I find this pathetic. As if the removal of the "dislike button" helps in any way, shape or form to create a more "inclusive" environment.
This reeks of the silencing of what is perceived as dissident voices, is all.
I've personally never hit a like or dislike button in my life, never will ... but found it interesting as a viewer to see the ratio. Even knowing this ratio is probably at least partially the outcome of bot farms and the like.
Disney, DC Comics and the like must've complained about bad ratings/unfavourable like/dislike ratios I guess..
Let's try and see this as a tangent/bastard child of Newspeak. What better way too have people forget about something than to remove its linguistic manifestation from the dictionary? Same thing here, except people's way of expressing their approval/refusal. Google and the like are literally after the ultimate "macdonaldification" of people - cattle unable to say "NO".
It's the exact same move that reddit made years ago, and it enabled a new era of online brigading and fake credibility/popularity for content shown on pages.
If you can't see downvotes on content, there is no way of determining how misleading the content may be.
It also disguises cases where a creator may be getting harassed or bullied by others.
The beginning of the end of being able to trust YouTube stats (as if that wasn't already an issue).
This is incorrect; Reddit still shows downvotes, as well as upvote/downvote ratio, even if you're on the "new" Reddit.
If you're on the old Reddit, some subreddits used CSS to hide the downvote button, which I think is dumb, but you can easily get around that by going to the new Reddit long enough to downvote.
You don't even need to suffer through "new Reddit" long enough to hit the downvote - you can just disable custom CSS on old Reddit (to be clear this is an account-level setting on Reddit, not some browser hack you need to do yourself).
You can't see the ratio, but if a comment has a negative score you'll see it, which I would count as downvotes being "visible" -- certainly better than YouTube's terrible choice here. Downvotes are also surfaced when the ratio is ~50% with a "controversial" marker.
I don't believe reddits obfuscation efforts did anything but significantly hamper spammers who relied on accurate counting to be able to determine when their spam accounts were shadowbanned and the reach they had.
Reddit has always had a significant problem with downvote meaning "disagree" or "dislike", which according to redditquette is incorrect usage. In practice, the downvote is about enforcing groupthink on reddit and enforcing subreddit culture, and it is just as powerful today as it has ever been. Sort by "controversial" and you'll see.
> If you can't see downvotes on content, there is no way of determining how misleading the content may be.
There are many ways of doing so, with their own pros, cons, and time requirements. At best you're losing a relatively low quality way of quickly filtering out some videos.
> It also disguises cases where a creator may be getting harassed or bullied by others.
You seem to want it both ways. It is judging videos by this ratio that allows the harassment to have an effect.
There may also be an emotional component to it; a private downvote likely stings a lot less than a public one. It prevents people from piling on or joining in the harassment. And it can allow YouTube to shadowban harassing accounts.
For the people who could actually affect change, it hides nothing. YouTube and the creator both get to see the numbers, and in extreme cases, they'd be available for law enforcement.
> The beginning of the end of being able to trust YouTube stats (as if that wasn't already an issue).
> Why are dislike counts more credible than like counts?
Nobody said they are. I'd be about as upset if like counts went away but dislike counts remained. The point is having both counts gives you a very useful signal.
>I'd be about as upset if like counts went away but dislike counts remained.
Interesting thought: why do you never just see a dislike button anywhere?
>The point is having both counts gives you a very useful signal.
No. The point is that you really don't know how useful the signal is.
Dislike buttons actively encourage bad behavior, so can create skew in the signal. That's intuitive, and also backed up by YouTube's research; hence the change here.
That might tell you something, but it’s not always clear what. Could be brigading, sock-puppets, tribalism, or lots of other causes.
In the end, pretty much all such voting mechanisms are flawed in some way or another. More interesting are the comments left by known experts in the field.
>pretty much all such voting mechanisms are flawed in some way or another. More interesting are the comments left by known experts in the field.
Agree with this. I think downvoting is net-negative, as it encourages bad actors. Let relative upvotes determine quality. For disagreement, use comments. For content that violates terms or is misleading, etc, use flagging.
They want to hide how many people are agreeing with the user, so the user feels he is the minority.
The power of resisting something comes from feeling like many other people are also resisting it. Removing the dislike count means the user can't see if he is alone in his opinion or not.
It's a shit move designed to promote further self censorship and control over the minds of the users.
