Wikipedia

Wikipedia:Wikipedia Signpost/2015-09-09/Op-ed

< Wikipedia:Wikipedia Signpost‎ | 2015-09-09
The Signpost
← Back to Contents
View Latest Issue



Op-ed

DYK, or proudly displaying incorrect information on the Main Page with alarming regularity

Contribute  —  
Share this
By Fram

"Did You Know", or WP:DYK for short, is the spot on the main page below the Featured Article, where new or newly improved articles can be shown. At the moment, 16 articles are highlighted every day on the Main Page with a one-sentence hook each, which gives some 500 articles per month a short but highly visible stay in the spotlights. During the twelve hours a hook is normally on the Main Page, the MP is viewed some 5 million times, and the articles bolded in the hooks typically get a few thousand views each: the number of readers per hook will obviously be somewhere between those two figures.

DYK hooks are normally checked by at least four people before they reach the Main Page: the article nominator (often the person who created or expanded the article), the reviewer and promoter (two editors who check the article and hook on the individual DYK nomination page, transcluded on Template talk:Did you know), and the person (an administrator) who creates the DYK sections for the main page (Template:Did you know/Queue). That's four people for every hook, and at the moment 16 hooks per day. Shouldn't be too hard?

I have written a few DYKs, so I know that part of the equation. It isn't always easy to expand an article sufficiently to make it long enough for a DYK, and many articles have no interesting hooks to highlight them with. My first DYKs, back in 2007, were for Happy Hooligan and Winnie Winkle (which later became a GA as well), and more than twenty followed over the next years, on topics like Cromwell Dixon or Achilles on Skyros. I started to look more at other DYKs at the same time, and noticed some errors and problems. My first comments were in Wikipedia talk:Did you know/Archive 87#Template:Did you know nominations/Gibraltar Rock, Porongurups. Earlier archives contain many complaints from other editors about accuracy and errors. Over the next two years, I removed many hooks from the preparation area and from the main page for having (sometimes blatant) errors.

Many suggestions to improve DYK were proposed: not showing nearly every proposed article but only the best or most interesting ones, limiting the number of DYKs per person (to get more variety instead of dozens of hooks on very similar topics), adding more reviewers, banning people with many problematic (incorrect) DYKs from nominating more pages, even abolishing DYK altogether. Some things, like the suggestion that one shouldn't only look for a source confirming the hook, but also for sources contradicting the hook, were seen as a good idea, but (as the below examples will show) have not been implemented.

A few days' worth of DYK

The provincial administrative building in Pescara

Just look at what we have had on the Main Page over the last few days:


... that the partnership between model aircraft company Airtronics and transmitter manufacturer Sanwa helped increase the former's annual profit from less than $1 million in 1983 to $8 million in 1989?
— September 2, 12-24h, Sanwa Electronic

The article says the same, sadly the source[1] says something different: "Airtronics has 18 full-time employees and had sales of about $8 million in 1989" and "After Airtronics teamed with Sanwa, sales more than tripled – from less than $1 million in 1983 to $3 million in 1985 – and have climbed steadily since" (emphasis mine). The difference between sales and profit seems quite basic and essential ...

... that evidence of the first settlement of the Province of Pescara dates to 1500 BCE?
— September 3, 0-12h, Province of Pescara

This, like many other DYKs from this series of "Province of Italy" articles, is very dubious, and easily refuted when one looks a bit further. E.g. in Catignano (in the Province of Pescara), there lived people certainly as early as 4000 BCE[2]. It is even the namesake of a whole "culture", the Culture of Catignano, which has an article on the French Wikipedia ([3]) but isn't even mentioned in the Province of Pescara article. The idea that no people permanently lived in a fertile Italian region before 1500 BCE is highly dubious for anyone with some knowledge of the subject. But then again, on 30 August we also proudly claimed for twelve hours “that the Province of Lodi was first inhabited during the fifth century BCE” (from Template:Did you know nominations/Province of Lodi), which is equally nonsense (e.g. in Guado di Gugnano, part of Casaletto Lodigiano, bronze age remains of about 1300 BCE had already been found in the 19th century). The reason for his hook: again editors and reviewers not knowing the difference between a city and a province of the same name, apparently (source for the hook: [4] clearly discusses the origin of the city of Lodi, not the province, just like the origin of Pescara in that book is about the city, not the region ...)

The hook for the Province of Pavia in the next queue should equally be taken with a grain of salt, as that article again seems incapable of making the distinction between the current province and the ancient city of the same name. Oh, and in the very next queue, the hook about the Province of Savona (did I mention the lack of variety?): Genoa only destroyed the city once ...

