Misplaced Pages

Revision as of 02:16, 25 May 2009

This page is in its infancy. For helpful suggestions, ideas on layout, improvements, criticisms, etc., please make changes here or discuss at the talk page.

This page has been created to review, discuss, and propose changes to the current closure process of Misplaced Pages Featured picture candidates. The need for this discussion has arisen following complaints and suggestions raised at the FPC talk page in May 2009.

History

NB: This is not intended to be complete; however, please add relevant points if you wish to.

Prior to late 2007, a guideline of 2/3 supermajority was in place for the closing of Featured picture candidates. However, it is important to note, that discretion was always encouraged and a strict adherence to 2/3 majority rule was never part of the closing procedure.One of the earliest expressions of the closing procedure is found here. This was removed from the guidelines, presumably without input from FPC contributors. (If evidence is found that discussion took place, please add it here.) It was reinstated in May 2009. For a considerable length of time and also currently, a 2/3 supermajority is utilised consistently and effectively at Commons, with a high standard of results. Whether or not this is also appropriate for FPC has yet to be determined.

Despite several bot requests, an effective, shared solution for the semi-manual closing of FPCs has not been procured. Closing nominations can be a time-consuming and tedious process, resulting in a lack of contributors with the time or inclination to sign off the nominations. This has, in part, led to there being only a few dedicated users with the time and means to perform the closures on a regular basis.

Nominations are currently foundering without adequate review, possibly due to a simple case of lack of traffic, or possibly due to more sinister conditions. What was a flourishing project with a high level of activity has tailed off to only a handful of regular reviewers. Compare, for example, nominations from May of 2006 to a more recent page of nominations.

Concerns have been raised in the past over the current method of closing nominations, whether this be a personal issue or a process-related issue. A selection of links to these past discussions for those so inclined:

• Foxe's Book of Martyrs, April 2009
• Buckets of slop, April 2009
• Proposal for a new bot task, April 2009
• Older nominations requiring additional input from users, April 2009
• Asilidae Stichopogon sp., April 2009
• Suspended nominations, March 2009
• Holding nominations for more feedback, March 2009
• Please be more rigourous in reviewing images, March 2009
• Charles Darwin caricature, February 2009
• Final call for under-reviewed nominations, February 2009
• Closing Nominations, January 2009
• The criteria and the general problem of the FPC project, December 2008
• Signing off, December 2008
• A survey on FPC results, December 2008
• Hidden closings, December 2008
• Holiday warning, December 2008
• How long?, November 2008
• Half-assed closings, August 2008
• Napoleon's exile to Elba, August 2008
• American Tree Sparrow, August 2008
• "anonymous votes are generally disregarded, as are opinions of sockpuppets.", July 2008
• Polarlicht, July 2008
• clashes between promotion guidelines 5 and 6, June 2008

(Ok, that's become exceedingly tiresome!)

This over-long list, which only covers now to June 2008, hopefully gives a sense of the pressing issue that this has become. The project clearly needs your input and decision.

Summation of points

NB:I intend to add "references" to these points when I have the time, but welcome help

NB:Add whatever I may have missed

• The closing process is poorly explained and generally lacking in detail, leading to mistakes, lack of voluntary participation, and a lack of transparency
• Closing nominations is too tedious and time-consuming
• Regular closers have become the unimpeachable "authority"
• There is no stated supermajority, leading to an ill-defined consideration of "consensus"
• It is unclear if the closer !votes in the process of closure, or if the closer should refrain from personal opinion being used to decide a nomination
• Nominations hang around for a long time, sometimes gathering more opposes than they would otherwise
• Some candidates are not promoted due to a simple lack of participation
• If flaws in the image were unnoticed during review or are in the minority opinion, it is unclear if the closer should fail a candidate for said flaws, promote the image regardless, or if a comment should be made and the nomination left for additional input.
• If candidates require additional input for whatever reason, they tend to be left in a bottom-of-page wasteland. It may be more serviceable to relist them.
• It is unclear if "determining consensus" requires superior technical knowledge or if it can be determined by any careful and experienced editor
• It is not clear how it is decided which !votes are valid, and which are not

Proposals

As a first stage, we need to determine if changes are needed. Please sign below in the section that best describes you. If you're neutral, you can feel free to ignore this.

Leave all as is

Support change

• Isn't that obvious? -- Alvesgaspar (talk) 19:59, 21 May 2009 (UTC)
• Best to give a choice, :-) Maedin\ 20:10, 21 May 2009 (UTC)
• I'm sure this can be resolved. Papa Lima Whiskey (talk) 00:26, 22 May 2009 (UTC)
• Yes. The level of poor reviews means the old model doesn't work any more. MER-C 04:57, 22 May 2009 (UTC)
  • Maybe we'll just leave the discussion for later. This stage seems to be more about establishing that people are willing to work towards a new goal. Papa Lima Whiskey (talk) 08:42, 22 May 2009 (UTC)
• I agree that it needs some work. I've already put forward my view/proposal, which doesn't require an overly drastic change IMO. It'll re-state it below. Diliff | (Talk) (Contribs) 09:40, 22 May 2009 (UTC)
• Raw vote-counting is problematic, but less problematic than closers imposing their personal tastes as if they were superior to clear consensus. Durova 18:45, 22 May 2009 (UTC)
• The biggest problem isn't the closing procedure as far as what gets promoted and what doesn't. The trouble is that people feel cheated if an apparently unambitious nomination gets closed. When this happens experienced reviewers may leave and the quality and quantity of votes takes a nose drive, compounding the problem. Noodle snacks (talk) 12:37, 23 May 2009 (UTC)
  • Precisely! The arbitrarities affect both reviewers and nominators/authors and have preverse side effects on the credibility and effectiveness of the forum. -- Alvesgaspar (talk) 13:17, 23 May 2009 (UTC)
• I think there is room for improvement. Spencer 01:10, 25 May 2009 (UTC)

Which model?

