Let's test that theory, Reg style
Hypothesis: Upvote:Downvote ratio is approximately equal to fandroid:fanboi ratio.
Internet forums often use reader moderation to determine which comments are the best, but new research suggests that tallying up and down votes for online comments is a poor measure of those comments' actual quality. Oh, you may think you know who's brilliant and who's a troll in our forums, dear Reg reader – but according to …
Instead I got 90 downvotes and 15 up votes.
All I said was:
'Here on El Reg the theory appears to be correct, there always appears to be a disproportionate number of downvotes when anyone on here praises an Apple product.
The comments also tend to be more extreme in their nature. So there could also be a correlation between this and the Political Left here in the UK, as they too tend to shout the loudest, have a nastiness and are less forgiving. The Left wing bias is also prevalent and entrenched in the media (except of course the Daily Mail) including the impartial BBC.
Of course there is other evidence as well. Comments on the news articles tend to dip off after 1pm on Friday which is the end of the working week for the public sector which is generally considered left wing.
In conclusion, he may have a point, or it's just that a disproportionate number of immature trolling di**heads populate this site. In which case my theory falls to pieces.'
I mean how can that be offensive to anyone?
"I mean how can that be offensive to anyone?"
Because the people that actually use the vote button consistently are the ones you targeted :-/ Sort of like putting a mirror in the face of a raving lunatic (ie. Ballmer). I was 1 of your 15 up votes 2 weeks ago, and I still back your comment in full.
The voting system is so skewed and useless, it should of never been introduced. But like anything else, one person finds a new hook, and then everyone is hooked and using it. Technology sites very typically play follow the leader when it comes to viral social schemes, the Register is no different apparently, and "Gangnam Style" posting appears to be encouraged.
Doesn't take a genius to figure out the voting system is biased and worthless, just takes a moment of thought. If you take another moment of thought, then you realize "Voting" + "Anonymous Coward" puts the whole BBS in a position for trending and complete lack of competence. (BTW, I dislike AC posting more than the voting system, but that is another matter).
But hey, if all the cool kids are doing it...
"after 1pm on Friday which is the end of the working week for the public sector "
Don't know which public sector in the UK you are referring to but MY part ( >27 yrs experience) is busiest after 2pm on Friday and frequently works beyond stated "day" to finish the work
Try not to generalise if basing this on your personal prejudices or "I have a mate in the pub whose wife knows someone in the public sector"
Statistically for every Apple praising comment you will get a ration of 1:23 upvotes:downvotes.
The majority of El Reg Commentards appear to own Android phones, while the minority WinPhone and Blackberry owners keep quiet for fear of embarrassment.
Fandroids are generally the same kind of people who wear Burberry and drive cars with 'my other car is a Porsche' sticker on the bumper and read 'Which Mobile' dreaming of the day they can get hold of an iPhone.
They would all love to own an iDevice but their menial jobs and minimum wage hold them back.
Jealousy is a green eyed monster.
..while the minority WinPhone and Blackberry owners keep quiet for fear of embarrassment..
Ahem, as a Blackberry owner of sorts*, I feel I have to say that maybe, just maybe, we 'keep quiet' as we find there are better things to bleat about than our mobile phones, which, after all, are just devices we allow into our lives for the sole purpose of enabling other people to annoy us with inane requests at the most inopportune moments (oh, and ordering Pizzas)
*I say, of sorts, it's currently U/S as the damn'd screen on the thing cracked for no discernable reason. (No, it wasn't dropped, nor subjected to stresses unusual or otherwise, I woke up one morning and the thing had cracked overnight, so, yes maybe in the middle of a dream I punched the feck out of it or something, as deep down in my psyche I really hate it and want to join either the Church of Jobs or Page-Brin.).
Whilst I'm making up my mind as to whether I'll replace the Crack'dberry, I'm quite happily using an old (and rather pleasantly quite dumb) Nokia.
What appears at first glance to be a harmless story from the Department of the Bleeding Obvious turns out to be the colossal middle finger which El Reg has been dying to give its commentards all these years. Impressive. Now can you put the PHP back and stop messin' around.