... that the Syllabical and Steganographical Table was the first cryptography chart ever made?
— September 4, 12-24h, Steganographical table).

Is this claim true? Hard to tell, as no definition of what does or doesn't constitute a "cryptography chart" is given. Not surprisingly, since the only books that mention this concept are the source for the claim, and parrots of it[5]. Instead of using a lightweight "first of X" record book, an actual book on the history of cryptography may be more interesting and relevant. Something like "A Brief History of Cryptology and Cryptographic Algorithms", published by Springer[6], seems to indicate that cryptography charts have been published in the fifteenth and sixteenth centuries, so the one from the hook is not even the first published one, let alone the first ever made.

But that isn't the worst hook in that queue:

... that footballer Adam Stansfield, who died in 2010, is featured on legal tender banknotes in the English city of Exeter?
— September 4, 12-24h, Adam Stansfield)

This features a Good Article (another “quality” project on Wikipedia with many deep-rooted problems). The hook is sourced to the local Western Gazette – Yeovil [7], but the hook should have rung some alarm bells for everyone with, well, some general knowledge and a critical mind. As expected, the local banknotes issued in Exeter are not, I repeat not, legal tender, which is the exact opposite of what Wikipedia claims on the Main Page and in a Good Article ... [8], section "What can I do with Exeter Pounds", states quite clearly and unequivocally: "* Exeter Pounds are not legal tender, therefore customers do not have to accept Exeter Pounds in change."

So that's on a total of 40 "hooks", 40 sentences on the main page, at least 4 factual errors (6 if you include the Province of Pavia and Savona ones). An encyclopedia where one in ten sentences, fact-checked by at least four editors, is wrong, should not be called an encyclopedia.


Going back in time

What appeared on the main page in the hours and days before I composed this complaint (2–3 September 2015)? Well, looking at Wikipedia:Recent additions, I see that the previous Main Page DYK section informed us:

... that the capital city of the Province of Isernia was destroyed by an earthquake in 874, and its cathedral collapsed in another earthquake in 1805?
— Province of Isernia

The source for this erroneous statement gives the correct year for the earthquake, not 874 but 847, as does the article on Isernia (the city) and all other sources: no sources for 874 could be found. A simple (though quite important) typo, yes, but missed by at least four people who essentially had to check one sentence before it came on the main page.

A few days further back, we learn:

... that the Minicoy Island Lighthouse is the only lighthouse in India that has a postage stamp and a first day cover issued in its name?
— Lighthouse, Minicoy island

This will be news to e.g. Mahabalipuram lighthouse and Alappuzha lighthouse (also known as Alleppey lighthouse, should have a redirect or a move), which got a stamp and FDC in 2012[9]. Finding this takes the absolute minimum of effort: searching for India stamp lighthouse on Google[10].

I have now stayed away from Wikipedia for 8 months, not editing anything at all, for a variety of reasons. The highly unreliable information we present day in day out on our main page in the DYK section, which is not time-critical (it often takes a month between article nomination and front page appearance) and gets what should in theory be sufficient scrutiny, is probably the most visible aspect of the many failings of Wikipedia. If four people, usually all experienced editors and one of them an admin, can't get one sentence thoroughly fact-checked, then crowd-sourcing as we practice and preach it doesn't work. The system is much too reliant on individuals doing one or two critical tasks. WT:DYK has become a much quieter place since I left, not because the error rate has dropped (on the contrary), but because no one seems to really care, even though this is a significant part of the face we present to the world.

Basically, DYK is too often an endless supply of dubious "facts" to grace the Main Page, mainly created by experienced editors, always checked by experienced editors. If this is the best we can present, just imagine what nonsense gets inserted and kept day in day out in countless other articles.

Fram has been an editor since 2005 and an administrator since 2007. Fram is the author of 20+ WP:DYK articles and was active between 2012 and 2014 on the main DYK talk page. Fram has been on a long wikibreak since late December 2014.
+ Add a comment