Consensus vs opinion poll

• Now that Maedin had the courage to start this discussion and some symptoms have already been identified, we really should handle the opportunity, make a correct diagnosis and try to find a cure for the disease. As emphatically shown by the previous discussions above, the situation is not new. FPC is becoming less and less popular among WP users and the risk of obsolescence is real. Some of us have repeatedly alerted for the danger and even suggested the system to be made more transparent and friendly for newcomers. But all proposals were turned down (even those as innocuous as keeping the closed nominations in the FPC page for a couple of days) and most of the regulars didn’t even recognize there was a problem. Now, I hope most will agree it is time to seat down and discuss the issues calmly and with an open mind. The first subject I would like to propose for discussion is consensus. Together with democracy, the concept is often invoked, and sometimes thrown at others as a dogma, but seldom understood. Consensus means a general concord or agreement between parts, and it is usually achieved through a structured discussion and negotiation. That is not the case of FPC, where we normally have a collection of unrelated comments headed by a support or oppose summary, that is, an opinion poll. If we were to forbid those headings (as well as the signatures at the end), then an objective interpretation of the consensus by the closer would be much more difficult, if feasible at all. Maybe it would be possible to reach decisions by consensus in FPC if the most relevant issues concerning each picture were identified, addressed and discussed by all reviewers, and a decision met by agreement, with the help of a moderator. But the solution is obviously impractical. If we accept this limitation and recognize that the present model can be much improved with only some minor adjustments and clarifications, then maybe the process of obsolescence can be reverted. But first, people have to come down from their ivory towers and accept that all users of good will have the right of opinion and participation. Alvesgaspar (talk) 08:46, 22 May 2009 (UTC)
• This is what I think primarily needs change: The process should require that the closer looks for faults in the reasoning of voters or issues that exist in the image but were not raised, and if this is going to result in a controversial close, summarises them and asks for further input from voters over a period of a couple of days, to give those who were involved previously to do so again if necessary. I really don't think that the closer should take the matter into their own hands though without further discussion from the community - otherwise what is the point of everyone else's opinion (right or wrong)? Even if a few poor images scrape through, or good images fail, so be it. That's what the de-list and renomination process is for. Better to have too much discussion and even disagreement than one person ignoring consensus and deciding for all of us. Other than that, I don't see a problem with using the supermajority as a rule of thumb. It's the only objective method of measuring this vague concept of consensus, but as stated above, should be used along with subjective analysis to avoid poor and unjustified votes tainting a result. Diliff | (Talk) (Contribs) 09:49, 22 May 2009 (UTC)
  • Also, I suppose it should be mentioned that clearly, the closer dismissing any discussion as misguided that disagrees with his personal opinion is not acceptable. Ultimately, as long as the debate is reasonable and has a semblance of logic, the opinion of every individual should be respected. Not that I'm insinuating that it would or does happen. It's just covering all bases. Diliff | (Talk) (Contribs) 10:29, 22 May 2009 (UTC)
    • I largely agree with you there Diliff but the problem with a delist as it reverses the onus - so an image which scrapes through on 2/3 will easily survive delisting (assuming the same votes) because delist would require 2/3 to support the delist (ie have opposed the original nom). The better option would be for the closer to justify the closing of potentially contentious nominations by WIAFP? and the standard of images --Fir0002 10:20, 23 May 2009 (UTC)
• The one I'm thinking of is how the other featured content processes function. We still have support/oppose, but any significant concerns still relevant at closing time means the nomination fails, regardless of how much support there is. Perhaps we need to move towards a more open-ended discussion period. The closer has to acertain whether the picture meets FP? from the discussion and, if necessary, his own review. That's what I've been trying to work towards and that's why I failed those nominations. Making the criteria more explicit helps people to determine exactly why a nomination fails. MER-C 10:21, 22 May 2009 (UTC)
  • I see your point, but I just don't know how productive it would be to fail every nomination that had significant concerns... There will always be different interpretations of the guidelines. Some people see HDR images as inherently fake and will oppose for that reason, regardless of how tastefully done or how a greater dynamic range will aid EV. Some people are sticklers for extreme accuracy and won't tolerate any misrepresentation of the scene (eg removing elements by cloning) or post production work to improve aesthetics of the scene. We don't want guidelines that expand to address every issue that we could encounter during the reviews, so there will be those who can justify their opinions without going against the criteria. There should be room for this dissent, and as long as this dissent is in the minority, I don't see the problem with passing the image. Again, I'm not suggesting that it be a hard and fast rule of a supermajority, but that it should be a rule of thumb. Diliff | (Talk) (Contribs) 10:42, 22 May 2009 (UTC)

By significant concerns I mean the usual bunch of technical flaws, lack of enc and copyright problems. (The expanded criteria I drafted spell these problems out explicitly). Someone who opposes all HDR images wouldn't be given much weight, but someone who opposes due to poorly done HDR (if the image is not edited to fix this) will cause the nomination to fail. The supermajority is necessary but not sufficient for promotion - I must also be satisfied the image meets FP?. MER-C 14:54, 22 May 2009 (UTC)

