Wikipedia talk:Newbie treatment at Criteria for speedy deletion/Archive 3
 | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
| Archive 1 | Archive 2 | Archive 3 | Archive 4 |
Control test
As a control against the experiment, how about listing all new articles created in a short timescale and following them in a similar way to the WP:NEWT experiment. Take three 10-minute timeslots, eight hours apart from each other, list all articles created in those timeslots and table their progress. Of course, there could well be junk articles, obvious CSD candidates etc, as well as decent articles. You might be lucky enough to catch a few newbies creating articles though. The timeslot can be made bigger/smaller if necessary. The three spread over a day accounts for different timezones. Mjroots (talk) 16:33, 18 November 2009 (UTC)
- That would be interesting, but you'd probably need a lot more than half an hours worth to get a statistically valid sample of incorrectly deleted articles. If we are still getting a couple of thousand articles a day this would only get you about fifty articles, and most of them would either have been correctly deleted or correctly kept. Expanding the sample so you have a meaningful number of miscalls would mean there'd be an awful lot of crud to go through. Also you'd need some sort of bot to create the lists, especially the deleted ones. As we've learned from this project pointing out people's errors at CSD is something that needs to be done very gently, especially where there is any hint of greyness in the call (you particularly need to allow for tags that were correct but ceased to be after someone else salvaged the article). Lastly a big part of the project would be admin only as I suspect as many as half the articles involved will have quite correctly been speedy deleted. All that said, if someone did organise it I would take my turn at sorting through the data. ϢereSpielChequers 17:28, 18 November 2009 (UTC)
- I'd also be willing to put in some time on this, assuming the concerns raised by WSC above are addressed in the planning stage. Olaf Davis (talk) 18:56, 18 November 2009 (UTC)
- The "experiment" itself wasn't done with any thought to the scientific process, so doing a control against it and trying to make it legitimate isn't going to fix that. The assumption that started this was with bias so of course their results are going to return bias. If you want to properly conduct this experiment and generate any useful data, then you need a mandate and permission from the community and then the test needs to be set up in a properly scientific manner from all sides. Until that happens nothing good will come out of this.--Crossmr (talk) 21:13, 18 November 2009 (UTC)
- the problem and biggest hole with the idea of an experiment is the subjective question: what is a valid article? For example, people above are arguing that NW's article was not notable. You would have to have an objective criteria, agreed upon before, which most everyone would agree on. Ikip (talk) 22:44, 18 November 2009 (UTC)
- The criterion was that the article not meet any of the criteria for speedy deletion; meeting that is no more or less subjective than the criteria for speedy deletion themselves. (And not being notable is not grounds for speedy deletion.) +Angr 23:04, 18 November 2009 (UTC)
- (Edit Conflict) Alternatively you make that part of this test. I suspect that the proportion of new articles flowing in where people really disagree whether they should be deleted is quite low. But this sort of test could quantify that, in effect we'd be classifying the articles in two ways, what did happen to them and what should have happened to them, and the "should have" may include some where we conclude that the community is currently divided. Also since in order to include those subsequently deleted, you would be randomly preselecting your articles by computer, there would be no need to restrict this to three ten minute slots; you could instead do a one in n sample over a longer period such as a month. That would remove time of week bias from the articles created as you would be including stuff from various school hours as well as the post pub stuff. I suspect many ten minute slots have most of their speedies deleted by the same admin, so a 1 in n over a longer time period would avoid skewing this by who happened to be on patrol at the time. ϢereSpielChequers 23:13, 18 November 2009 (UTC)
- If the criteria for speedy deletion were sufficiently objective that they could be used to provide a definition of a a valid article, then the results would already be known. The whole problem is that interpretations differ and people may choose to be more liberal/conservative on borderline cases; that won't change just because you're looking at the article in a different setting. The most frequently used CSD are far too subjective to use for anything remotely scientific. Personally, I would abandon the idea of "was the deletion correct" except for exceptional cases. 1) You avoid the possibility of the project devolving into "naming and shaming" and 2) If the article would be deleted at PROD or AFD anyway, the issue is purely academic. What the project should focus on is communication with the new user and timing issues (time between creation, tagging, patrol, deletion, etc.), as these can actually be measured in a mostly objective way. Mr.Z-man 23:30, 18 November 2009 (UTC)
- I've done tests a year or so back with the deletion log. It will be better done taking groups of related criteria at a time. If you select only a1, a3, a7 & G11, about 500 deleted articles per day, with perhaps 10 for rescuable or even adequate articles, another 10 deleted for reasons that don't apply, but articles unlikely to be rescued. That's 10 a day that we lose, 3,650 a year. There are maybe twice that number of copyvios that could be rewritten into adequate articles for notable subjects. The articles don;t worry me that much--it's the 4,000 contributors we lose--I think many of the copyvio people could produce adequate articles if they were helped, but most would not bother. Frankly, I'm not interested in getting a scientific figure, but in instructing the individual admins who do repeated wrong deletions (every one of us who does many of them do a few rare wrong ones)--privately if possible. I've given up on converting the admins who delete when they could rewrite if they did a little work at it.. DGG ( talk ) 23:45, 18 November 2009 (UTC)
- BTW, the reason i did not pursue it was that most of the bad deletions came from well-known admins, and I wasn't up to that kind of trouble. One guy I did follow up, we became good friends, but it didn't change how he does things. Informal approaches are not always what is needed. DGG ( talk ) 23:57, 18 November 2009 (UTC)
- I've done tests a year or so back with the deletion log. It will be better done taking groups of related criteria at a time. If you select only a1, a3, a7 & G11, about 500 deleted articles per day, with perhaps 10 for rescuable or even adequate articles, another 10 deleted for reasons that don't apply, but articles unlikely to be rescued. That's 10 a day that we lose, 3,650 a year. There are maybe twice that number of copyvios that could be rewritten into adequate articles for notable subjects. The articles don;t worry me that much--it's the 4,000 contributors we lose--I think many of the copyvio people could produce adequate articles if they were helped, but most would not bother. Frankly, I'm not interested in getting a scientific figure, but in instructing the individual admins who do repeated wrong deletions (every one of us who does many of them do a few rare wrong ones)--privately if possible. I've given up on converting the admins who delete when they could rewrite if they did a little work at it.. DGG ( talk ) 23:45, 18 November 2009 (UTC)
- If the criteria for speedy deletion were sufficiently objective that they could be used to provide a definition of a a valid article, then the results would already be known. The whole problem is that interpretations differ and people may choose to be more liberal/conservative on borderline cases; that won't change just because you're looking at the article in a different setting. The most frequently used CSD are far too subjective to use for anything remotely scientific. Personally, I would abandon the idea of "was the deletion correct" except for exceptional cases. 1) You avoid the possibility of the project devolving into "naming and shaming" and 2) If the article would be deleted at PROD or AFD anyway, the issue is purely academic. What the project should focus on is communication with the new user and timing issues (time between creation, tagging, patrol, deletion, etc.), as these can actually be measured in a mostly objective way. Mr.Z-man 23:30, 18 November 2009 (UTC)
- (Edit Conflict) Alternatively you make that part of this test. I suspect that the proportion of new articles flowing in where people really disagree whether they should be deleted is quite low. But this sort of test could quantify that, in effect we'd be classifying the articles in two ways, what did happen to them and what should have happened to them, and the "should have" may include some where we conclude that the community is currently divided. Also since in order to include those subsequently deleted, you would be randomly preselecting your articles by computer, there would be no need to restrict this to three ten minute slots; you could instead do a one in n sample over a longer period such as a month. That would remove time of week bias from the articles created as you would be including stuff from various school hours as well as the post pub stuff. I suspect many ten minute slots have most of their speedies deleted by the same admin, so a 1 in n over a longer time period would avoid skewing this by who happened to be on patrol at the time. ϢereSpielChequers 23:13, 18 November 2009 (UTC)
- The criterion was that the article not meet any of the criteria for speedy deletion; meeting that is no more or less subjective than the criteria for speedy deletion themselves. (And not being notable is not grounds for speedy deletion.) +Angr 23:04, 18 November 2009 (UTC)
- the problem and biggest hole with the idea of an experiment is the subjective question: what is a valid article? For example, people above are arguing that NW's article was not notable. You would have to have an objective criteria, agreed upon before, which most everyone would agree on. Ikip (talk) 22:44, 18 November 2009 (UTC)
- The "experiment" itself wasn't done with any thought to the scientific process, so doing a control against it and trying to make it legitimate isn't going to fix that. The assumption that started this was with bias so of course their results are going to return bias. If you want to properly conduct this experiment and generate any useful data, then you need a mandate and permission from the community and then the test needs to be set up in a properly scientific manner from all sides. Until that happens nothing good will come out of this.--Crossmr (talk) 21:13, 18 November 2009 (UTC)
- I'd also be willing to put in some time on this, assuming the concerns raised by WSC above are addressed in the planning stage. Olaf Davis (talk) 18:56, 18 November 2009 (UTC)
"the reason i did not pursue it was that most of the bad deletions came from well-known admins, and I wasn't up to that kind of trouble." spoken like someone who really knows how wikipedia works. Made me smile, thanks. I would be willing to conduct such a test. A person would have to have admin rights though, correct? Again, I could create a scrip with some help from Tod, which scraped all of the speedys the moment they are tagged. Another problem is this: regardless of the results, most editors behavior will not change at all. It is almost as if we are going to conduct this research only for our only gratification, and no other reason. Ikip (talk) 00:31, 19 November 2009 (UTC)
- Since we've outdented @DGG - it's the 4,000 contributors we lose Do you have any evidence that we lost 4000 contributors? You have some anecdotal evidence that we might be losing 3650 articles a year that might be rescuable. How many have been recreated? I'm not interested in getting a scientific figure Then I'm not sure how we can go anywhere with this discussion. Until someone can objectively demonstrate that there is in fact a problem with CSD that is significant, I can't see much compelling reason to adjust policy or run cloak and dagger "experiments" based on a bias we have no evidence that exists and that assumes bad faith of new page patrollers. If individuals are constantly making errors, educate. But we're now 4 days into this and no one has been able to objectively demonstrate that there is in fact a problem here.--Crossmr (talk) 00:52, 19 November 2009 (UTC)
- I am assuming that 4000 wrongly deleted articles =4000 eds. who never come back. The new eds. who write several articles all at once are rare, and usually if their stuff gets speedied, it deserves to be. There are some testimonies from people who continued anyway, but I assume that if they feel the need to actually say it, it's because it's pretty rare. but yes, that's something we could look at. More important, I want to emphasise I am not assuming bad faith of either the NPPs or the deleting admins. It does occur sometimes (someone speedy-tags an article about a singer or politician they dislike, or an article about a rival, or written by someone they dislike) buts that is quite uncommon. Almost always the NPPs and the admins who get things consistently wrong are either going too fast, or have a misconception about policy. Both are in perfectly good faith trying to help the encyclopedia, however ineptly. Even admins who insist n deleting articles that they know don't fit the criteria are general thinking that their ignoring the rules is justified, as if IAR is the normal policy, not the exception. The question is whether they will change if it is made clear by the community that we see it otherwise. I hope all of them will, but just going by probabilities, I doubt they all will--and then we need to decide what to do about it. DGG ( talk ) 02:07, 19 November 2009 (UTC)
- "Until someone can objectively demonstrate that there is in fact a problem with CSD that is significant" as I said, regardless of the results, most editors behavior will not change at all.Ikip (talk) 04:10, 19 November 2009 (UTC)
- I am assuming that 4000 wrongly deleted articles =4000 eds. who never come back Wow.. that is a pretty dangerous assumption. You've got nothing to back that up and you'd want to propose policy change based on that? I never said you were assuming bad faith, I said this "experiment" was. It was started with the premise that the NPP were doing something wrong and causing a problem. It was a biased bad faith unscientific test. Several people have said that, several people have quit NPP because of it. So trying to compare any data to this is pointless. The real issue here is trying to figure out if there is an issue with CSD at all. We simply can't do that with the data in front of us. Assumptions and half guesses don't tell us anything. I've got no objection to softening templates or making them more attractive. Nor do I have an issue with increasing education or asking people who regularly do NPP to use tools to benefit them and the community. I see no issue with making sure article creators are notified of other editors concerns with their newly created articles. But as you said that was 10 articles out of 500. That is only 2% of articles each day that are deleted improperly and had a hope of possibly staying around. That isn't a significant number and would indicate that 98% were deleted without issue. That seems to me to be a process that doesn't require a major overhaul. A tune up for friendliness perhaps, but it doesn't seem to be a situation that needs sweeping policy change.--Crossmr (talk) 04:50, 19 November 2009 (UTC)
- "I said this "experiment" was [assuming bad faith]. It was started with the premise that the NPP were doing something wrong and causing a problem." Assuming that someone is making a mistake is not the same as assuming that they are acting in bad faith. I don't recall anyone in the set-up of this project ever claiming the latter.
