% support seems like grounds for some compromise on the actual number of articles as well. Agree that the technicality of keeping the extension on shouldn't be an issue. Ocaasi 02:13, 8 September 2010 (UTC)
• Conditional Support: As someone who did not vote (having a mixed opinion--I found option 2 to be palatable if not optimal, but option 4 to be so poisonous that voting keep with the possibility that that would be the result that I couldn't in good conscience be lumped into the same general category), I support this notion adding that it is okay for PC turned on exactly those pages it should be most useful on--pages that have had a relatively even mix of approved and rejected pending changes. As has been said over and over, if most to all of the IP edits to a page are vandalism, semi-(or full) protection is better. If only one or small number of edits are vandalism, open editing is better (requiring Vandalism patrol and article watchers to revert as normal). So keep the function on, turn it off for all articles, but allow admins to turn it on. Admins would have to make their best judgment about when PC is better than semi-protection. This isn't really substantially different from what an Admin (should) do now, given that they shouldn't semi- an article if the IP edits are actually helpful. If in doubt, the admin can err on the side of PC, but then requests to remove the PC should also be taken seriously. Qwyrxian (talk) 02:22, 8 September 2010 (UTC)
• Oppose. The developers haven't said that it would be infeasible, only that it would be some work for them. More to the point, if we allow Pending Changes any further, it will never go away; there will always be a cry to keep it around "just a bit longer". There isn't consensus to do that, and in any case it would be harmful to do that, so the proper response is to shut PC down as was supposed to happen three weeks ago. — Gavia immer (talk) 02:30, 8 September 2010 (UTC)
• Support remove PC from all articles until everything is fixed then have another trial/ poll.Sumsum2010 · Talk · Contributions 02:32, 8 September 2010 (UTC)
• Oppose 30 day extension, support finishing the trial as close to on the original schedule as possible. --Yair rand (talk) 03:20, 8 September 2010 (UTC)
• Clarification The place to put caveats or ask for it to be left on for a class of articles is on another proposal. This is meant to end the trial "softly" while we figure out what to do - essentially buying us time so that the developers won't have to turn the feature on and off, another separate proposal should be made, concurrent to, or following this one, if you wish to see a further trial. Triona (talk) 03:24, 8 September 2010 (UTC)
• Comment. It seems unlikely that the very strong opposition will compromise. Which leaves either working around them, waiting for an admin to push this in a direction, or overlooking the 65% support. If the soft-close option went forward, I would prefer to have a plan for a future trial in place, to avoid chances of a similar delay scenario. What about a new 2 month trial to start within 6 months, details to be determined? Ocaasi 03:37, 8 September 2010 (UTC)

• Everybody already agreed to stop this thing after it had run two months. Why should we have people trying to set conditions on honoring that agreement? What makes it reasonable to say "We won't stop the trial like we agreed to when we started unless you agree to let us start another one within 6 months"?—Kww(talk) 03:44, 8 September 2010 (UTC)

Because, at this point, I suspect many people are willing to accept the support shown for PC as strong enough for PERMANENT usage. Even contemplating rejection of PC, with its significant majority support, at this point is compromise from that group. What compromise is being offered by the opposers?? BigK HeX (talk) 03:49, 8 September 2010 (UTC)

We don't even let individualt become admins at that level of support.—Kww(talk) 03:56, 8 September 2010 (UTC)

Of course, but my main point is that the supporters of PC seem quite willing to compromise, whereas it's very unclear where opposers would give any ground. BigK HeX (talk) 04:00, 8 September 2010 (UTC)

I opposed PC, but I am willing to compromise (see my support for this proposal below). BigK Hex, you seem to be reacting to specific users who happen to have been vocal early in this discussion. Please do not jump to conclusions about all those who opposed PC. Revcasy (talk) 14:18, 8 September 2010 (UTC)

It's because they lied. It's really as simple as that. If the trial had stopped at the agreed upon time, I would be happily discussing how to make it better for the next trial. We can't talk about how to make the next trial better until people honor their word and stop the first one.—Kww(talk) 04:06, 8 September 2010 (UTC)

They "lied" ... or did "they" give up considerable ground in not taking the results as enough support to make the system permanent? In any case, for whatever reason, if there's no discussion about where to give ground, the only alternative is for the groups to ossify into "significant majority will do" vs "we won't discuss anything until we get our way". Seems pretty sub-optimal. BigK HeX (talk) 04:12, 8 September 2010 (UTC)

Flat out, BigK, they either lied or they neglected the end date of the trial. They promised us two months. We're going on three. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:15, 8 September 2010 (UTC)

• The more I think about this the more I have to wonder if ArbCom can't force the end of the (now overextended) trial. It doesn't help that admins and editors have been requesting or honoring requests for PC as they pop up on WP:RPP. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 03:54, 8 September 2010 (UTC)