I have paid a click farm to give me 1000 up-votes for this post, a bargain at 0.005 USD per up-vote. These up-voted will of course promote millions (if not billions) of other up-votes from gullible ElReg easily swayed commentards.
I look forward to my paid for popularity, should be starting any time now....
It seems that the commentard/readership demographics also help. Over here, a different effect comes into play; good comments usually get upvoted, bad comments get downvoted and it is usually fanboi/shill stuff getting really downvoted.
But an interesting thing is that the later your comment gets in, or the further down the pile it is, the less up/downvotes you get. Or it might even get zero votes. That's mostly because it's rare that commentards will ever read all the comments and will start skimming if it is large enough.
But no, I haven't seen the balancing out stuff going on in here. Those that get balanced votes (like say, 10 up 10 down) are usually like that because their comments actually do have the same number of people for and against the comment, not because of a balancing out thing.
Well, El Reg is full of some rather smart individuals. Even some of the biggest asses on here regularly contribute intelligent, thought provoking and sometimes funny things.
But at places where the great unwashed frequent, like Yahoo! News, the results of this study do seem more accurate. A nice follow up study would be to determine if a narrower range of subject matter had an impact on the results vs a general current events venue.
I have to completely disagree, there is no correlation between a posts intelligence and its up votes on this website.
The most up voted comments on the Register contain a combination of snark, anti Microsoft sentiment, anti Apple sentiment, pro Google sentiment or pro Linux sentiment.
The Register attracts a considerable amount of blind faith followers. For example, The Register posted a news article on the morning Microsoft officially cut the price of the RT in the UK stating that Microsoft had cut the price in the US but had no plans to cut it in the UK. The highest rated comments was;
"Are MS aware that the instant they lower the price of something in America the entire world is aware of it thru' something called 'The Internet'. They might want to get onboard, there's load of good info on it like-
Tidal wave of hatred on Windows 8 months before it went on sale, demanding a start menu but just adding the button for the .1 upgrade
Screwing consumers over first sale rights and demanding once a day online verification when releasing a new games machine
Getting into bed with NSA then releasing Cloudy office and always on kinect.
Who, in the UK, Australia etc are going to buy a full priced RT when it's For Sale in USA?"
We have snark, we have anti-Microsoft AND we have factually incorrect.
Yeh, but sometimes people put themselves in that position.
For instance, you started off very well with 4 typical sentiments found here, you were stating the obvious truth. However, you wound up sounding to be in a very strong stance that aligns you closely to MS. So, even though you didn't post negatively, you still took a side personally. But, I'll just assume that in general you're a lover not a hater :-).
I'm a lover of factually correct posts ;-)
But you've highlighted the main issue with the voting system in that the vast majority of forums, The Reg included, have a very strong "you're either with us or against us" tribalism. Pointing out errors is quite often seen as an attack against their camp.
" But an interesting thing is that the later your comment gets in, or the further down the pile it is, the less up/downvotes you get. Or it might even get zero votes. That's mostly because it's rare that commentards will ever read all the comments and will start skimming if it is large enough."
I think t's more to do with the fact that once people read an article and its associated comments, they generally don't come back - especially as there is no way to track comments/threads.
The later you post, the less amount of people read it.
I am not sure if you are being facetious or if you don't know this, but the bronze, silver, and gold badges are not dependent on any number of votes
I don't know if you are being anal or if you don't know this, but you're totally wrong.
The qualifying thresholds for badges are:
Bronze (about 650 commentards qualify)
Operated forum handle for more than one year, and more than 100 posts in the last 12 months.
Silver (about 200 commentards qualify)
Silver badge holders meet bronze requirements and have more than 2,000 upvotes.
Gold (awarded to 10 commentards)
This discretionary badge is awarded by Reg staff, and we are starting off with 10 people who have been very helpful - to us, through news tips and beta testing, for example - and to their fellow readers, through their posts.