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.
  • Every time such complaints are made about DYK, I look at the current FA and invariably find that it is possible to nitpick it in the same way. For example, the current FA blurb states, "Hughes is generally thought to have died after his Spitfire was struck by flying debris from a German bomber that he had just shot down." This is misleading because there is no definite explanation of what happened and there are several plausible theories (see Australia's Few and the Battle of Britain). And note that such articles about Australian fighter aces are a recurring feature at FA and so we typically get one each month. In other words, we have a similar pattern of repetition and deviation.
[Inserted] Pinging Ian. He's traveling, though. Andrew, have you ever raised any of your concerns over the many months that any of these articles are being reviewed, at PR, GAN, FAC, and sometimes A-class? Reviewers' views are treated respectfully, in my experience. - Dank (push to talk) 20:42, 11 September 2015 (UTC)[reply]
Tks for the ping, Dan. I'm certainly happy to discuss things but I think the wording is supported by Australia's Few and several other sources -- in fact I corresponded with the author of Australia's Few to help get it right. DYK or an article lead is not the place to go into detail about every possible explanation beyond the one most commonly suggested -- that's for the main body of the article, where the various speculations have been touched on. Cheers, Ian Rose (talk) 21:04, 11 September 2015 (UTC)[reply]
  • I still beg to differ. The ADoB states, "The confusion that surrounded the crash remains unresolved." Aces High states "More recent research appears to indicates that he was in fact shot down by Bf109s" There's no consensus in these sources; just a variety of possibilities. Anyway, that's water under the bridge. Looking at the blurb for today's FA, we have "The hijackers crashed the aircraft into the western side of the Pentagon at 09:37 EDT..." Sources such as The 9/11 Encyclopedia state that the time was actually 09:37:46. This exact time appears in the body of the article but it has unaccountably been truncated in the lead and blurb. It's not clear why we should be rounding the time off but, if we're going to do this, it should be to the the nearest minute, which is 09:38 not 09:37. My point, of course, is not to find fault with these particular articles but to demonstrate that it is possible to nitpick any and all of our content, just by taking a hard look at it. Andrew D. (talk) 22:49, 11 September 2015 (UTC)[reply]
  • The time is debated on my talk page. Meanwhile, the main page has moved on. Today's FA howler is "Unlike in much of his work, Waugh did not..." The double preposition makes me wince, especially as the topic is a work of stylish prose. See the Economist's style guide for more on that. Andrew D. (talk) 16:52, 16 September 2015 (UTC)[reply]
The answer to all this is found at the bottom of the main page – the disclaimer which states, "WIKIPEDIA MAKES NO GUARANTEE OF VALIDITY ... Please be advised that nothing found here has necessarily been reviewed by people with the expertise required to provide you with complete, accurate or reliable information." Perhaps we should state this more openly and emphatically as we are never going to achieve 100% perfection. You might still think that a 10% error rate is too high but please compare with the following recent headline: Nearly 2/3 of psychology papers cannot be trusted. Scholars such as Prof. Ioannidis reckon that a remarkably high number of peer-reviewed papers are wrong – as many as 80%. We seem to be doing reasonably well compared to them. Andrew D. (talk) 12:59, 7 September 2015 (UTC)[reply]
It's comparing apples and oranges in any case, but you are comparing 2/3 papers containing an error with 1/10 sentences containing an error (in this very small sample). Fram (talk) 09:51, 9 September 2015 (UTC)[reply]
  • Well, the famous/infamous Nature study suggests that rates of error are endemic in Wikipedia and in the most prominent encyclopedia. [11] [12]. Moreover, Nature admits its reviewers probably made some mistakes, themselves (published mistakes in peer review . . . sounds familiar). [13] (last paragraph - interestingly, it calls them "honest errors" - perhaps, they mean unitentional?). Human failing being what it is, it can be hard to know how much perfection is the right amount but yes Wikipedia is bad - sometimes seemingly endlessly, so -- nonetheless, it is a worse shame if any thinking person does not expect that to some degree. Between the hope and idealism and the despair and reality, each person may decide.
The editors' work criticized above should be directly contacted and asked to respond (like Nature did). If DYK needs a Final Review Board, and there are 2-3 people like Fram was, willing to do it, just create it - but that, of course, will also have errors. -- Alanscottwalker (talk) 16:59, 7 September 2015 (UTC)[reply]
That Nature "study" was very poor science. No one should take it seriously. Thank you for an excellent report, Fram, and it's good to have you back. --Anthonyhcole (talk · contribs · email) 18:32, 9 September 2015 (UTC)[reply]
Is there a science of mistake in written work (erratumology? lathosology?) All Nature did was get "experts" to blind-read articles and find mistakes - what would it mean to take that seriously? Alanscottwalker (talk) 13:12, 12 September 2015 (UTC)[reply]
  • Worse is when the DYK tries to trash a living person - without any evidence at all. This one (pulled) was unforgivable. It was blind luck that I happened to look at the queue before it hit the Main Page. Black Kite (talk) 20:33, 9 September 2015 (UTC)[reply]
  • Sorry, but there's enough problems with the DYK process without this sort of ridiculous nit-picking. The difference between sales and profit? Oh noes! Come on. If you don't like it, don't read it, but to talk down to everyone involved in the process for minor, and in most cases questionable, problems seems beyond the pale. Maury Markowitz (talk) 20:04, 11 September 2015 (UTC)[reply]
