A single edit could change the accuracy of an article completely but the ratings remain unaltered, so what's the motive? If you want to encourage knowledgeable people to contribute then just say so.
Encouraging people to pass casual, unqualified judgement in this way will at best demonstrate the value of an article according to popular opinion, and for a reference work that could be counterproductive.
This post was posted by VanishedUser sfkdjljakldj, but signed as JoeMCMXLVII.
Ratings expire after (any) 30 edits, so they do not remain "unaltered" forever.
A few editors have said they are offended that mere readers are being permitted to express an opinion on their work, but I can't actually see any of our regular editors screwing up an article just to get a higher rating from "casual, unqualified" readers. Can you?
I'm not offended by you or anyone else. I'm just someone who thinks the facility is irrelevant and for the reasons I stated.
I'll add that 30 edits can be a long time to display obsolete information.
This post was posted by VanishedUser sfkdjljakldj, but signed as JoeMCMXLVII.
I agree that any set number of edits is a limited approach. Three edits might result in major changes; thirty edits might be nothing more than 15 pairs of vandalism and reversion. Furthermore, half the ratings might have been made while one of those 15 vandalized versions were up.
I don't agree that the tool is irrelevant to the broad goals of the WMF. I think it may prove to be far more effective at turning readers into editors than anything else we've tried. Passively reading "This page needs to be fixed" has not proven to be very effective. If you start with someone who is already thinking about the article and interacting with the page (by rating it), and then we've already got their attention when we say, "You know, you create an account—you could edit this article—you could try this out", and we've got their focus on the article's strengths and weaknesses.
Might it be irrelevant for experienced editors identifying articles that need work? Quite possibly, but nobody's ever promised that the ratings would be useful for that purpose. In fact, the devs have specifically said that's not a validated use of the tool.
WhatamIdoing, when were expiring rating introduced? Has the tool worked like that from the start? I only become aware of it recently so I was partly misunderstanding how it works.
It's been described in the documentation since at least this spring, but I don't know exactly when the feature was first introduced.
Don't make it too complicated. The only thing that matters is that anons who never edited an article can vote Justin Biber up and George W. Bush down.
Nicely put, Subfader.
This post was posted by VanishedUser sfkdjljakldj, but signed as JoeMCMXLVII.
So out of the 3 million+ articles on en.wiki, Justin Bieber, George W. Bush and maybe 100 other high traffic articles could be considered "1 offs" because their survey results will be skewed by fanboys/haters. That is still an infinitesimally small fraction and even if these are high traffic articles, that doesn't diminish the potential usefulness that this tool can offer to the project across the board.
The editors who watch Bieber and GWB will not be making significant changes to those articles based on the AFT tool. However, the editors to articles like the 1984 Summer Olympics, Prince Edward Island, Piston valve, Aging of wine, Royal Scottish Academy, Bamboo, Australopithecus etc might be inclined to look more closely at those articles if they start receiving a pattern of low ratings in comparison to similar articles.
I think this is a key point that critics are missing when they try to bring up "crap" ratings on articles like Bieber and GWB. Articles like that make up a very, very small percentage of Wikipedia and they should not be held as a benchmark for how new tools or ideas are implemented on the project.
"I think this is a key point that critics are missing when they try to bring up "crap" ratings on articles like Bieber and GWB."
I think the key point they're missing is that determining the true quality of the article is not a validated use of the tool. Experienced editors should be evaluating articles according to the English Wikipedia's extensive policies and guidelines, not according to the responses from recent readers. So far as I can tell, that's what our editors are actually doing. I haven't seen anyone yet say "I've got to totally re-write this Good Article, because some readers gave it a low rating!"
Great how you accept the fail of the tool and how much you don't care that good articles stay 1/5 rated by abusers forever for every visitor since (as you pointed out) nobody would edit the article to change the rating (except WMF staff of course).
I do not consider a disconnect between the reader's rating and the experienced editor's rating to be a failure of the tool.
I consider your insistence that reader and editor ratings much match to be a sign of your failure to read and understand the documentation for the tool, which explicitly says that low ratings from readers do not mean that there is any problem in the article.
I see a lot of negative feedback about the whole concept and I see a lot of nit-picking in response. You won't make a bad idea work by trying to discredit its critics.
This post was posted by VanishedUser sfkdjljakldj, but signed as JoeMCMXLVII.
Well I am new to this discussion but here are the criticisms that I've seen.
The fourth criticism is the only one that really has meat to it but it is a multi-faceted issue. On one hand, there is a lack of understanding in how data research/surveys are conducted and the limitation of the "laziness principle". However, the tool can, indeed, be improved but that is the point behind studying the results of a launch--you can't see where to improve until you see how it is being used.
While all the criticisms are worth discussion (which is obviously being done here), I haven't yet seen a single criticism that is worth completely vetoing or shelving this project.
Is there something I missed?
I'll expand on your point number 4). A tool that does not provide sufficient data for editors is utterly useless.
And I didn't suggest that any one criticism should warrant vetoing the project. That's your defence to the various sweeping criticisms and it's irrelevant. What I am suggesting is that there is so much conceptual criticism and so little response of a worthwhile nature that the project seems like it's dead already but simply won't lie down.
This post was posted by VanishedUser sfkdjljakldj, but signed as JoeMCMXLVII.
Yes I was summing up the various sweeping criticism in that posts.
But let's look at point 4. You state "A tool that does not provide sufficient data for editors is utterly useless."
Now as I said before that issue is multi-faceted mostly because the idea of "sufficient" and "useful" are subject. The tool already provides some data and, again as noted before, any knowledge of survey taking and market research can see immense value in the kind of data already being provided since it is formatted to show patterns of response.
Also, the idea at the moment is to provide data about what the readers think and then figure out how to make it useful to editors at larger. So while you may feel that it is not sufficient and useless now, the appropriate response is to make recommendations for improvement (which many have) and to give it time to see how the tool develops and improve. It is the then part in the equation that I have to say is surprising that a vocal few don't seem to want to wait for.
I mean the first computers were pretty "useless" oversize calculators. You can only imagine what would have happened if we shelved THAT idea so early on in the beginning under the idea that "A tool that does not provide sufficient data for users is utterly useless."
Fair enough. About the collected data: I think the 4 criteria are ok and other data (like who voted / when / how "trustworthy" the user is) can be generated later. Dunno how much data is stored with each vote tho.
