Hacker News Reader

Comment by notadoc | Hacker News Reader

An impressive amount of information on Wikipedia is just outright wrong, misleading, clearly biased, or astroturfed. At the rate in which I encounter inaccurate data on Wikipedia that I can identify, it makes me suspicious of nearly everything on the site. It's an interesting experiment to allow the masses to create and edit their own facts and it can provide a broad (albeit often inaccurate) overview of some topics, but it's hardly a substitution for a legitimate source let alone encyclopedia.

Nonetheless, I know plenty of people who take it as legitimate and cite it as if it's an authority.

View on HN

This is a rare opportunity to cite the Gell-Mann Amnesia Effect: "You open the newspaper to an article on some subject you know well. In Murray [Gell-Mann]'s case, physics. In mine, show business. You read the article and see the journalist has absolutely no understanding of either the facts or the issues. Often, the article is so wrong it actually presents the story backward—reversing cause and effect. I call these the 'wet streets cause rain' stories. Paper's full of them. In any case, you read with exasperation or amusement the multiple errors in a story, and then turn the page to national or international affairs, and read as if the rest of the newspaper was somehow more accurate about Palestine than the baloney you just read. You turn the page, and forget what you know."

http://www.goodreads.com/quotes/65213-briefly-stated-the-gel...

I'm shocked - there is no Wikipedia article on this!

I've been pointing this out about journalism for years, and it's awesome that it has a name. However, I still want a name for just the first part of the effect - the part where you notice an article about your area of expertise is laughably wrong. And you never forget it and conclude that journalism is just terrible.

There's a similar saying referred to as Knoll's Law of Media Accuracy: "Everything you read in the newspapers is absolutely true except for the rare story of which you happen to have firsthand knowledge."

There's also another related topic on TVTropes: http://tvtropes.org/pmwiki/pmwiki.php/Main/CowboyBebopAtHisC...

I'm not sure that's a Wikipedia problem, as opposed to a general sources problem. Wikipedia at least often shows proponents and critics of a given idea, while the average source just presents its own position and doesn't provide any sort of context. I'm not entirely sure what you mean by "legitimate source", I don't think I've ever seen one. Often an old book has more bias and inaccuracies in it than a blog post. I recall reading a discussion on how The Rise and Fall of the Roman Empire is no longer perceived as accurate, and few people will call that an illegitimate source. Information is hard.

All Wikipedia does half the time is reference such sources, so it's about as biased as the sources are, except it tends to combine them, and there's nothing forcing it to do so in a way that represents the true balance. That's up to the content editors. The only thing I can suggest is, if you see a problem, try to correct it.

Also, pretty much every page these days has a warning on it.

Wikipedia is great for basic coverage of a fairly broad range of topics, such as physics, psychology, linguistics, religion, biology, computer science, etc. As long as you're sticking to information that's mostly perceived as fact, as opposed to opinion. I have yet to find a source that rivals Wikipedia in this regard. Wikipedia had a huge, huge influence on information transfer for many people, I think we forgot how big of a deal it is because it's been around for so long. It's absolutely invaluable to many people.

I don't believe it's Wikipedia's job to identify answers to opinionated stances, so don't expect it to be "accurate" on things like controversial history, politics, design, etc. Those things are usually not covered accurately by anyone.

Curious, what areas do you find are usually flat out wrong?

The math sections seem to be pretty good. A friend found a flaw in a crypto algorithm though.

I've eventually learned to limit my usage of the Wikipedia to science, distant history, and synopsis of works of popular culture while avoiding like plague current politics and ongoing events.

...but thats on english wiki only. As pole understanding english, I despair for my natives trying to use polish wikipedia, because here politics creep everywhere. I remember what a disaster polish's article on Big Bang was before its ridiculous state was called out and went viral in sceptic circles as being plainly antiscientific. Until then it's contents were more fittingly titled as "philosophical and religious criticisms of Big Bang theory". Remnants of those may be found in Article's discussion which filled with philosophical and religious debate about Big bang.