MER-C, your closures have divided the FPC community and committed contributors are on the verge of departing. Please stop defending your actions; they're obviously problematic and this has been brought to your attention before. Durova 19:18, 22 May 2009 (UTC)

I disagree with you Durova - MER-C has every right to defend his actions. Even a murder is given that privilege. And I totally agree with the approach MER-C takes to closing images. --Fir0002 10:20, 23 May 2009 (UTC)

In featured article processes, editing csn deal with concerns. How am I supposed to deal with people saying that they dislike an image? Shoemaker's Holiday (talk) 12:20, 23 May 2009 (UTC)

• IMO we should stick with 2/3 majority consensus and the closer should respect the opinions of the voters. The closer should however, be given the right to disregard/give less weight to opinions of newbies. --Muhammad(talk) 04:56, 23 May 2009 (UTC)
• The majority of the problems can be solved if all of us participate and air our views in the nominations. Take for instance the lighthouse picture. It had a clear majority and only after reopening the nomination did it get opposition votes by users. If these votes had been made earlier on during the process of the nomination, then we wouldn't be facing any problems. --Muhammad(talk) 04:56, 23 May 2009 (UTC)
• IMO under no circumstances should we descend to a vote count. If someone makes a vote on flawed reasoning in light of WIAFP? and current standards there vote must be disregarded. If the consensus has failed to pick up a significant flaw, the closure should either not promote (in obvious cases) or alert the voting community of this flaw and seek further input --Fir0002 10:20, 23 May 2009 (UTC)
  • Has there ever been a case where this was decided in favour of promoting the picture? There were some absolutely bizarre opposes to a couple of my recent nominations which sunk couple noms. Were they ignored? No. In practice, this only serves to give the closer an arbitrary excuse not to promote, even when the reason for this is weak (vis: Rob Roy, where a scene from a book was said to be unencyclopedic for illustrating the book) Shoemaker's Holiday (talk) 12:41, 23 May 2009 (UTC)
    • My issue with the concept of "flawed reasoning" is that it usually only applies to opposes. If you oppose, you have to explain, defend, and discuss, but if you "think it's pretty", you say "per nom" and your vote is counted because there is no "flaw" in your "reasoning". If we give weight and are actually going to consider each !vote on its own merits, then we need to level the bar for opposes and supports. As it is, this is skewed in favour of promoting images, an issue which I could see getting more apparent once FPC has wider participation. Maedin\ 15:25, 23 May 2009 (UTC)
      • That's true in a sense, but is also part of the reason that Opposes effectively carry higher weight, i.e., needs 2/3 Supports for promotion in the basic interpretation, which means an Oppose effectively counterbalances two Supports. A valid consequence of what you say is that there is unrecognised 'flawed reasoning' in a number of noms, yet Support per nom is considered valid regardless (see here for example and consider the validity of the supports). --jjron (talk) 17:29, 23 May 2009 (UTC)
        • Very good example of poor supports, and no real response to the issue of EV which I think had some merit. Without getting bogged down in the specifics of that, I have often thought that drawings/paintings don't do a great job of illustrating a story (unless it's the story's original art and is indelibly connected, perhaps) because it is an interpretation of the subject and doesn't illustrate it precisely, thus EV is limited to illustrating the author - IMO. There are so many nominations which scraped through and are only now being given the attention they originally deserved and that is a shame. Anyway, back to the issue at hand - It would be ideal if we could force all voters to state a detailed opinion referencing the criteria, but on the other hand, we don't want to scare newbies away by making it too difficult to participate. There are a number of negative aspects to raising the expectation level of FPs. In the past, it really was more of a beauty contest with only a vague nod to EV and while it is a good thing that we've become more focused on the criteria and increased our technical standards, it no doubt has made voting far more difficult for newbies. And that is the crux of the problem right there: how do we make FPC inclusive while also maintaining high standards? The more difficult that contributing becomes, the more necessary an experienced closer is to make sense of the nomination. Diliff | (Talk) (Contribs) 22:08, 23 May 2009 (UTC)
• For all intents and purposes we should stick to the process as it currently operates. As for what should be done with noms that the closer determines issues with, I basically agree with what Diliff says above, or as I described in detail here (which suggested a solution over a month ago, but only three people bothered to look at). --jjron (talk) 16:12, 23 May 2009 (UTC)

Which majority ?

• I'm starting this section but not making a specific proposal for now. Please notice that the explicit definition of a majority in the guidelines does not automatically mean that raw vote counting is to be used. That will come later... Alvesgaspar (talk) 18:54, 22 May 2009 (UTC)

• In other processes such as AFD, FAC, etc. closers have worked off a nuanced understanding of supermajority. Rationales that clearly run counter to the established criteria get discounted and exceptionally well-reasoned rationales get extra weight, especially if the salient points are not rebutted. So (at FAC for instance) and Oppose, not notable enough to get featured doesn't count because notability is not a criterion for featuring an article. Raw supermajority rules at commons due to language barriers, and as a distant second choice I would support its implementation at en:wiki FPC. But only as a last resort against bias closures. For some time we've had closers who--to put it bluntly--abuse the process to force their own will upon the outcome. That would lead to formal corrective action at other processes. We've been patient and tolerant here, with the end result that the FPC process itself is in turmoil--all over a handful of middling-good images, which ought not to divide us. If there's a rational way to implement the standard en:wiki norm I'd strongly prefer that. Namely, in instances where a 2/3 supermajority is not implemented the closer notes impartial reasons for weighting and/or discounting reviews. Closers who disagree personally with an outcome are expected to respect that process even if its result is (in their eyes) wrong. Remedy is to be sought through the delisting process, not through closures. Durova 18:56, 22 May 2009 (UTC)