> You can still dislike videos to further personalize and tune your recommendations
Not long ago in history, such a statement would have been totally incomprehensible, I mean completely nonsense. Even linguistically, there's so much baked into this. For starters, that there is an algorithm which which the viewer has a relationship with, and that the viewer wishes to further refine that relationship by expressing preferences. Yet, the interpretation of those preferences (by servers) are held within a black box.
It is interesting to put things into perspective like that, but I'm not quite sold on the idea that you need to know these background details (of algorithms and black boxes) to understand that expressing your preferences (dislikes) to a business may lead them to tailor what they serve you.
It turns out YouTube didn't see much benefit in providing a way to (metaphorically) shoot content creators. Taking the bullets out seems like a positive step for the community as a whole.
Helps to know who are the people that don't buy whatever it is that Google is pushing. At the very least serve them carefully selected content to change their ways.
Maybe you are right and he didn't read the article, but I think his points still stand.
This is a ridiculous decision. When someone puts something into the public space and everyone dislikes it, that should visible to anyone. This is just going to make ridiculous conspiracy videos and other harmful material like racist propaganda seem more legit than it is viewed by the rest of the public.
Now tell me more about the gray comments here on HN. If this system wasn’t designed to promote brigading and hive mind reflex… well… unintended consequence?
Often people will post comments which are inflamatory, off-topic, advertising, etc. and 'the community' can mark them down.
The parent poster saying "I've never hit a like or dislike button in my life" is a little bit like saying "I've never thrown litter or picked up litter in my life". Both understandable, one respectable, the other a bit weird - if you like the environment you are in but leave all the litter collecting to others and don't contribute to it, what's that saying about you and your one-sided use of the environment?
If you want HN or YouTube to have stuff you like, and not be overrun by the wild west of the internet, and there are no paid editors like the newspapers of old, not adding to the collective voting is like saying "everyone else, moderate this for me into a place I like, thank you".
> "I've never hit a like or dislike button in my life"
The Prime Directive protects lesser-evolved, unprepared civilizations from the dangerous tendency of well-intentioned starship crews to introduce advanced technology, knowledge, and values.
And even then, HN's guidelines say the downvote button is for comments that don't add to the discussion. Even a comment with incorrect information can add to the discussion (though admittedly much less often than an accurate comment), so it's not really about accuracy, either.
(But I will admit that I often downvote things simply because they are inaccurate.)
Why should that be up to the viewer to deduce from likes or dislike count?
The videos that people fully watch, have a positive comment section, and receive many more likes than dislikes are the videos that should result from searches.
Or how about repair videos. There are countless diy fixes for cars, computer, cell phone repair, and countless others. I use the dislike count to see if I am wasting my time watching a bad repair video. That doesn’t mean I won’t watch it but that I am more aware I may need to skim through or fast forward to key parts like the actual removal of a part or something. Anyways I think this is a negative change.
I think this is the common problem with rating systems though. Youtube is saying that the dislike count and ratio are used to unfairly pile on and reduce the real signal of the video. Instead some videos get brigaded for various reason ("I don't like this person" gets conflated with "I don't think this content is useful")
There was a user-submitted recipe app that had a similar problem, where rating system on recipes is challenging because does it imply that the recipe was incorrect, the food tasty, the writing poor? It is too hard to get that signal from a single point of "average rating"
Don't be silly. Of course YouTube can be relied on to properly censor videos themselves. They will make sure you never see anything that may make you uncomfortable. Your complete lack of faith is astonishing. (sarcasm)
> that should visible to anyone. This is just going to make ridiculous conspiracy videos and other harmful material like racist propaganda seem more legit than it is viewed by the rest of the public.
Report videos that violates Youtube's guidelines, and start your own platform if you take issue with the engagement tools they offer. Arguably, the data shows that outrage drives unhealthy engagement [1], and this appears to reduce outrage driven engagement, no different than HN flamewar detection and other mechanisms to encourage more civil discourse.
EDIT: Youtube isn't the internet, nor "commons", it's a single web property. Lots of other forums for your speech (including your own Mastadon, Discourse, or Peertube instance).
"What we learned from the experiment:
Those in the experiment could still see and use the public dislike button, but because the count was not visible to them, we found that they were less likely to target a video’s dislike button to drive up the count. In short, our experiment data showed a reduction in dislike attacking behavior 1. We’ve also heard directly from smaller creators, and those just getting started with their YouTube channel, that they are unfairly targeted by dislike attacks. Our experiment data confirmed that this behavior does occur at a higher proportion on smaller channels."
