Wikipedia talk:Good article nominations/Future of GAN Backlog Elimination Drives

Welcome to the discussion page of this Request for comment. Here, you can discuss all the proposals made during the duration of the request. Proposals by Hahc21 Proposal 1 Introduce a qualifying process: Each user willing to participate selects and article from the Good article nomination list, does the review, and submits it for consideration by the drive coordinators. If, after the submission is revised, coordinators agree that it is of enough quality, the user is allowed to participate. Discussion Proposal 2 Review limits: Each participant is bound to a limit of 5 reviews per day (between new and opened reviews). This means that any user cannot have more than 5 opened reviews at the same time. If a user violates this rule, they may be disqualified from the drive. Note that this does not limit the maximum number of reviews that a user can do per day, but the number of opened reviews. Reworded proposal below Review limits: Each participant is bound to a limit of 5 reviews per day and they cannot have more than 5 opened reviews at the same time (between new and open reviews from the previous day). If a user violates this rule, they may be disqualified from the drive. Discussion Just want to clarify, if I have five reviews on the go, then I can not start reviewing another one until I close one of them? AIRcorn (talk) 23:28, 3 November 2012 (UTC) I have reworded the proposal. Now it read as this: You cannot have more than 5 opened reviews, between reviews you started the previous day(s) and are still in progress and new reviews you pick up that day. Also, you can pick up a maximum of 5 reviews per day. Example: Day1 i took 5 reviews. Day 2 I closed 2 reviews so I have 3 left. Then, I can only pick up two reviews more and, until I close the other 3, I can't pick up more reviews. If i close all the reviews from Day 1, I can pick up up to 5 reviews on Day 2. — ΛΧΣ21™ 23:51, 3 November 2012 (UTC) Isn't that redundant? If someone can only have five reviews open at a time, then they can't open more than five in one day anyway. Also say a reviewer takes on average one week to conduct a review, then they will only be able to get through twenty reviews in a month with this limit placed. Some reviews take much longer than that for various reasons, for example articles on vital topics, editors on wikibreaks, unexpected excess work, the need to double check at other noticeboards (pictures, reliability of sources etc), time to gain access to non-internet based sources and so on. I fear that this limit could potentially reduce the quality of the reviews as competitive editors will look to rush nominations they probably shouldn't. AIRcorn (talk) 23:57, 3 November 2012 (UTC) Well, you may be right. That's why I asked community imput to see what they thought about it as a proposal. We can further discuss this. I proposed this as a limit for persons who make fast reviews just to have the highest number of reviews of the drive. This will limit them to only have 5 opened reviews so that they are forced to review only five. Also, by stating that they can only make 5 reviews a day, we limit the number of possible bad reviews we can handle daily and find potential problems fast. This way, we identify 5 mediocre reviews from a single user rather than 25, as happened in the previous drive... just my thoughts on it. (PD: I am neutral. if it's supported or opposed i will be fine XD) — ΛΧΣ21™ 01:09, 4 November 2012 (UTC) Seems a bit harsh to disqualify someone if the reviews are good, but since you say "may" I assume there is some descretion applied. The only way to know if it will increase or decrease the quality will be to try it, and as it has quite a bit of support at the moment we might well find out. AIRcorn (talk) 00:25, 4 November 2012 (UTC) Well, it will be discretionary. If an incredible reviewer opens a sixth review on a single day, of course we won't be disqualifying them from the drive, as it will be disruptive, and this principle may be applied only when it showcases that it will prove useful. — ΛΧΣ21™ 01:02, 4 November 2012 (UTC) How is it "of course"? If I'm a reviewer, why would I try to open a sixth review if by doing so I risked disqualification? Discretion is a double-edged sword. Maybe you need to include something in the drive's rules that says something like, "if all your reviews have stalled and are waiting on action from the nominators, we're happy to grant permission for you to start a new one upon request".BlueMoonset (talk) 04:15, 5 November 2012 (UTC) Perhaps a safety valve might be in order? Perhaps a reviewer could be allowed to have up to an additional five on hold, for a total of ten? Or would it make sense to be allowed to always start one new review a day up to some maximum limit of open reviews even if they're at five or above? (Note that any second new review in a day would depend on there being fewer than five reviews open at that time.) These might help to limit the likelihood of people closing reviews prematurely. BlueMoonset (talk) 00:34, 4 November 2012 (UTC) Interesting. I may work on an additional clause to add this. — ΛΧΣ21™ 01:02, 4 November 2012 (UTC) Well. I have given it an intense session of thought and i consider t may not be neccessary. We introduce the 5 review limit as a type of enforcement. I mean, a race