Still nobody replied how high the percantage of anon and 0 edit users is to experienced editors.
I haven't seen a response to that question either but an ideal would probably be a much higher percentage of anon or low edit users since the tool is aimed to get more a reader's perspective then an editors perspective.
The dataset (as explained in the documentation; did you read it?) differentiates between registered and unregistered users (a set that includes experienced editors who don't happen to be logged in at the time they rate the article), but it does not record the number of edits made by the editor (which may be zero for a registered account, and thousands for an IP editor).
I don't have the time to read this ENTIRE thread, but I have to say that I agree with some of the points that have been brought up here. As far as I can see the ratings are pretty much useless. I've seen some well written articles with good sources getting low ratings simply because 10-year-old girls don't like the character an actor plays on TV, or conversely, I've seen some stubby articles getting five stars from multiple raters simply because people loved an actor's work 50 years ago. Granted, this seems to apply more to the "celebrity" bio articles more than the "science" articles about nuclear power, etc, but I've noticed this seems to be a theme running throughout MANY celebrity bio pages and not limited to just Justin Bieber and George Bush. I'm not an "exepert", but it's painfully obvious people are simply using the ratings to vote on how they feel about the SUBJECT of an article and there's really no reflection whatsoever on the merits of the article itself. Not to mention the "highly knowledgeable" option is just a joke. I could dub myself "highly knowledgeable" about quantum physics, but it doesn't make it true. From what I have read of this thread, it looks like the ratings aren't being taken too seriously by anyone who matters, so I guess that's good news, but then what's the point of having the ratings at all? Just to trick a few new people into editing? It seems to me that anyone who would be tricked into believing the ratings are any indication of what an article needs would be the LAST people you'd want editing the article.
On WP some WMF guy explained how AFT works and how to stop vandalism by fan / hate votes. It all makes sense now. --Subfader 19:11, 20 September 2011 (UTC)
Lol.. I have to admit I've been tempted to do that myself, but I decided not to based on principal. Not because it's "wrong", but because I made a decision when the tool first showed up (and obvious vandalism ratings started soon after) that I wouldn't vote on ANY article I'd had a significant hand in writing (aside from minor cleanup - correcting typos, redirecting backlinks, etc). I wanted to give the tool some time to see how I liked it, how it worked, etc (I honestly didn't have any opinion about it one way or the other when it was first implemented). If I was "rigging" the votes then I wouldn't be able to do that, so I decided I'd wait to see how things played out before I came here because I wanted to be able to be 100% impartial. What irks me about this page is how obnoxious a couple of the people here are - implying our hard work, observations and experiences are "infinitesimal" to the value of Wikipedia, etc. I think they know what they can do with their new "tool", so I won't even bother lowering myself to their level.
Well, it is interesting to follow this project over months and seeing how the WMF people play boss is indeed very scary. I did follow the test phase of FlaggedRevs on German WP with much interest and actually they learned from it and took consequences. But AFT? Nothing. Instead, as you mentioned, they try to sell us this useless feature as the next big thing which WP cannot live without. And non-related people like critics and scientists who follow the development of WP have just another reason to turn the back on WP when they see that good articles are rated 1/5 by people (visitors don't know that haters and fan abuse the AFT and how easy it is to do so). I could continue forever why AFT is bad... I have not seen one advantage of AFT here yet. Someone link me to it please. I know the FAQ yeah but it reads like a fake review for an overprized tool nobody actually wants or needs.
Exactly. It reminds me of what bugs me when editors will tag an article with a "biased" template without explaining why on the talk page (even well-meaning, well-respected editors will do this routinely). It casts a shadow over the ENTIRE article, and sends a flag to readers that there's something specious about the article as a whole (whether the average reader has personally found any problems with it, most readers will think - "well, it looks good, but there must be SOMETHING wrong with it"). And I don't want to hear this b#llcr@p about me whining just because I personally "disagree" with the voting results. I posted SEVERAL examples (taken from the dashboard page yesterday) of good articles that were rated "B-Class" or higher (presumably by people who actually KNOW something about Wikipedia's GA standards), but had mysteriously received the "lowest" ratings by casual "users", but this was summarily dismissed. I mean the Barack Obama article is CONSISTENTLY among the lowest rated articles on the dashboard almost EVERY SINGLE day i've looked, and that's a FEATURED article for Chrissakes.. It's like there are people here who are DETERMINED to bury their heads in the sand - no matter how many people report problems or how many examples we give they just keep telling us "it's infinitesimal". What they can't seem to wrap their little heads around is the fact that the Justin Bieber and Barack Obama articles are representative of THOUSANDS of articles that DON'T show up in the dashboard. I'm not pretending to be a mathematician, but it stands to reason - if half the articles voted lowest in the dashboard on any given day are celebrity pages with good to very good "quality scale" ratings then any reasonably sane person could (and would be naive not to) infer that 50% of all Wikipedia articles being rated are celeb pages getting the same meaningless votes that have no bearing on the merits of the article whatsoever. The dashboard is a random sampling of 50 articles that clearly shows the TYPES of articles that are most frequently vandalized. I just don't understand what the point is in creating a "Feedback" page, when it's obvious nobody here wants any honest feedback. If this was really about feedback, then our contributions wouldn't be dismissed as "infinitesimal" and we wouldn't be accused of simply coming here because of "sour grapes" about an article we've written receiving low ratings. That's simply not the case. In fact - I'd like to invite anyone who seems to love this tool so much to link to two articles they've written here (and not just articles they've added one or two sentences to, I mean an article they wrote from SCRATCH, or took-up as a "stub") so I can see their work and evaluate their Wikipedia "expertise" for myself. It can be any subject - anything - it doesn't matter - I'd LOVE to see what you've written, and I'll be more than happy to post two of my recent articles I've written as well. Anyone.....? Anyone.......?
You still don't seem to get it: AFT is not trying to determine the True™ quality of the articles. AFT is collecting casual readers' personal opinions, and trying to turn some of them into brand-new editors.
Most casual readers are totally uninformed about what makes a decent encyclopedia article. Something with a lot of blue numbers that lead to personal blogs is often going to be called "well-referenced" by some of these readers. A stub that happens to contain the one tidbit of information they want is going to be called "complete" by some of these readers. But that's okay, because our smarter editors know that if they want information on the actual quality of the article, use the 1.0 team's assessment system. AFT is not trying to duplicate that.