Honest question, what alternative do you use instead? HN mostly links to news article, and I would question if current politics and ongoing events are more accurately written by news media and with less bias. I don't know about polish news companies, but its not uncommon elsewhere that news companies are openly biased towards one political party or bias their articles towards their main customer group (UK example from Yes, Minister: https://www.youtube.com/watch?v=DGscoaUWW2M).

We have books or research papers, but meta research generally acknowledge that researchers include bias into their work, often linked to the opinion of those that funds the research. Meta research is thus generally a bit better, especially the meta-meta-research papers, through it generally takes a quite political contested topic for that to happen which then further increase the risk of bias by the meta researchers selecting results that favors their side.

In terms of news, I use Haystack (http://www.haystack.tv). It aggregates YouTube videos from various media sources daily, much of it American (then again, that might be intentional). You can also pick and choose specific sources and topics you want to follow, even cast to Chromecast. The developer(s) are pretty responsive too and the app is free.

My only frustration is that it only aggregates videos on YouTube. If you want to get video news segments from media companies that publish their videos to a proprietary CDN or hosting service, you can't use it with Haystack.

Topics of contention between two nations (or worse, between two political factions inside a nation) are usually suspect. Even if every sentence is meticulously sourced, you can't tell if someone is "lying by omission." Basically, the tone of the article is determined by which country (or faction) can muster more people with free time and passable English skills, which may not correspond to actual expertise on the topic...

I agree that math/physics sections are pretty good. It's difficult for knowledgeable people to disagree on these matters...

> Even if every sentence is meticulously sourced, you can't tell if someone is "lying by omission."

This is a problem for literally every source on Earth, and persists even among the most prestigious newspapers, books, etc.

> This is a problem for literally every source on Earth, and persists even among the most prestigious newspapers, books, etc.

That's true but practically meaningless. Some sources are dramatically more reliable and accurate than others. I can't completely trust my 4 year old nephew about physics, nor can I completely trust a leading physics text book, but that doesn't make those sources similarly reliable.

What we're discussing here is where Wikipedia falls on that continuum. I agree with the GP; there is so much deception by omission in WP that I don't trust it. For example, I was just looking at (American) football player Peyton Manning's article; it completely omitted a major sexual assault/harassment allegation, one about which there was a court settlement, book, major news coverage, etc.

I mean explicitly that Wikipedia does as least as good on lying by omission as the most reputable sources (NYTimes, the Wall Street Journal, whatever).

Surely there must be some significant differences? It's hard to believe they are all the same. Claiming they are all the same is a strategy of propagandists (I'm not saying you are one; I'm saying it's a dubious approach). 'It's all the same' is the opposite of truth and accuracy, which require discernment; it the justification of liars (again, not the parent).

Anyway, the parent's claim isn't my experience, but now we're just one person on the Internet disagreeing with another.

I'm not saying their all the same. I'm saying Wikipedia is generally better (though of course far from perfect), without trying to make difficult-to-quantify claims about how much better.

The Biographies of Living Persons policy is rather strict, because of the potential for libel lawsuits. The Wikipedia foundation can barely afford to keep Wikipedia running as it is, let alone deal with hundreds of libel lawsuits.

> It's difficult for knowledgeable people to disagree on these matters...

See eg "Speed of light" for an example of the complications that people get into in writing an article that's readable by the general population, but which is also correct for the 5% of wikipedia editors who have a good grasp on this.

See this thread for an example: https://news.ycombinator.com/item?id=13468651

The math sections are okay up to what a typical mechanical engineer would know. My own sub-field is very poorly covered.

Do you submit corrections?

Do you? Seems like there should be a xkcd covering this. The usual case as near as I can tell is something like: Make a minor correction to an obvious error, including a link to a source and other wikipedia articles. Edit gets immediately reverted by whoever has declared themselves the guardian of that particular page. Talk page lights up with all sorts of tangential discussion of whether or not this particular subtopic should be corrected or deleted or reworded or something else. Someone else tries making a modification, which again is immediately reverted. People who enjoy the drama have a self-interested desire to perpetuate it. If you are lucky you could go back 6 months later and there's maybe a 50% chance that the original has been corrected (that's if the original topic hasn't come into contact with the non-notable deletionists).