  • If you don't think there's a major difference between sales and profit then I have a bridge to sell you (It's actually just a plank of wood, but both things cross water). I'm not sure how I feel that you don't think there's an important difference between the two when you claim to work at a hedge fund; I guess you actually just day-trade penny stocks as a hobby (the difference between the two is nit-picking).
  • It's not a minor point he's making- literally the only thing the reviewer/poster is supposed to do is see if the sentence has a source in the article that says the same thing. Apparently, 10% of the time, all they do is check if there is a little [1] after the sentence. That's kind of a problem. --PresN 20:43, 11 September 2015 (UTC)[reply]
  • Oh, wow. If you are arguing that that is a minor error I think you have a WP:COMPETENCY problem and should not edit Wikipedia. Ever. DreamGuy (talk) 21:36, 12 September 2015 (UTC)[reply]
Sigh. The minor error is not sales vs. profit, the minor error is that the wrong term made it into the hook. That's minor, because the error is not going to cause the Wikipedia to fail, which appears to be what Fram is trying to say. On the contrary, misunderstanding what people are trying to say and then suggesting they be kicked off the project is precisely the sort of thing that is causing the project to fail. In any event, given the multiple reviews the article gets as it stands, I suspect that any remedy that attempts to "fix" this problem will ruin the whole process. So which is it, do we accept a certain number of errors will get through, or make the process so rigourous that nothing ever passes? Maury Markowitz (talk) 19:45, 13 September 2015 (UTC)[reply]
Obviously this or any single error will not cause Wikipedia to fail, speaking of strawman arguments... Symptoms of failure are not causes of failure. Apart from that, after getting a to-be-expected thrashing for your opening post, you now try to make a completely different point and claim that we are misunderstanding your position. Perhaps, if your opinion is "if you don't like it, don't read it", then perhaps you could just have skipped this op-ed instead of reading it and replying to it. If I have included "minor, and most cases questionable, problems", feel free to bring them here for discussion. Your first attempt backfired quite badly. As for your final question, if those are the only choices: with the current error rate (and the similar error rate going back for years), my vote is for option B. Fram (talk) 07:00, 14 September 2015 (UTC)[reply]
The following is copied from the current version of the article: "As a result, Airtronics increased its profit from less than $1 million in 1983 to $3 million in 1985, and to $8 million in 1989 with a staff of 18 employees." Cited to a source that says: "Airtronics has 18 full-time employees and had sales of about $8 million in 1989. [...] After Airtronics teamed with Sanwa, sales more than tripled--from less than $1 million in 1983 to $3 million in 1985--and have climbed steadily since." The wrong term did not just make it into the hook, Maury Markowitz. Andreas JN466 06:43, 15 September 2015 (UTC)[reply]
  • Fram is not talking about the rate of errors here but the types of errors. The article writer can misread, misunderstand or mistranscribe the source; or the source may itself be wrong. A leading cause of error seems to be the requirement to paraphrase. Let's not kid ourselves that DYK articles are more error prone than average. They get checked multiple times. And part of the reason for having DYK, and sending articles to it is to have them checked. All our quality control relies on editor review. That has its weaknesses. Hawkeye7 (talk) 21:30, 11 September 2015 (UTC)[reply]
    Speaking of which, why is the draft banner still on this page? Hawkeye7 (talk) 21:33, 11 September 2015 (UTC)[reply]
  • I have just discovered that a reference book I have been using, produced by respected publishers Methuen, has the wrong date of death for someone. They were out by one year. Shall I throw the whole book out or just make a correction on that page? I wonder how many people were involved in the production of the book? Philafrenzy (talk) 01:11, 12 September 2015 (UTC)[reply]
    • Philafrenzy nails it. Fram has demonstrated what it's like to throw the baby out with the bathwater. Most surprising, this fallacious line of reasoning appears to have escaped the editors of the Signpost, who should have prevented its publication unless Fram rewrote his editorial to conclude with suggested recommendations for improvement and change. Without that, this is simply delicious fractal hypocrisy, humorously meta-demonstrating the very problem Fram complains about. Viriditas (talk) 01:20, 12 September 2015 (UTC)[reply]
      • Amusingly, our article Surrey Theatre actually already had the right date (1901). Philafrenzy (talk) 01:26, 12 September 2015 (UTC)[reply]
      • As you are well aware, I have made enough recommendations to improve DYK, as have others, and I listed a few in the op-ed. And could you perhaps indicate which "fallacious line of reasoning" that is? I am not asking to throw ouw the baby with the bathwater, the hooks are all there is to DYK, they are the stagnant pool of odiferous bathwater you wouldn't want to put a baby in. Of course, as so often was the case when I pointed out these errors on the DYK page, shooting the messenger is easier than actually tackling the issues at hand. See also Prioryman's message below for more of the same, though without the pseudo-intellectual veneer. Fram (talk) 19:19, 13 September 2015 (UTC)[reply]
        • Sorry, Fram, but this piece utterly fails the basic, requisite structure of a five paragraph essay, and as an opinion piece, it offers no suggestions for improvement or concluding remarks other than the childish and absurd "no one really cares". Seriously, Fram? Seriously? I suppose this is acceptable if it was written by a young child, but the big boys and girls expect a bit more than a plaintive cry for sympathy. Viriditas (talk) 21:47, 13 September 2015 (UTC)[reply]