It's not about appeasing Disney, it's about discouraging shitty behavior.
> because the count was not visible to them, we found that they were less likely to target a video’s dislike button to drive up the count
How is this any different than users targeting a video with lots of likes to drive up its count, though?
If YouTube were hiding both the public like and dislike counts, I wouldn't give this a second thought. But to say "seeing counts influences people" as a reason for hiding dislikes, and not extending that same argument to likes as well just seems a bit shallow.
Public counts are useful because they provide a gauge of perceived quality. If I see a video with lots of likes and no dislikes, it's fair to assume I'll like that video. If I see a video with lots of likes and I can't see the dislikes, I have absolutely no idea whether I'll like it or not; maybe it's a really good video and everyone liked it, or maybe the majority of people disliked it but I just can't tell. If you're removing my ability to gauge quality anyway, then why not just remove the like count as well?
This change is essentially shifting from "three out of five dentists prefer Trident" to "three dentists prefer Trident". The loss of information is so substantial that the remaining information is essentially useless.
I don't think the ratio of likes to dislikes is a reliable indicator of quality. It probably will be for certain categories of video, but not for anything polarizing like a review of an Apple product or anything even remotely politics-adjacent.
Totally agreed, that's why I wanted to call out that it's a measure of perceived quality.
If there's really great documentary about the International Space Station, I'd expect flat-earth believers to think that it's rubbish, that's just part of being human.
Conversely, if I saw a documentary about abortion (or some other very polarized topic) which had a 9:1 like-to-dislike ratio, the difference between that ratio and my expected 1:1 ratio would tell me that lots of people thought that the video was high-quality, even those who didn't share the same ideological views.
"Dislike attack behavior" could just be a video that is widely disliked. It doesn't necessarily mean there's any sort of coordinated attack on it, so your "shitty behavior" comment doesn't really fly.
So people watch it without the prejudgement and bias of knowing what the wider group thinks and dislike on their own judgement. That seems like a simple solution to coaching people into good curation patterns and behaviour.
It will likely so give some good results in the background for YT and media partners but it is allowed to be both.
If nobody can see the dislike number, what is the point of disliking a video? It doesn't make the video less likely to appear on feeds. Having the number makes disliking a video feel like you've actually done something rather than just piss in the ocean.
This seems like the wrong target to me. Why not go after the user?
If there's a user that is going around disliking a ton of videos, especially on a single channel, and/or without watching some minimum time or % of that video, it seems likely that user is being abusive; so perhaps shadow-ban that user's votes?
Similar with new users.. disallow use of dislike completely until they've been around a while and built up some reputation/history (by watching, upvoting, etc). If the user is possibly a bot (I assume there's algorithms to determine confidence of bot vs real user), require more verification.
I'm sure there's a reason they're not taking this approach, but I find it hard to believe they're choosing they way they are based on the reason they're giving here.
If I can tell just by the dislike count if a video is a fake tutorial or w/e, I'm likely to leave immediately. If I'm on my phone (don't have adblock), I'll have closed the video before the preroll ad even lets me skip it.
Now, I'll have to sit through it and make youtube some money before I can determine if a video is any good or not.
Edit: Think about all those scam channels on YT. Disabling likes/dislikes looks fishy, but now they don't have to worry about that since only the likes are visible. This is a massive win for the "GTA online free money"-type channels.
> I don't see how removing the dislike count will in any way increase the click rate.
The rationale is, Google noticed that people use the dislike count to help decide whether or not to watch a video.
If you can't use dislikes to decide whether or not the video you're considering is actually worth watching, you're more likely to click simply to find out.
In order to find out, you first need to watch the ads.
That's all. Someone at Google simply just put two and two together and realised that the visible dislike count was costing them in ad clicks.
Was it politically motivated? Probably not.
Will it have profound political effects? Absolutely!
That's one argument, but on the other hand if you can't vet the videos by yourself before watching, the average quality of the videos you watch will drop, decreasing the incentive for you to spend time on the site.
Isn't autoplay the normal, default setting? Do a significant percentage of users have autoplay disabled?
Disabling autoplay on an ongoing basis isn't an option for me because I won't log into Google services, except to check who is still mailing my gmail address on the long tail of de-googling.