• Here is my proposal for the text regarding the majority rule:
  • An image listed here for xx days with a qualified majority of favorable opinions in its favor and five or more reviewers in support (including the nominator’s) will be added to the WP:Featured pictures list. Qualified majority is here regarded to be a two-third proportion of founded and independent support opinions. Note that anonymous opinions are disregarded, as are those of sockpuppets. If necessary, decisions about close candidacies will be made on a case-by-case basis.
  • What’s different here from the present version? First, the concept of consensus was dropped as it is not applicable to this kind of polling. It was replaced by the figure of qualified majority, with the usual meaning of 2/3 majority. Second, the minimum number of favorable opinions was raised to five, to avoid the horrible 4/2 case; Third, the about referring to the polling period was dropped (but this is another discussion, below). Fourth, the concept of founded and independent opinion was introduced, meaning that an opinion to be valid must be justified and not the facsimile of some other opinion. For example, opinions like "Support - Beautiful, I like it", "Support", "Support – Per nom" or "Oppose – Agree with someuser" should not be considered. Finally, I think that the opinions of anonymous users should never be considered. As for the power of discretion of the closer, I think that should be the subject of a separate discussion and we might not yet be ready for it. Alvesgaspar (talk) 22:29, 22 May 2009 (UTC)
    • Where User X has made a good assessment of an image it is entirely appropriate to vote "per User X". Qualified majority = 2/3 majority, it should equal a majority of votes which provide an accurate assessment according to WIAFP?. Also disagree with raising the min to five - there are too few votes these days as it is --Fir0002 10:23, 23 May 2009 (UTC)
      • I think we'd best stick to four for now: The problems have driven a *lot* of voters away. It's make the process be a bit of a timing lottery, with nominations at certain periods failing simply because people are on holiday. Shoemaker's Holiday (talk) 12:45, 23 May 2009 (UTC)
        • I'm not sure the exact wording will make a huge difference, e.g., leaving 'consensus' in or out, but not that fussed. Agree with Fir that valid 'per User' votes are OK when done properly, but I do find things like 'per Nom' aggravating when the nom hasn't even stated a reason (which I saw recently), or 'per User' when 'User' hasn't given a valid reason themself, so that does have some merit. Also concerned that you're again opening up some of the very problems you're worried about, i.e., that the closer gets to decide what is and isn't a founded and independent opinion. FWIW 'votes' of anon users are never considered, but very occasionally they raise valid points that are worthy of discussion. Tend to agree with raising the bar to 5 supports (have always thought it should be nominator + 4 supports), but with the understanding that this is not another criteria that someone then uses to decide to wade through all current FPs to retrospectively put up for delist all the ones that got promoted with 4. --jjron (talk) 15:35, 23 May 2009 (UTC)
        • A random example (just the first one I stumbled across) where 'Support per nom' makes no sense, as the nominator never gave a reason, simply stating what the picture was. This was promoted based on two 'supports per nom' and one 'support' with no reason. Why aren't these getting put up as being closed improperly? Is the system biased to promoting? Is it just that those not getting their noms promoted are raising the stink? --jjron (talk) 17:41, 23 May 2009 (UTC)
• I like the idea of 5 support votes. wadester16 | Talk→ 20:46, 24 May 2009 (UTC)

How long ?

• I'm starting this section but not making a specific proposal for now. Should we define a precise polling period (to be always followed except in well-defined cases) or not? -- Alvesgaspar (talk) 18:54, 22 May 2009 (UTC)
• Has this been dealt with adequately? The 'holdover' option seemed to have gotten misused for a while, but as far as I've noticed it seems to have settled down. Bear in mind that I never put images into it and more or less boycotted it after the problem arose. Seven days as standard term has usually worked fine. And if there aren't enough reviews to close then a holdover makes some sense; holdovers of images that already had sufficient participation has been the problem. Durova 19:36, 22 May 2009 (UTC)
• I think the "vote" period should be adhered to to the minute. Otherwise, it leaves open the possibility of closers waiting for a time that suits their personal bias on the nomination, i.e. taking advantage of stochastic fluctuations in the same way a stock broker might. Please note that this doesn't mean that a closer would have to be up at some ungodly hour, just that "votes" and opinions given after the official period be discounted when closing. If someone then thinks that some information should have been taken into account, they can renominate for either promotion or delisting, as applicable. As for what the duration should be, 7 days is fine with me. Papa Lima Whiskey (talk) 20:51, 22 May 2009 (UTC)
  • PS: I believe mediawiki templates can be used to indicate how much time is left on a nomination, or whether it has expired. Papa Lima Whiskey (talk) 20:53, 22 May 2009 (UTC)
• I agree with Papa Lima Whiskey that the evaluation period should be fixed, as it is the only way to avoid arbitrary closings when the outcome is not clear. Maybe special cases should be considered as for example nominations not gathering, at the end of the period, a sufficient number of opinions to reach a decision. Alvesgaspar (talk) 22:38, 22 May 2009 (UTC)
  • And in those cases, you could again extend by a fixed period. :) Papa Lima Whiskey (talk) 09:54, 23 May 2009 (UTC)
• Totally oppose any strict time frame. A good image will be a good image for 21 days as it will for 7 days. If an extra second or an extra hour or an extra day or an extra week will change the outcome then that extra time is crucial. This project is about determining the best images on WP, and if extra time is necessary to determine consensus then that's the way it should be. This is not some kind of sport or competition in which the nominator should be demanding every technicality they can to get their image promoted. They should nominate it and allow the community, in its own time, to assess whether it is worthy of the FP star. --Fir0002 10:30, 23 May 2009 (UTC)