- I agree that "4000 wrongly deleted articles => lost editors" is a pretty big assumption; verifying that would require a fairly well-thought-out experiment with actual controls (how many people create one article then leave even if it wasn't deleted, for example?) Olaf Davis (talk) 11:59, 19 November 2009 (UTC)
- Actually the language on the page is can we avoid the problem problem implies people are doing something wrong or that their behaviour is bad. A much better statement would be a very neutral "A series of tests to see how the CSD criteria and templates are applied to newly created articles by new users"--Crossmr (talk) 21:46, 19 November 2009 (UTC)
- Yes, I agree that would have been a better wording because it doesn't explicitly assume anything. Still, I maintain that what was assumed was not the presence of bad faith; neither 'doing something wrong' nor 'causing a problem' require bad faith. Olaf Davis (talk) 10:42, 20 November 2009 (UTC)
- Actually the language on the page is can we avoid the problem problem implies people are doing something wrong or that their behaviour is bad. A much better statement would be a very neutral "A series of tests to see how the CSD criteria and templates are applied to newly created articles by new users"--Crossmr (talk) 21:46, 19 November 2009 (UTC)
- I am assuming that 4000 wrongly deleted articles =4000 eds. who never come back. The new eds. who write several articles all at once are rare, and usually if their stuff gets speedied, it deserves to be. There are some testimonies from people who continued anyway, but I assume that if they feel the need to actually say it, it's because it's pretty rare. but yes, that's something we could look at. More important, I want to emphasise I am not assuming bad faith of either the NPPs or the deleting admins. It does occur sometimes (someone speedy-tags an article about a singer or politician they dislike, or an article about a rival, or written by someone they dislike) buts that is quite uncommon. Almost always the NPPs and the admins who get things consistently wrong are either going too fast, or have a misconception about policy. Both are in perfectly good faith trying to help the encyclopedia, however ineptly. Even admins who insist n deleting articles that they know don't fit the criteria are general thinking that their ignoring the rules is justified, as if IAR is the normal policy, not the exception. The question is whether they will change if it is made clear by the community that we see it otherwise. I hope all of them will, but just going by probabilities, I doubt they all will--and then we need to decide what to do about it. DGG ( talk ) 02:07, 19 November 2009 (UTC)
- A pedant writes - actually, it's 2%, not 0.2%, but even so, it's a very small minority. And as I've said before - the best way to stop people tagging articles for speedy incorrectly is to fix some of the CSD criteria so they're not so easily misinterpreted (what is a claim of importance - how do you define that?) Also, to add CSD categories for the most often badly tagged articles that are clearly non-encyclopedia - namely obviously made up stuff, neologisms, and duplicate pages). Black Kite 12:11, 19 November 2009 (UTC)
- Good idea but I don't think tweaking the CSD tags alone are going to be enough: I seriously think we need to be more rigorous about tracking how Twinkle is (mis)used. Perhaps admins need to be even more proactive in warning editors who are serially mistagging articles for speedy deletion, and even imposing blocks in extreme cases, for misuse of CSD. After spending a couples of days monitoring and in some cases removing CSD tags, I'm alarmed at how WP seems to be de-evolving into a race to delete, rather than create. CSD is being widely misused in non-blatant cases. Shawn in Montreal (talk) 21:06, 19 November 2009 (UTC)
- Two examples from the last hour or so alone: a CSD for for blatant hoax where the tiniest bit of WP:BEFORE work prove on Google proves the strain exists; and this diff from User:Alastair Rae (who's been repeatedly warned about improper CSDs) for db-spam where no spam text existed. Shawn in Montreal (talk) 21:14, 19 November 2009 (UTC)
- You can pick out anecdotal evidence all day long. It does nothing to prove there is a problem. That was the entire problem with this "experiment". I see the article wasn't deleted which shows the system works. A single individual isn't deleting these. As for the spam one, I don't feel like something has to be written in a promotional tone to be spammy. Wikipedia tracks high on google the moments articles are created. Simple creation of articles is enough to show up in search results so certain articles on certain subjects which may lack notability can be seen as spammy.--Crossmr (talk) 21:46, 19 November 2009 (UTC)
- Two examples from the last hour or so alone: a CSD for for blatant hoax where the tiniest bit of WP:BEFORE work prove on Google proves the strain exists; and this diff from User:Alastair Rae (who's been repeatedly warned about improper CSDs) for db-spam where no spam text existed. Shawn in Montreal (talk) 21:14, 19 November 2009 (UTC)
- thanks, fixed that.--Crossmr (talk) 21:46, 19 November 2009 (UTC)
- Then this gets back to the question of making the CSD tags more stringently worded, I guess, because that's clearly not what G11 was designed for, regardless of your "feelings." Shawn in Montreal (talk) 22:30, 19 November 2009 (UTC)
- Clearly there is a division of opinion as to the amount of mistagging at CSD and therefore whether we have a problem, this test would could establish that. But if anyone has doubts about the scientific robustness of the methodology discussed at the beginning of this thread please explain your concerns - we may be able to address them in the design stage. As for "I see the article wasn't deleted which shows the system works." That's only partially correct, sure the article wasn't deleted incorrectly, but it was tagged incorrectly, and incorrect tags can be very bitey. ϢereSpielChequers 22:07, 19 November 2009 (UTC)
- As I said above, you're never going to be able to get anything remotely scientific if you use the CSD themselves as the criteria for the test. The most used criteria (A7 and G11) are simply too subjective to analyze with any degree of reliability, and you run into other issues as well. 1) You're looking into the decision in hindsight. One of the major problems I see with the goals here is lopsidedness. Before it even starts, its introducing a bias. The goal seems to be only looking for articles deleted improperly; it should also look at articles that perhaps should have been deleted but weren't. It needs to decide what the outcome should have been, then compare that to the actual outcome, right now its sort of doing the opposite, its taking articles that had a certain outcome, and only looking to see if the outcome could have possibly been different. Its taking a sample, only considering one particular anomaly, and ignoring everything else. Which leads to issue 2) the lack of context. In a study like this, you're choosing a small selection of articles and you have all the time in the world to review them. "Out in the field" patrollers have to review every article created by non-autoreviewers and they have to do so on somewhat of a deadline. Things like copyvios and attack pages need to be dealt with immediately and articles drop off the back of the new pages queue after 30 days. Mr.Z-man 23:49, 19 November 2009 (UTC)
- I've already pointed out that real data can't be used as a control against this experiment because this experiment wasn't run in a scientific manner. You don't have mandate or permission to perform any further "tests". So all we can do is analyze existing data if it exists, but DGG has apparently done that a little bit and shown that by his estimates roughly 2% of articles are incorrectly tagged. That's not bad. You also have done nothing to shown there is an issue with templates being "bitey". But I have said at least half a dozen times that if you think templates are too bitey they can be revamped to be cuddlier. Sorry, you haven't provided a single piece of evidence to indicate that there is any kind of serious problem either in chasing people away or tagging articles too quickly. Do certain individuals need more education? Probably. But I don't see anything here that shows there is a problem with the system itself.--Crossmr (talk) 03:13, 20 November 2009 (UTC)
- Nope, he stated that 2% are incorrectly deleted. Any admin that patrols CSD can testify that the % of incorrect tags is way higher, perhaps as much as 50%. --ThaddeusB (talk) 04:32, 20 November 2009 (UTC)
- Which again shows the system works. Admins review and decline the incorrect ones. However it should be apparent at this point that anecdotal evidence won't benefit this discussion. I'll state it for the 7th or 8th time that if the tags need to be reworked to be cuddlier that can be done. It seems like some people are pushing for some kind of sweeping policy change with zero useful evidence to back it up. But really so what if articles are incorrectly tagged? No one has shown that it has any significant impact on new users. No one has been able to demonstrate there is a problem here at all.--Crossmr (talk) 04:57, 20 November 2009 (UTC)
- I agree that this test wasn't scientific or all that useful. However, I'm not sure how exactly you can prove people leave (or don't) b/c their article was incorrectly speedy tagged. People who leave and never come back don't exactly tell us why on their way out. Anyway, I agree that the only way to draw somewhat meaningful conclusions is to look at a large sample of real data. --ThaddeusB (talk) 05:05, 20 November 2009 (UTC)
- No. The best we could do is some kind of exit survey. if a user has set an email address and all their article space contributions are to articles that they created that were deleted improperly I suppose we could attempt an e-mail. It should be easy to rig a bot to check for articles created by new users (say less than X edits) that were tagged for CSD within 24 hours of creation and who have edited in the last 30 days and propose a survey to them just to see how they felt about the experience. There are also a host of other stats we could try and run to get a much better picture of what is happening with CSD. But if anyone here thinks CSD is bitey then I'd invite them over to the template page for redesign/rewording discussion.--Crossmr (talk) 08:06, 20 November 2009 (UTC)
- I agree that this test wasn't scientific or all that useful. However, I'm not sure how exactly you can prove people leave (or don't) b/c their article was incorrectly speedy tagged. People who leave and never come back don't exactly tell us why on their way out. Anyway, I agree that the only way to draw somewhat meaningful conclusions is to look at a large sample of real data. --ThaddeusB (talk) 05:05, 20 November 2009 (UTC)
- Which again shows the system works. Admins review and decline the incorrect ones. However it should be apparent at this point that anecdotal evidence won't benefit this discussion. I'll state it for the 7th or 8th time that if the tags need to be reworked to be cuddlier that can be done. It seems like some people are pushing for some kind of sweeping policy change with zero useful evidence to back it up. But really so what if articles are incorrectly tagged? No one has shown that it has any significant impact on new users. No one has been able to demonstrate there is a problem here at all.--Crossmr (talk) 04:57, 20 November 2009 (UTC)
- Nope, he stated that 2% are incorrectly deleted. Any admin that patrols CSD can testify that the % of incorrect tags is way higher, perhaps as much as 50%. --ThaddeusB (talk) 04:32, 20 November 2009 (UTC)