to have the highest number of reviews is discouraged and barnstars will be given only to 5, 10 and 25 reviews. Also, there won't be leaderboard. So, I think that each reviewer might take the time enough to make a good review instead of going and making 10 regular reviews per day. We want to reduce the backlog, yes; but we want quality reviews. Barnstars are designed to encourage participation, not a race for first place. I have always thought that we need this type of events to gather interested parties and do a collaborative effort that is not usually done every day. — ΛΧΣ21™ 19:00, 4 November 2012 (UTC) This 5 limit thing is a joke. Why would you restrict someone from doing reviews when there's a backlog. What's the point of calling it a backlog emlimination drive then? --Dom497 (talk) 20:25, 4 November 2012 (UTC) With the drives being held each 3 months, we won't have to add the word "elimination" anymore. I may change it with this: "GAN Backlog reduction drive". And also, I have made my homework and it may work. With 10 reviewers, if each of them review 5 articles per day, all of them would review 1500 articles in 30 days. That is like 4 times the backlog. Last drive had 47 participants (722 reviews) and the rate was 15.3 reviews/participant. The current backlog is 465 articles and, with the limitation of older reviews first, I guess we will reach our goal with 20 participants (half of the last drive's). — ΛΧΣ21™ 21:18, 4 November 2012 (UTC) Doesn't your calculation assume that each review gets closed in one day? That is highly unrealistic and more likely to cause issues than having multiple reviews open. AIRcorn (talk) 03:44, 5 November 2012 (UTC) Well, it is half-unrealistic. If real, we'd have 1500 reviews, but we had 722. Of course, this happened with 47 reviewers, not 10. If we go real, I'd say that each user may close a maximum of 2 reviews per day. That means that 10 users may close 20 reviews per day. In a month, that would be 600 articles, which is far less that the previous drive's final count and more than what we have right now as a backlog. Of course, during that month, I'd say that 250 nominations might be added (this happened during last drive). So, we'd hacve a net reduction capacity of 350 articles. If the backlog is at 465 articles, we would have 115 articles left for review. I guess these numbers are way more close to reality. — ΛΧΣ21™ 04:11, 5 November 2012 (UTC) (edit conflict) These are going to be old reviews, making it more likely that the nominator is not immediately available to respond quickly when the review begins. Since a seven-day waiting period is pretty much the minimum you can limit it to before ending a review for non-response, a reviewer could be stuck in limbo in fairly short order if the luck is bad. Having no safety valve at all to the five-total limit is looking less attractive. As I recall, the reason behind introducing the limit was because people last time were giving brief, complete reviews to more than five a day, or grabbing ten or more reviews at a time, and letting them sit for days before starting them: both sorts of behavior need to be discouraged. (The five new a day also limits the damage from a problematic reviewer.) The current proposal seems likely to have unintended consequences, as reviewers may be up against the limit of five, all stalled reviews awaiting action by the nominator, and unable to take on more reviews. That is not a good thing. BlueMoonset (talk) 04:15, 5 November 2012 (UTC) Oh. BlueMoonset, you have showed me something I haven't considered before. Now I see your point. I agree that we should introduce a safety valve to avoid users being stalled because nominators did not respond to issues raised at time and thus rendering the reviewer unable to take more reviews. Oh damn. Well, developing the safety valve would not be an easy task if we want it to work properly, but I will put all my efforts into it if the proposal is approved. — ΛΧΣ21™ 04:26, 5 November 2012 (UTC) Proposal 3 Disqualifying process: Each user with more than 3 removed reviews may be disqualified from the drive. Disqualifying process: Each user with more than 5 removed reviews may be disqualified from the drive. (reworded) Discussion Proposal 4 - Part 1 Drive timespan: Drives will last for 15 days. Discussion Proposal 4 - Part 2 Drive timespan: Drives will last for one month. Discussion Proposal 4 - Part 3 Drive timespan: Drives will be held each three months. Discussion Proposal 4 - Part 4 Drive timespan: Drives will be held each six months. Discussion So now that the three-month one has been redefined as three months from the end of the previous drive, or every four months assuming the one-month drives proposal continue to win, does this one get redefined as six months from the end of the previous drive, which would mean every seven months? Seems confusing to me, and much simpler if it starts every six months so it's the same two months every year. BlueMoonset (talk) 03:19, 10 November 2012 (UTC) Well, this is not redefined because the time is much longer. So, with this proposal, drives will be in month 1 and 7, as an example. — ΛΧΣ21™ 04:20, 10 November 2012 (UTC) Then you're being inconsistent in your usage, unless you originally meant every three months as starting in months 1, 4, 7, and 10, as an example, and have only just today changed it to 1, 5, and 9. BlueMoonset (talk) 04:33, 10 November 2012 (UTC) I originally meant 1, 5, and 9. I never meant 1, 4, 7, and 10 for the three-month proposal. — ΛΧΣ21™ 04:40, 10 November 2012 (UTC) I believe you. But the problem is that you were using three months in one way and six months in another, and it's only logical for people to assume that the same methodology applies to both. BlueMoonset (talk) 04:51, 10 November 2012 (UTC) Yes; I understand your point and I di not realize it until now. Thaks for the comments :). — ΛΧΣ21™ 04:54, 10 November 2012 (UTC) Proposal 5 Drive scope: Drives will only be aimed to review the oldest nominations. Older nominatios are those with more than one-to-two months on the queue. Nominations with less than a month should be excluded from the drive untill all older nominations are reviewed. Discussion While I don't like discouraging review of new articles, I think it is good to encourage reviewing older articles, so I have an alternative idea: it would be a competition for only those interested in participating. For the old articles, each article is worth however many days it has been sitting in the queue (156 days old = 156 points), and whoever does the most of those would win a special award. Of course, all of those articles reviewed would need to be checked over to qualify. --Tea with toast (話) 02:22, 5 November 2012 (UTC) Well, I thought of this idea before, but then I came with the conclusion that it'll be too hard to apply unless each reviewer does the job to count each nomination's oldness and add it to the drive page. I mat discuss this later at WT:GAN to see if we reach consensus. — ṞṈ™ 03:33, 5 November 2012 (UTC) I don't think it would be difficult to do. Users can just go to Wikipedia:Good_article_nominations/Report to look to see how old an article is. Moderators can use a tool like this one: [1] to check for accuracy. --Tea with toast (話) 04:23, 5 November 2012 (UTC) Proposal 6 - Part 1 Drive and barnstars, Part 1: Barnstars will now be given for users who reviewed 5, 10 and 25 nominations. There will be neither leaderboard or a number-one position race, and disruptive competence for holding the highest number of reviews is discouraged. Discussion Oops, sorry I didn't see this talk page earlier. I made a comment here about replacing "award for most articles" with something else like most articles with 5k+ word count. --Tea with toast (話) 02:11, 5 November 2012 (UTC) The way the WP:GOCE drives deal with getting people to review very large and wordy articles is to have a special competition for who does the most articles that have more than 5000 words. For example, current nominees like Uttarakhand, and Antinomian Controversy would qualify. (Go to the toolbox on the side menu and click "Page size" to reveal how many words are in an article). On second thought, this may not be as feasible for GAN since GOCE deals with 1000s of articles and we only have a few hundred, so there aren't as many of them. Yeah, I think the best way to deal with jumbo articles is to lump them with the other undesirables that for whatever reason end up in the queue for months. Please see my comments at Prop 5 for another idea. --Tea with toast (話) 03:19, 5 November 2012 (UTC) Well, i like this idea and, after thinking of it for several days, I have polished it to this: We can make a list of the oldest reviews and discretionarily award a barnstar to users reviewing them. Of course, this barnstar[s] will only be given if the review is checked and it is considered of high quality. Otherwise, the user won't be the recipient of the award. — ΛΧΣ21™ 01:31, 8 November 2012 (UTC) Pretty sure you mean "competition" or "competing", not "competence". I recommend modifying "for holding" to "to hold". BlueMoonset (talk) 18:37, 6 November 2012 (UTC) Proposal 6 - Part 2 Drive and barnstars, Part 2: Barnstars may be discretionarily given [by coordinators or drive reviewers] to those users who reviewed many old nominations, as well as to those who did high quality reviews. To give these barnstarts, a mini-vote may be held by coordinators and reviewers approving or dissaproving the award. [may be held means that such vote is not mandatory but optional.] Discussion Proposal 6 - Part 3 Drive and barnstars, Part 3: Each user may submit up to two reviews for consideration to receive a special barnstar which will be awarded to the best three reviews of the drive. A vote may be held with coordinators, drive reviewers and participants voting for their favourite review in order to select the three reviewers who will receive the award [each reviewer may receive only one award]. Discussion A new approach to the drive format I am obviously quite late to this RFC, so I'll just drop this here for light consideration: Would it be worthwhile to, in place of quarterly or semi-annual drives, instead target a specific backlogged queue for a brief period once a month? i.e.: If the Sports and recreation queue were to be this month's target, send notices to regular reviewers and editors heavily involved in the sports queues (as nominators or reviewers) asking them to do only a small number of reviews - say 2 or 3 in a week's time (with emphasis on oldest outstanding noms). If you can get 10 editors to average 3 reviews in that time, you could probably drop the backlog from 80 to 60 in a week. Then the next month, focus on music articles, then