Also, your assertion that people's opinions will be swayed by the ratings is either nonsensical or shows an appalling level of disrespect for the readers. If you read an entire article, and you think it's pretty good, and then you see that it's been rated low, are you so stupid and intellectually insecure as to chuck your whole opinion out the window and adopt the view on the page? And if you're not, then why do you keep saying that everyone else is?
AFT is collecting casual readers' personal opinions, and trying to turn some of them into brand-new editors. Most casual readers are totally uninformed about what makes a decent encyclopedia article. If you read an entire article, and you think it's pretty good, and then you see that it's been rated low, are you so stupid and intellectually insecure as to chuck your whole opinion out the window and adopt the view on the page? And if you're not, then why do you keep saying that everyone else is?
I couldn't have said it better myself. Once again you've made my point for me. What exactly is the point of a ratings system that lures only the most "stupid and intellectually insecure" readers into editing? Anyone who is smart enough (and knowledgeable enough about Wikipedia's policies) to know what an article does and does not need isn't going to pay any attention to this useless tool. So what's the point of it??
By the way, I see there are no takers among the people on here who keep defending this tool to post a link to a single article you've actually written. I think that speaks volumes about the "expertise" behind the logic defending this tool. Know-it-all "editors" who've never actually written a single article.
I couldn't have said it better myself. Once again you've made my point for me. What exactly is the point of a ratings system that lures only the most "stupid and intellectually insecure" readers into editing?
No where do I see WhatAmIdoing describe readers as stupid and intellectually insecure. Rather, she was asking if YOU believe that they are and would formulate their own personal opinion on an article and then chuck that opinion out the window because they saw a low rating.
Should we use this reply above as affirmation that you do, indeed, hold the opinions and capabilities of Wikipedia's readers in such low esteem?
Just as an FYI, Common Sense:101, any producer who makes a product (even a free online encyclopedia) and doesn't value or seek the opinions of its consumers will not have the most promising of futures with people continuing to use said product.
So is that the equivalent of "I can't defend my point so I'm going to take my ball and go home"?
Which level of discourse are you having issues with?
Me asking Crakkerjakk if that characterization of readers accurately described his/her view of them?
Or me point out the common sense truth in business/life that any producers who ignores/doesn't see the views of the consumers using their product/services is not going to be in good shape for long?
WhatamIdoing has CLEARLY stated the purpose of the tool is to get new editors interested in editing, and yet he/she turned right around and basically admitted that only an idiot would pay any attention to the ratings. Result = Only an idiot will be lured into editing by this tool. I'm just following his/her own logic.
Once again, you're having trouble reading, so I'll try and spell it out for you one more time. I thought asking for ratings/feedback from readers was a GREAT idea when the tool first showed up. However, what I'm not interested in is whether people like Justin Bieber's new girlfriend or not. That's what these numbers give us on a VAST number of articles (sorry, you still haven't PROVEN this little 1% statistic you seem to have pulled out of your @$$.
No, Crakkerjakk, you are lying about what I said.
Many people might choose to rate articles.
Only an idiot or insecure person would (1) think "Great article" after reading it, (2) look at the accumulated ratings and see that they're low, and then (3) decide "I guess I was wrong; it must be a bad article, because everybody else hates it".
This is the process you say happens all the time. That is the only possible mechanism by which the accumulated ratings can harm readers' perceptions of the articles you write.
It is my belief that the readers do not have this problem: It is my belief that readers do not radically change their perception of an article based on other people's ratings. It is therefore my belief that the tool is not recruiting readers with this problem.
LOL...WhatamIdoing, admirable effort but Crakkerjakk's argument has been exposed so much with him pushed in a corner that his only retort is going to be "But NO! You can't REEEEEAAAAAAADDDDD!" ;)
Or he may try to mix thing up and drag my mythical alcoholism over to you and call you a drunkard :P Stay tune.
Yeah, right.. I, and the landslide of other editors reporting the same problems here, have ALL been "backed into a corner" by the two of you. Keep on trying to convince yourself. Unfortunately for you - Two people repeating the same old desperate excuses dozens of times doesn't outweigh the posts from dozens of people complaining about the numerous problems with the tool.
FYI - If you actually HAD read my posts, you would have KNOWN what I'd said the FIRST time and I wouldn't NEED to keep pointing out your obvious deficiencies. But seriously though, nobody wants a creepy resident wino lurking on this board just waiting to troll every single person who posts. The first step is admitting you have a problem. I suggest you get some serious help.
I know I said I wanted to avoid creating another Wikimedia account but I was just too charmed by that moniker Crakkerjakk!
Yeah, yeah, I know this was your go-to ad hominem and you seemed to be having fun with the flaming, name calling and incivility but the name was just too cute!
Now this is the AFT talk page and I'm still very willing to have a civil discussion about how the tool and dashboard can be improved.
But I do have to give you due props for the witty nickname. It got me to bite the bullet and finally create an account here :)
I'm sorry if your own twisted logic has backfired, but if anyone is "lying", it's you. You've mentioned at LEAST a hundred times here that this tool is WONDERFUL because new users will view the ratings and will be magically "inspired" to edit the articles that the tool indicates are in such desperate need of their help. You've also stated that only an idiot would pay any attention to the tool's results. The only logical conclusion is that only an idiot would be inspired to edit because of this tool. Now you're changing your logic to say - "the VAST MAJORITY of new editors will be smart enough to disregard the tool". Ok, fine - let's say that's true - Then this now new and improved "logic" of yours only proves how useless the tool is.
Getting back to the subject of what I was saying originally about users being "misled" by editors arbitrarily tagging a page with "bias" templates (with absolutely NO explanation) and/or by viewing this ratings tool. I admit I didn't expand on my point enough, so I'll try and explain myself better..