Not in my world it's not. I'm pretty sure FF has disabled autoplay as well as default. The only time autoplay is allowed is when listening a "full album" play list. But that's only after I've scrolled the list to make sure some asshat hasn't pushed in a rando video below the first page.
I think you could see it on the page, even before the video started playing (i.e. while you're still watching the ads). But maybe I remember wrongly.
Still. Far more likely to quit in the first 10 seconds than you would be to stick around for a couple of minutes and find out. Let alone click a related link.
There actually are browser extensions and userscripts that show a like/dislike bar on the thumbnails. Let's say goodbye to them. Likes per view will be the only metric they'll be able to calculate.
Nothing is stopping anyone from storing likes and dislikes offsite. Now that YT is opting-out of being a source of record for dislikes, it's pretty straightforward for those sorts of tools to step in.
No, that's just not believable.
YouTube (and Google more generally) are full of people who believe they need to control the public conversation for the public's own good. They've gone to the best schools, lived cosseted lives in safe, sparkly clean neighborhoods, and never had to shower after work instead of before work. They're totally out of touch with what regular people think, yet believe they know better than regular people and have not only a right, but a duty to guide the public conversation in what they see as the right direction.
YouTube people have spent years censoring inconvenient thoughts (e.g. the lab leak hypothesis), juicing rankings in favor of mainstream media, and outright banning their ideological opponents.
If they believe that they're just making the internet less "toxic", it's because they don't differentiate between people with different value systems and monsters to be silenced.
There is zero chance this lately instance of censorship is profit driven. It's yet another favor Google is doing for an establishment that's rapidly losing credibility with the public and control of the narrative.
You have to load the page to see the dislike count. They already got ad revenue by the time you see it.
I am sure they have better data than anybody as to how often a video is loaded for the first time and disliked within a few seconds of starting.
While once in a while this could be caused by a bad video start it probably mostly isn't caused by the content of the video.
By looking at behavior of users and pattern matching it probably wasn't hard to see that certain users were "making the number big".
Additionally as the article details they likely targeted those users with a new "don't show the number" feature and saw an improvement in the unwanted behavior (unnecessary dislikes).
The reality is that while dislike is fantastic from a quality filter standpoint human behavior likes making numbers big and sometimes they do weird things you have to correct for.
This is some next-level conspiracy. Like most, it's feasible and perhaps in the realm of possibility, but there are probably 30 explanations above it in the "that makes sense" queue.
Notably, their rationale they posted. Isn't it possible that they both see a positive outcome for creators not to get votebombed, and remove negative publicity from corporate partners like Disney?
I challenge anyone to tell me they do NOT think that brigading downvotes/dislikes is a thing on forums and Youtube.
> I challenge anyone to tell me they do NOT think that brigading downvotes/dislikes is a thing on forums and Youtube.
The discussion is about whether YouTube should be hiding the existence of these occurrences and the potential ramifications of such a policy. I haven’t seen anyone denying the existence of brigading.
Out of curiosity, what is the criteria for dislikes to be considered brigading? If I share a video containing hateful comments towards a community of people and they share that video amongst their community, what do you think the outcome of that will be? I would bet a high percentage of viewers from that community will dislike the video. Is this brigading? Why or why not?
It's because their explanation is nonsensical on its face. How is a dislike "harassment"? How are genuine groups of people downvoting something an "attack"? YouTube isn't like HN where likes and dislikes are summed. People disliking a video doesn't change how many people liked it, both are visible.
This seems to come from a culture in which disliking someone, disagreeing with them or telling the they're wrong is the worst offense possible. In no way is it comparable to actual harassment, so of course people speculate about alternatives.
If someone seeks out content that they wouldn't ordinarily interact with just so that they can dislike/downvote it or post nasty comments - that's an attack.
If someone gets recommended a video and they don't like it, disliking is then constructive criticism.
But there are groups out there that actively look for particular types of content to dislike - without even watching it.
YouTube knows how someone arrived at a video - if they wanted to, they could only count likes and dislikes from people who had the content recommended to them. Or they could drop likes/dislikes from incomplete views. Actually I think they already do that, I recall reading something to that effect years ago.