history, etc. The commitment is intended to be small; Just a couple reviews. It allows new reviewers to wet their feet without causing a big splash if there are problems, and regular reviewers would face only a small additional burden. My theory is that we could create a process that continually refreshes the queues while remaining lightweight enough to avoid the boom-bust cycle of reviews that seem to follow the backlog drives. Thoughts? Resolute 17:19, 8 November 2012 (UTC) I quite like the sound of this. Maybe worth raising at WT:GAN? Sarastro1 (talk) 22:25, 10 November 2012 (UTC) Can do. I might let this RFC run its course first, however. No sense creating multiple discussions until we know how this shakes out. Resolute 22:47, 10 November 2012 (UTC) I like the idea of small focused drives. AIRcorn (talk) 01:52, 13 November 2012 (UTC) Actually I like it too. Could we develop this as a future substitute for drives? — ΛΧΣ21™ 02:25, 13 November 2012 (UTC) Moving goalposts and other problems I'm a little concerned that several of the proposals have changed at least once during the duration of this Rfc. I'm pretty sure that this shouldn't happen; what about the people who commented on the old proposal and are not aware that it is changed. I'm also a little concerned that several problems have arisen over the lack of clarity in several of the questions. Perhaps it would have been better to have this checked before it went "live". We currently have a proposal which states "and disruptive competence for holding the highest number of reviews is discouraged". A user has pointed out that this does not actually make sense in a section above, but the wording remains unchanged. A final point is the location of this Rfc. Usually, a Rfc is held in project space, but this one is in user space. Is there any particular reason for this? It is perhaps unfairly creating an impression that the user is "in charge" of both the process and the Rfc. And who will close the Rfc? This probably needs to be established quite quickly, and it probably should not be the person who created it. Sarastro1 (talk) 22:33, 10 November 2012 (UTC) Well. I have been sure that the wordings that were controversial, when made, I left a note to all users who voted. Also, I won't close it. Any uninvolved user may do it. Although I designed the proposals, I have no personal vote for any of them. If they are all rejected, I have no problem; same as if they were all approved. Also, I have not seen any user saying nothing about "competence" in proposal 6.1. I have changed it with "competition", which is the accurate word. — ΛΧΣ21™ 22:43, 10 November 2012 (UTC) And, the reason why this was held in my user space is because i dod not find the right place (WP:GAN or WT:GAN). — ΛΧΣ21™ 22:44, 10 November 2012 (UTC) I also raised similar concerns on your talk page. The recent change of the passing percentage takes it too far, especially this late in the game. It gives the appearance that you are adjusting it to push through proposals you are interested in. AIRcorn (talk) 01:51, 13 November 2012 (UTC) Actually I was about to ask here which would be the perfect percentage. I am fine with 70%, but I believed it was too high :S — ΛΧΣ21™ 20:10, 13 November 2012 (UTC) And adding a bit of background, the only proposal I may be interested is the first one, "should we have drives, or not?". I have no personal interest in the rest of the proposals. — ΛΧΣ21™ 20:14, 13 November 2012 (UTC) That one is quite concerning, particularly using an alternate account which not everyone might realise belongs to the same user. I would suggest that this is reverted. I also wonder if it should be raised at WT:GAN in case anyone is unaware that changes have been taking place. Sarastro1 (talk) 20:03, 13 November 2012 (UTC) I changed it back. Actually, I am neutral in this RFC. I have no personal feelings about the proposals and I just made them to see which ones community likes the most. If they are approved, or rejected, it is the same for me. I am just trying to help GAN people to finish the drives-yes drives-no debate that has been taking place since May. — ΛΧΣ21™ 20:12, 13 November 2012 (UTC) You sort of shot yourself in the foot a bit with the 70 percent rule, as it is seldom that black or white (for example proposal 6 part 1 is currently outside the threshold, but at least two of the opposes are supports or partial supports). AIRcorn (talk) 02:22, 14 November 2012 (UTC) Well, actually I considered that 70% was the amount of community support needed, like RFA. We don't want proposals passed with 50% support, 50% oppose; that's not consensus in my opinion XD — ΛΧΣ21™ 02:29, 14 November 2012 (UTC) There is no magic number. I have seen 50-50 outcomes be called consensus. It all depends on the strength of the arguments. I understand why you might think that though, editors on the wrong end of a close decision will inevitably point to the percentages as a reason for no consensus. AIRcorn (talk) 02:40, 14 November 2012 (UTC) Yes, I understand your point [and that's why ArbCom members are elected with 50%, to give an example]. If we take the strength of arguments, proposal 6.1 is passed, although we still have two days left. Although, I think we'd need another discussion at the talk to determine the timing between the drives, as this won't get consensus here... IMO. — ΛΧΣ21™ 02:53, 14 November 2012 (UTC)