The door's over there. Close it on your way out. Have a downvote for your troubles too...
I don't think upvotes on posts as AC count toward the silver badge as I'm way past 2000. Either that or I have a hardcoded bronze badge after a less than civil argument I had with Trevor Pott about the original implementation of the system. Either way, I cba to manually count them to be sure.
There needs to be a tongue-in-cheek icon.
"What do you think, dear reader? Is moderation of online forum comments to be believed? Or can comment moderation be used to sway public opinion like a virtual Pied Piper, herding online readers like a pack of gullible lemmings to the proverbial cliff? Naturally, here at El Reg we encourage you to make up your own mind."
One way to test the theory is to not display the number of up or downvotes any comment gets to any reader until or unless they choose to up or downvote that comment. That way they can't be swayed by what's already happened, and can only be voting as a result of what the comment itself actually says.
It'll be a difficult thing to analyse because it would be impossible to truly know how the voting would have gone if the numbers were displayed as now - but if the study is correct (and if it also applies to El Reg readers) then at the very least the number of votes cast over the period of the experiment would drop off.
(Are you able to compile statistics about how many up/downvotes are cast per week/month/whatever?)
But that would result in a high number of junk rates from people wanting to see what the voting is like for a particular post.
HotUKDeals ran into that problem when they ran a promo that involved members having to rate X many deals and users were either randomly voting deals as hot or cold or they were voting based on whether thy liked/disliked the product ant not whether it was a good deal for hat product or not.
"But that would result in a high number of junk rates from people wanting to see what the voting is like for a particular post."
Have three buttons - upvote, downvote, show votes. You can show votes without voting, but you can't vote once you've done so; once you upvote or downvote, you see the results.
I'm curious if they have a way to segment this by whether or not the post already had a vote. Oftentimes I feel like people don't want to be the first one to vote either way on something, but are totally willing to chime in either direction once they see voting happening. A kind of steam or critical mass effect. No real data to support that, obv.
If that were correct, then the tendency to upvote no matter what would be explained by the fact that comments on the site are overall better than neutral (votes don't need to be zero sum) and that they cracked a lot of vote virginities.
Mysteriously this study reveals that people do follow a herd mentality - fancy.
Unfortunately you(*) and I might be free thinking, free speaking and intelligent beasties of some sort but collectively we all behave like wankers.
'Tis the way of the world and why basic mod systems on free chat networks fail to work effectively.
Goo+Book work reasonably well - err - better by using the friend filters: you have to agree to "give and receive" in general. This means you get to filter out people who are not like you which is nearly a simile for people who you don't like.
In the (nearly) totally free speech setups like el Reg, /. int al you don't get that instant filter. Hence you get the rabid Eadons (**) that you can't run away from.
One day t'internets will discover how to work with people properly - at the moment it's shit.
* - Yes: YOU, you tosser
** - I am a dyed in the wool Gentoo flavored penguin - to mix a metaphor - but I hate uninformed dogma.
Internet forums often use reader moderation to determine which comments are the best, but new research suggests that tallying up and down votes for online comments is a poor measure of those comments' actual quality.
Well, all they had to do was read xkcd on the TornadoGuard rating.
This isn't about quality; it's mostly about opinions. And yeah, sometimes people don't agree with yours and sometimes they do. But to link ratings to quality is a bit far fetched to begin with; because wouldn't it take quality to judge quality? So what exactly are the requirements on "personal quality" before one can start voting?
Usually it only takes filling out a registration form.
...This isn't about quality; it's mostly about opinions. And yeah, sometimes people don't agree with yours and sometimes they do. But to link ratings to quality is a bit far fetched to begin with; because wouldn't it take quality to judge quality? So what exactly are the requirements on "personal quality" before one can start voting?...
The Greeks experimented with Democracy over 2000 years ago.
They reported that it worked ok, but it had the great disadvantage that it was effectively mob rule, and a good orator could easily whip up a frenzy and cause a sensible, well thought -out policy determined by the experienced elite political leaders to be rejected in favour of a wild and dangerous adventure.