Involving Arb-com runs the risk that they will see 65% as sufficient support to keep the feature going while it is improved. It's not necessarily in the interest of opposition to bring in outsiders. They might not share your assumptions. Ocaasi 03:59, 8 September 2010 (UTC)

Not so, Ocaasi. ArbCom cannot abrogate or rewrite policy, especially a policy as central as WP:Consensus. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:02, 8 September 2010 (UTC)

Template:Editnotices/Page/Wikipedia:Requests for page protection might help with the issue of people continuing to grant requests.—Kww(talk) 04:06, 8 September 2010 (UTC)

Since, there seems to be debate about how consensus would apply here, I don't fully agree with those implications. If what you were suggesting was that obvious, Admins would have likely weighed in on the poll already. That they haven't means that consensus is undetermined or close enough to warrant clarification/further discussion. Since the compromise ideas only addresses a continued trial as opposed to permanent usage or expansion of the feature, I'm not sure you would have the reception from Arbcom you want. Maybe better to come up with a reasonable alternative, at least among those willing to do so. Also, ArbCom could take weeks or longer (and they're crazy busy), which could be worse than a compromise. Ocaasi 04:19, 8 September 2010 (UTC)

I do not think there is an active administrator that has not already put in his two cents. (Admins are not permitted to judge consensus in discussions or strawpolls they've taken part in.) I know, however, that ArbCom has not touched this trial, and that arguing for an extended trial when the first one has already overstayed its welcome is akin to holding someone hostage. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:22, 8 September 2010 (UTC)

I'm not holding anyone hostage. I'm suggesting that support is significant enough to warrant a compromise. I think proponents of PC are being reasonable by not overly advocating the support which exists both here and on the Closure page, including the considerable work towards issues that need improvement. Since you oppose the feature entirely, I doubt you are as focused on the fairly accessible changes which are likely to happen as the feature develops. But I suspect others see these as part of the discussion, and as a meaningful part of why an ongoing trial would be acceptable. Ocaasi 04:40, 8 September 2010 (UTC)

I really don't understand your stance at all. What did "two month" mean in the phrase "two month trial"?—Kww(talk) 04:44, 8 September 2010 (UTC)

It meant two months, then determine consensus. We're trying to do that. Turning off 1500 articles before consensus is established is a bit inconvenient considering that no harm is being done while the poll wraps up. Two months meant short. I don't know what specifics were agreed to in terms of disabling extensions or the timeframe for switching off PC on articles. It's not unreasonable to want to wipe the slate clean while the discussion goes on, but then again, if the discussion has significant support to continue the trial, which it does, turning off the feature on all articles would be premature or counter to the discussion. It's an inconvenient gray area due partly to the technical realities of software and developer priorities. Attempts at compromise are focusing on practical alternatives to keep things moving without wasting energy or overstating opposition. Ocaasi 05:18, 8 September 2010 (UTC)

Very strange ... usually "two month trial" means "a trial of two months duration." I hope you can at least understand why many of us believe we were deceived.—Kww(talk) 05:28, 8 September 2010 (UTC)

• Support – One important advantage of this course is that it would leave the PC logs from the trial available, whereas disabling the extension would cause the logs to vanish, and any future future discussions would be based on even less hard data than this one has been. – Smyth\^talk 08:51, 8 September 2010 (UTC)
• Support. I've no objection to new trials, hell, I've no objection to just keeping PC, but this way seems the best method to achieve consensus. Before any new trial, however, I'd like to see us decide far better on what to do post-trial - otherwise this will just repeat, and the failings of the community as a whole will be blamed on the majority - again. TFOWR 09:13, 8 September 2010 (UTC)
• Support. Although I would say that the soft shutdown could be performed in less than 30 days. I think we should then think carefully about a new trial - what software and policy changes we want and how the new trial will work. See #A way forward?. I think there are some good ideas for changes at Wikipedia:Pending changes/Closure#Working Summary (unofficial). It also lists some things we may want to try to quantify with a well-designed trial. I also agree with TFOWR on the need to think in advance of the new trial what the post-trial process will be. Yaris678 (talk) 11:57, 8 September 2010 (UTC)
• Support I opposed the continuation of PC, and contrary to what supporters seem to be alleging about those who opposed, I am willing to work toward a compromise in the interest of obtaining consensus. Revcasy (talk) 14:19, 8 September 2010 (UTC)
• Neutral as long as it's not being used, I don't have a problem with it being a checkbox on a menu somewhere. On the other hand, since many people (including me) have complaints about the current implementation as much (if not more so) than the feature itself, I see little harm in removing the feature at least until it's been improved. Perhaps one of the improvements the dev team should consider is making it easier to enable or disable the feature. :) Xtifr _tälk 17:48, 8 September 2010 (UTC)
• Consensus does support leaving the implementation in dormant state but PC should be removed from articles without delay - see my analysis. Cenarium (talk) 23:08, 8 September 2010 (UTC)
• (Conditional) Support Since we agreed on a two-month trial i agree to soft-disabling the extension - partly to preserve PC logs, partly to reflect that the majority (Based on headcount) would support leaving the feature active, and partly because it was agreed that was to be a trial rather then a full-fledged implementation (Discussions tend to sour if the respective sides won't trust each others word anymore). My condition, however, is that both sides have to work together in order establish a clear way forward from here. This means that the opposer's issues should be addressed where reasonable and possible, and that supporters involve themselves constructively in this process. Once we have an agreeable list of issues that should be addressed these should be improved on before we have a final test \ debate as to the future of this feature. I would explicitly point out that "Constructive" equals walking to a middle ground, rather then demanding "An immediate implementation based on headcount and arguments" or an "immediate halt to usage and development because the proposal would never be accepted anyway \ because opposing arguments are stronger then support arguments". If anything, i hope that we can keep red tape and rushing\stalling tactics out of this. Excirial ^{(Contact me,Contribs)} 20:21, 9 September 2010 (UTC)