          • Oh no, and I was so aiming to write a school essay in five paragraphs! Boohoo, this dreadful news just ruined my whole week... Opinion pieces don't have to offer suggestions for improvement, I have given those (as stated in the piece) at WT:DYK in the past, but some of the entrenched DYK regulars resist all criticism or attempts to improve things. The purpose of the piece was not to suggest improvements, but to identify a problem spot which has gone unnoticed or uncared about for too long. First identify problems, then find the causes of the problems, and then look for remedies. My piece aimed to do the first bit, not all three in one go. And Viriditas, you may find "no one really cares" (actually "no one seems to reallly care") childish and absurd, but people like you, or Prioryman, or Maury Markowitz above, do give a very good impression of not caring one bit about having basic errors passing DYK review easily and regularly. Fram (talk) 07:00, 14 September 2015 (UTC)[reply]
            • Fram, the basic, requisite structure of a five paragraph essay necessitates an introduction, narration, affirmation, negation, and a conclusion, and this structural component is found in every good editorial or opinion piece. So, yes, opinion pieces most certainly do have to offer suggestions for improvement as part of their structure, and without it, your essay is just a lengthy whinge, with no rhyme or reason. A closer reading, however, reveals projection on your part, as the only person that doesn't care here is you. This piece demonstrates that you care not one whit for the reader, for without a negation and conclusion, there's no purpose to your complaint. Furthermore, this piece illustrates a lack of social intelligence on your part. You've insulted every editor who spends their valuable time reviewing DYKs and tracing errors, and you've shown little to no awareness of your faux pas. I think the real reason you are no longer active on Wikipedia is because you don't understand interpersonal communication. Viriditas (talk) 21:58, 14 September 2015 (UTC)[reply]
              • What I produced was what is described here, especially "a simple, declarative op-ed (“policy X is bad; here’s why”)". That your replies focus on perceived procedural problems with the op-ed, and on the writer of it, instead of on the actual contants of the op-ed, and the underlying structural problems of DYK, is quite telling. Baseless speculation on why someone chooses not to edit here any longer, and using that to attack someone, is a (mild) personal attack though, and a rather poor ad hominem argument in any case. So, do you care about the number of errors that get on the main page through the DYK process, and if so, do you have any suggestions on how to change this? Perhaps anything from the op-ed paragraph "Many suggestions to improve DYK were proposed: not showing nearly every proposed article but only the best or most interesting ones, limiting the number of DYKs per person (to get more variety instead of dozens of hooks on very similar topics), adding more reviewers, banning people with many problematic (incorrect) DYKs from nominating more pages, even abolishing DYK altogether. Some things, like the suggestion that one shouldn't only look for a source confirming the hook, but also for sources contradicting the hook, were seen as a good idea, but (as the below examples will show) have not been implemented." (from that op-ed that should "offer suggestions for improvement" but didn't, in your version of reality: or should I, to meet the strict criteria of what an op-ed must be to be acceptable to you, have somehow suggested new solutions instead of listing the ones proposed in the past, as if these weren't worth considering any longer?). Feel free to continue your some line of argumentation as in your previous posts here, but don't expect further answers in that case; replying to strawman and ad hominem arguments isn't entertaining or fruitful for very long. If you want to discuss DYK, then I'm all for it though. Fram (talk) 10:12, 15 September 2015 (UTC)[reply]
  • There is a simple answer: wikipedia works based on the famous say about "all bugs are shallow". Why not enough eyeballs at DYK? answer is simple *BORING*. Nobody really peruses them. It is not like in goode olde dayse when every DYK was a pearl: you could even use them as pickup phrases :-). Now: "DYK that the mountain degu can obtain sufficient moisture from its food to satisfy its water requirements?" Sad. (Language on stilts ("sufficient", "moisture", "water requirements"; Nothing special; many small animals can do without drink; Generated by a DYK-craving author (September 9: 4 DYKs; sheesh, be more modest, colleague))) Staszek Lem (talk) 01:53, 12 September 2015 (UTC)[reply]
  • To err is human, and all publications are full of errors. One of the most frequent problems I encounter when writing articles is conflicting sources, which means some of the "reliable sources" are certainly wrong. I've been reading a well received book by a highly regarded Harvard historian, and found errors every few pages, but I still learn a lot from it. The best thing about Wikipedia is whenever you find something incorrect, you can fix it instantly. But let's not kid ourselves, Wikipedia is full of junk, POV, and falsehood, and at the current rate of improvement (with simultaneous degradation - many GAs become junk after a few unsourced IP edits), it will become obsolete long before it becomes perfect. Just click the "Random article" (gasp, it's linked from the main page) a few times and you'll see what I mean. -Zanhe (talk) 09:26, 12 September 2015 (UTC)[reply]
    • You seem to suggest that because stuff isn't perfect we shouldn't do quality control, Zanhe. You ticked off on Template:Did you know nominations/Province of Lodi. The citation for the hook, as it turns out, is not offline at all--here it is. I don't wish to argue that it's not a reliable source or that print sources don't make errors, but hey, first of all it says "settled", not "inhabited", and one could argue they're different. More importantly, and I guess Thine Antique Pen didn't notice it either, that paragraph isn't talking about the province but about the city. And that's a huge difference: a claim about settling the city is now made to apply to the area that city is in, an area in--as Fram I think said--a fertile area in the heart of Italy, over 300 square miles big. (I see now Fram also linked the source, and commented on the city/province point.) So, sorry, but you ticked that off way too quickly and way too easily, and I think the same happened to Template:Did you know nominations/Province of Pescara. Someone with over 200 DYKs should know better, and someone with that much experience should know that some articles need to be looked at closer than others. Drmies (talk) 16:48, 15 October 2015 (UTC)[reply]