Anyone know when C2 became a javascriptified dumpster fire? Seems like that's a "Day the Music Died" event that should have a date associated with it.

(Not OP) I have in the past, and my experience was that the talk page was civil, my changes were discussed and we came to a common agreement as to what the change should be. Nothing like the fiasco you're describing.

You still haven't answered the question though; do you submit corrections?

>You still haven't answered the question though;

I was never asked a question until now (N.B. the original posters).

> do you submit corrections?

I haven't for years (as in over 5+ years). I'm sure someone will point out it is different now.

> The usual case as near as I can tell is something like:

Based on this I'm not sure you have edited yourself, despite retorting the parent. First off articles don't have guardians; [0] just because someone disagreed with you, it doesn't mean they disagree with everyone or on every article, you're making a huge generalization. Second, I'm not sure why you see talk page discussion as a bad thing; when two parties disagree on something, they usually discuss and try to reach a consensus. [1] Do you believe you're above that or something?

> Someone else tries making a modification, which again is immediately reverted.

This is called edit warring [2] and shouldn't be done. You shouldn't just try to force your revision in after it has been disagreed with.

All you're doing in your comment is painting up some illusory image to discourage someone from trying to engage the system themselves. How about you let them edit and see for themselves if what you said is true? Based on my experience Wikipedia isn't anything like that at all, in fact most articles aren't watched enough for anyone to care how you tweak them.

[0]: https://en.wikipedia.org/wiki/Wikipedia:Ownership_of_content

[1]: https://en.wikipedia.org/wiki/Wikipedia:Consensus

[2]: https://en.wikipedia.org/wiki/Wikipedia:Edit_warring

Yeah, that's a hopelessly optimistic viewpoint, I fear. I have many hundreds or more contributions, but stopped a few years ago.

Tried coming back and immediately ran into edit wars on a list of fastest production cars, when legions of people were eager to get the Tesla on there on the strength of a press-release of an upcoming release. Rules warring, silent reversion, you name it.

Wikipedia Review (though not without its share of cranks, and now largely dead), shares a pretty good history of the more sordid history of WP.

When complaining about inadequate behavior from other Wikipedia editors, it would be very useful to point out concrete links to edits and reverts. This gives a chance to other interested parties to fix the article and contribute to the discussion.

Used to, circa 2011. Got tired of rules lawyers and deletionists.

Anecdotal, but what here isn't now: I used to edit as well after noticing how incredibly wrong some of the less popular but obviously wrong articles were. Until one day I made some anonymous corrections with proper sources properly cited. I would see my edits immediately undone. I went back and re-did my edit, rewording it to be clearer, and explaining my citations, assuming I was at fault and if I simply corrected MY mistakes the edit would go through. It did not, and this time I was notified I had attempted to vandalize said page and would be banned from making further edits if I persisted. I went on to create an account and try and argue my points civilly in the talk page, to no avail, and being attacked by the article's caretaker (who by the way knew ABSOLUTELY NOTHING about the article's underlying facts/science, anything). After this I left the editing of wikipedia to people with far more patience than I, and I really believe after seeing other talk pages with glaring errors on the wiki article, that this is the modus operandi for many of these "caretakers". You step on their territory and they shoo you away as quickly as possible(they get so bad and pedantic with their arguing they call themselves "wikilawyers"). Still upsets me to this day that when I see an obvious mistake such as a bad date, misspelling or other easily identifiable misinformation I can't be bothered to do more than wonder about what the poor soul who tried to correct it went through. /rant

> Until one day I made some anonymous corrections with proper sources properly cited. I would see my edits immediately undone.