Compromising

High Traffic BLP's

Just the BLP's with more than two vandalisms per day.?--intelati^(Call) 02:23, 8 September 2010 (UTC)

There's not really support for that. It could easily be in the thousands. Ocaasi 02:25, 8 September 2010 (UTC)::Just making a suggestion to get people thinking...

I like the suggestion, I just think it's further from consensus than the discussions happening above. (Thanks for the ip checkk. I'm not logged in... I removed the note). Ocaasi 02:37, 8 September 2010 (UTC)

Recent Changes and watchlisting should cover those; those articles on which they do not should be semiprotected anyway, since PC would presumably backlog even further. Septentrionalis PMAnderson 02:31, 8 September 2010 (UTC)

It would probably alert users that don't usually edit those articles to the problem?.--intelati^(Call) 02:32, 8 September 2010 (UTC)

So will Recent Changes; the latest change will be an anon. Given the response time problem, it is likely to be a toss-up which will backlog further - but in either case semi-protection is the answer. Septentrionalis PMAnderson 02:35, 8 September 2010 (UTC)

Compromise -- Med, Socks, 200, and 1

I could get easily get onboard with a complete shutdown of the current trial and rejection of the current request for PC (despite its majority support), with a new trial across a very limited number of articles. I'd ask that PC be allowed in:

1. GA's and FA's within WikiProject:Medicine that have been vandalized;
2. the very limited number of articles where sockpuppetry is already known to be an extensive problem;
3. no more than 200 additional articles for statistics-gathering purposes. If the 200-article limit is reached, an article must be removed from PC-protection before a new one can be added. Also the articles will meet the following criteria:

1. have already been PC-protected at least once under the original trial
2. are low-traffic articles with less than 40k hits per month

4. and, finally, a single high-traffic BLP [decided before trial] that will run with PC-protection on and off, alternating each week.

I propose that we decide a few general acceptance criteria beforehand, so we know what we're looking for in the future discussion. After these are decided, the trial then lasts 120 days, after which time PC protection is removed from ALL articles. One week will then be given for metrics to be compiled, and only then do we follow with a poll which indicates general acceptance/rejection of PC.

1. Support, as nominator. BigK HeX (talk) 02:50, 8 September 2010 (UTC)
2. Support A reasonable compromise. More data /experience will help all of us decide if we wish to keep PC or not. Doc James (talk · contribs · email) 02:56, 8 September 2010 (UTC)
3. Oppose. By its very nature Pending Changes is defeated by most sockpuppets (whom are of the autoconfirmation-buster variety and thus would bypass it), and high-traffic BLPs were tried and had to be taken off the trial because the vandalism and BLP violations overloaded the ability of reviewers. This is akin to saying that a house is still standing after you've torched it. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 03:13, 8 September 2010 (UTC)
4. Opposed. It's over, the compromise you've created only serves to extend a trial which has already been breached by 4 weeks, and by rights should have finished well before now. BarkingFish 03:22, 8 September 2010 (UTC)
5. Oppose. Two months are not three months. It has had it's chance. Jarkeld (talk) 09:58, 8 September 2010 (UTC)
6. Oppose for several reasons: pending changes isn't useful against sockpuppetry, partly because sockpuppets can pass the autoconfirmed threshold and partly because the reviewer won't necessarily recognise the account as a sockpuppet; no-one has ever produced a consensus for trialling articles by quality instead of by vandalism level; we have plenty of statistics from the original trial and if we need more then leaving the feature on for a few hundred articles isn't going to help; and a single BLP is largely useless for gathering statistics. Hut 8.5 12:27, 8 September 2010 (UTC)