Except that, during the Kronheim nomination, it was left open for days past its time, seemingly for no better reason than MER-C dislikes sets. It may, however, be worthwhile to let noms without a quorum a period of extra time. Shoemaker's Holiday (talk) 12:47, 23 May 2009 (UTC)

This spring we've seen some pretty extreme cases where nominations that already had clear consensus to promote were kept open as long as a month, until one or two opposes nudged them into a gray area--at which moment they were promptly closed as not promoted, without additional time for more supports. Yet during the same period low-turnout FPCs that could justifiably have gone into the 'more time' holding pen were getting closed without promotion immediately. That's clear gaming of the system on the part of one closer, and watching it unfold is a blow to the morale of even the hardiest contributor. Durova 16:35, 23 May 2009 (UTC)

You mean 'this autumn'? ;-) --jjron (talk) 17:06, 23 May 2009 (UTC)

Durova, would you mind linking to a couple examples? A nom open for a month seems pretty extreme. wadester16 | Talk→ 20:45, 24 May 2009 (UTC)

• The current 7 days if fine, but trying to set it to the second is silly; we're talking approximately 7 days as has been typically working for years (with minor erroneous exceptions as noted above which have pretty much been remedied). No exceptions need be made for those with lack of votes per Lack of votes = Lack of interest = Not promoted. --jjron (talk) 15:15, 23 May 2009 (UTC)
  • If it's because of lack of interest, it should be easy for people to come out and say so. Papa Lima Whiskey (talk) 21:41, 23 May 2009 (UTC)
    • Not true, but I'm not going to argue it here as it's not really relevant to the closing issue. I will shortly start a new discussion on FPC Talk discussing this and other things in detail. --jjron (talk) 11:56, 24 May 2009 (UTC)