- Clearly there is a division of opinion as to the amount of mistagging at CSD and therefore whether we have a problem, this test would could establish that. But if anyone has doubts about the scientific robustness of the methodology discussed at the beginning of this thread please explain your concerns - we may be able to address them in the design stage. As for "I see the article wasn't deleted which shows the system works." That's only partially correct, sure the article wasn't deleted incorrectly, but it was tagged incorrectly, and incorrect tags can be very bitey. ϢereSpielChequers 22:07, 19 November 2009 (UTC)
- Then this gets back to the question of making the CSD tags more stringently worded, I guess, because that's clearly not what G11 was designed for, regardless of your "feelings." Shawn in Montreal (talk) 22:30, 19 November 2009 (UTC)
- Good idea but I don't think tweaking the CSD tags alone are going to be enough: I seriously think we need to be more rigorous about tracking how Twinkle is (mis)used. Perhaps admins need to be even more proactive in warning editors who are serially mistagging articles for speedy deletion, and even imposing blocks in extreme cases, for misuse of CSD. After spending a couples of days monitoring and in some cases removing CSD tags, I'm alarmed at how WP seems to be de-evolving into a race to delete, rather than create. CSD is being widely misused in non-blatant cases. Shawn in Montreal (talk) 21:06, 19 November 2009 (UTC)
- A pedant writes - actually, it's 2%, not 0.2%, but even so, it's a very small minority. And as I've said before - the best way to stop people tagging articles for speedy incorrectly is to fix some of the CSD criteria so they're not so easily misinterpreted (what is a claim of importance - how do you define that?) Also, to add CSD categories for the most often badly tagged articles that are clearly non-encyclopedia - namely obviously made up stuff, neologisms, and duplicate pages). Black Kite 12:11, 19 November 2009 (UTC)
- I agree with you: things like copyvios and attack pages do need to be dealt with immediately. And patent nonsense and vandalism, too. But it's the rush -- indeed, the race -- to delete a multitude of potentially notable articles, not within the first 30 days but the first 30 seconds, that may be getting out of control and discouraging newcomers. Shawn in Montreal (talk) 23:59, 19 November 2009 (UTC)
- They're tagged, not deleted in 30 seconds. The only thing close was an article that was deleted after 2.5 minutes whose content was this [1]. No assertion of notability, nothing. It was also a recreation of previously deleted content. The rest of the articles that were deleted were deleted hours into their life. Not seconds.--Crossmr (talk) 03:13, 20 November 2009 (UTC)
- Mr Zman made a very good point above "The goal seems to be only looking for articles deleted improperly; it should also look at articles that perhaps should have been deleted but weren't." Quite true
wp:NWTwp:NEWT was only ever intended to look at the fate of articles that should not have been deleted. This new test, by looking at the fate of a random group of submitted articles would cover both scenarios - underkill and overkill. ϢereSpielChequers 11:06, 20 November 2009 (UTC)- Except you have zero permission nor mandate to conduct any new test. Frankly I still think a very hard look needs to be taken at the conduct that has led us to here and 6 new page patrollers quitting. What we can do is set up a list of data we'd like to procure from already existing and genuine data that could be gathered by a bot, to see if there is anything even close to a problem. In addition, we can go to the template pages and discuss softening/redesign of the templates themselves if you feel they are too bitey.--Crossmr (talk) 11:37, 20 November 2009 (UTC)
- "set up a list of data we'd like to procure from already existing and genuine data that could be gathered by a bot" Is actually a fairly good summary of the new test proposed on this thread. As for the template review I agree with you on that one though its not one of my personal priorities, and doubt if I will get involved in that myself unless it doesn't happen this year. ϢereSpielChequers 12:01, 20 November 2009 (UTC)
- Except you have zero permission nor mandate to conduct any new test. Frankly I still think a very hard look needs to be taken at the conduct that has led us to here and 6 new page patrollers quitting. What we can do is set up a list of data we'd like to procure from already existing and genuine data that could be gathered by a bot, to see if there is anything even close to a problem. In addition, we can go to the template pages and discuss softening/redesign of the templates themselves if you feel they are too bitey.--Crossmr (talk) 11:37, 20 November 2009 (UTC)
- Mr Zman made a very good point above "The goal seems to be only looking for articles deleted improperly; it should also look at articles that perhaps should have been deleted but weren't." Quite true
- Crossmr - It doesn't look like anyone has refuted this point you've made multiple times, but in actual fact neither "permission" nor "mandate" is required before editors take steps that violate no written policy. Additionally, your repeated assertions that the experiment is useless because it is not "scientific" takes a correct assessment of the methodology and comes to the wrong conclusion - the results can't be extrapolated for statistical purposes, we can't use them to estimate the actual rate of incidence of any particular problem, but they can certainly give people an idea of the problems that are out there. The PARC study and others have suggested problems, we've heard many more anecdotes in blogs and press, and this experiment serves to confirm for Wikipedians the types of difficulties new users might experience. It usefully frames and focuses debate on those processes and policies that could use tweaking. Perhaps more importantly, it gives participants and readers an idea of how folks in this area react even to constructive criticism - allowing participants to mold and refine future efforts to minimize discontent. Nathan T 17:28, 20 November 2009 (UTC)
- If they're talking about creating more disruptive socks to do more tests, they need permission and mandate. They had neither for this experiment. To get a bot to get data on actual deletions, no, they don't need anything for that. But the original post was about using that data as a control against this one and I pointed out that was pointless. Then WereSpielChequers started talking about a new test, and frankly after this I don't really trust him. So I have no idea what he meant by "new test". These results don't do anything to frame anything. The only thing we know for sure about what came from this is that a bunch editors go together and drove out 6 new page patrollers. That is the only thing we can say for certain. Anecdotal evidence does nothing to prove there is any problem here that requires any wide spread policy change. All it does it show someone wants to complain about what happened to them. There are hundreds or possibly thousands of articles created a day. DGG has said just 500 deleted a day I'm sure much more than that created each day. A handful of people complaining because their articles got tagged or deleted means nothing really. DGG pointed out through his analysis that only 2% of deleted articles might have been bad deletions. We'd need to verify that properly, but that isn't alarming. We also have no useful evidence to demonstrate that there is anything overly bitey about the CSD process, but I have suggested that those concerned head to the template page to discuss a redesign. This seems like a very reasonable next step with an utter lack of usable evidence to do anything else. But I haven't seen anyone interested in change over at Template_talk:Db.--Crossmr (talk) 00:56, 21 November 2009 (UTC)
- For clarification. This thread started as a proposal by Mjroots, "how about listing all new articles created in a short timescale and following them in a similar way to the WP:NEWT experiment" my comments in this thread are about that proposed new test. ϢereSpielChequers 04:19, 21 November 2009 (UTC)
- If they're talking about creating more disruptive socks to do more tests, they need permission and mandate. They had neither for this experiment. To get a bot to get data on actual deletions, no, they don't need anything for that. But the original post was about using that data as a control against this one and I pointed out that was pointless. Then WereSpielChequers started talking about a new test, and frankly after this I don't really trust him. So I have no idea what he meant by "new test". These results don't do anything to frame anything. The only thing we know for sure about what came from this is that a bunch editors go together and drove out 6 new page patrollers. That is the only thing we can say for certain. Anecdotal evidence does nothing to prove there is any problem here that requires any wide spread policy change. All it does it show someone wants to complain about what happened to them. There are hundreds or possibly thousands of articles created a day. DGG has said just 500 deleted a day I'm sure much more than that created each day. A handful of people complaining because their articles got tagged or deleted means nothing really. DGG pointed out through his analysis that only 2% of deleted articles might have been bad deletions. We'd need to verify that properly, but that isn't alarming. We also have no useful evidence to demonstrate that there is anything overly bitey about the CSD process, but I have suggested that those concerned head to the template page to discuss a redesign. This seems like a very reasonable next step with an utter lack of usable evidence to do anything else. But I haven't seen anyone interested in change over at Template_talk:Db.--Crossmr (talk) 00:56, 21 November 2009 (UTC)
- Crossmr - It doesn't look like anyone has refuted this point you've made multiple times, but in actual fact neither "permission" nor "mandate" is required before editors take steps that violate no written policy. Additionally, your repeated assertions that the experiment is useless because it is not "scientific" takes a correct assessment of the methodology and comes to the wrong conclusion - the results can't be extrapolated for statistical purposes, we can't use them to estimate the actual rate of incidence of any particular problem, but they can certainly give people an idea of the problems that are out there. The PARC study and others have suggested problems, we've heard many more anecdotes in blogs and press, and this experiment serves to confirm for Wikipedians the types of difficulties new users might experience. It usefully frames and focuses debate on those processes and policies that could use tweaking. Perhaps more importantly, it gives participants and readers an idea of how folks in this area react even to constructive criticism - allowing participants to mold and refine future efforts to minimize discontent. Nathan T 17:28, 20 November 2009 (UTC)
Questions that could be answered with stats
Moving forward on the other issue of dealing with analyzing CSD and what questions could be answered by gathering statistics. These are a list of things that I think might be helpful in finding out.
- How many articles are created each day on average
- How many articles are deleted under CSD each day on average
- How many articles are tagged for CSD but are not deleted
- How many articles are created by new (less than 100 edits) editors are tagged each day?
- How many articles are created by new editors are deleted each day?
- How many new editors that have articles tagged but not deleted make edits to other articles at least 1 week after the tagging?
- How many new editors that have articles deleted make edits to other articles at least 1 week after the deletion?
- How many new editors that have created articles but that were neither tagged nor deleted make edits to other articles at least 1 week after deletion?