It's all irrelevant anyway - their stated justification puts "harassment" first. Given the poisonous cancel culture that exists inside Google it's very obvious what their real motives are and the rest is thus all open to question. If it was 10 years ago Google's only justification for this sort of change would have been to improve the utility of the site, and we'd have believed it because that'd have been consistent with their other actions. In 2021? Not so much.
I 100% believe that high dislike counts on ideologically favored videos is the real reason. This includes both White House sources and official propaganda regarding Covid-19 and vaccines. (It is undoubtedly propaganda IMO, though note that that does not necessarily make it wrong)
The question to ask is not what reasons could they have done this for, but why did they do it in this particular way at this exact time?
It's a hypothesis completely unsupported by any evidence, but confidently stated as if it were self-evidently true. This is how conspiracy theories start. Later, someone vaguely remembers reading somewhere (here) that "the White House asked YouTube to get rid of dislikes," then still later it becomes a story on Fox and Friends: "Why is Biden making YouTube get rid of the like button?" By that time, someone's off the cuff supposition has become part of the dank ocean of fact-free memes that have polluted our discourse.
as the post itself points out this has nothing to do with dissident voices but with preventing dislike attacks and 'ratio-ing' as is common on websites that have these mechanisms in place. Downvote bombing content if anything distorts discourse and makes it harder to figure out what people actually think. There's a reason this very website caps downvotes at -4.
What I find cringeworthy is this keyboard warrior mentality of likening an incentive change on YouTube videos to dissident activity, as if downvoting some cat video on the internet is like publishing Samizdat in the underground of the Soviet Union, maybe pipe down a little bit
like downvoting a cat video is the only type of video being downvoted. If a "tutorial" is bad, people can downvote it. If the next view sees lots of downs, they can avoid it and go to the video saving them time, and no longer rewarding with ad share for a worthless video.
>If the next view sees lots of downs, they can avoid it
the exact point is that this is terrible because the downvotes aren't representative of anything. two examples. Go to Youtube right now, pick a 5 minute daily covid update. I'm not talking about op-eds, just factual reporting about the stats. Chances are it has been downvoted 80% because there's a bunch of crazies on that site downvoting every news video.
another example, Thandiwe Newton years ago made some political comments 4chan didn't like, so they started to bombard every westworld video with her in it. I saw those ratios and decided to not watch the show for months. Turns out is actually great and it was just some right-wing internet mob being angry about her.
Nobody ever downvotes shitty tutorials, because the only people who passionately downvote youtube videos are crazy. I can't tell you how many crappy tutorials with wrong info in it i've seen, entirely upvoted.
>the exact point is that this is terrible because the downvotes aren't representative of anything.
The video owner can disable comments. Why not allow them the decision on showing/hiding this data? Clearly, there are people that are in favor and some that are against. Allow the "publisher" to make the decision. Seems like a decent compromise.
> this is terrible because the downvotes aren't representative of anything
Speak for yourself?
I watch football highlights on YouTube. Due to copyright many videos are just clickbaits with perhaps a few seconds of real footage, or are missing games or whatever. The up/downvotes are a reliable measure in that niche.
Oh I could not give a dead flying flamingo about likes and dislikes on YT.
It's the intent behind all this that I find repulsive. As someone justly mentioned above, this all comes down to uninterrupted flow of content towards (less) able to resists psyches.
you nailed it on the head. some players want to control public opinion rather than changing their business model or philosophy closer to what the people want
Actually it was Sony who recently had their trailers spammed with dislikes by Spider-Man stans complaining for the delay in the release of the second trailer for “Spider-Man: No Way Home”.
In reality they have been moving towards this for a while. I've been hit by one of their A/B experiments that turns off the visual ratio bar, so it takes a second to figure out the ratio (if it's bad, I normally skim over the comments).
This reeks of the silencing of what is perceived as dissident voices, is all.
I've personally never hit a like or dislike button in my life, never will ... but found it interesting as a viewer to see the ratio. Even knowing this ratio is probably at least partially the outcome of bot farms and the like.
Disney, DC Comics and the like must've complained about bad ratings/unfavourable like/dislike ratios I guess..
Let's try and see this as a tangent/bastard child of Newspeak. What better way too have people forget about something than to remove its linguistic manifestation from the dictionary? Same thing here, except people's way of expressing their approval/refusal. Google and the like are literally after the ultimate "macdonaldification" of people - cattle unable to say "NO".
(edit typos :-) )
Have a great day all :-)