There were a lot of complaints about stupid wars caused this way...
There's no icon for 'learning from history', so I'll get my coat...
I've always thought that most up/down votes are generated by trolls, robots, or that dreaded combination of the two: the marketing department. Those votes generated by people who have actually read and thought about the article or comment are quite rare. It's always been the case that just because something is "popular", that doesn't mean it's any good at all. Then again, just because something is unpopular also doesn't mean it's any good either.
Ah yet another worthless study into something that isnt going to change the world. In fact I find studies like this a little insulting, and given the nature of its subject matter I wonder why the "worthy scholars" couldn't find something that really matters to probe into. If I were a forum owner I'd be pretty pissed off about them "gaming" my forum in the pursuit of some worthless theory.
The thing about forum votes is that when I vote it is because I actually have an opinion on the topic - agree with the post of not it is still a valid one - and feel that I want to express it despite having nothing worthwhile to add to the existing comments, So rather than just generate "white noise" I vote in accordance with my views.
There is of course the bottom line to this sort of thing - the old chestnut about "lies, damned lies & statistics" - I think that applies quite well in this case.
So it looks like all you need to do is be very contentious in your comments. The people who disagree with you will seed you with down votes and that will cause the people who agree with you to up vote you and that will lead to a feedback loop of aproval.
Sent from my Blackberry, still the best smartphone on the market.
While I'm writing this comment, all the rest of the comments in the thread have exactly the same number of upvotes and downvotes. People seems to be divided on this subject. :0)
I've got a proposal for the ElReg forums: Votes could be hidden by default in all comments, and the readers could have a button to show the votes on a single comment, and another one to show the votes for all comments. This would lower the bias a little bit, methinks.
Or, alternatively, up and down votes could be shown only 48 hours after the original article has been posted.
"While I'm writing this comment, all the rest of the comments in the thread have exactly the same number of upvotes and downvotes. People seems to be divided on this subject. :0)"
Hmm. All of the posts have a 1:1 vote ratio. Statistically unlikely, particularly when some posts have 6 or 7 of each.
Someone seems to be having a bit of fun.
Mrs. Thatcher was aware of this herd mentality in the 1970s. That's why she wanted to put into law that secret ballots should be held when voting on strike action. Before that, votes on strike action were done in the open where peer pressure no doubt had a very large effect.
But to more important things, El Reg commentards. Why not do the thing where you only get to see a post's up/down vote count after you have voted on it. To alleviate the problem of random voting just to see the vote count, provide another button to reveal the count but on the understanding that the reveal-er can no longer vote on the post.
I wish the newshounds in Murica were like you guys. This study is being reported via multiple outlets, but the others are cringeworthy dullards by comparison. And it is like this pretty much all the time.
I don't always agree with El Reg, but I always enjoy the read.
I like to pride myself on individual thought, and not caring what the mass thinks, and yet I find myself tempted to tailor oposts to what I think will get upvotes and avoid downvotes, even though I like to think that a reasonable number of downvotes is a good thing.
It feels to me like a variation of the peer pressure principle: one is motivated to post what one thinks the masses want to hear, and not necessarilly what one believes to be correct. I'm not at all sure that its a healthy trend for society if we all start to try and talk and think exactly the same things. Apart from anything else I thought I'd grown out of that a bit wheen teenage years departed!
"the so-called wisdom of crowds can often be misleading."
It's called a general election and it falls in to two camps: better the devil we know versus this guy has got to be better than the last guy. Amped up more for Americans where they appear to vote based note upon personalities. In most cases, the actual attributes and policies count for less; and the public are swayed by so-called analysts and experts. In most cases, we could question how many people voting actually understand what they are voting for.
or maybe it's just that where ever they did this, people tend to post things that appeal to the readers and they then tend to upvote? Cause that's what "negative votes getting voted against and positive ones being enforced" sounds like to me.
As for a control group, as if anyone could control anything on an online forum ...
I remember warning people of the Rasberry Pi hypes and how it wasn't really powerful enough for a HTPC and all I got was downvotes...