      • @Drmies, Cwmhiraeth, and Thine Antique Pen: No, what I meant is no amount of quality control will be able to catch all mistakes and satisfy all critics. Wikipedia's scarcest resource is good editors, and constantly criticising prolific contributors on minor issues with no real long-term consequence is counterproductive. After all, the more work you do, the more likely you make mistakes (I've already dramatically slowed down my contributions lately, after watching another long-term editor being hounded off Wikipedia by multiple SPA attackers). As for Lodi and Pescara, I obviously would have suggested changes if I had access to the sources, but I ticked them off based on the AGF principle because I didn't. DYK rules do not require sources to be online and do not require reviewers to track down offline sources (which is often impossible). -Zanhe (talk) 19:23, 15 October 2015 (UTC)[reply]
        • I've had similar discussions before. Of course we can't catch every mistake, but we can do a better job. What irks me is the "DYK doesn't require" this or that. DYK doesn't require lots of things; it doesn't actually require good writing or an exciting topic or decent coverage of the subject in the article. Or solid sourcing. That doesn't mean that an article we put on the front page shouldn't have all those things. And there is a bit of slippage between "if I had access to the sources" and "ticked them off on the basis of AGF" (you don't have to link that for me): we do have access to that source, and AGF doesn't mean one doesn't have to look for it. And yes, I believe DYK reviewers should improve the article whenever they can (TAP knows I'm a believer), and in this case that wasn't hard. Drmies (talk) 19:44, 15 October 2015 (UTC)[reply]
Although in no way an excuse, this article was expanded by Thine Antique Pen as part of the Stub contest, with no intention of it becoming a DYK. That it did so was my fault, as I asked TAP if he objected to my expanding it further and putting it forward for DYK. So I added the geography section and a bit more history and nominated it. Cwmhiraeth (talk) 19:50, 15 October 2015 (UTC)[reply]
Cwmhiraeth, that it went through DYK with errors is not your fault, and you didn't write that section. BTW, it's nice to see you're still around. I feel like a dinosaur sometimes. Drmies (talk) 19:58, 15 October 2015 (UTC)[reply]
  • The quality of this piece would have been marginally improved if it had been published in 20 point Comic Sans throughout. Prioryman (talk) 09:32, 12 September 2015 (UTC)[reply]
  • Interestingly, Fram is not immune to error. The article Adriano in Siria was written by Fram and nominated for DYK with a careless error which was fortunately picked up by the reviewer (me). Cwmhiraeth (talk) 09:34, 12 September 2015 (UTC)[reply]
  • In which case that's an example of the system working correctly. Black Kite (talk) 09:46, 12 September 2015 (UTC)[reply]
  • And of course, that was an error in the article, not in the hook. So your reply kind of completely misses the point of the op-ed. Fram (talk) 19:19, 13 September 2015 (UTC)[reply]
  • And yet, it doesn't really miss the point, because at the end of the day, the hook links to the article, error-prone or not. The accuracy of the article comes first, which helps insure the accuracy of the hook. So far from missing the point, Cwmhiraeth has nailed the solution -- the solution you failed to provide. Viriditas (talk) 04:22, 17 September 2015 (UTC)[reply]
  • Brilliant. But for those of us less capable of mind-reading, could you explain that "solution" that Cwmhiraeth provided to avoid regular errors in DYK hooks on the main page? And perhaps explain why it didn't work, if it has been in place for so long already? From here, it somehow looks as if that solution, whatever it is supposed to be (find a glaring typo errors and ignore the ones that actually take some source-control to find? Something like that?), isn't really that good. Oh wait, I get it: if people, instead of making the effort to check that one sentence is error-free, would make the effort to check that the complete article is error-free, then DYK would be error-free. I'ld like to retract my "brilliant" and replace it with "mind-bogglingly stupid". If you can't find the people, the time, and the competency to check 16 sentences a day, then how would you propose your "solution" would work in practice? There is a reason I didn't provide this "solution" (make the articles error-free), as you would have rightly ridiculed it. I somehow don't think @Cwmhiraeth: intended this to be the solution for DYK errors you read in it, but I'll let them state their own position on it. Fram (talk) 06:56, 17 September 2015 (UTC)[reply]
We are all human and prone to error and my point really was that it is risky criticising others when one's own efforts are not perfect. However, Fram did perform a useful service to the DYK project during 2014, drawing people's attention to problems with hooks that others had missed, and as a result the quality of the DYKs was raised. Fram may not think so, but it has had a lasting effect on me. Cwmhiraeth (talk) 08:40, 17 September 2015 (UTC)[reply]
Thanks. That's basically the reason why I didn't make an op-ed criticizing people, but one criticizing the process by pointing to problems with the end product. I'm not certain that the quality of DYKs in general has been raised, but that's not a criticism of you. And my articles contain plenty of errors, though hopefully mainly typo's and outdated information. Fram (talk) 09:03, 17 September 2015 (UTC)[reply]