Let's take the Barack Obama article as an example (I know everyone is sick hearing about the Obama page, but it's just for the purpose of an example). It's a featured article. Now I would say I know about as much as the AVERAGE American about the man. No more, no less, just AVERAGE. I have a pretty good idea about the man over the last 3 years, but for the first 40-something years of his life, I'm most definitely NOT anything CLOSE to an "expert". He seems like a nice enough guy, I don't have anything against him personally, but I just haven't spent all my time researching the man's life (nor do I care to). So I come to the article, I read the entire thing, and then I'm asked to "rate" the article on FOUR criteria. Again, I'd say I have just about the "average" American's knowledge about the man. I would say I'm COMPLETELY unqualified to vote on three of the four criteria. First question, "Trustworthiness": (which is basically asking if the sources are "reliable". I can honestly say I haven't the faintest idea. I mean checking/verifying the 300+ notes cited on the page would be the equivalent of reading an entire BOOK about him, and that's not even counting the books the article cites as references. I'm not (nor is the "average" reader) going to sit there and check EVERY SINGLE source to judge the article's "trustworthiness". If I wanted to spend WEEKS reading stacks of books about the man I wouldn't be visiting Wikipedia to look him up. Next question, "Objectivity": yet ANOTHER question I, as an "average" reader, would be COMPLETELY unqualified to judge unless I checked every single one of the 300+ sources/references cited on the page. I mean aside from some GLARING bias, how would I know if it's "objective" unless I read EVERY SINGLE SOURCE the article cites to compare the article to??? Next question, "Completeness": Once again, how the hell do I know if it's "complete"??? Aside from some HUGE omission, I would have to spend WEEKS checking EVERY SINGLE source/reference to know if something I thought was important was left out. The ONE criteria I'd say I'm reasonably qualified to judge is how "Well-written" the Barack Obama article is.. Now, as a Wikipedia editor, I know it's reasonable to believe what I read on the Barack Obama article is true/fair/complete because: #1 - It's a FEATURED Article which means I can reasonably expect that all the claims and sources to back them up have been thoroughly vetted. #2 - The page is LOCKED from "casual" (even possibly well-meaning) editors who know NOTHING about Wikipedia's criteria. and #3 - There are a THOUSAND trusted Editors watching the page every minute like hawks. All indicators which the "average" reader most likely knows nothing about.
This all leads into what I was talking about before - If someone were to tag a "bias" template on the Barack Obama page (and there wasn't a "featured" star on the page, which I, being an experienced enough editor, KNOW Wikipedia does NOT hand out lightly), then YES, I would probably think to myself even BEFORE reading the article - "Wow, this "bias" template has been her for six months, so I better take EVERYTHING I read here with a grain of salt". I'm not even close to being an "expert" on Barack Obama, so I wouldn't have any way to judge first-hand unless I decided to become an "expert" and devote the next six months of my life to researching his life story. It definitely makes the "average" person question the "trustworthiness", "objectivity", and "completeness" of what they read. The exact same thing applies to this tool - Yes, experienced Wikipedia editors (who aren't in denial) know about the disproportionately high number of fan/hate votes the tool is recording across THOUSANDS of pages, but it may very well not even cross the "average" reader's mind that 90% of the low votes are from 14-year-old nazi skinheads who hate Obama and haven't even READ the article. Unless someone is an "expert" on the subject of an article they're reading (although, I don't know why anyone would be reading a Wikipedia article they already knew EVERYTHING about), they're just an average person, they're NOT going sit there and check out 50, or 100, or 300 sources on a page (many of which are two or three or four pages long EACH). Many readers are just going to look at the fan/hate ratings and assume that SOMETHING must be wrong. This doesn't necessarily mean a person is an idiot. It just means that the average reader isn't going to do the HOURS, or DAYS, or WEEKS of research to verify all the sources. Unless a person is an "expert" about the ratings tool AND the subject of the article, this fan/hate tool has the very REAL potential to make it seem like there's something specious about the entire article, no matter HOW good it actually is on the quality scale. Of course I have no proof how that will affect anyone who has been tricked into editing as a result of viewing this tool, but my point was about how the average READER thinks of Wikipedia after they READ a page. Viewing the tool, they may disregard very GOOD information as "myth", or take an article filled with lies and inaccuracies as being COMPLETELY trustworthy, unbiased and complete. Aside from being useless in the attempt to lure even remotely "qualified" editors (as your own logic has concluded), my LARGER point was that it has the potential to be HIGHLY misleading to the READERS with regards to how they should evaluate the accuracy of an article. If the purpose of the tool is to get people to take Wikipedia more seriously, I'd say it's MUCH more likely to have the exact OPPOSITE effect.
I encourage you to look at every single comment I've made, and let me know if you find any comment that says I think very many readers will "view the ratings". I believe you will find several indicating my belief that the vast majority of readers will not view the ratings.
This is one of the laws of the web: every click loses a substantial percentage of the audience. Most readers will never click through to view the ratings. The people who look at the ratings are editors like you. The vast majority of readers will not be "misled" by the ratings, because readers will not ever bother to even look at the ratings.
AFT is collecting casual readers' personal opinions...
No, it's collecting abusive fake / fan / hate votes by logged out spammers and people who misunderrstand the tool. Nothing else. Trends of these over time may be interesting but do not add any value to WP content.
...and trying to turn some of them into brand-new editors.
Trying to turn abusers into editors is a nice goal but pure waste of time.
If you read an entire article, and you think it's pretty good, and then you see that it's been rated low, are you so stupid and intellectually insecure as to chuck your whole opinion out the window and adopt the view on the page?
No, I would think the rating is bad or was abused by questionable people. I maybe would wonder why the responsible people do not improve the rating tool to avoid such. Surely I would wonder what the goals of the tool actually are if fake / fan / hate votes by anonymous users is obviously accepted. I then may read the FAQ and find out that AFT has no purpose except development timelines of abusive votes becoming "interesting". Maybe I would end up on this epic feedback page and find out that all of my observed disadvantages and the uselessness of this tool have been reported here in full detail with lots of suggestions how to improve it but everythings has been greatfully ignored by the officials. I would surely be surprised how officials interact with feedback reporters and try to silence them in order to push AFT.
But you're right, I would not edit any article.
...are you so stupid and intellectually insecure...
I've read such personal attacks so often by you now, I wonder if YOU are. Are you working for WMF? I hope not.
Most casual readers are totally uninformed about what makes a decent encyclopedia article
So let them rate the quality of encyclopedia articles!
Most casual readers are totally uninformed about what makes a decent encyclopedia article
So let them rate the quality of encyclopedia articles!
Better yet - Let's use this as a way to get them to EDIT it!!
I think it's been made abundantly clear the level of intellect we're dealing with here. They've basically admitted they've never written a single Wikiepdia article and their desperately flawed logic defending the tool shows they don't know the first thing about how to improve the encyclopedia. It's been amusing, but I'm afraid I'm going to have to stop wasting my time expecting to have an intelligent discussion here, but I WILL be taking the matter to the appropriate people to report the obvious problems EVERYONE else who has come here has reported with this beloved tool.