Depending on how many samples we get in these categories it might give us some idea of how tagging or deletion affects people's contributions to wikipedia. Something important to distinguish is between normal new users who only ever make a couple edits and that is it and editors that might potentially be chased away. I've been to plenty of articles where I see a new user had made some perfectly fine contributions at one point, perhaps was even welcomed, and 6 months a year later they still hadn't made any further contributions.--Crossmr (talk) 05:52, 21 November 2009 (UTC)
- Q1 Wikipedia:Size_of_Wikipedia and Wikipedia:Modelling Wikipedia's growth have stats on the growth in the number of articles on Wikipedia and in particular the slackening in its growth as fewer new articles are now being added, but this is net growth in articles and excludes deleted ones (and some of the graphs stop in 2007). Q3 Articles tagged for CSD but not deleted will include both incorrect tags and salvaged articles, I would be very cautious about combining them together. Q5 Is unlikely to be helpful unless you can subdivide them by deletion type, and controversially by whether these deletions were "correct" or not. Also if you want to measure things on a daily basis you need to be aware that the daily figures vary on a weekly cycle, comparing a Saturday and a Wednesday could be misleading. ϢereSpielChequers 12:55, 22 November 2009 (UTC)
- Hence the need for gathering proper sample sizes and analyzing them. Yes, some of those might need to be subdivided, but I would suggest that until we have stats for this we can even begin to suggest there is any problem with the CSD process at all in driving away new users.--Crossmr (talk) 13:52, 22 November 2009 (UTC)
- Well the stats do seem to show a fall in the number of new articles accepted. I'm interested in knowing whether this is a sign of the project reaching maturity or of increased barriers to participation (or quite possiblly a bit of both). As for whether the CSD process is driving away new users, IMHO one key statistic is the percentage of articles that don't merit deletion but that are deleted. We had an estimate above that 2% of articles are incorrectly deleted. I'm more than happy for the attack pages, CVs and biographies of non-notable pets to be deleted. But I would like to know; When a newbie submits a valid article, what is the chance of it being deleted? ϢereSpielChequers 14:58, 22 November 2009 (UTC)
- The problem with that is that its extremely difficult to create a definition of "valid article" that isn't just based on the feelings of whomever is compiling the statistics. If such a thing could be determined objectively, then we wouldn't need CSD, and we might not even need humans to patrol new pages. Mr.Z-man 18:15, 22 November 2009 (UTC)
- Well yes I can agree that would be difficult. But for the purpose of a test like this I would hope it would be possible to categorise a group of articles into three: Delete, Keep, and disputed. Though that would require enough admin volunteers to have more than one set of eyes on each article. ϢereSpielChequers 18:34, 22 November 2009 (UTC)
- I'd like to keep the subjectiveness to a minimum as much as possible. We should only collect data on users who created articles which weren't blatantly disruptive. Attack pages, or random nonsense, other languages, etc. Only data on users who had articles turfed for being non-notable or unsourced. I'm not sure whether users who promote should be included. If someone is coming here to copy and paste their bands bio, or companies pr release, chances are their interest in wikipedia is related solely for promotion and not building an encyclopedia, so even if they weren't turfed its unlikely they'd ever do anything but edit related articles, and those numbers could skew any results. Basically I want to see if there is a significant difference in return editing by people who have had articles turfed for being non-notable vs those who haven't. If we have a hint of difference, then we might want to take a harder look and see of those people who had articles tagged and turfed, were they here to promote something or were they making a genuine article we needed. That will require some subjectiveness, but I hope we can at least get through a first round of deciding if there is a possibility of a problem before we get to the subjectiveness.--Crossmr (talk) 03:02, 23 November 2009 (UTC)
- The problem with that is that its extremely difficult to create a definition of "valid article" that isn't just based on the feelings of whomever is compiling the statistics. If such a thing could be determined objectively, then we wouldn't need CSD, and we might not even need humans to patrol new pages. Mr.Z-man 18:15, 22 November 2009 (UTC)
- Well the stats do seem to show a fall in the number of new articles accepted. I'm interested in knowing whether this is a sign of the project reaching maturity or of increased barriers to participation (or quite possiblly a bit of both). As for whether the CSD process is driving away new users, IMHO one key statistic is the percentage of articles that don't merit deletion but that are deleted. We had an estimate above that 2% of articles are incorrectly deleted. I'm more than happy for the attack pages, CVs and biographies of non-notable pets to be deleted. But I would like to know; When a newbie submits a valid article, what is the chance of it being deleted? ϢereSpielChequers 14:58, 22 November 2009 (UTC)
- Hence the need for gathering proper sample sizes and analyzing them. Yes, some of those might need to be subdivided, but I would suggest that until we have stats for this we can even begin to suggest there is any problem with the CSD process at all in driving away new users.--Crossmr (talk) 13:52, 22 November 2009 (UTC)
Question to Werespielchequers
- This is a break-away from Control Test that went off-topic for that section
- At this point we're 6 days in and you've pretty much made sure as not to comment on the behaviour of this group to this point. You haven't provided anything to prove a single benefit from what went on here and the result has been a lot of ill will, 6 new page patrollers quitting, and even though you've removed the articles from the list, some inappropriate articles were created by people in this group. Since you continue to ignore those points in your replies, I'm going to assume you agree with that assessment.--Crossmr (talk) 04:55, 21 November 2009 (UTC)
- For Christ's sake, just step away, take ten deep breaths, and come back to the computer. Goodness. I haven't read a post by you on this talk page that isn't laden with hate-filled wrath directed at NEWT and/or a user involved with the project. I can somewhat understand the former, but frankly I am sickened and quite fed up with you ABF'ing established users and administrators. It is not going to get you anywhere, so please calm down and discuss your thoughts civilly. Thank you. —Ed (talk • majestic titan) 05:11, 21 November 2009 (UTC)
- At this point we're 6 days in and you've pretty much made sure as not to comment on the behaviour of this group to this point. You haven't provided anything to prove a single benefit from what went on here and the result has been a lot of ill will, 6 new page patrollers quitting, and even though you've removed the articles from the list, some inappropriate articles were created by people in this group. Since you continue to ignore those points in your replies, I'm going to assume you agree with that assessment.--Crossmr (talk) 04:55, 21 November 2009 (UTC)
(outdent)I'm plenty calm thank you. You might want to try your own assumptions of good faith. I think this project was a giant mistake, and in calling it out I feel I'm doing what is best for wikipedia. it has damaged new page patrol, and the person responsible for it seems unwilling to defend or discuss its actions. There is nothing uncivil about pointing out facts. There is no bad faith assumption about what I said. If you can dispute any of these facts, please do so.
- The users involved in this "experiment" created alternate accounts with neither permission nor mandate
- The tests were carried out under the assumed bias that what new page patrollers was doing was a problem
- Several users have felt some of the articles created were borderline cases
- some users involved in this project created articles which were inappropriate for wikipedia
- 6 New Page patrollers have quit over this groups actions
- The group has provided nothing usable that could be used to actually form any conclusions about the process
I've taken plenty of deep breaths over the last 6 days while I've waited for WSC to explain or give some justification for what went on here. But the best we've gotten is one user saying "Some people do this and some people do that". Which was as true before it began as it was after. I'm sorry, but I just don't roll over and accept it when a bunch of people get together and chase away a bunch of good faith users from a project that needs users.--Crossmr (talk) 05:22, 21 November 2009 (UTC)
- Cossmr states here:
- "There is no bad faith assumption about what I said."
- Cossmr stated above:
- "I said this "experiment" was [assuming bad faith]." 11:59, 19 November 2009.
- "Problem assumes the NPP are doing something wrong and that assumes bad faith." 08:44, 17 November 2009.
- Ikip (talk) 20:59, 22 November 2009 (UTC)
- Cossmr states here:
- Apologies if that seems abrasive, but I feel what went on here was pretty abrasive as well. I think regardless of their intentions before this started to which I've assumed nothing, there has been very serious consequences for what the people involved in this "experiment" did. So while some people might want to move forward and just pretend that didn't happen, I cannot in good conscience and thinking about the good of wikipedia just pretend or ignore what the editors involved here did.So while I will happily tone it down, let's not pretend that there aren't still questions here that need answers and those questions have nothing to do with CSD.--Crossmr (talk) 06:20, 21 November 2009 (UTC)
- I'm still not clear as to why you say people need 'permission or mandate' to create alternative accounts. Obviously if some people used those accounts to create inappropriate articles then that's abusive, but needing permission just to create such accounts is a separate question - and I can't see any reason one would need special permission. Olaf Davis (talk) 13:26, 21 November 2009 (UTC)
- Because if you look under Wikipedia:SOCK#Legitimate_uses_of_alternate_accounts You won't see anything that covers this situation. not to mention that the editors are intentionally trying to make "bad" edits and behave as noobies and make mistakes to appear new and fool New page patrollers into believing they are so as to be treated as a new editor would be treated. For a long-term editor to start doing that, is in my opinion, disruptivie in itself to begin with. We had editors intentionally introducing formatting and spelling errors when they obviously knew better. And it is quite likely as it happened this time that editors involved in this process may end up creating inappropriate articles which happened a few times. So in order for any kind of further tests to go forward in that vein they would need the permission of the community to permit editors to edit in that environment since it would be disruptive. This "experiment" didn't have permission and there were editors who were part of it that were disruptive. Even the simple fact that legitimate uses doesn't cover this scenario makes their use of the socks disruptive in itself.--Crossmr (talk) 14:12, 21 November 2009 (UTC)
- That's not entirely correct. For years the section WP:SOCK#LEGIT started thus: "Multiple accounts have legitimate uses. For example, prominent users might create a new account to experience how the community functions for new users." Some time this year the passage disappeared, and it was recently restored in slightly more general form: "Alternate accounts have legitimate uses. For example, [...] longterm users might create a new account to experience how the community functions for new users." Hans Adler 14:41, 21 November 2009 (UTC)
- Experiencing how the community functions for a new user and impersonating a new user are two different things. To me the first one implies changes might be made to the system so that there are different welcome pages, sign-ups, etc that might exist. You also don't need to intentionally introduce spelling errors and formatting errors to experience the system as a new user. Also as I pointed out below the first 4 users I checked all clearly violated the policy by editing the articles with both accounts. It doesn't exactly give me the confidence that they fully understood WP:SOCK before they began this.--Crossmr (talk) 14:46, 21 November 2009 (UTC)
- That's not entirely correct. For years the section WP:SOCK#LEGIT started thus: "Multiple accounts have legitimate uses. For example, prominent users might create a new account to experience how the community functions for new users." Some time this year the passage disappeared, and it was recently restored in slightly more general form: "Alternate accounts have legitimate uses. For example, [...] longterm users might create a new account to experience how the community functions for new users." Hans Adler 14:41, 21 November 2009 (UTC)
- And if you further read WP:SOCK you'll note that anyone with an alternate account cannot edit the same page with both accounts. Contributing to the same page with multiple accounts: Editors may not use more than one account to contribute to the same page or discussion. And yet, here[2] we have Skomorokh doing so, here [3] is WereSpielChequers violating policy, NVO here [4], and here [5], Stormie here [6], etc..etc all violations of standing policy. This is disruptive.--Crossmr (talk) 14:19, 21 November 2009 (UTC)
- You have hijacked a section that was supposed to discuss a way of getting data without using socks. You insist on talking in this section exclusively about ways of getting information that involve socks. That is disruptive. Hans Adler 14:44, 21 November 2009 (UTC)
- Actually it was proposed as how to gather data to compare against the data gathered by socks. Since the current data gathered by socks isn't scientific and Werespielfchequers was talking about new tests, I felt it relevant to mention that. I'm happy to move this part to another section, I have no objections to that. But regardless of what section its in, it doesn't change the fact that all of those users violated the policy.--Crossmr (talk) 14:50, 21 November 2009 (UTC)
- They did not. They did what they did in good faith in order to help Wikipedia. They discussed it openly before they did it. And they notified Arbcom. The policy did not explicitly cover what they were doing, it said neither that it's allowed nor that it isn't. You happen to believe that it shouldn't be allowed, and your interpretation of the policy is influenced by that. A lot more people believe it should be allowed, and their interpretation of the policy is influence by that.