Turns out a year later I was right. I never owned one, but I had a friend who told me how he tried but it isn't that great and his numerous friends having 2-3 Pi lying around not know what to do with them.
I told you so is what I'd like to say to the world, but hey most people are sheep.
On a more serious note, up or down votes never matters. It's only a measure of popularity, if your argument is sound then you still stand a chance of getting your voice, no matter how small it maybe, heard, that's all that matters at the end of the day.
+1/-1 never stopped me from making my voice heard and I'm sure it never stopped real trolls, so you have to ask yourself, wtf is the point of it and why did people adopt such a measure just because some other site has it? If you ask me not much thinking was done before it was implemented.
About the Pi - I just discovered RaspXBMC. That and an HDMI to VGA adaptor and an old 17" 1280x1024 LCD monitor and suddenly there is something more I can do with my media. I try to avoid 720P as my 1.8GHz netbook doesn't like it much; the Pi copes and barely even warms up. Given an Android's mobile of similar spec stutters like hell on HD, it's quite impressive what the Pi can play. I also like how you just switch SD cards and it is doing something completely different. I find it to be friendly, flexible, and simple to use (having everything on SD which can be easily changed is great). Yes, sure, there are better options, but how many cost ~£30?
"I remember warning people of the Rasberry Pi hypes and how it wasn't really powerful enough for a HTPC and all I got was downvotes...
Turns out a year later I was right. I never owned one."
Just cause your mates can't do it....you should borrow one off your mates and try yourself. RPI with Openelec works just fine for me. it struggles with 10+GB MKV files but if you're trying to stream something a little less extreme it's the dogs.
If you didn't already know that 90+ percent of popular opinion by the vocal minority was usually wrong, all you need to do is look at Manning, ASSange and Snowden to see that despite the ignorance of many, these criminals will still be convicted and punished for their crimes. Though the naive and gullible view them as heroes, they are in fact just grandstanding criminals headed for a long prison stay.
Though the naive and gullible view them as heroes, they are in fact just grandstanding criminals headed for a long prison stay.
Come on man. Even a long diatribe from Hitler himself (channeled via paranormal radio) about semitic races and the malfeasance of The Other Collectivism is not going to bring us out of our midsummer torpor. Be nice.
You mean like when El Reg recently had a list of the "most popular" (or some such) comments at the bottom of an article (before you clicked on the Comment button) which pretty much guaranteed that they were the ones most likely to get votes?
I quite liked the preview of comments at the end of articles. Especially since access to forums.theregister.co.uk* is blocked by my employer. All that's needed is to remove the voting buttons from the preview. If you want to vote, you have to trudge through the rest of the commentary outpourings.
*Can El Reg change it to fora.theregister.co.uk? I know it's pretentious but it might take a while for the powers that be to notice and block it.
So up votes tend to attract more up votes, and down votes tend to attract more...up votes! You would get the same result if it were simply the case that votes tend to attract more votes, and the majority of votes tend to be positive. It's not unlikely that votes attract more votes: I for one tend to scan comments and read only the ones that have more votes (up or down): I assume the low-voted ones aren't interesting. No evidence of herd mentality there.
Nothing to see here. Move on.
On consideration, I find I have quite a complex algorithm for upvoting and downvoting on El Reg, more so than other sites, due to El Reg's voting system requiring a wait for two page reloads to vote. As a result, I only go to the trouble if I feel the post is worth voting on.
I'll upvote a post if:
1) The post makes me laugh or spit coffee/lunch over my keyboard or monitor.
2) The post expresses a political opinion I strongly agree with (regardless of its factual content or literary coherency - the political alignment with my own is sufficient cause for an upvote.)
3) The post has good literary merit or posits its owner's case with aplomb.
4) amanfromMars posted it and I cannot decipher or understand what he posted, or if I simply give up trying part way through.