I think it's been made abundantly clear the level of intellect we're dealing with here. They've basically admitted they've never written a single Wikiepdia article and their desperately flawed logic defending the tool shows they don't know the first thing about how to improve the encyclopedia.
I know it is frustrating when other editors don't accept your viewpoint prima facie and actually ask you to extend your logic beyond the small scope that you want to deal with. It is tough when people look behind the curtain and question while we should limit our focus and attention to only <1% of the screen and ignore the colossal mountain of OTHER articles, OTHER experiences and OTHER viewpoints when most of your argument is based on them only seeing what you want them to see.
But it is healthy to have civil discourse, though truthfully, it becomes increasingly less civil when folks start dismissing the "intellectual level" of people who simply disagree with them and start assuming omniscience over their editing history and experience.
It is healthy to have civil discourse, though truthfully, it becomes increasingly less civil when folks start dismissing the "intellectual level" of people who simply disagree with them and start assuming omniscience over their editing history and experience.
Right back atcha. I couldn't have said it better myself. Try taking some of your own advice, and we wouldn't have the problems we see on this page.
I don't have a problem with editors who feel the tool is working in the areas of Wikipedia they're most experienced in. I RESPECT their work, and would NEVER tell someone their experience was "infinitesimal". I've already stated that I'm sure the tool DOES work on many types of articles that are far less likely to be vandalized. However, by the clear statistics - those types of articles only make up about HALF of the encyclopeida's daily traffic as a whole (not very good averages).
I know you're still clinging to this 1% myth you've been desperately trying to perpetuate, but (short of cataloging all 3 million + articles) the BEST evidence we have is the dashboard and, I'm sorry, your absurd assumption that the articles showing up in the dashboard are the ONLY ones experiencing this HUGE problem isn't holding water with anyone else here (except WhatamIdoing).
The only thing that "frustrates" me is seeing how many people were summarily dismissed when they posted the problems they were experiencing here. The difference is the average person will just leave when met with an obnoxious user, whereas I'm not inclined to be so readily dismissed as "infinitesimal". A little advice - when someone posts here, they don't need you, or anyone else, to "rate" the feedback they leave on the ratings tool. It's not necessary for you to leave a snotty comment on EVERY SINGLE criticism of the tool. You've made it clear you love this tool, the vast majority of people here disagree with you - DEAL WITH IT.
Seriously, guys: you all need to log off for a while, take a breath or fifty, and find a cat to pet or something.
This thread has clearly entered the "emotional flamewar" stage, and should just cool down.
"The only thing that "frustrates" me is seeing how many people were summarily dismissed when they posted the problems they were experiencing here."
Just make sure that while the problem are being discussed here (where people can provide arguments/evidence/personal opinions/etc to show why they agree/disagree with each other), each specific request is also posted on bugzilla, where the (technical details of a possibly) implementation can be discussed. On bugzilla the requests will have its priority set to help coordinate developer efforts and eventually be resolved as FIXED or WONTFIX (or something else) depending on their decisions after analysing both the technical requirements and the desirability of the feature being requested.
AFAIK, anything which is reported on bugzilla will eventually be considered by some developer and anything else (e.g. problems posted only on talk pages of some wiki) have little chance of being noticed during the development of MediaWiki and its extensions...
"would NEVER tell someone their experience was "infinitesimal"."
Hey! What a coincidence! I would 'never tell someone their experience was infinitesimal either! :)
But since we're on the same page, let me give you some advice as someone who has walked this road before.
Even though you would never tell someone that their experience was infinitesimal, you do have to be careful when you are having a internet disagreement, especially if you are successfully countering their points. Because even though everything is written plainly in text and is saved in the history, when the other person starts losing their footing and feel like they're not making progress with their arguments, they're going to take whatever you said and then SAY that you called their experience infinitesimal and keep repeating that mantra, over and over again and pray that no one will actually read the exchange and see that other person is just blowing smoke. Now, granted, given the "laziness principle" of humanity, that is not a crazy bet to make.
Yes, I know you can point out to them (repeatedly) how you never called them, their experience, their observations or anything else beyond a category of articles as "infinitesimal" but, again as someone who has went through this pony show, it probably won't help because at this point, strawmen and misrepresentations of the other side is all they got going for them.
So to wrap up. Good for you! It is important to maintain civility in discourse and not belittle the contribution of other editors by calling their experience "infinitesimal" or other ad hominem attack but I just want you to be prepared for when people still try fabricate their own ad hominem by blatantly and repeatedly misrepresenting what you said. It's tough. It's frustrating. But stay the good course and not sink to their level of grasping straws.
Though, admittedly, it doesn't hurt to let out a little steam via sarcasm but stay above the belt with more irony and understatement than jibes.
AFT is collecting casual readers' personal opinions... No, it's collecting abusive fake / fan / hate votes by logged out spammers and people who misunderrstand the tool.
For <1% of Wikipedia's 3 million+ English articles.
Interestingly, according to w:en:Maternal_death, <1% of women in the developed world die during pregnancy. Small but still an unfortunate number, yet, I'm not sure you would find many advocates saying that we need to get rid of that whole "Getting pregnant/having babies" things based on the unfortunate outcome affecting <1% of the cases.
============ Most casual readers are totally uninformed about what makes a decent encyclopedia article So let them rate the quality of encyclopedia articles!
So I suppose that we should only let automotive engineers who know the ins and out of car production rate vehicles, only professional chefs with 5 star restaurant experience that have mastered a particular cuisine do Yelp ratings, only Academy Award winners should do IMDB ratings and only prostitutes and call girls rate hotels?
"...votes by anonymous users is obviously accepted"
That is not exactly the case. Votes by anonymous users are registered with an explicit indication that they come from anonymous users in the field aa_user_id (see more about the data being collected). So, if someone wants to take into account only ratings given by registered users, they can just filter those votes which are not marked like that. If someone else wants to compare the ratings from anons to the ratings from registered users, they can used the very same field from the tables to distinguish both kinds of ratings.
In other words, anyone looking into the data being collected by AFT can decide by itself if votes by anonymous users should be accepted or not for the analysis they are doing.