- Your examples are not real examples of abuse either. "Editors may not use more than one account to contribute to the same page or discussion." This rule obviously has unstated restrictions. Otherwise an operator of a bot account would not be allowed to edit any page that their bot has operated, and vice versa. Same problem for accounts used by admins for security reasons when working from a public computer. The precise extent of this rule is worth discussing, but it seems clear to me that starting an article with a legitimate sock, and then cleaning up with a reference to this project in the edit summary, is not what is meant by this rule. Some editors didn't mention the project, and I accept that this takes them into a borderline zone. But it was still done in good faith. Hans Adler 15:08, 21 November 2009 (UTC)
- These are not bot accounts, nor are they security accounts. many people have complained about what they've done and 6 new page patrollers quit over it, so I think you'll find that lots of people also disagree with it not just me. While I never said they started the project in bad faith, the results haven't been good. There is no useful data, 6 NPP have quit, some of the socks were used inappropriately, and there are a lot of "borderline" things happening all over that created numerous issues. Acting in good faith doesn't excuse people indefinitely. I might not be angry at them because I assume they were trying to help, but they still have to be responsible for the consequences of what happened.--Crossmr (talk) 15:26, 21 November 2009 (UTC)
- I suspect we are going to have to disagree as to how useful the results are, I think we have a fascinating amount of information, and I've learned a lot from this test. As for WP:SOCK, the relevant part of the policy is "Contributing to the same page with multiple accounts: Editors may not use more than one account to contribute to the same page or discussion in a way to suggest that they are multiple people. Contributions to the same page with legitimate alternate accounts is not forbidden (e.g. editing the same page with your main and public computer account or editing a page using your main account that your bot account edited)." So the question is looking at User:Dahsun or the edit summary from this edit, do the edits by User:WereSpielChequers look like they were trying to hide a connection to Dahsun or WP:NEWT? I honestly don't think so, but clearly they do to Crossmr, so I've now edited User talk:Dahsun to add a link there. ϢereSpielChequers 17:25, 21 November 2009 (UTC)
- You haven't provided anything to back up your assertion that this was useful. Only opinion and anecdotal evidence. You can't argue policy change on anecdotal evidence. I still don't agree that these socks were legitimate accounts. Experiencing the site as a new user is not the same as impersonating a new user to the degree that some users introduced bad edits and inappropriate articles to wikipedia, that is disruptive. So is chasing people away from a much needed task. Since 6 people were chased away and we have actual clear evidence of that, the actions taken by the users of this project while using socks was clearly disruptive, regardless of their intent. And I would also point out that that description was just changed by Sowhy [7] without discussion. Sowhy violated the policy at [8], [9], and [10]. While I agree that if a bot edits an article the admin still can. Bot edits are autonomous. A criteria is set up and they edit lots of articles. In this case you and several other editors deliberately edited the articles with two accounts. There is no exception in the existing policy for that.--Crossmr (talk) 01:22, 22 November 2009 (UTC)
- I suspect we are going to have to disagree as to how useful the results are, I think we have a fascinating amount of information, and I've learned a lot from this test. As for WP:SOCK, the relevant part of the policy is "Contributing to the same page with multiple accounts: Editors may not use more than one account to contribute to the same page or discussion in a way to suggest that they are multiple people. Contributions to the same page with legitimate alternate accounts is not forbidden (e.g. editing the same page with your main and public computer account or editing a page using your main account that your bot account edited)." So the question is looking at User:Dahsun or the edit summary from this edit, do the edits by User:WereSpielChequers look like they were trying to hide a connection to Dahsun or WP:NEWT? I honestly don't think so, but clearly they do to Crossmr, so I've now edited User talk:Dahsun to add a link there. ϢereSpielChequers 17:25, 21 November 2009 (UTC)
- These are not bot accounts, nor are they security accounts. many people have complained about what they've done and 6 new page patrollers quit over it, so I think you'll find that lots of people also disagree with it not just me. While I never said they started the project in bad faith, the results haven't been good. There is no useful data, 6 NPP have quit, some of the socks were used inappropriately, and there are a lot of "borderline" things happening all over that created numerous issues. Acting in good faith doesn't excuse people indefinitely. I might not be angry at them because I assume they were trying to help, but they still have to be responsible for the consequences of what happened.--Crossmr (talk) 15:26, 21 November 2009 (UTC)
- Actually it was proposed as how to gather data to compare against the data gathered by socks. Since the current data gathered by socks isn't scientific and Werespielfchequers was talking about new tests, I felt it relevant to mention that. I'm happy to move this part to another section, I have no objections to that. But regardless of what section its in, it doesn't change the fact that all of those users violated the policy.--Crossmr (talk) 14:50, 21 November 2009 (UTC)
- You have hijacked a section that was supposed to discuss a way of getting data without using socks. You insist on talking in this section exclusively about ways of getting information that involve socks. That is disruptive. Hans Adler 14:44, 21 November 2009 (UTC)
- Because if you look under Wikipedia:SOCK#Legitimate_uses_of_alternate_accounts You won't see anything that covers this situation. not to mention that the editors are intentionally trying to make "bad" edits and behave as noobies and make mistakes to appear new and fool New page patrollers into believing they are so as to be treated as a new editor would be treated. For a long-term editor to start doing that, is in my opinion, disruptivie in itself to begin with. We had editors intentionally introducing formatting and spelling errors when they obviously knew better. And it is quite likely as it happened this time that editors involved in this process may end up creating inappropriate articles which happened a few times. So in order for any kind of further tests to go forward in that vein they would need the permission of the community to permit editors to edit in that environment since it would be disruptive. This "experiment" didn't have permission and there were editors who were part of it that were disruptive. Even the simple fact that legitimate uses doesn't cover this scenario makes their use of the socks disruptive in itself.--Crossmr (talk) 14:12, 21 November 2009 (UTC)
- I'm still not clear as to why you say people need 'permission or mandate' to create alternative accounts. Obviously if some people used those accounts to create inappropriate articles then that's abusive, but needing permission just to create such accounts is a separate question - and I can't see any reason one would need special permission. Olaf Davis (talk) 13:26, 21 November 2009 (UTC)
- Obviously others disagree with you on the need for permission/mandate, and additionally on whether the results of this project were useful. You're essentially filibustering the process now with irritable comments, repeating yourself as though that is going to be persuasive. A few points:
- It's unfortunate that some people were uncomfortable with the project, particularly with any criticism or identification of errors. A defensive response is human nature, we should expect and plan for it, but it isn't an ideal reaction to criticism and doesn't mean that criticism should be barred.
- The results don't need to be scientific to be useful, a point you haven't addressed.
- The project, its format and its results don't need your endorsement in order to provide a teachable moment both for some patrollers and for people interested in the "newbie" experience here.
- Your approach of pure criticism is starting to wear thin; if you want to have an impact, you might consider a different strategy or a break from the discussion. Nathan T 01:50, 22 November 2009 (UTC)
- and obviously some people agree. Good faith doesn't mean you can break all kinds of rules. It doesn't mean you can reword a policy mid-discussion to cover what you did in the past. Without being scientific they aren't useful. They do nothing more than provide opinion. We all had opinions before this started. I never said they needed my permission. I said they need the communities permission to create socks which violated policy, which some clearly did. Both in creating inappropriate articles and causing disruption. Sorry but not one person here has defended that driving away 6 new page patrollers wasn't disruptive.i've also suggested plenty of things that can be done. Did you miss my questions section above? Everyone wanted to talk about that, I proposed some questions, no one has responded. I also suggested rewording the templates, so no I haven't had a pure criticism approach. In fact I even went ahead and started a section on the page after linking it Template_talk:Db-meta#Making_the_template_more_inviting, no here seemed interested in contributing to it thus far.--Crossmr (talk) 01:58, 22 November 2009 (UTC)
- "Without being scientific they aren't useful. They do nothing more than provide opinion. We all had opinions before this started." I don't see that this follows. Five years ago I knew very little about how Wikipedia works; now I know a great deal more. That's not the result of any scientific test on my part, but of simple observation and experience which have been processed by a horrendously un-scientific and biased machine called the human brain. I could equally say "all I have now are opinions about Wikipedia but I had opinions about it five years ago so the experience was useless", but that's clearly not true. Your argument that a scientific test would have been more use than a non-scientific one is certainly reasonable, but there is no a-priori reason that a test needs to be scientific to be useful. Most fact-finding processes engaged in by humans are extremely unscientific, but many are useful nonetheless. Olaf Davis (talk) 18:54, 22 November 2009 (UTC)
- Obviously others disagree with you on the need for permission/mandate, and additionally on whether the results of this project were useful. You're essentially filibustering the process now with irritable comments, repeating yourself as though that is going to be persuasive. A few points:
- I completely agree with Nathan. Did you also notice that you are the only one who keeps insisting we need a "mandate" etc.?
Please stop this borderline trollingSee below —Ed (talk • majestic titan) 21:04, 22 November 2009 (UTC); saying that it is "starting to wear thin" is an understatement—it's just plain annoying. Is it possible for you to contribute constructively to these discussions? Thanks. —Ed (talk • majestic titan) 06:24, 22 November 2009 (UTC)- Edit: yes I do see the questions above, but you are still responding in this section rather abrasively. Lastly, at the six NPP's: please correct me if I am wrong, but didn't four of them say that they would come back when this project is over anyway? That's different than retiring. —Ed (talk • majestic titan) 06:27, 22 November 2009 (UTC)
- Edit 2: also, arguing that this was against policy wastes words, considering what is at WP:SOCK#LEGIT. How else does one experience being a noob unless they pretend to be a noob? @"Good faith doesn't mean you can break all kinds of rules" - why is there a page entitled "ignore all rules" then? —Ed (talk • majestic titan) 06:31, 22 November 2009 (UTC)
- Now I'm a troll? your last post wasn't overly civil, and now this. Perhaps you should take your own advice and step back. The editors involved in this all want the community to extend them lots of good faith and let them be, and yet anyone who questions them is insulted, has bad faith assumed of them, and is generally attacked in a variety of ways. IAR isn't a shield. You can't just invoke it and think it ends a discussion. IAR means you can edit in good faith without needing to know the particulars. You are however, still responsible for your actions, and if you're called on them you have to defend them. So far, if we take out all the insults and beating around the bush, the only thing we've had explained is that the results of this tell us some people do one thing and some people do another thing. That is it. That is all that I've seen offered here. For a test that drove away 6 new page patrollers that is frankly disgusting. The only other defense of the points I've brought up has been one member's introduction of a rewording to the policy to make sure it covers one of the things that went on here. Even if 4 come back, assuming they can believe these tests are genuinely over, you've still lost 2 permanently and for the time being 4 have stopped. That's 6 editors right now, not doing the job they did because of what the editors here did. Pretending to be a new user doesn't mean going out of your way to intentionally be disruptive with your edits. not every new user is disruptive. There is a big difference. You can be a "new" user simply by creating a new account. It doesn't require you make inappropriate articles or edit those new articles in a disruptive manner by intentionally creating bad articles.--Crossmr (talk) 07:44, 22 November 2009 (UTC)
- Edit 2: also, arguing that this was against policy wastes words, considering what is at WP:SOCK#LEGIT. How else does one experience being a noob unless they pretend to be a noob? @"Good faith doesn't mean you can break all kinds of rules" - why is there a page entitled "ignore all rules" then? —Ed (talk • majestic titan) 06:31, 22 November 2009 (UTC)
- Edit: yes I do see the questions above, but you are still responding in this section rather abrasively. Lastly, at the six NPP's: please correct me if I am wrong, but didn't four of them say that they would come back when this project is over anyway? That's different than retiring. —Ed (talk • majestic titan) 06:27, 22 November 2009 (UTC)
- Cossmr, to effectively make changes as you propose, and inspire editors to help them mount a study, usually a person has to be more toned down and neutral in his responses. I would suggest letting other editors who are more selective in what they say proceed with this. Ikip (talk) 21:32, 22 November 2009 (UTC)
Accusations of trolling against User:Crossmr
As one of your involuntary guinea-pigs, I am breaking my silence because you guys have crossed the line here. User:Crossmr is simply asking the questions that need to be answered and his insistence is not "trolling" but simply pointing to what must be terribly uncomfortable and embarrassing for most of you. Fact is that he is right, and most of you are wrong. Your dance around the hot pot reeks of evidence for the fear that should you admit to your trespassing you could well end up being held accountable for it.