  • While it's true that errors are inevitable in a project of this type, and that DYK adds a level of quality control that would not otherwise exist, I agree with Fram that there can be little excuse for the propagation of errors in DYK hooks that appear on the main page - at least, not at such a rate. A DYK hook, after all, is no more than a single sentence that needs to be fact-checked - and yet, too often, DYK reviewers cannot even achieve this much. Why not? Well, a lot of DYK reviews are now, of necessity, QPQ reviews, meaning a review that must be done before one's own DYK nomination is accepted. It is pretty clear to me, as a longstanding DYK administrator, that some QPQ reviewers simply don't apply themselves to the task. And why should they, when they know there are no consequences for errors, even egregious ones? So long as we do not establish greater accountability for QPQ reviewers, I think we can be assured that issues in the DYK section such as that identified by Fram will persist. Gatoclass (talk) 11:50, 12 September 2015 (UTC)[reply]
  • You forgot to mention the proposal to switch DYK to use Good Articles instead of new articles. I don't remember when it was discussed, but it was presented as a way to decrease the amount of completely bogus information in DYK. Kaldari (talk) 21:20, 12 September 2015 (UTC)[reply]
That's a terrific idea. Too bad that, given Wikipedia's inability to change, that will never happen. ResMar 23:43, 12 September 2015 (UTC)[reply]
You raise an interesting point. What exactly is the deal with the stubborn, inflexibility of the site demographic? I'm looking at the main page right now and it's like looking back in time a decade. Do people realize the rest of the Internet has moved on? Viriditas (talk) 01:14, 13 September 2015 (UTC)[reply]
 Viriditas: You're touching on a well-known, well-entrenched, and ultimately likely lethal problem with the movement: there is a total lack of flexibility on the part of the editors because of the massive investment that the senior editors who have come to dominate this site have in keeping it as it is or as close as possible to as it is today. The main page looks to be a decade old because it is a decade old. DYK has the same issue. ResMar 21:17, 13 September 2015 (UTC)[reply]
@Resident Mario: I don't understand why anyone in their right mind would want to keep this site "as it is". Presentation and design evolve over time as we discover new ways to do things. Can anyone name a top website that hasn't changed in the last ten years? Something is seriously wrong here. Viriditas (talk) 22:22, 13 September 2015 (UTC)[reply]
Drudge Report. Like wiki, it's simple on its surface. Unlike wiki, it's also simple behind the scenes. Alaynestone (talk) 23:59, 16 September 2015 (UTC)[reply]
That's a terrible idea, based on the barely-reviewed GAs that are routinely submitted to DYK. Too many articles get their green pluses with little or no commentary and little sign that the reviewer has checked the article against the GA criteria. The whole point of DYK is to reward new content creation so eliminating newly created content from the process is a non-starter. I'm never impressed by people that wish there was a Grand High Poobah overseeing a project and, by golly, they're exactly the person for the job even though they dislike the actual work involved and no longer soil their hands with such tasks. - Dravecky (talk) 03:29, 13 September 2015 (UTC)[reply]
The RFC (Wikipedia:Did you know/Good Article RfC) discussed whether recent GAs should be added to the class of articles suitable for a DYK appearance. The proposal passed. --Anthonyhcole (talk · contribs · email) 16:49, 13 September 2015 (UTC) (I put this in the wrong spot - pinging User:Kaldari. --Anthonyhcole (talk · contribs · email) 17:05, 13 September 2015 (UTC))[reply]
(e/c) That was approved. And Fram's OpEd says one of DYK mistakes was from a good article. Alanscottwalker (talk) 17:01, 13 September 2015 (UTC)[reply]
  • The author has identified the failing of this process is that Wikipedians (the nominator, the reviewer, and the admin) aren't doing their jobs. Why does this happen? There's pressure to publish on time. Chris Troutman (talk) 04:17, 13 September 2015 (UTC)[reply]
    • If you care to look into publishing times, you'll see something is published two weeks and even month after deadline (used to be 5 days). Staszek Lem (talk) 19:18, 13 September 2015 (UTC)[reply]
    • "publish on time"? DYK has been repeatedly asked to reduce its publication rate in an attempt to catch some of these issues, there is a real objection to it. Why DYK feels a need to cycle quicker than any of the other parts of the main page is a mystery, a bit like the myth that DYK is there to encourage new editors to contribute. Evidence please? The Rambling Man (talk) 20:46, 13 September 2015 (UTC)[reply]
      Articles have to be nominated within seven days of creation (or achieving GA status), but they can be worked on for as long as the creator wishes in the Draft space or User space. The days when the review had to be completed in as many days are long gone; nowadays articles sit on the queue until the reviewers are satisfied. So there is very little deadline pressure. Hawkeye7 (talk) 22:41, 13 September 2015 (UTC)[reply]
      • And nowadays articles sit in the queue because the process has been bureaucratized to extremes. In the past the only three questions were asked: Is is a new article? Is it more than a stub? Is the hook catching the eye?. Done. No paranoia to have an error in the article: this is wikipedia, not cast in stone; the crux is "INTERESTING". If something is interesting, the reader will be ..er.. interested and do more research (and I would not care about passive trivia consumers, who just read and bitch). If an interest is provoked, chances are the person would want to share her interest of excitement, may be even with wikipedia. Staszek Lem (talk) 00:57, 14 September 2015 (UTC)[reply]
And that process changed because of articles like these - well, ones in the "real press" anyway. You can reminisce all you want, but those days are dead. Maury Markowitz (talk) 10:32, 15 September 2015 (UTC)[reply]
Dead, not without the self-serving efforts of self-nomming "DYKes of Main" Staszek Lem (talk) 17:17, 16 September 2015 (UTC)[reply]