If the tool currently doesn't (and it currently doesn't) provide an interface for its users to select/compare/filter ratings depending on its prefered criteria (e.g. to get rid of anonymous users' ratings by [x] marking a small checkbox somewhere), such a feature should be requested on MediaWiki's bug tracker. This was requested for the dashboard (bug #30334), but something similar could also be requested for the rating form which is displayied on articles (anyone who wants to improve AFT can report this on bugzilla if it hasn't being already reported). Besides, interested users can also vote on open requests which they think really needs to be fixed.
Subfader is not looking at the data; I'm not convinced that he even knew that the data existed.
Subfader's notion of what happens with AFT is apparently based on the <50 articles listed on the dashboard and a handful of articles he's read. He can't be upset about "his" articles being dissed, though, because in the last five years, Subfader has managed just 28 mainspace edits on the English Wikipedia and created zero articles there.
Yes, I didn't know abou teh existance of that data, so what? Kill me? I bet you didn't know about http://toolserver.org/~catrope/articlefeedback/ either! If you did you would have posted the link 50 times by now as you did with the FAQ in your arrogant know-it-all-wannabe posting style... Also it's not my fault that I didn't know about that ressource. The plain fact that the page was only linked now, months after this epic feedback page started, is just a proof of the poor communication or ignorance by the officials.
Instead of making crazy assumptions about me, my wiki knowledge and my activity on other MW project accounts, answer some of the questions on the FAQ talk?
No one said that your work was infinitesimal but, rather, that the amount of articles that would be adversely impacted by "silly ratings" is infinitesimal compared to the sum of the 3 million+ articles on en.Wikipedia.
A HUGE part of my work on Wikipedia is monitoring pages that are highly trafficked by children and teens. So YES, when you tell people reporting problems here that their observations are "infinitesimal" you're being dismissive of THEIR WORK. I admit child star pages get a disproportionately high number of vandals (which is why I started watching them), but anyone who views the dashboard can see it's not just child stars - it's clearly a sweeping cross section of bio and "celebrity culture" articles in general (yes, for the purposes of this discussion Barack Obama is one of many who falls into the category of "celebrity"). It's obvious to anyone who understands how to read numbers that each bio article that shows up on the dashboard represents THOUSANDS of others getting the same TYPE of fan/hate votes - they simply aren't getting rated as OFTEN so they will never show up there. That's where I came in - After giving the tool a couple of months to see how it worked, I came here to report that it's NOT "infinitesimal" (after seeing you had repeatedly attempted to dismiss/discredit other editors coming here, using this exact same line, like some broken record). It's happening on dozens of pages I'm watching - and it's common sense to infer that each page I'm watching represents hundreds of pages that I'm NOT watching. If you think celebrity bio pages, film pages and awards pages don't make up at least HALF of the daily traffic on Wikipedia then you're seriously delusional. The tool doesn't work and nobody here is going to suddenly change their minds and start drinking the kool-aid just because you tell them to ignore what they're seeing FIRST HAND.
And I didn't even say that your observations were "infinitesimal"....
Martyrdom doesn't work when you're doing your own flogging. No one is saying the things that you are claiming.
Again, to repeat, the amount of articles' (noun: inanimate object that is not a person, trait of of a person or action being done by said person) that would be affected by "silly ratings" of people using the tool inappropriately is infinitesimal.
So far you are neglecting to answer that point or explain why we should let an issue that affects less than a fraction of a percent of the 3 million+ articles on en.wikipedia be a major reason in shelving this tool.
It's like saying because some people use their cell phones to talk/text while driving (very likely a higher percentage than those who abuse the AFT tool) are using their phones improperly ergo we should ban the production or use of ALL cell phones period. It doesn't make sense to let a problem that affects a very, very, very, very, very small amount of articles have undue weight on the decision to shelve the entire tool.
You're imperious attitude doesn't really work when you've repeatedly demonstrated that you can't read. I mean honestly, this is the ENGLISH MediaWiki right?
Again, to repeat, the amount of articles (noun: inanimate object that is not a person, trait of of a person or action being done by said person) that ARE affected by "silly ratings" of people using the tool inappropriately are NOT infinitesimal, and repeatedly pretending that they are does NOT make it true.
So far you are neglecting to understand what I have repeatedly stated previously. I did not recommend that Wikipedia should "shelve the tool" on ALL pages, but LOOK AT THE DASHBOARD - yes, that's an infinitesimally small number of articles, but it gives you a clear "sampling" of what is happening on ALL articles across Wikipedia. Fan/hate votes on celeb related articles makes up a HUGE number of inexplicable ratings on the dashboard, and celeb related articles make up a HUGE number of overall daily TRAFFIC on the encyclopedia. Expecting someone to cite all 3 million + pages individually in order to prove this to you before you accept the obvious statistics is ridiculous. Of the feedback I've seen here, the only thing that's "infinitesimal" is the number of people who seem to think this tool IS working. You haven't PROVEN to me that silly ratings on celeb culture articles are "infinitesimal", and that the tool is working on 99.9999999% of the encyclopedia yet either. So go ahead, cite the 2.999 MILLION articles where it's working.
I'm doing simple math here - If only half the ratings are people HONESTLY voting on the merits of the articles (which is being generous) then I'd rather not have ANY ratings, but that was NOT my suggestion. I recommended the tool not be used on the categories of pages that it's been PROVEN thousands of people are fan/hate voting on. Unfortunately for you, this includes People, Films, TV shows, etc, etc - which you know full well is a HUGE part of the encyclopedia. If you really thought it was infinitesimal then why would you have a problem with the ratings being removed on those pages? Instead you know FULL WELL that would yank the tool from HUNDREDS of THOUSANDS of pages viewed by MILLIONS of readers every day, so you can drop your debunked "infinitesimal" myth. Nobody is buying it.
The problem with your cute little cell phone analogy is that cell phones are USEFUL. This tool is NOT. It doesn't make sense to let a very, very, very, very, very, very, very, very, very, very, very, very small number of "developmentally challenged" people defending the tool (and who have never written a Wikipedia article in their entire LIFE) have undue weight on the decision to keep it.
You're imperious attitude doesn't really work when you've repeatedly demonstrated that you can't read.
You know, I get the idea behind "Truthiness", but outside of beckoning Beetlejuice, repeating something incessantly doesn't make it true. So while you may wish to push other's people views or them noting your errors aside under the auspice that "they can't read", it really isn't going to add strength to your argument or support your point.
Though I see there is at least some progress in you no longer claiming that people are calling your work/observations/etc infinitesimal. Victimhood is so unsexy and counterproductive.