The vehement denial that something about this so-called "experiment" was inherently unethical and morally wrong is reminiscent of the numerous unblock requests along the lines of "Me? No! Idindoitwasntme!". "Permission or mandate" are indeed required and essential for the type adventure this group has embarked on; if you believe and keep arguing they aren't, you are giving a future carte blanche to every vandal and any disruptive editor who could and probably will subsequently claim to be conducting an inquiry for the benefit of wikipedia. Of those who understand that, User:Crossmr is the only one with the nerve to participate in this discussion. Thus far, I have kept my mouth shut because the wikilaywering that has been going on here over the past days was completely foreseeable. Leave Crossmr alone and take a good hard look at yourself/ves. Disgusting isn't even strog enough a word. Choyoołʼįįhí:Seb az86556 08:54, 22 November 2009 (UTC)
- Most of the criticism by Crossmr is valid – the problems mentioned are part of the reason why the WP:NEWT project has stopped creating articles, and it looks like the questions are now being answered. I can only see one accusation of trolling, which is is a reference to Crossmr's claims that the project did not have permission or mandate (things that you have also mentioned). If you think that permission and mandate should be necessary, maybe you could propose a policy or guideline for it, as it doesn't appear to be disallowed by any policy that currently exists. snigbrook (talk) 14:14, 22 November 2009 (UTC)
- I also object to the characterisation of Crossmr as a 'borderline troll'; I think (s)he's raising perfectly good-faith objections. But I don't think your post is justified, Seb az86556. Perhaps those of us who don't see anything inherently unethical about NEWT are fools for thinking that; if someone convinces me of that then I'll thank them for setting me straight. But I doubt anyone's going to change their mind because you accuse them of dancing around in fear of admitting trespassing, and I don't see any evidence above that people are doing that rather than defending a genuine (but perhaps woefully mistaken) point of view.
- Let's discuss the merits of the project, not the motives of the discussers, if we can. Olaf Davis (talk) 19:07, 22 November 2009 (UTC)
- Holy cow, my comment was misinterpreted slightly, though I can (now) see how it was. I wasn't attempting to refer to Crossmr as a "troll"; it was an attempt to describe what it looked like he was doing on this page... but it was also very poorly phrased. My apologies.
- On another point: Olaf, thank you for the wonderfully phrased post above. I agree wholeheartedly with it, and will try to disengage from Crossmr. Regards, —Ed (talk • majestic titan) 21:04, 22 November 2009 (UTC)
- User:Seb az86556, by "breaking your silence" do you mean breaking your silence since you posted an ANI 7 days ago, on 15 November 2009? Which was moved here by Thaddeus, Wikipedia_talk:Newbie_treatment_at_CSD#Magic_pen_.28software.29_.26_Huntsville_.28game.29? In which you said "That's what I am wondering as well." when Cossmr said: "if you've got direct evidence of all this and it seems that page is quite clear, these user are not blocked why?" Ikip (talk) 21:09, 22 November 2009 (UTC)
- I consider disengaging for a week something very commendable, if that's what happened. (I didn't make any research, just responding to what is stated here.) Hans Adler 21:14, 22 November 2009 (UTC)
- I meant the latter. The involuntary discovery of this "experiment" had upset me to such a degree that I deemed it likely for me to engage in serious incivility; I thus quit English wikipedia for an unknown amount of time to gain distance from it. I believe the necessary amount of time has passed. Having stayed away for a week, I am naturally calmer, yet stand by my point that this project is completely repulsive and, to repeat, unethical. Does that answer your question?
- User:Seb az86556, by "breaking your silence" do you mean breaking your silence since you posted an ANI 7 days ago, on 15 November 2009? Which was moved here by Thaddeus, Wikipedia_talk:Newbie_treatment_at_CSD#Magic_pen_.28software.29_.26_Huntsville_.28game.29? In which you said "That's what I am wondering as well." when Cossmr said: "if you've got direct evidence of all this and it seems that page is quite clear, these user are not blocked why?" Ikip (talk) 21:09, 22 November 2009 (UTC)
- Secondly, I have been asked to explain my use of the word "unethical" (see above). Though my explanation might seem too lofty and philosophical for the scope of this project, I will try to elaborate since I cannot think of any other way:
- In ethics, a moral agent (the wikipedia article has a "multiple-issues"-tag so I won't link to it) is a thinking being that has been giving the opportunity to make autonomous decisions while being fully aware of the situation and circumstances at hand. Not making a person aware of all variables of his/her situation is equal to lowering that person's status to less than that of a moral agent. Only animals, newborns, and the feeble-minded aren't regarded moral agents and can thus be deprived of becoming fully aware of all circumstances regarding their actions. Treating individuals who are not members of these groups in such a way is thus unethical.
- In every "experiment," be it in psychology, sociology, or any other related fields, one has the moral obligation to inform all participants of their circumstances, as well as giving them a choice not to become part of it. While it is not required to explain the nature or methodology of the test, a phrase like "This is an experiment, would you like to participate?" is elementary. Every psychology-student in the first semester learns this right away.
- In the course of this and other discussions (which I have been following without commenting), there were attempts to draw parallels to companies spying on their employees or teachers tricking their students. This point is easily refuted: I can think of only few companies whose standards include being the torch of morality in the dark abyss of indecency; rather, companies will look into and exploit any legalistic loophole to set up sweatshops in Southeast Asia or dump oil into rivers without telling the local population about it. As for teachers, their students are forced at gunpoint (metaphorically) to sit in the classroom and are helpless victims of such trickery.
- While there are no wiki-policies against behaving unethically, users of this site should not act in the spirit of the aforementioned companies. As the saying goes "just 'cause it's legal, doesn't mean it's right." Both here and in the real world, I am amazed that some people need laws and regulations in order to become aware of an appropriate amount of decency and ethical behavior. I therefore see the above suggestion of my engaging in an attempt to change policy as futile. The only sentence I recall that could be changed is the rule that says "there is no 'common sense' on wikipedia." I always understood it as "there is no 'common sense' on wikipedia with regards to the content of articles and the claims made therein; however, do not abandon your 'common sense' and general moral standards for the rest of this project." Choyoołʼįįhí:Seb az86556 23:17, 22 November 2009 (UTC)
- Your theory about all experiments needing to satisfy certain specific criteria makes no sense. In the real world there is no clear demarcation between experiments and non-experiments. You don't need the agreement of everybody who might be in New York tomorrow if you want to make an experiment to find out whether people there are likely to donate to beggar who is dressed like a banker. You don't need to consult an ethics commission before you launch an April Fools' Joke. Candid Camera type programmes are not necessarily unethical without asking their victims for general permission first. Hans Adler 23:46, 22 November 2009 (UTC)
- Request for clarification: Candid Camera is entertainment. April Fools' Jokes are jokes. Do you see WP:NEWT as an entertainment program or a joke? (In that case, I'd be willing to laugh and forget about it.)Choyoołʼįįhí:Seb az86556 00:19, 23 November 2009 (UTC)
- The New York example is obviously not meant as a joke, the others were just things that came to my mind when thinking about well known "experiments" that contradict you. As a mathematician I find it hard to believe that someone could seriously think that exactly the same thing could be ethical if you do it as a joke and unethical if you do it in order to gather experience so that you can improve a process. Yes, Candid Camera done as medical or psychological research would need approvement by an ethics commission. But that's because of the general rules for research in those high-risk fields. I deny that for experiments with Wikipedia more ethics approvement is needed than the usual on-wiki peer review before starting them, with reasonably large participation. Hans Adler 01:03, 23 November 2009 (UTC)
- Obviously, we do not share the same set of ethical rules. I do think approval is needed. Had it been the type of situation where I was informed in a manner of "here's what we did, think about it and have a nice day," I could have accepted that. Instead, people informed me in a manner of "we are right, you are wrong, your concerns are junk, shut up." The user who trapped me into this has yet to leave a single comment on this page. I will agree to disagree. Choyoołʼįįhí:Seb az86556 01:43, 23 November 2009 (UTC)
- "than the usual on-wiki peer review before starting them" - except of course that NEWT never got that as far as I'm aware. While shows like Candid Camera may not get permission beforehand, you can be sure that they do before they put the results on-air. The same would be the case with the New York example. If you were going to publish anything that could identify one of the participants, you should get their permission first and you certainly should give back any money they give you. I didn't hear of any cases of users being contacted in private before the results of the experiments on them were shared here. Mr.Z-man 02:00, 23 November 2009 (UTC)
- I accept your second point (distinction between experimenting and publishing non-anonymous results) as valid. I didn't think of this aspect since it has no direct connection with WP:SOCK, but it makes sense. By the usual on-wiki peer review I mean that you propose something in a fairly frequented place and get opinions on it, and only proceed if there is a consensus that it's OK to do so. That happened. Whether it should have been announced on WP:CENT (was it?), or whether the result of the peer review was deficient, are other questions. But this wasn't a sneaky project organised by a secret cabal. Hans Adler 13:41, 23 November 2009 (UTC)
- Where did it happen? I don't think I remember seeing it. Mr.Z-man 17:21, 25 November 2009 (UTC)
- The discussion was at Wikipedia:Requests for comment/new users#Lets all create an extra account ϢereSpielChequers 23:55, 25 November 2009 (UTC)
- Where did it happen? I don't think I remember seeing it. Mr.Z-man 17:21, 25 November 2009 (UTC)
Hired as a greeter at Wal-Mart
My question is this, if you were an employer at Wal-Mart, hiring for a greeting position, and you googled some of these editors above, would you hire these editors based on their comments above? (see comments in next section) Criteria for Speedy Deletion and New Page Patrol is usually the very FIRST interaction that the outside world has with wikipedia editing. The abrasive and combative comments above, make me question if it is not a net positive for wikipedia that some of our volunteers are now doing something else.
I am heartened by those editors who were reviewed by this project, and took it as it should be: a good faith evaluation of CSD and a suggestion to be a little more careful.
- Wikipedia_talk:Newbie_treatment_at_CSD#If_you_want_bitey_treatment_of_a_newbie_from_a_new_pages_patroller...
- Wikipedia_talk:Newbie_treatment_at_CSD#One_advantage_of_this_over_passive_observation (please correct me if I am wrong)
Ikip (talk) 21:26, 22 November 2009 (UTC)
- Q. What pays less than greeter at Wal-Mart?
- A. NPP at Wikipedia.
- --Curtis Clark (talk) 23:42, 22 November 2009 (UTC)
- To follow the logic, however, one does'nt hire a VOLUNTEER with anger-management issues to watch young children. 192.58.204.226 (talk) 19:31, 25 November 2009 (UTC)
Personal attacks and other comments
It is obvious there will be no change here, collapsing conversation, welcome to archive
| ||
|---|---|---|
|
I asked User:the_ed17 to apologize for his comments, and he was smart enough to do so, above. Thank you. User:the_ed17 is actually the THIRD person to use the word troll in relationship to this project:
Then there have been all of the other comments, which are not very constructive. Please note, the majority of these comments come from editors who are currently the face of wikipedia, greeting new editors everyday. The vast majority of all negative comments come from editors opposed to this project.
Ikip (talk) 21:27, 22 November 2009 (UTC)
|
Invitation
How about putting this debate into the Signpost in a structured format? I'd like to propose a piece for them along the following lines: get one editor from each side of the NEWT issue into chat (gchat or Skype) and moderate a debate on the merits of NEWT between them for the newsletter. Disclosure: I participated in NEWT before the major complaints really arose. If you're interested please contact me. Looking for one editor from each side. Respectfully, Durova369 01:49, 22 November 2009 (UTC)
- Hi Durova, I'm happy to take part in such a skype call if you organise it. ϢereSpielChequers 08:26, 22 November 2009 (UTC)
an study that was done outside of Wikipedia about editing by new/infrequent editors
An article on a study that was done outside of Wikipedia Wikipedia approaches its limits: The online encyclopedia is about to hit 3m articles in English – but growth is stalling as 'inclusionists' and 'deletionists' fight for control. Excerpts:
- Chi's team discovered that the way the site operated had changed significantly from the early days, when it ran an open-door policy that allowed in anyone with the time and energy to dedicate to the project. Today, they discovered, a stable group of high-level editors has become increasingly responsible for controlling the encyclopedia, while casual contributors and editors are falling away. Wikipedia – often touted as the bastion of open knowledge online – has become, in Chi's words, "a more exclusive place".