I understand where you're coming from. I've edited for items such as the citation having nothing to do with the content, and seen it reversed and gone through the same thing. Sometimes something as simple as "show me where I can find this in your citation" brings more than civil discourse -- it brings obvious animosity and is countered with straw man attacks or other logical fallacies. I think people have psychological and emotional needs, and some have a need to feel important, dutiful and powerful. This is where they satisfy that need.

Another problem is I think people in authority--sometimes very smart people--draw conclusions that they haven't properly thought about.

Ah, yes some areas are lacking. I am more concerned about correctness though. Do you ever find math material that is completely wrong on Wikipedia? Stuff like number theory, abstract algebra, and category theory seem to hold up when I cross reference. I'm in no position to qualify myself as mathematician though, I just enjoy mathematics a lot.

Really? I find many math topics to be covered extremely poorly, and presume a lot of prior knowledge without linking to that knowledge.

Plenty of philosophy articles are also very poor. For example the page on Deconstruction, though its better than it was ~4 years ago.

https://en.wikipedia.org/wiki/Deconstruction

>Really? I find many math topics to be covered extremely poorly, and presume a lot of prior knowledge without linking to that knowledge.

At what level, an ELI5 level or the level of people who are actually going to look that stuff up?

I think all articles should be at formers level since people with a deeper knowledge will be using sites, books, etc where that deeper knowledge is assumed. If they aren't already they can get started by reading reading the sources in the article

It is an encyclopedia, not a collection of tutorials. That seems more appropriate for something like wikibooks. Otherwise, every single article would be in inordinately long with a lot of redundancy. I like it better in which pre-requisites are often linked to other pages, so you can brush up on the pre-req's if you don't understand them.

wikipedia is a reference, reference material isnt teaching material.

Many of them seem to be written/edited by people who are more interested in using an equation editor than in trying to explain the topic. I don't necessarily expect ELI5 level but I wonder how many people who don't already understand the topic in depth can decode most of those math pages.

They dive into a lot of jargon and rarely attempt to provide any real context that a non-specialist can understand.

I understand that it's hard to do and it also reflects the fact that Wikipedia pages are wildly variable in the audience they're written for.

It's an encyclopaedia. A reference. It's not intended that anyone can open any random page and understand all of it.

> Plenty of philosophy articles are also very poor

Try the Stanford Encyclopedia of Philosophy; it's amazing and authoritative, written by the experts. And free (but takes donations!). You'll never look up a philosophy topic in Wikipedia again.

https://plato.stanford.edu/

I find that quality on technical and mathematical subjects varies wildly.

I remember once looking up a very specific topic, just to be sure that I don't misremember a certain simple but crucial core relation. What I found was a page that was so shockingly disorganised that I had to wade through 5 pages of mindbendingly complexified, quasi-obscurantist... drivel... just to extract that central piece of information that really should have been mentioned in the bit above the TOC, or at least easy to find.

So, YMMV.

> At the rate in which I encounter inaccurate data on Wikipedia that I can identify

Why haven't you tried to address those inaccuracies? I could sympathize if you said you did and were rebutted, but sounds to me like someone complaining that their local beach is full of trash when they themselves never bothered to help clean it.

> At the rate in which I encounter inaccurate data on Wikipedia that I can identify, it makes me suspicious of nearly everything on the site.

If you don't do that for every source you use, you might be suffering from the Gell-Mann Amnesia effect.

> but it's hardly a substitution for a legitimate source let alone encyclopedia.

A Nature study (http://news.bbc.co.uk/2/hi/technology/4530930.stm) found it to be comparable in accuracy to Brittanica.

> Nonetheless, I know plenty of people who take it as legitimate and cite it as if it's an authority.

Encyclopedias are tertiary sources, they should be used as reference works, not sources in and of themselves.

>Why haven't you tried to address those inaccuracies?

Trying to edit wikipedia is a hellish experience. The vast majority of articles have one or more de facto owners who will do everything they can to preserve control over their turf. They tend to be experts in rules laywering which is how they got their power in the first place. So "addressing inaccuracies" typically involves 5 minutes of editing and a month-long war afterwards.