But you still wish to handcuff all discussion on the use of this tool to celeb/pop culture article which do, INDEED, account for an infinitesimal percentage of articles on en.wikipedia. While I duly acknowledge that issues with those articles are reasons to discuss how the tool and dashboard can be improved, I can't buy into the idea that any discussion or gauging of the usefulness of the tool should be relegated to the impact on this <1% of articles. As demonstrated amply elsewhere, it is nonsensical to take drastic measures that impact a large scope based on the results of <1% of anything.
======== The problem with your cute little cell phone analogy is that cell phones are USEFUL. This tool is NOT. It doesn't make sense to let a very, very, very, very, very, very, very, very, very, very, very, very small number of "developmentally challenged" people defending the tool (and who have never written a Wikipedia article in their entire LIFE) have undue weight on the decision to keep it.
1.) Personal opinion on the usefulness of cell phones/AFT. Some may share your personal opinion on one, both or neither. Whose personal opinion is valid? Still doesn't discount the point of the analogy that having a small percentage of "abusers" misusing a tool is not reason enough to completely banish/shelve said tool. It is reason to have a rational, civil discussion on how to improve the tool.
But rational discussion is what leads to things like cell phone driving laws and hands-free headsets. Irrational, uncivil discussion leads to people name calling their opposition and advocating the complete shelving of a tool instead of working to improve something.
2.) Your whole pontification about people who disagree with you having never written a Wikipedia article has been demonstrated as false (and is a very weird dick measuring contest BTW). Which begs the question, do you still need a saddle if you're trying to beat a dead horse?
Let's be clear, I'm not portraying myself as a "victim". I'm merely pointing out that you've been an obnoxious snot to numerous other users. I know, the truth hurts. Go cry into your jug of wine about it, nobody here cares.
I see you're still pushing this 1% myth. Once again, nice try, but this isn't about my PERSONAL opinions. I've cited pages showing up on the DASHBOARD numerous times.. Oh right, I forgot, those are just the dashboard's "personal opinions".
And yeah, I just checked those pages you CLAIMED to have written.. The only thing those links showed is your apparent OBSESSION with alcohol. NO username provided, NO revision page that shows what YOU'VE written. Just you taking credit for HUNDREDS of other editors' work.
As you've repeatedly demonstrated by INSISTING that I've advocated the "complete shelving of the tool" - you CLEARLY have problems reading. I would suggest putting down your jug of wine, try sobering up for a few days, the discussion will most likely make a lot more sense to you then.
Crakkerjakk, you've got a math problem. You're assuming that 50% of the Dashboard = 50% of Wikipedia's traffic. That only works if all of the four million articles are equally likely to be rated strangely. This simply isn't true.
Quit relying on the Dashboard as your sole source of data. Go get the real data from the server and have a look, okay?
I'm not the one with the math problem. Repeatedly citing the traffic statistics only PROVES my point. Yeah, I know, you want people to only view the top 30 articles (HALF of which are "special" pages: the main page, the random article page, the search page, the 404 error page, etc, etc, etc), but what looking at that convenient little number you're clinging to does NOT show is that at LEAST half of ALL the top 1,000 pages are celebrity/pop culture pages: actors, singers, politicians, billionaires, films, tv shows, albums, video games, etc, etc (all pages it's been PROVEN are the MOST likely to be voted on for how much people like, or dislike, the SUBJECT). That's just a FACT. Expecting me to list all the HUNDREDS of celeb/pop culture pages from the top 1000 traffic statistics here before you open your eyes and admit the truth is simply absurd. The reason I keep using the dashboard as a simple example is because it's easy for ANYONE who comes here to glance at and get the basic the idea. If you remember, I originally came and reported on the problems on the pages I'm watching. After you tried to discredit me by saying "those are just YOUR 'favorite' articles", I used the dashboard as ANOTHER example. Once again, you tried to pretend like those votes were just an "anomally". So then you threw out the article traffic page (which CLEARLY shows at least HALF of the pages are celeb articles).. It's like Subfader has said, this whole discussion is a vicious circle. Any which way you slice it, the statistics come up the same, but you just keep desperately trying to dance around the numbers. Sadly for you, NOBODY is falling for it.
Where we may be having some miscommunication is with exactly WHAT numbers we're talking about. To clarify - I'm not saying celeb/pop culture articles make up half of all articles on Wikipedia. I admit, in my boredom with repeatedly correcting people's misrepresentations of what I've said here I may have misspoken (mistyped) that here somewhere, so let me clarify. I'm saying half of all daily TRAFFIC is readers viewing celeb articles. In case you're still confused - Let's say only 10% of the encyclopedia is celeb/pop culture articles (I don't know the exact number - it may be 20%, it may be 30% - but being generous, let's just say 10%). What ALL the statistics tools show is that every time ONE person views a non-celebrity article, TEN people ARE viewing a celebrity article. I'm talking about the number of daily READERS, not the overall number of ARTICLES.
I've seen no stats tools that show this. I've been asking around, and nobody seems to be aware of any stats tool that actually provides traffic stats by general subject. If you've got one, then I've got several people who want to know where you've been hiding it.
Let's say that 10% of Wikipedia articles are about celebrities, and that half of the high-traffic articles are about celebrities (and, therefore, that half of the high-traffic articles are not about celebrities). That suggests that half of the traffic from the high-traffic pages is about celebrities (and half is not).
But we have said that celebrities are only 10% of Wikipedia articles. So celebrities get half the traffic—from a small subset of articles. But celebrities are only a small fraction of the overall articles: 10% of articles account for 50% of high-traffic articles, so they account for less than 10% of the moderate and low-traffic articles. This means that the other 90% accounts for more than 90% of the traffic on non-high-traffic pages.
The net result is that the 10% of articles about celebrities account for more than 10% of page views—but probably less than (probably much less than) 20% of the total traffic, because their "extra" views on the top end are dwarfed by the 90% of normal articles. (If we assume a simple distribution, the celeb articles (assumed to be 10% of total articles) have basically no chance of account for more than 15% of page views/traffic, but I have no idea how well the simple distribution conforms to reality.)
What you're missing is this: The English Wikipedia gets 5.7 billion page views a month. All of the top 1000 articles (excluding special pages) accounts for less than 10% of the traffic. That means that the highest-traffic celebrity pages account for less than 5% of total page views—and their effect on the other 90% of traffic is disproportionately low (because they are overrepresented in the top-traffic articles, they are automatically underrepresented in the rest of the traffic categories).