- One of the measures the Parc team looked at was how often a user's edit succeeds in sticking. "We found that if you were an elite editor, the chance of your edit being reverted was something in the order of 1% – and that's been very consistent over time from around 2003 or 2004," he says.
- Meanwhile, for those who did not invest vast amounts of time in editing, the experience was very different. "For editors that make between two and nine edits a month, the percentage of their edits being reverted had gone from 5% in 2004 all the way up to about 15% by October 2008. And the 'onesies' – people who only make one edit a month – their edits are now being reverted at a 25% rate," Chi explains.
- In other words, a change by a casual editor is more likely than ever to be overturned, while changes by the elite are rarely questioned. "To power users it feels like Wikipedia operates in the way it always has – but for the newcomers or the occasional users, they feel like the resistance in the community has definitely changed."
- While Chi points out that this does not necessarily imply causation, he suggests it is concrete evidence to back up what many people have been saying: that it is increasingly difficult to enjoy contributing to Wikipedia unless you are part of the site's inner core of editors.
stmrlbs|talk 21:26, 22 November 2009 (UTC)
- Stmrlbs, nice job, another article to add to my list, Wikipedia:Notability/RFC:Reevaluation#Journalists as I wrote above though, these articles will be dismissed completely by some editors, as any evidence, no matter how definitive will be. Attacking the sender: the journalists are ignorant, etc., just as they attack the editors of this project as disruptive, etc. See the predictable reaction I got in posting links to those articles. Ikip (talk) 21:38, 22 November 2009 (UTC)
- Here's another interesting quote from that... "There's no place on Wikipedia that says: 'Want to become a Wikipedia editor? Here's how you do it.' Instead, you basically have to really become part of that community and pick it up through osmosis and have the tradition passed down to you." —Ed (talk • majestic titan) 21:42, 22 November 2009 (UTC)
- WOW that was a really incredible article. One of the best I have ever read on the subject Thank you Ikip (talk) 21:47, 22 November 2009 (UTC)
- The study is trying to figure out why edits have plummeted.[17]:
- ...the total percentage of edits reverted increased steadily over the years. The total percentage of monthly reverted edits...has steadily increased over the years for the all classes of editors...
- ...More interestingly, low-frequency or occasional editors experience a visibly greater resistance compared to high-frequency editors [see the top two reddish lines, as compared to other lines]. The disparity of treatment of new edits from editors of different classes has been widening steadily over the years at the expense of low-frequency editors.
- We consider this as evidence of growing resistance from the Wikipedia community to new content, especially when the edits come from occasional editors.
- PDF:[18]
- The greater resistance towards new content has made it more costly for editors, especially occasional editors, to make contributions. We argue that this resistance may have contributed, with other factors, to the slowdown in the growth of Wikipedia. These data appear consistent with the hypothesis that the “deletionists” may be increasingly outnumbering the “inclusionists” among the administrators. (citing Lih, A. (interview), The Battle for Wikipedia’s Soul, The Economist (magazine), March 6th 2008.)
- Ikip (talk) 21:57, 22 November 2009 (UTC)
- Here's another interesting quote from that... "There's no place on Wikipedia that says: 'Want to become a Wikipedia editor? Here's how you do it.' Instead, you basically have to really become part of that community and pick it up through osmosis and have the tradition passed down to you." —Ed (talk • majestic titan) 21:42, 22 November 2009 (UTC)
- Except that it's missing two massive elephants in the room. (1) The likelihood of sporadic editors edits being revert is clearly going to be higher, because they are likely to be less au fait with policy and guidelines, and are therefore more likely to add spurious, trivial, badly formatted or unsourced material (and that's not counting vandalism and nonsense). (2) Of course the number of new articles is going to slow down. As the number of articles increases, there are obviously going to be fewer notable subjects to write about (or at least, there are in the areas that most editors write about). Given these two quite obvious points, the "study" is stating the obvious. Black Kite 23:53, 22 November 2009 (UTC)
- "Badly formatted or unsourced material" is no reason to revert—just because they don't know anything about the black hole that is MOS or don't know how to cite sources doesn't mean that we should revert them. If it seems plausible, why not take a a second to Google search it? (especially if you are catching the edit through your watchlist, meaning that it is probably to an article you like.) —Ed (talk • majestic titan) 00:41, 23 November 2009 (UTC)
- Yes, and I do that - but I was talking about material that is either original research, synthesis or plainly so trivial it doesn't belong. Furthermore, a lot of badly formatted material is likely to fall into that category too; any edit that breaks a page or is really badly written is likely to be reverted as it degrades the page. OK, many editors will say "Ah, I see what they were trying to do there" and fix it - but a lot will get reverted. Incidentally, unsourced material is a reason for reversion, especially if it's contentious or in a BLP. Black Kite 00:51, 23 November 2009 (UTC)
- Yes, no editor is under any obligation to allow unsourced material to stay on any page. There is no requirement that they tag it and leave it for a week, a month, a year, or attempt to clean it up or find sources for what was added. Especially if the information that is added is of lower quality it might not be readily apparent as to why that information is relevant to the article. It would probably be a good idea for vandal tools to ensure that anyone reverted or warned for something that isn't blatant vandalism like "Kyle's mom sux!" is first welcomed before they're warned.--Crossmr (talk) 02:53, 23 November 2009 (UTC)
- (edit conflict) (@Black Kite) - okay, good. :-) I do understand your other points, but even though unsourced material is a reason for reversion, it doesn't mean that you have to revert, which is what I fear some do (not you, just a general statement). Think of it from an anon's perspective: you add information to an article that has no citations, but you are reverted because your new material is "not sourced". Hypocritical much? :-) —Ed (talk • majestic titan) 03:02, 23 November 2009 (UTC)
- To be honest, if the article's unsourced already, I doubt if anyone's going to revert an addition that's unsourced itself, unless it has other problems. I'd be looking at the article rather than the addition ... Black Kite 18:50, 23 November 2009 (UTC)
- (@ Crossmr) - I completely agree. I am certainly not trying to imply that there is an "obligation" to leave it, but I am attempting to say that any uncontroversial information added about, say, a WWI aircraft doesn't have to be reverted, either. Am I making my point well? I can't really tell if I am or not. :-) —Ed (talk • majestic titan) 03:09, 23 November 2009 (UTC)
- It doesn't have to be reverted. But sometimes I've reverted unsourced information because an article is already in dire straits and more unsourced info isn't going to help us fix the problem (this is usually on articles that already have section/article tags on them indicating that there is a sourcing problem). Sometimes I've left it alone if there isn't a serious issue already existing on the page. I guess the point is if people are under no obligation to leave it, we shouldn't be getting on them for removing it. How they remove it might be another issue, but the removal itself shouldn't be chastised.--Crossmr (talk) 07:18, 23 November 2009 (UTC)
- (edit conflict) (@Black Kite) - okay, good. :-) I do understand your other points, but even though unsourced material is a reason for reversion, it doesn't mean that you have to revert, which is what I fear some do (not you, just a general statement). Think of it from an anon's perspective: you add information to an article that has no citations, but you are reverted because your new material is "not sourced". Hypocritical much? :-) —Ed (talk • majestic titan) 03:02, 23 November 2009 (UTC)
- Yes, no editor is under any obligation to allow unsourced material to stay on any page. There is no requirement that they tag it and leave it for a week, a month, a year, or attempt to clean it up or find sources for what was added. Especially if the information that is added is of lower quality it might not be readily apparent as to why that information is relevant to the article. It would probably be a good idea for vandal tools to ensure that anyone reverted or warned for something that isn't blatant vandalism like "Kyle's mom sux!" is first welcomed before they're warned.--Crossmr (talk) 02:53, 23 November 2009 (UTC)
- Yes, and I do that - but I was talking about material that is either original research, synthesis or plainly so trivial it doesn't belong. Furthermore, a lot of badly formatted material is likely to fall into that category too; any edit that breaks a page or is really badly written is likely to be reverted as it degrades the page. OK, many editors will say "Ah, I see what they were trying to do there" and fix it - but a lot will get reverted. Incidentally, unsourced material is a reason for reversion, especially if it's contentious or in a BLP. Black Kite 00:51, 23 November 2009 (UTC)
- Personally, I'm surprised our article creation rate is as high as it is. As articles become more comprehensive, it also increases the chances that content being added will be trivial, irrelevant, or duplicated. Wikipedia wasn't nearly as visible in 2003-2004 as it is now, I wouldn't be at all surprised if the actual vandalism rate doubled or tripled since then. According to m:History of Wikipedia, the English Wikipedia had 77 million hits in February 2004 and made the Alexa top-40 in November 2005. Last month (according to [19]), the English Wikipedia had more than 6 billion hits and has an Alexa rating of 6. Mr.Z-man 00:43, 23 November 2009 (UTC)
- Though I haven't been editing let alone new page patrolling long enough to compare current spam levels to those of earlier years; I subscribe to the theory that as we are a much more important site than we were a few years ago, and even the most luddite business knows that, the amount of spam has almost certainly increased. Since total edits seem to be fairly stable at ten million every 6 or 7 weeks that implies to me that non-spam edits are falling. If that's the case then unless vandalism and unhelpful edits are falling, total helpful edits are probably less frequent. I see an increased level of reversion of newbie edits as evidence for this theory. But if someone wants to go through a random set of newbie edits to check this theory scientifically I would be open to changing my view based on those results. However to avoid confusion I would suggest doing that as a separate project Wikipedia:Newbie treatment at Recent Changes. ϢereSpielChequers 09:36, 23 November 2009 (UTC)
- "Badly formatted or unsourced material" is no reason to revert—just because they don't know anything about the black hole that is MOS or don't know how to cite sources doesn't mean that we should revert them. If it seems plausible, why not take a a second to Google search it? (especially if you are catching the edit through your watchlist, meaning that it is probably to an article you like.) —Ed (talk • majestic titan) 00:41, 23 November 2009 (UTC)
- Except that it's missing two massive elephants in the room. (1) The likelihood of sporadic editors edits being revert is clearly going to be higher, because they are likely to be less au fait with policy and guidelines, and are therefore more likely to add spurious, trivial, badly formatted or unsourced material (and that's not counting vandalism and nonsense). (2) Of course the number of new articles is going to slow down. As the number of articles increases, there are obviously going to be fewer notable subjects to write about (or at least, there are in the areas that most editors write about). Given these two quite obvious points, the "study" is stating the obvious. Black Kite 23:53, 22 November 2009 (UTC)
Wall Street Journal: Volunteers Log Off as Wikipedia Ages
Mentioned in this month's Signpost: Wikipedia:Wikipedia_Signpost/2009-11-23/In_the_news, excerpts:
"In the first three months of 2009, the English-language Wikipedia suffered a net loss of more than 49,000 editors, compared to a net loss of 4,900 during the same period a year earlier..."
"Today, [Wikipedia]'s rules are spelled out across hundreds of Web pages. Increasingly, newcomers who try to edit are informed that they have unwittingly broken a rule -- and find their edits deleted, according to a study by researchers at Xerox Corp...In 2008, Wikipedia's editors deleted one in four contributions from infrequent contributors, up sharply from one in 10 in 2005, according to data compiled by social-computing researcher Ed Chi of Xerox's Palo Alto Research Center."