I'll downvote a post if:
1) The post has even the slightest whiff of political correctness about it.*
2) The spelling, grammar and/or syntax is so bad as to make the post indecipherable (amanfromMars is treated oppositely here: he gets a downvote if his post is actually decipherable!)
3) The argument of the post is so ill-researched, fallacious or delusional as to expose the poster as a retarded fuckwit.
Fanboi/fandroid arguments get nothing. I hate Apple myself, because of their control-freakery and egregious litigiousness, but I won't upvote a post simply because it's an Apple hater and I won't downvote one simply because it supports them. My fanboy loyalty to IT companies died with the Commodore Amiga.
*Political correctness: I won't downvote a post simply because it expresses a political opinion I disagree with. For example, you could call for abolition of minimum wage or billion-year copyright extensions and I wouldn't downvote you for that even though I disagree with it. But any hint of the poster supporting feminism (note: by this I mean focus on women's rights alone), affirmative action, advocacy of censorship of dissenting views, or anti-white/anti-male dogma WILL get a downvote, guaranteed, regardless of the validity of the argument or how well the post might be written.
If I feel the politically-correct post is aggravating enough and it pisses me off sufficiently, I'll actually go to the trouble of going to the poster's public post history page and sequentially downvoting their other posts until my temper calms down (David W, Vladimir Plouzhnikov, Oolons and NomNomNom have all been victims of this at various times!) I know that's juvenile, and one should expect more from a 46 year old adult, but I find it therapeutic and it makes me feel better... So to the guys listed above: If you suddenly find your last 50-odd posts all have at least one downvote regardless of topic, you can feel gratified that you've pissed me off enough to waste half an hour attacking your post history!
Consider yourself upvoted.
It'd seem NomNomNom has experienced exactly what you describe at least twice then. Matt Bryant, Jake, Eadon and RICHTO/TheVogon have suddenly leapt to mind too, having experienced similar, although more for clause 3) reasons than clause 1)... Obviously!
AC for petty juvenile fear of petty juvenile reprisals.
Ahhh.... TheVogon. Yes, the MS shills and outright conservatards get an instant down vote most of the time, though some of them have actually earned upvotes every now and then. (Really, REALLY rare.) Probably the only ones that have recently earned the shitstorm downvoting would be the obvious trolls like BornToWin and the evil-Eadon twin that roams the comments section: mmeier.
Also using AC as the users mentioned are sure to engage in said shitstorm and I'm too tired to deal with retards today.
Is the balancing of votes on here because people are being up-voted for a decent comment and then other people are trying to skew it by adding an equal number of down-votes, or are all the comments actually complete tosh and are being down-voted, and the people trying to skew it are up-voting?
I think the only way to figure it out is to vote this comment up if you think people are purposefully down-voting and vote it down if you think they are up-voting , then we can compare the amount of up and down votes to see if we are adding down or up votes to change the up or down votes.
Paris, because she likes it both ways…
This is true, but I believe that commendations are in order since you seem to have no problem accomplishing the feat all by yourself.
I think a more astute interpretation of your propensity for receiving down-votes, rather than assuming everyone else is wrong and you are right, is that you are actually too stupid to see that your attitude, your reasoning and your method of delivery constitute a wholesome embodiment of failure.
What particularly annoys me is that within your interminable drivel, on every occasion in which you try to employ the use of well-known language constructs, figures of speech or idioms, you do so incorrectly. This quite frequently results in the meaning of your words being unclear or in fact completely the inverse of that which you intended. It also has the side-effect of demonstrating your lack of education and wilful ignorance of the language as well as your lack of capacity for logical thought.
This may represent an ancillary reason for your accumulating volume of downvotage.
I was surprised when ElReg decided to introduce the voting scheme, given the authors and readership I assumed it was a common understanding that popularity of an opinion is not an indicator of it's accuracy, insight, truthfulness or any other desirable characteristic. It's an indicator of it's popularity.
I also assumed past experience of slashdot had shown the pitfalls of such a system in this type of community.
Apparently not...Was it because 'crowdsourcing' came in vogue?
Biting the hand that feeds IT © 1998–2019