trout

For serious discussion of the "major" forum for Wikipedia criticism and how it fails.
Post Reply
User avatar
CrowsNest
Sucks Maniac
Posts: 4459
Joined: Tue Feb 27, 2018 4:50 am
Been thanked: 5 times

trout

Post by CrowsNest » Mon Aug 13, 2018 12:35 am

I like trout. Always makes good points. Should post more often. He probably doesn't because when he does, he is either ignored, talked down to, or is otherwise reminded why it was a waste of his time to post on a supposed critic site.

His latest effort is a simple but convincing experiment to show that 80% to 90% of Wikipedia articles are not being actively monitored, by posting error reports on their talk pages, and listening to the sound of crickets.

Kingsindian aside, and excepting dykslyver too because he's probably not being serious, all the Wikipediocrats can think of doing, is offering up reasons why that's not a bad thing.

According to their brains trust, it's OK because.....

-there aren't enough editors to watch five million articles

-watchers may simply not reply because they're OK with the article as it is

-those pages are probably not the ones people read anyway

Is it really that hard for these people to agree with his basic premise that it you call yourself a volunteer curated encyclopedia and you amass five million articles, then not noticing or caring about error reports placed on 90% of those pages is a problem?

trout isn't reporting a phenomenon that should be a surprise to anyone who spends time actually observing Wikipedia. Wandering their ruins, you notice article talk page posts that have never been responded to all the time, even on articles being actively edited. Which is something he could look into, to see what percentage is abandonment, and what is disinterest, or even lack of awareness there is even a message. But it is no surprise to see the brains trust making replies that suggest they don't really do that at all, that this is all brand new information to them, and so they're only now putting their minds to it. And even then, it's just basically guesswork.

Kingindian's comment is arguably even worse. He acknowledges the problem, but as is tediously familiar, proposes a solution that is never going to be accepted, namely slash the number of articles down to what they can monitor. Is it too hard for him to give up on this never-ending fantasy that Wikipedia can somehow be reformed?

User avatar
Dysklyver
Sucks Critic
Posts: 391
Joined: Thu Jun 21, 2018 10:14 am
Has thanked: 8 times
Been thanked: 24 times

Re: trout

Post by Dysklyver » Mon Aug 13, 2018 1:07 am

There is no chance of Wikipedia being able to curate such a mass of mostly crappy pages. There are just about enough editors to keep an eye on a few thousand pages, tops. And as is frequently noted, this is not even enough to keep their Featured Articles up to date, let alone the entire encyclopedia. And as editors leave, this get worse, and worse. The best that can be done is to vaguly monitor changes in the hope that this catches vandalism, but there simply aren't and never will be the millions of active editors needed to maintain the encyclopedia.

User avatar
The End
Sucks Fan
Posts: 245
Joined: Mon Jul 31, 2017 7:45 am
Has thanked: 1 time
Been thanked: 1 time

Re: trout

Post by The End » Mon Aug 13, 2018 4:05 am

"In the long run, volunteers are the most expensive workers you'll ever have." -Red Green

"I am a dark bouquet of neuroses..."
- Jerry Holkins, Penny Arcade

User avatar
Strelnikov
Sucks Admin
Posts: 1041
Joined: Fri Feb 24, 2017 11:25 pm
Has thanked: 395 times
Been thanked: 251 times

Re: trout

Post by Strelnikov » Mon Aug 13, 2018 6:10 am

Dysklyver wrote:There is no chance of Wikipedia being able to curate such a mass of mostly crappy pages. There are just about enough editors to keep an eye on a few thousand pages, tops. And as is frequently noted, this is not even enough to keep their Featured Articles up to date, let alone the entire encyclopedia. And as editors leave, this get worse, and worse. The best that can be done is to vaguely monitor changes in the hope that this catches vandalism, but there simply aren't and never will be the millions of active editors needed to maintain the encyclopedia.


I've tested en.Wikipedia for vandalism; they have 'bots that will remove obvious vandalism quickly from an IP account. If I wanted to be a jerk, I could probably concoct some horrendous nonsense in the history of Mexico article ("Groucho Marx dated the wife of the President of Mexico in the 1920s* before she was married") and it might not be found for years, especially if I avoid using the words that trip the 'bots (probably "shit", "piss", "fuck", "cunt", "cocksucker", "motherfucker", and "tits" plus "fart" and a bunch of sexual terms.)


* He was married to Ruth Johnson, whom he drove to booze because Groucho had a knife for a tongue. They divorced in 1942.
Still "Globally Banned" on Wikipedia for the high crime of journalism.

User avatar
Dysklyver
Sucks Critic
Posts: 391
Joined: Thu Jun 21, 2018 10:14 am
Has thanked: 8 times
Been thanked: 24 times

Re: trout

Post by Dysklyver » Mon Aug 13, 2018 3:45 pm

Strelnikov wrote:
Dysklyver wrote:There is no chance of Wikipedia being able to curate such a mass of mostly crappy pages. There are just about enough editors to keep an eye on a few thousand pages, tops. And as is frequently noted, this is not even enough to keep their Featured Articles up to date, let alone the entire encyclopedia. And as editors leave, this get worse, and worse. The best that can be done is to vaguely monitor changes in the hope that this catches vandalism, but there simply aren't and never will be the millions of active editors needed to maintain the encyclopedia.