The closer's role

• There is a large spectrum between the present model of a all-mighty closer and a vote-counting bot. I believe the best solution is somewhere between these two extremes. Also, we should discuss the closer's profile. Should he/she be some sort of image specialist or, as Maedin suggests, just an experienced user in administrative tasks? -- Alvesgaspar (talk) 10:15, 23 May 2009 (UTC)
• It is crucial that the closer is relatively competent in assessing images according to WIAFP?. I'd suggest 3 months as a minimum experience/participation in FPC level for a closer --Fir0002 10:37, 23 May 2009 (UTC)
  • Not at all. The closer's role is not to assess the images, or to interpret WIAFP, but to close the poll in accordance with the agreed guidelines. Alvesgaspar (talk) 12:08, 23 May 2009 (UTC)
    • But you're contradicting yourself. You've just said "...the best solution is somewhere between these two extremes." But if the closer is unable to assess images at all, then how can they judge if the reasons given are valid? --jjron (talk) 14:57, 23 May 2009 (UTC)
      • I didn't precise which was my position in that long spectrum... IMO only in rare and crisp clear situations should the closer judge the substance of the comments (and eventually disregard them, in borderline cases). A good example is when the reason given for an oppose is insufficient resolution and that is obviously false. Alvesgaspar (talk) 17:34, 23 May 2009 (UTC)
      • "Relatively competent at assessing images" - at assessing what images? With digital photography that might be a realistic expectation, although there are doubts about its desirability. The temptation to act as a super-reviewer is the chief source of our problem. With historic subjects and media there are few things more frustrating than a closer who pretends at competence. Durova 16:44, 23 May 2009 (UTC)
        • And that's a whole other problem, not just for closers but for voters. We try to apply WIAFP? to a large variety of media, and it's not easy to do so in many cases because those criteria most readily fit digital photos. Like it not, in many cases people end up voting more on a gut feeling than according to the criteria. And nominators claim to want votes on their images, but don't like it when those votes go against them. --jjron (talk) 16:59, 23 May 2009 (UTC)
• Another issue. MER-C needs to release his tool for wider use. If he refuses to do so someone else should write a closing bot. Making the process straightforward will encourage many closers and reduce problems considerably. MER-C has previously been concerned about people making mistakes during closures. One could say the same about Misplaced Pages, but look at how successful the model actually is. Noodle snacks (talk) 12:36, 23 May 2009 (UTC)
  • Why should he refuse to release the tool? I believe that MER-C's goals are very much the same as ours. Furthermore he could well continue to be the closer, if the consensus dictates so. Alvesgaspar (talk) 13:22, 23 May 2009 (UTC)
    • IIRC he went on holiday and someone (can't remember who) made a bit of a mess of things in the mean time. The mess was relating to the numerous pages that have to be updated when an image is promoted, not the actual decision made at closure. Noodle snacks (talk) 13:39, 23 May 2009 (UTC)
      • FWIW MER-C hasn't refused to release his closing tool, he just restricts who can access it based on proven competence with manual closing of noms. The (IMO) valid reason is that if the tool messes up then the person using it better be damn competent with fixing up the mess that it could create. (And incidentally it wasn't "a bit of a mess" that was left by the random closer, it was an absolute train wreck that took us - and by 'us' I mainly mean me - hours to tidy up). --jjron (talk) 15:04, 23 May 2009 (UTC)
        • Thanks for that. Mistakes are less likely if the process is simple and mostly automated though. Noodle snacks (talk) 00:03, 24 May 2009 (UTC)
• IMO the closer's role is to interpret consensus. However, at times interpreting consensus may include being able to evaluate the image so that they can accurately evaluate opinions expressed, thus they should be proven to have at least serviceable skills in this area through prior involvement. In some very close cases it is almost inevitable that the closer's final decision essentially becomes a de facto vote, which isn't such a bad thing IMO. They also need to be able to decide when for example a single oppose may sink a nom - a case that always comes to mind is a DNA illustration a couple of year's back that had a long list of supports before I pointed out about half-a-dozen factual errors which destroyed its EV and meant it rightfully couldn't be promoted despite all the support. --jjron (talk) 16:24, 23 May 2009 (UTC)
  • Yes, I agree that when something is factually incorrect, support votes should be ignored and the nomination closed. But those are exceptions to the rule. I think it's a bit dangerous when you give the closer essentially far more power than a single vote, by giving them the ability to nullify any vote they wish to. I think it would be okay only if (as I said below under 'another possibility') they did it sparingly and only with an explanation that would assure transparency. Diliff | (Talk) (Contribs) 16:36, 23 May 2009 (UTC)
    • No argument from me with that. --jjron (talk) 16:52, 23 May 2009 (UTC)
      • (ec, disagreeing) What happens with historic nominations is a significant minority of nominations get sunk without good reason. Someone who doesn't know the medium or the history makes an off-target comment, and if the comment is posed in an authoritative manner then other people who also don't know the medium or history either get swayed, sometimes including closers. If one responds to say "yes, photochrom is an artificial color process" then threaded discussion results which drives away reviewers. If one doesn't respond because the objection is ridiculous then the failure to respond generates a risk that the closer (who didn't read the bluelink either) will refuse to promote due to 'unanswered substantive objections: color looks artificial'. The regularity with which this dynamic occurs is already a problem, so let's not carve it in stone as if it were a good thing. Durova 17:00, 23 May 2009 (UTC)
        • Pardon my ignorance - what is ec? Anyway, this seems more an issue of the voting process and the expectations of nominators rather than closing (if the closer works to interpret consensus as I say). I will comment on this elsewhere sometime soon (as this page is focussed on closing). --jjron (talk) 17:17, 23 May 2009 (UTC)
          • We edit conflicted (apologies for the delayed reply). Durova 01:31, 24 May 2009 (UTC)
        • Durova, would you mind linking to a couple examples of that in the past months? I can't honestly believe that a nom has been closed due to "artificial coloring" on something that is itself an artificial coloring. I think the closers here know what photocroms are by now. Ready to place foot in mouth upon proof. wadester16 | Talk→ 20:44, 24 May 2009 (UTC)