Consolation prize

If it's any consolation, this Signpost has an article on the 2-millionth article in the Swedish Wikipedia, which led me to a much more embarrassing ongoing situation with their millionth article of 2 years ago.[14] Fram (talk) 13:22, 15 September 2015 (UTC)[reply]

  • I don't speak Swedish either but that issue seems related to a point made above – that much scientific literature is debatable or just wrong. Taxonomy is especially fluid and seems about as reliable as our category system. See Taxonomic inflation which explains that "Clearly, there are big problems with species lists." Andrew D. (talk) 17:20, 16 September 2015 (UTC)[reply]
hey Fram, are you still blocked from bugzilla? have they extended it to Phabracator? this kind of critique is counter-productive: the signal to noise is so low, people tune it out. however, you are a good poster child for the community; anyway, thanks for all the dramaz. Duckduckstop (talk) 21:45, 17 September 2015 (UTC)[reply]
And the relevance of bugzilla/phabricator to this op-ed is...? For someone complaining about the signal-to-noise ratio, you surely aren't giving the best example here. Oh wait, could the fact that a long time ago I blocked your other account User:Duckduckgo as a username violation have anything to do with your appearance here with a post which has nothing to do with the op-ed but is only poisoning the well about the author of it? Nah, no one would be that childish. Fram (talk) 14:48, 18 September 2015 (UTC)[reply]


The Signpost is looking for new talent.
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Wikipedia_Signpost/2015-09-09/Op-ed&oldid=1193875250"