"Ultimately, it was decided that Ms. Paley's comics were suitable for the site. Samuel Klein, a veteran Wikipedian who serves on the board of trustees, intervened and restored her contributions. Mr. Klein says experiences like Ms. Paley's happen too often. Mr. Klein says that the Wikipedia community needs to rein in so-called deletionists -- editors who shoot first and ask questions later."
Jimmy Wales: "...if the community has become more hostile to newbies, that's a correctable problem." Ikip (talk) 02:37, 25 November 2009 (UTC)
The data collection phase of this project is now over
Seven days ago I paused the data collection phase of this project as it seems clear to me that there is enough concern about what we were doing that to continue collecting data would be divisive. As there has been almost no objection to that pause, and still some concern from opponents of the project, I have called an end to the data collection phase, and would advise anyone thinking of resuming this to read this talkpage carefully and propose any resumption in such a way as to address the concerns that some have raised. I hope that regardless of our views as to the way in which this data was collected, unless and until someone proposes resuming data collection we can now focus discussions on the implications of that data, rather than its collection method. ϢereSpielChequers 13:31, 23 November 2009 (UTC)
- Good call, I think. Olaf Davis (talk) 13:50, 23 November 2009 (UTC)
- Thank you. As one who is happy to receive constructive criticism on my talk page, but who takes offense at being set up with marginal situations specifically designed to trip up good-faith volunteers, I am glad to see this experiment come to an end. Now that the mines have presumably been cleared, I will resume my contributions in my small way. -- Tom N (tcncv) talk/contrib 17:46, 23 November 2009 (UTC)
- I thought the whole fake newbie thing was just being done to illustrate some kind of point. You cannot seriously propose to actually study any "data" derived from such an unsound (and that is putting it very tactfully) methodology? Any conclusions that anyone tried to draw from this "data" would be so devoid of credibility as to do more harm than good to the goals of this project. CIreland (talk) 09:47, 28 November 2009 (UTC)
Project to save our oldies
I read here [20] that experienced editors and patrollers are leaving Wikipedia in droves. Time to start a new project? *I KID, I KID* <>Multi‑Xfer<> (talk) 04:16, 24 November 2009 (UTC)
- Apart perhaps from the headline, that's actually a fair balanced article. It covers three interrelated phenomena, readership, newbies and retention of veterans. We are still growing in terms of Internet readers, albeit only growing at 20% per annum which is less than the average for the Internet. We have become less welcoming to newbies, and we have a declining volunteer community. Its a good summation of much that we discussed at Wikimania in Buenos Aires.
- Personally I suspect that our growing more slowly than the Internet in general is a combination of our weakness in China and possibly an increase in people accessing our data via mirror sites which don't count in our stats. The decline in the community is the weakest part of their article and an area where I feel we lack research, yes we know that we are no longer recruiting newbies as fast as we lose longer term editors. But we don't know whether the turnover amongst longserving editors has increased, or whether the decline is merely the result of our already high turnover and the driving away of the newbies. This project was a response to the problem of our having become less welcoming to Newbies, and I think we have disproved the wildest claims such as "guaranteed your article will be deleted within 7 days". But we have identified a number of things that could be improved and also some of the things that divide us. To my regret it also lead to friction and I lost the trust of at least one fellow editor. As for starting a new project to reduce our loss of experienced editors, actually we have two: User:Juliancolton/Project and strategy:Main Page - though the latter also covers several other things as well, including the fact that fewer newbies are staying around after their firsts few edits. ϢereSpielChequers 09:36, 24 November 2009 (UTC)
- The only mention of patrollers:
- "In 2005, journalist John Seigenthaler Sr. wrote about his own Wikipedia write-up, which unjustly accused him of murder. The resulting bad press was a wake-up call. Wikipedians began getting more aggressive about patrolling for vandals and blocking suspicious edits, according to Andrew Lih, a professor at the University of Southern California and a regular Wikipedia contributor."
- Most of the article was dedicated to the way that treats new editors, "Increasingly, newcomers who try to edit are informed that they have unwittingly broken a rule -- and find their edits deleted..." I don't see anywhere that experienced editors are leaving. Ikip (talk) 03:07, 25 November 2009 (UTC)
- It was the section "Volunteers have been departing the project that bills itself as "the free encyclopedia that anyone can edit" faster than new ones have been joining, and the net losses have accelerated over the past year. In the first three months of 2009, the English-language Wikipedia suffered a net loss of more than 49,000 editors, compared to a net loss of 4,900 during the same period a year earlier, according to Spanish researcher Felipe Ortega," (my emphasis). So unless someone can dispute that study, or something happens to indicate that things have changed, not only do we have a declining number of active admins but we have a dwindling volunteer community. I don't know whether the turnover of existing editors has increased or not, but it seems to me obvious that if we could slow the loss of existing editors down to the level of new editors joining, it would be one way to stop the community declining in numbers. It would also be very useful to know whether the turnover amongst existing editors has altered. However I agree with you that our main problem seems to be the way we drive newbies away, as though we still have circa 200,000 new editors a month, the small proportion who stick around is actually falling. ϢereSpielChequers 14:04, 25 November 2009 (UTC)
- The only mention of patrollers:
In the comment section of that WSJ article, there is a Henry Grimmelsman who writes:
| “ | I tried to publish an article once about a relative of mine, an uncontroversial, long-dead Catholic bishop, since every other bishop in the history of that diocese had a page on the site. It was deleted for no apparent reason and since then, I haven't had any desire to try to contribute again. | ” |
The bishop in question is presumably Henry Joseph Grimmelsmann, Bishop of Evansville (1944-1965). That article was started 24 September 2009 and has never been deleted. There is, however, a page Henry Joseph Grimmelsman that was deleted 6 April 2008 by User:Jmlk17, stating "A7 (bio): Real person; doesn't indicate importance/significance".
Could an administrator with access to deleted content please check what this article looked like before it was deleted? Deleting an article on a Catholic bishop as not "indicat[ing] importance/significance" seems a bit odd. Jmlk17 seems to have done nothing on Wikipedia since June 2008, so there is no point in asking him about it. --Hegvald (talk) 06:38, 26 November 2009 (UTC)
- Deleting this article under criterion A7 was an incorrect decision. The article is more comprehensive than the existing article, and contains several explicit claims of significance, such as
- In 1920, he was appointed Professor of Scripture and Hebrew ...
- he wrote two ... textbooks for those studying to be priests.
- He also gave the Nihil obstat approval of the 1941 official Catholic version of the New Testament in American English.
- Source: http://wifasso.com/Bishop/Overview.htm
- As expected, it's not perfect. It doesn't maintain an NPOV, and gives its references using prose. Still, it should never have been speedily deleted. decltype (talk) 07:19, 26 November 2009 (UTC)
- Can you add the text to the talk page of the current article to be incorporated? I'll look at it. --IP69.226.103.13 (talk) 10:03, 26 November 2009 (UTC)
- It seems like WereSpielChequers is already on it. I would simply have done a history merge with the existing article to take care of attribution, but there are other means as well. decltype (talk) 10:10, 26 November 2009 (UTC)
- Oops, as it turns out, the deleted article is a substantial copy of [21], which is already used as a reference to the other article, which means that the original should have been deleted under criterion G12 (copyright violation). Grimmelsman complained that the article was deleted "for no apparent reason". If the article had been handled correctly, perhaps things would have worked out. decltype (talk) 10:25, 26 November 2009 (UTC)
- I've Emailed an apology to the Newbie, and restored and redirected the article. Agreed if its Copyvio, that would have been a legit delete, and a history merge is not needed. But that emphasises the importance of using the right speedy criteria. Sometimes copyvios are submissions by the author, and a message that says we respect the copyright is very different from "this subject doesn't belong here". ϢereSpielChequers 10:53, 26 November 2009 (UTC)
- Could someone redelete the copyvio revisions (all except the redirection) as if the article were tagged with {{copyvio-histpurge}}? Thanks in advance. Flatscan (talk) 03:45, 27 November 2009 (UTC)
- Agreed. This also served as a another reminder to myself to not act too hastily in speedy deletion matters, seeing as how my original assessment that the article "should never have been speedily deleted" was incorrect. Under regular circumstances, I probably would have checked this, though, especially for such a large article created directly in mainspace. decltype (talk) 11:02, 26 November 2009 (UTC)
- I've often stepped in to save an article only to have it turn out to be a copyvio, or to be deleted at AFD. I've learned that one has to know when to cut your losses when trying to salvage articles, and I know from my deleted contributions that there've been articles I declined speedies on that subsequently were deleted. If I'm not sure about an article I like to hope that categorising it and fixing typos makes it easier for someone else to make the right call. My concern about NPP is that I fear some patrollers have a "what speedy category does this belong to" mentality. Thats why when I review a candidates contributions at RFA I'm more tolerant of the occasional mistagging if they sometimes use prod, hotcat or other options rather than just speedies. ϢereSpielChequers 12:35, 26 November 2009 (UTC)
- I was mostly just having a bit of a joke with y'all but I agree that this is a very good article. If people are leaving the project, it isn't just because some newbies feel bitten. Experienced volunteers are also quitting, not to mention sysops. If you ask me, it's because of a combination of things: 1) Rampant vandalism that becomes worse rather than better, 2) Deliberate harassment of longstanding editors by people like Grawp, 3) A lack of the project to evolve over time. I do enjoy my time here but I also see Wikipedia as moribund regarding policies, missions, etc. I personally think significant changes need to be made here in a number of areas, but none of this will happen because the community cannot agree on it and it's always easier to just leave it alone rather than evolve. In the discussion forum for the WsJ article, someone makes a very interesting connection between Wikipedia and AOL back in the early days of the Internet. <>Multi‑Xfer<> (talk) 17:46, 26 November 2009 (UTC)
- I've often stepped in to save an article only to have it turn out to be a copyvio, or to be deleted at AFD. I've learned that one has to know when to cut your losses when trying to salvage articles, and I know from my deleted contributions that there've been articles I declined speedies on that subsequently were deleted. If I'm not sure about an article I like to hope that categorising it and fixing typos makes it easier for someone else to make the right call. My concern about NPP is that I fear some patrollers have a "what speedy category does this belong to" mentality. Thats why when I review a candidates contributions at RFA I'm more tolerant of the occasional mistagging if they sometimes use prod, hotcat or other options rather than just speedies. ϢereSpielChequers 12:35, 26 November 2009 (UTC)
- I've Emailed an apology to the Newbie, and restored and redirected the article. Agreed if its Copyvio, that would have been a legit delete, and a history merge is not needed. But that emphasises the importance of using the right speedy criteria. Sometimes copyvios are submissions by the author, and a message that says we respect the copyright is very different from "this subject doesn't belong here". ϢereSpielChequers 10:53, 26 November 2009 (UTC)
- Oops, as it turns out, the deleted article is a substantial copy of [21], which is already used as a reference to the other article, which means that the original should have been deleted under criterion G12 (copyright violation). Grimmelsman complained that the article was deleted "for no apparent reason". If the article had been handled correctly, perhaps things would have worked out. decltype (talk) 10:25, 26 November 2009 (UTC)
- It seems like WereSpielChequers is already on it. I would simply have done a history merge with the existing article to take care of attribution, but there are other means as well. decltype (talk) 10:10, 26 November 2009 (UTC)
- Can you add the text to the talk page of the current article to be incorporated? I'll look at it. --IP69.226.103.13 (talk) 10:03, 26 November 2009 (UTC)