   Level 2 pending changes would be extraordinarily useful against sockpuppets on certain low-traffic articles. I know of quite a few articles that have been placed on full page protection because of "excessive sockpuppetry", which would opened up for far more free editing with PC. BigK HeX (talk) 20:06, 8 September 2010 (UTC)

   Yes, level 2 pending changes would eliminate my first objection, if they can be implemented. What I'm concerned about is articles like Witton Albion F.C. where pending changes not only didn't deter User:Tile join from attacking the article but where his edits were reverted because of sourcing issues and not because of sockpuppetry or the fact he's a banned user. I suspect pending changes are only going to be useful in countering sockpuppetry on articles where the socks only do blatant, obvious vandalism. Hut 8.5 22:18, 8 September 2010 (UTC)

7. Oppose. I am highly opposed to blanket-applying pending changes to GAs or FAs, even within a single wikiproject. GAs and FAs do not necessarily receive more - or even an equivalent amount - of vandalism as articles which have not been through the review processes. Pending changes is not a "quality" control mechanism; that is, reviewers do not necessarily have the specialized knowledge to know whether edits are accurate or not - their job has been to determine whether an edit is blatant vandalism or not. We'd therefore be adding an additional level of bureaucracy and effort - an edit would be checked by a reviewer, who'd look to see if it was vandalism, then be rechecked by someone with content knowledge, who would have to also check if it was vandalism (some edits that people without the specialized knowledge think are vandalism aren't), and then check whether it was accurate. Why add the second step? If pending changes has consensus to continue, then GAs and FAs should be treated as any other article - if PC would help reduce vandalism on that article, then apply it, if not, leave the article alone. Karanacs (talk) 13:23, 8 September 2010 (UTC)
8. Oppose I prefer the soft shutdown of PC, followed by improvements in the UI, followed by another discussion about a new trial at a future date. Discussing a new trial here and now is not ideal in any case. Revcasy (talk) 14:24, 8 September 2010 (UTC)
9. Weak and conditional support Oppose. However, I share Jeremy's reservations about sockpuppetry: actual sockpuppet problems tend to be named accounts, which can easily autoconfirm. If this is intended to deal with IP's which rotate from day to day, which is a problem with the service provider, range-blocks are preferable. Septentrionalis PMAnderson 19:17, 8 September 2010 (UTC)
   • I can only support this on the basis of the promise to shut it down. If this is not shut down as promised, I think there will be a case for proposing user bans. Septentrionalis PMAnderson 19:44, 8 September 2010 (UTC)
   • Having just had the joy of dealing with an article with two revisions, one acceptable and one not, I will support this, weakly, when the interface is improved, and not until then; until then, a weak oppose. Septentrionalis PMAnderson 15:40, 9 September 2010 (UTC)

   Yes, banning you if you keep spouting bad faith accusations over and over might be refreshing. ♫ Melodia Chaconne ♫ (talk) 22:09, 8 September 2010 (UTC)

10. Point 1. goes against the clause of the protection policy that articles should not be protected preemptively, it would require a massive consensus to overturn this policy. Cenarium (talk) 23:01, 8 September 2010 (UTC)

    I'm pretty sure it's not a problem to limit Point 1 to articles that have been vandalized. BigK HeX (talk) 03:09, 9 September 2010 (UTC)

    I actually have a question about point 1: Aside from it being Doc James' WikiProject of choice, what is so special about WP:WikiProject Medicine that gives it such preferential treatment? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 03:27, 9 September 2010 (UTC)

    I'm fairly sure Medicine articles already get somewhat preferential treatment (though not to the extent of BLP's or anything). I know they have their own version of (at least) one of the Big WP guidelines (can't think of which one off the top of my head). BigK HeX (talk) 03:59, 9 September 2010 (UTC)

    It might be this one: Wikipedia:RS#Medical_claim. Not sure if it was this one or another that I'd seen. BigK HeX (talk) 04:02, 9 September 2010 (UTC)

    That's not special treatment; that's an application of the rest of reliable sources. Wikipedia:No disclaimers in articles implies that medical articles should not receive any treatment which would suggest that they are reliable; they aren't - and PC won't make them so. Septentrionalis PMAnderson 15:38, 9 September 2010 (UTC)

    Not to mention that we explicitly disclaim the notion that we are a substitute for the opinion of a licensed doctor. So I ask again: Why does WikiProject Medicine receive such preferential treatment? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 18:31, 9 September 2010 (UTC)