WatamIdoing, you've got a "math problem" (your words). You're assuming that if "half of the high-traffic articles are about celebrities" = "half of the traffic from the high-traffic pages is about celebrities ". That only works if the "traffic" for "high-traffic articles" is a "simple distribution" (your words). This simply isn't true and you admit you have "no idea" about its validity.
Quit relying on your imagination as your sole source for assumption. Go get the real data from the server and have a look, okay? You can then calculate the page views for celebrity pages (celebrities being a set which you claim you can quantify) versus page views per non-celebrity pages.
Bottom line: your previous post just makes assumptions to make the faulty syllogism that
Which is of course trivial. Worse is you use that conclusion to jump to a different vague conclusion "probably less than (probably much less than) 20% of the total traffic" using an improperly incorporated "dwarfed" claim. The tip off is your repeated use of weasel words in "basically no chance" which you "have no idea how well...conforms to reality. Keep in mind that if you've "seen no stats tools" regarding the 90% claim, then those "stats tools" don't exist for your <20% fabrication. Using the irrrationality of the feedback rating tool, why not make a survey GUI to ask people's feedback on what is the celebrity-to-non-celebrity page view ratio -- after all, Wikipedia claims that feedback is an effective way of quantifying information, right?
Actually, I did pull the page view data for the high-traffic articles, and that does appear to be a fairly simple distribution, so that assumption appears to be valid.
The assertion that 10% of articles are about celebrities was yours; I merely used your example. About 25% of articles in the English Wikipedia are about people, and the majority of these are not what we would call "celebrities", but 10% is not a completely unreasonable guess for a back-of-the-envelope calculation.
WhatamIdoing, you've got an integrity problem. Your replied to the poster who used IP address 208.54.5.202 with the false accusation that "The assertion that 10% of articles are about celebrities was yours", but that poster never made that faulty assertion. Worse than that lie, WhatamIdoing used that assumption to arrive at a trivial claim, then jumped to the unjustified weasel word conclusion that the assumption "is not a completely unreasonable guess" (i.e., it *is* an unreasonable guess) by claiming "the majority" of about 25% of Wikipedia articles are about celebrities. Depending on your weasel words "about" and "majority", the percentage of celebrity articles ranges from 0 to less than some "about" amount over 12.5%. Hence, the 10% assumption you used is unfounded, and lying about who originated that 10% assumption to cover your use is bad faith.
Even easier to debunk is your "simple distribution" claim, which you now qualify as "fairly simple distribution" (i.e., *not* a "simple distribution", whatever that weasel term means). Clearly the page views for all high traffic pages aren't equal, but perhaps you'll fabricate the claim they're "fairly" equal). Despite your claim of "no idea" about such statistics, you now claim "I did pull the page view data" and jumped to the conclusion the distribution of page views for "high-traffic articles...appear[s] to be a fairly simple distribution". But what "distribution" metric did you apply to weasel the "fairly simple" judgment? Did you look at the highest traffic article with the least page views and subtract that from the value for the most-viewed article, then just claim 'oh, that difference is less than the WhatamIdoing-standard "fairly simple" value used by the statistics industry? Again, you have a "math problem" (your words), as you haven't provided the mathematical definition of "simple distribution" (maybe you should author a wikiarticle) and then mathematically justify how the "page view data" meets your math definition.
You also have a math problem regarding your much less than 20% claim for the total page views for celebrity pages, as it is still unjustified. But of course you didn't address that fallacy in your reply. Best wishes formulating another rationalizing reply (I'm confident it won't include the stats for page views per celebrity article, despite your indication that you can identify them)! 208.54.5.202 21:23, 3 October 2011 (UTC)
Your account is no longer blocked, and I encourage you to start using it again. Alternatively, we could ask Checkuser to look into it your concern. Would you like me to do that?
Page view data is trivially available for the top 1000 pages. I ran the numbers by manually sorting the data for one day into two categories: "biographies" and "all other articles". (I excluded about two dozen non-article pages in the process, and assumed that all of the most-viewed biographies could be reasonably counted as "celebrities"). You could do the same, if you happen to have some the basic skills and reasonable access to software.
You do seem to be having trouble understanding what I'm posting. For example, I said that the majority of the English Wikipedia's biographies are NOT about celebrities.
As I was the user who was able to post (wasn't blocked) the preceding 30 September & 3 October replies and I don't have an account, the WhatamIdoing false claim "Your account is no longer blocked" is another blatant lie. And my bait to that user "Best wishes formulating another rationalizing reply" was successful, in that WhatamIdoing now has posted a laughable veiled threat to have my posts investigated by Checkuser with the misworded veil that the check would be for the reason "look into it your concern." BTW, to which of your math or integrity problems are you referring regarding my "concern"? Yes, I would like a Wikipedia authority to look into your math and integrity problems.
Also, the WhatamIdoing fabrication created when he claimed that his post "majority of the English Wikipedia's biographies are NOT about celebrities" is an "example" that someone else "seem"s to misunderstand is another falsehood--you'd need content that someone (who you fabricated is misunderstanding) wrote in order to exemplify the misunderstanding. It appears WhatamIdoing has trouble understanding what he writes, but posts it anyway. The old 'post first, analyze later' philosophy!
Nevertheless, you haven't provided the metrics for your two terms "simple distribution" and "fairly simple distribution", so the terms are not mathematically understandable. And since those metrics are needed to assess your faulty statistical analysis, your failing to provide them is another "integrity problem". Bottom line, your false "you could do the same" claim isn't possible for the same assessment without your "simple distribution" metrics (and the specific day, which you inaccurately assumed is representative of the general case).
WhatamIdoing, "you've got a math problem" (your words). "You're assuming that" (your words)--in "the top 1000 pages"-- that "all of the most-viewed biographies"="celebrities" in the top 1000. This simply isn't true. Worse, it fails to consider all the celebrities not in the top 1000 for that small sample of an unspecified "one day".
Quit relying on misdirection and evasion, and state the "real data from the server...okay?" (your words): the page view total for the total pages you claim you identified for "celebrities" in the top 1000 -- which will help establish whether your conjecture--"(probably much less than) 20% of the total traffic" is for celebrities--is bunkum. But of course you will instead fabricate more false straw men (or perhaps threats) to try to misdirect readers.
Wow - I didn't think debunking the worst Article Feedback Tool defender (worst at defending) would be this much fun!