And that doesn't even begin to address that wikipedia's rules are insane to begin with. Just the rules covering acceptable usernames are almost 4000 words long (https://en.wikipedia.org/wiki/Wikipedia:Username_policy).

> Trying to edit wikipedia is a hellish experience. The vast majority of articles have one or more de facto owners

I'm so tired of this meme. No one owns an article. [0] You say "de facto", but just because someone disagrees with you on something, it doesn't mean they disagree with every edit or do it only for that article. The vast majority of articles barely have anyone watching them.

> who will do everything they can to preserve control over their turf.

The fact that you describe it as "turf" only further proves that you have this mentality of the article being a battleground.

> They tend to be experts in rules laywering which is how they got their power in the first place.

What power? Even admins on Wikipedia can't overrule consensus and discussion. What's this about rule lawyering too? They just understand the community norms better. I don't walk into a fancy restaurant and complain about "rules lawyering" after I get kicked out for breaking the dress code. Once someone tells you about a rule, you should read up on it, not boisterously claim that you should be above the community rules.

> Just the rules covering acceptable usernames are almost 4000 words long

Seems simple enough to me:

> This page in a nutshell: When choosing an account name, do not choose names which may be offensive, misleading, disruptive, or promotional. In general, one username should represent one person.

Those 4000 words should only matter if you don't find the above to be sufficient explanation.

[0] https://en.wikipedia.org/wiki/Wikipedia:Ownership_of_content

> [0] https://en.wikipedia.org/wiki/Wikipedia:Ownership_of_content

Are you seriously trying to persuade someone who has doubts about the accuracy of Wikipedia content with a link to a Wikipedia page?

This is a page on internal Wikipedia policy, so there is no question about the accuracy here. The parent comment was pointing out that, in principle, Wikipedia policy is that contributors do not "own" content on Wikipedia in the sense of having a right to refuse edits to content that they contributed.

On the contrary: Wikipedia users are just as much (if not more) subject to their own wishful thinking and dirty politics when writing internal policy pages as they are when writing regular articles. Especially that there is no requirement to cite sources for the former. The policies as written may fail to reflect how editing actually looks like; hell, they might even not be actual policies at all. A single person could slip a passage in a policy page that supports their own opinion without anyone else noticing, and later cite that passage as if it were authoritative. (Or imply authoritativeness by simply placing a mere 'essay' in the Wikipedia namespace.)

Nobody cares what should happen 'in principle' unless the principles are enforced — which they are often not. See for example how the enforcement of the verifiability policy looks like.[0]

[0] https://en.wikipedia.org/wiki/Category:Articles_lacking_sour...

> I'm so tired of this meme. No one owns an article

Over 1000 hits for "ownership" on ANI: https://en.wikipedia.org/w/index.php?title=Special:Search&pr...

> What's this about rule lawyering too?

Nearly 1000 hits for "wikilawyerng" on ANI: https://en.wikipedia.org/w/index.php?title=Special:Search&pr...

As Tycho Brahe[0] once said: 'because I don’t have time to babysit the Internet'.

Maybe because the moment he would try to actually go and even, say, enforce the verifiability policies that already exist, an angry crowd of self-important ignoramuses would stop him in his tracks shouting 'Stop right there, you filthy deletionist! The article is notable!' and drag him to the dramaboards to have him blocked. And I should know, I have had the same thing happen to me on Commons.

> A Nature study (http://news.bbc.co.uk/2/hi/technology/4530930.stm) found it to be comparable in accuracy to Brittanica.

Are you seriously referring to that one over-ten-year-old study which used a non-representative sample of articles?

[0] As in, the writer of Penny Arcade, of course.

>> someone complaining that their local beach is full of trash when they themselves never bothered to help clean it

Here you're shouting others down by trying to shame them. Please don't do that. Being unwilling or incapable of cleaning up the messes others make does not disqualify one from criticizing the mess makers.