        • See Misplaced Pages:Featured picture candidates/Tunis, Tunisia 1890s: unanimous support (4-0), not promoted. It was one of your closures, actually, and your closing note asserted the artificial color objection had weight. Happened last month. About 20% of my nominations tank in similar ways (different issues, same story). Sometimes for legit reasons, but the really disappointing thing is that this occurred on what would have been our first Tunisian FPC. It is not easy to get good historic material about Africa, so each time we burn through one of these it reduces a scarce pool of source material. Durova 22:12, 24 May 2009 (UTC)
          • Yea, I can see your points, but the dramatic shift in color (in the original) is, admittedly, a cause for concern. Either the original was done poorly or aged poorly. Additionally, it did not meet the requirements for passing; if you want to play numbers, you had 3.5 supports. Any other much more clear examples? wadester16 | Talk→ 22:27, 24 May 2009 (UTC)
            • And what you are doing right here is attempting to justify overturning a unanimous candidacy, by letting your own personal tastes intrude. That's the core problem that is eating away at the morale of this process. Why hold FPCs at all? How about we dismiss with the formalities and I submit my work directly to your user talk? :P Durova 22:30, 24 May 2009 (UTC)
              • Ouch! "...with three or more reviewers in support (excluding the nominator(s))" is definitely not related to my personal tastes. This was closed for not meeting the minimum number of supports (3.5 < 4.0). I added the other comments, making them incidental, so others would completely understand the reason for the closing. I don't think one can argue with my closing, really. So does this mean you don't have another, clearer example? wadester16 | Talk→ 22:48, 24 May 2009 (UTC)
                • Look at how you presented the argument above. That wasn't your main reason; that was the afterthought. The primary argument was your own personal taste. And the only reason it failed by the backup criteria you now cite is because you refused to keep it open long enough that it could succeed. You refused to do that even though its support was unanimous (the only thing short of full support was one weak support, no opposes whatsoever). The closure itself, although wrong, was less offensive than this hemming and backtracking. Because now you're trying to argue on principle and the principles don't hold together. I formally request that you retire from closing FPCs; this is a case in point of exactly what's gone wrong with this process. Durova 23:15, 24 May 2009 (UTC)
                  • Are we talking about the same nom here? Not enough time? I asked for more reviews after 7 days and 3 hours, at which point you only had 2 S and 1 WS. This was before the issue about asking/waiting for more votes (which I thought—and still do—was a favor to the nominator). Then, you finally got another vote after 2 days, the content of which was "Looks better, great pic" (then the question is raised as to how much weight I should even give that vote). I closed it four days later, which gave you a remarkable 14 days, 11 hours of open nom; and it still didn't have four full supports. This was absolutely not an incorrect closure. FPC is for WP's greatest pictures. Lack of interest = not one of WP's greatest images. I enjoy and very much respect your skills at restoration, but just because you do one doesn't mean it's a shoe-in for FP. Same goes for Shoemaker's Holiday. You ask for this new "director", but have now explicitly asked the two main closers to recuse themselves forever. Who do you expect to step up to that position? So far you have offered me minimal reason as to why I should retire from closing noms in that you have not offered another, clearer example, as requested above. If I'm doing such a poor job, please prove it. wadester16 | Talk→ 23:50, 24 May 2009 (UTC)
• (outdent) I ask for either clear criteria or the election of a director who will follow the same impartial standards that are normal and routine at other processes. The fact that the two most active closers at this process refuse to do that and try to justify overturning consensus on the basis of personal taste, is why we have the current problem. Although reluctant to join Shoemaker's Holiday in his boycott, I am prepared to join him if this discussion does not prove fruitful soon. Nominations were kept open as long as a month during the time when you closed that unanimous support as a failure. It was clearly on the verge of passing. Your own statements confess that personal taste intruded. That is unacceptable. Durova 00:17, 25 May 2009 (UTC)
  • I will tell you now that I am extremely impartial and you are blowing this one example out of proportion. I have closed more than 80 noms now, and you have used the greyest of grey examples as a reason for me to stop closing outright? I am only human and therefore not perfect, but I can almost assure you that you will find no examples of me placing preference or feelings above quality, consensus, and the FP criteria. Find me a nom unrelated to you that fits this false accusation of "overturning consensus on the basis of personal taste." Unfortunately, you and Shoemaker's Holiday are taking all of this way too personally and must realize that, while your work is respected and appreciated at a great level here on WP, your works are not guaranteed an FP star. I'm sorry that you contribute in a realm that many find uninteresting in many cases, but these issues aren't my fault (nor MER-C's). The rules are the rules: 4 supports, 67%+, with no valid opposes or comments that point out flaws that may detract from the possibility of a candidate's passing, all within a 7-day time period. I asked you to prove to me, and everyone else here, why you really think I shouldn't close. What is unacceptable is your unfounded accusations against a good-faith editor who has put in much time and effort to keep this place running smoothly. I'm honestly flabbergasted in your comments to me here; I've looked up to you since I started taking part at FPC, seeing you as an impartial, unbiased, and reasonable contributor. Your actions here have begun to degrade that view. wadester16 | Talk→ 01:02, 25 May 2009 (UTC)
  • You appear to mistake a clear expression of a conclusion with a request for advice. I am not in need of your counsel regarding significance to attach to these matters, and until this discussion have gone to extremes to assume good faith in the face of long term doubts. Shoemaker's Holiday now has a partner in his boycott: I withdraw from FPC until changes are implemented in this process. Durova 01:24, 25 May 2009 (UTC)
    • My goal is not to scare you away from the place, but you make claims of putting taste over rules and you "going to extremes to assume good faith in the face of long term doubt." You've never made this known to before and one generally expects evidence to go along with claims like these. So far you are lacking substantially in that department. wadester16 | Talk→ 02:03, 25 May 2009 (UTC)
      • Your goals are as irrelevant as your attempt to establish the standard of evidence by which to 'permit' my withdrawal. Adieu. Durova 02:16, 25 May 2009 (UTC)