I've tested en.Wikipedia for vandalism; they have 'bots that will remove obvious vandalism quickly from an IP account. If I wanted to be a jerk, I could probably concoct some horrendous nonsense in the history of Mexico article ("Groucho Marx dated the wife of the President of Mexico in the 1920s* before she was married") and it might not be found for years, especially if I avoid using the words that trip the 'bots (probably "shit", "piss", "fuck", "cunt", "cocksucker", "motherfucker", and "tits" plus "fart" and a bunch of sexual terms.)


* He was married to Ruth Johnson, whom he drove to booze because Groucho had a knife for a tongue. They divorced in 1942.


The bot's use AI now to score your edit against known vandalism by comparing it to a complex machine learning thingymabob called ORES. It effect this means that as long as you write a grammatically correct sentence with a reference from an autoconfirmed account it will sail past.

For example if you get autoconfirmed, find a random article and insert something like the following:

In modern times, it has been shown that druidic rituals are universally not equivalent to neo-pagan druidic tetatal rituals which are performed throughout Europe by adherents of the pseudo-realist cult of the druids.<ref name="Paxillus">Paxillus involutus adminus bolluckus, Oxford University Press, Oxford, pp. 518–520</ref> Furthermore it has been shown that adherents of the adminus sect have not performed human sacrifice since {{circa|910}} except in tetatal rituals.<ref name="Paxillus" />


It will survive for weeks before being discovered and removed with the edit summary "worthless citation".

User avatar
AndrewForson
Sucks Critic
Posts: 266
Joined: Sun Feb 26, 2017 7:56 am

Re: trout

Post by AndrewForson » Mon Aug 13, 2018 4:02 pm

An alternative is to boast on a forum like this that you have written nonsensical articles on topics like "Chromatic homotopy theory" or "Redshift conjecture". As far as I can tell, these are perfectly legitimate articles, written by an expert, but they might be hoaxes, and almost nobody else on Wikipedia could possibly tell. The object would be to get the Wikipedians to delete perfectly legitimate information and watch hilarity ensue.

User avatar
Dysklyver
Sucks Critic
Posts: 391
Joined: Thu Jun 21, 2018 10:14 am
Has thanked: 8 times
Been thanked: 24 times

Re: trout

Post by Dysklyver » Mon Aug 13, 2018 4:49 pm

AndrewForson wrote:An alternative is to boast on a forum like this that you have written nonsensical articles on topics like "Chromatic homotopy theory" or "Redshift conjecture". As far as I can tell, these are perfectly legitimate articles, written by an expert, but they might be hoaxes, and almost nobody else on Wikipedia could possibly tell. The object would be to get the Wikipedians to delete perfectly legitimate information and watch hilarity ensue.


I once wrote an article about smelly farts and that survived at least a week!

User avatar
CrowsNest
Sucks Maniac
Posts: 4459
Joined: Tue Feb 27, 2018 4:50 am
Been thanked: 5 times

Re: trout

Post by CrowsNest » Mon Aug 13, 2018 7:08 pm

trout wrote:I'm not actually using "Random article" to find these articles as you seem to be implying. I've just been adding notes as I find them. Looking at the articles where I'm the last non-bot to edit the talk page I see a very big Hollywood film, a well-known TV news anchor, a big piece of geography of the USA, a famous actress .. it goes on and on.
Poetlister wrote:Obviously, if trout is not working at random then it would be dubious to use his results for statistical purposes.
Stupid fuck. "As I find them" would be the most representative way of measuring the responsiveness of Wikipedia to reader generated error reports. You want more robust results, then ask more readers to do it.

That is the purpose of Wikipediocracy, right? Shining lights at Wikipedia's underbelly? As so often happens, he seems to interpret it as holding a torch for Wikipedia.

User avatar
CrowsNest
Sucks Maniac
Posts: 4459
Joined: Tue Feb 27, 2018 4:50 am
Been thanked: 5 times

Re: trout

Post by CrowsNest » Mon Aug 13, 2018 7:16 pm

Midsize Jake wrote:
trout wrote:I'm not trying to solve Wikipedia's problems. I'm adding the comments to confirm the observation that most articles are unwatched.
I don't want to discourage you in any way, but I'd be really surprised if that weren't the case, considering how obscure most article subjects are and how little anyone cares about non-controversial articles in general, given that "non-controversial" probably describes (in line with what you say above) at least 90% of the total, and more likely, 99% or more. So, as the other replies seem to indicate, most of us probably just take this for granted - but nevertheless, you're correct in implying that it would be helpful to have some means of showing the extent or magnitude of the problem statistically. (At least I think that's what he was implying.)
Jesis Christ. Discourage him from what? Believing Wikipedia was great? Anyone who has the basic respect for trout to actually remember his posts, knows he doesn't think that. Christ, if that's too much work for the site admin, then that much is obvious from the post itself. He's set out to confirm the observation.

What must discourage him, is the very obvious fact that the members of Wikipediocracy don't seem in the least bit interested in putting a hard number on it, only in explaining it away, and indeed their resident statistician wants everyone to think the findings are dubious.

Fuck these people.

User avatar
Dysklyver
Sucks Critic
Posts: 391
Joined: Thu Jun 21, 2018 10:14 am
Has thanked: 8 times
Been thanked: 24 times

Re: trout

Post by Dysklyver » Mon Aug 13, 2018 10:45 pm

I for one would like to see a number put on it, and there is a WMF grant request in the pipeline to get an "abandoned articles" metric.

But my estimate on the number was >99.9%

Maybe I am off by a few points of a percent, but I doubt it's going to be cheering up the Wikipedians if this was proven.

Post Reply