My point with the example was to show that Wikipedia is a public resource. It's not "that other guy's" responsibility to fix it. Every editor is a volunteer. If you have the energy to complain about a public resource that you benefit from for free, that is maintained by volunteers, you absolutely should be shamed for not chipping in yourself and instead writing about it elsewhere. As the famous line goes: "patches welcome".

Except that WP has its own arcane laws and policies used to great effect by regulars to ensure no outside voice is given a say in truly contentious articles. e.g. If you try to get involved, you are a "single purpose account". If you contest the veracity of a source, it turns into a meta debate where you are expected to provide an alternative view that is nevertheless considered a reliable source by everyone there who thinks otherwise.

It's a trap, because it ensures the only people who direct it are the ones who also have time to get fill the site with unimportant trivia and camp out. Unsurprisingly, anything to do with the current culture war is a lost cause.

>> "patches welcome"

Here we are in a topic that actually documents the hostility people encounter when participating in Wikipedia, a topic that provides overwhelming anecdotal evidence among the comments that people are being actively discouraged from contributing to Wikipedia by hostile admins and topic campers (something I have personally experienced) and you pretend the actual problem is lazyness.

You deserved your downmod.

Anything that touches on politics or religion is a tire fire, but the math and science articles aren't bad. I mostly use Wikipedia to help me find actual resources.

Science is political now because evidence, studies, and results can impact policy, so it's deeply polluted.

Personally I've found many medical and scientific articles to contain vast amounts of misinformation and disinformation. I would recommend not relying on any of it, it is at the whims of whoever bothers to edit.

I now recognize Wikipedia as the original 'fake news' or 'fake information' on the Internet. Just like all those news stories people pass around, nobody cares about Wikipedia's accuracy. It's the same phenomenon, and possibly (I have no evidence) Wikipedia conditioned people to accept unreliable information.

I was about to write the same thing. Many articles have fake information.

For instance, someone might edit an article to say that a person was a known political extremist. Someone else might write an article (not on Wikipedia) saying the same thing (after having read the Wikipedia article). Years later, if the information is questioned on Wikipedia, then editors will add a reference to the off-wiki article, and everyone will be happy. Circular fake news, with truth going down the plughole. The entropic heat death of information.

With the widespread adoption of the term "fake news", I'm wondering what you mean by "fake information" as opposed to wrong, or false, or mistaken, or in more subtle cases, from a different perspective (think "freedom fighter" vs "terrorist"). "fake" to me implies purposeful deception. Is that what you mean here?

It doesn't have to be deliberate, purposeful, or even conscious - though it often is. An exaggeration can be treated as fact. That fact can again be exaggerated, until the words bear no relation to reality. Wikipedia practically invented this mechanism.

I do think in that case that fake is not the right adjective to use. Every definition of fake implies intent to deceive. If you're trying to encompass the meanings you describe above, fake doesn't appear to cover it.

You are right. But I do think that what is now called "fake news" is generated by a similar mechanism. I don't know what term to use. Wrong or false don't quite cover it. Perhaps 'unmoored' - a boat can be unmoored accidentally or deliberately, or by outside events - a big wave.

Yeah, I'm not sure what the appropriate term is here, either. Perhaps contentious? disputed?

I'm still trying to get my head around what fake news is actually trying to describe. I'd like it to refer to intentional deception, but I think it's being used much more broadly than that. I fear that it's effectively meaningless at this point, other than as a generic insult.

Wikipedia policy is such that slanderous claims like that about living people should be (and usually are) immediately reverted.

What does this mean in practice? Editors on Wikipedia are not a balanced cross-section of the world. I would guess they at 90% male, nerdy, middle class, and disproportionately liberal. This will be reflected in their articles, and the leeway given to certain 'facts'. And god help you if there is an article edited about you after you die - then the gloves are off and slander is fine.

I know the policy; I've seen it ignored many times. Also, that applies only to articles about living people.

The Internet says ...

Reply on news.ycombinator.com