Another Possibility

Implement some form of supermajority, allowing discretion in borderline cases (eg 4/2 nominations). The delist process could then be used without prejudice if there is a problem (ie no "just got promoted" votes). This would iron out the crappy images and no one would feel cheated. Featured pictures don't really recieve much additional traffic until POTD, which is usually ~9 months from the promotion date iirc. I don't believe a poor quality FP floating around for a week or two taints wiki's reputation. Photographs and scans are generally inherantly factual and neutral when captioned appropriately. Noodle snacks (talk)

• I don't know if this would only complicate the process unnecessarily, but perhaps there could be a section at the bottom where completed nominations sit for a couple of days where others can verify the result of the nomination. However, no further votes would be counted. The closer could briefly explain how they reached the conclusion of the nomination, including which votes/reasoning they discounted, etc. The entire process would then be transparent and if anyone had an issue with the result, they could raise their concern before it is closed and all various pages updated. I know this increases bureaucracy and process, but I really think the more transparent the process, the less problems we'll have. Diliff | (Talk) (Contribs) 16:03, 23 May 2009 (UTC)
• Fully agree with Diliff that closed noms should stay for a couple of days in the page before the final decision is executed. This period could be used to scrutinize the closing and, if necessary, to adjust the decision. I also agree that, in borderline cases, the closer should justify his decision. Well, I suggested this some time ago but got an overwhelming (still poorly founded) opposition. Good to see that people is now in the mood to discuss everything. Kudos to Maedin and to the regulars who joined the discussion Alvesgaspar (talk) 16:34, 23 May 2009 (UTC)
  • I kind of like Diliff's suggestion too. I had agreed some time ago that I would like to see the closed noms stay on the page for a while (say a couple of days or until the next closing) - whether this was before or after the 'full' closing process had happened would need to be decided. Must say I'm not a fan of the closing decision/process becoming the seed for more regular extensive debates though, which is a risk if they're only provisionally closed at the first stage - the closing process is already arduous enough. And it is fair that borderline cases are justified by the closer (I have always done so myself when closing). --jjron (talk) 16:48, 23 May 2009 (UTC)
    • Yeah, I know that there will be a few nominations that will be hotly contested if the result is close, and I do pity the closer who has to deal with it. Ultimately, somebody has to decide one way or the other though, and inevitably there will be some unhappy 'interested parties' - we can't completely avoid that. At least they'll be able to have their say, which is something that has never happened prior to the closure before. And I agree that a lot of borderline nominations are closed with justification, but MER-C (and I'm not singling him out because of any great fault of his own necessarily), by virtue of the volume of closures that he deals with, generally does not. I think that he's going to either provide better justification or share the load a little more. It's the only way forward. Diliff | (Talk) (Contribs) 16:55, 23 May 2009 (UTC)
• Agree Placing recently closed noms would just be one minor addition to the long checklist involved in closing. I would suggest that the recently closed nom stays for 2 days and a bot (maybe Dustybot?) could remove it exactly after 48 hours. In this case, being correct to the second could be good, because we could end up with closers mistakenly removing closed noms too early (the suggestion of removing them at the next closing doesn't work because they are erratic). wadester16 | Talk→ 20:56, 24 May 2009 (UTC)

I see that the proposals on this page include barring entire classes of articles, such as novels, from ever having an FP attached to them, because of some false understanding of EV that claims that images of scenes from novels do nothing to add to their articles, and other proposals that do nothing but make sure that we'll be driving people out of the process for years to come, like claiming that an oppose with the slightest merit should undo all supports an image has.

Again, no image has ever been promoted against supermajority: opposes are never discounted. Even if they're so blitheringly stupid that they make your jaw drop. Is the point of this process to fix non-existent problems, with the side effect of aggravating ones that really do exist, or is the point to actually fix the problems? At the moment, it appears to be the former. Shoemaker's Holiday (talk) 09:56, 24 May 2009 (UTC)

Can we kindly refocus onto actual, documented problems, instead of making proposals to deal with non-existent problems that would make the real problems worse? Shoemaker's Holiday (talk) 11:53, 24 May 2009 (UTC)

Sarcasm is really helpful! wadester16 | Talk→ 20:56, 24 May 2009 (UTC)

You do seem to have a nack of taking complex arguments, removing all the nuances that do justice to that complexity, then summarising what is left in such a way as to completely misrepresent the original arguments. I assume that you're refering to one of my responses when you say that we've proposed to bar literature articles from having FPs. That's definitely not what I said and I don't think anyone else did either. But what I will say is that some subjects are inherently more difficult to illustrate to FP standard than others, and I would include literature in that. Diliff | (Talk) (Contribs) 21:23, 24 May 2009 (UTC)

The fact is this: any response other than support, no matter how spurious, runs a good chance of driving away other reviewers simply because it isn't a support. People suppose the objection is valid, or simply don't want to get caught in 'drama', and move on. Then closers assert the power of discretion to hold over some nominations but not others, and to overrule even unanimous or near-unanimous consensus based upon any objection at all--which as time goes on looks more and more like a closer's repertoire of cover stories for WP:IDONTLIKEIT. Discretion has been misused. It's been misused blatantly and for a long time, and when closers get questioned their own justifications invoke personal taste as an excuse for overturning consensus. The answer to that is to cut down on discretion. If we cut down on discretion drastically, point your fingers at the closers who do this and who still make excuses for it. We're at the point where a prolific contributor is openly boycotting FPC. Others less prolific people are probably doing the same and just not talking about it. Let's get down to brass tacks and fix this: either elect an FPC director or implement a much more restrictive closure standard. Durova 22:44, 24 May 2009 (UTC)

Summary report (as of May, 24)

• Less than three days after this discussion was started and we already have a significant participation and even some consensus on certain matters. Here is my reading of what has been discussed so far. I tried to remain neutral (and ignore punctual discussions) but maybe some important points were not correctly addressed. Please feel free to make the necessary changes.
  • Everybody appears to agree that changes are needed, as no user signed below the Leave all as is title. Still this might be not 100% true as some opinions seem to defend the present status quo, when discussing particular issues.
  • Most agree that some supermajority should be explicit in the guidelines. No alternatives were presented to the 2/3 solution.
  • No consensus as whether the minimum number of support opinions should remain as it is or raised to five.
  • Most agree with the present 7-days evaluation period. Whether it should be fixed or only approximate is not yet clear. The possibility of extending this period in particular cases (e.g. lack of opinions) was raised but not discussed in detail.
  • A proposal was made to create a section at the bottom where completed nominations sit for a couple of days where others can verify the result the closer could briefly explain how they reached the conclusion of the nomination, including which votes/reasoning they discounted, etc. This was supported by three users and opposed by none.
  • Most agree that the present closing model has to be adjusted as, in some cases, does not reflect the opinion expressed by the reviewers. A consensus seems to be growing that only in exceptional and/or borderline cases should the closer use his/hers discretion to disregard user’s opinions. Also, these actions should always be documented and justified. It is still not clear what reasons could the closer invoke to make use of such discretion: only the guidelines or also WIAFP-related ?
  • An important point was raised on the inadequacy of WIAFP to deal with other than digital photos, like Illustrations and restores. Alvesgaspar (talk) 15:09, 24 May 2009 (UTC)