Starter Kit to browse Web in 2024/2025 BUT BETTER! (Still Must-have)
Starter Kit to browse Web in 2024/2025 BUT BETTER! (Still Must-have)
better version of: https://lemmy.world/post/21210097
<3
You're viewing a single thread.
the return dislike plugin is just stupid. it's a community database now, so rather than being based on the actual number of dislikes on youtube it's based on the dislikes of the people who have the plugin.
it used to be that it actually got the real numbers but youtube removed that endpoint so now it's just a misanthropic echo chamber.
It's better than nothing. Also I'd probably weigh the opinion of people who have the extension higher than of those who don't.
Is it? personally, when i think of "people who want to see the dislike bar" my mind equates that to "people who want to dislike", and that's not a group of people i want to interact with.
To me it wrings more of "people who want to know before they waste 20 minutes whether this video is clickbait, just actually false, rightwing conspiracy theories, or has some massive editing flaw that makes the video pointless.
For that, I prefer DeArrow. Gets you community sourced, informative titles for videos (and removes clickbait thumbnails as well)
Tbh the removal of clickbait thumbnails made YouTube less enjoyable for me. I don't really watch those over the top clickbait youtubers anyway, so I've found that this isn't too big of a deal.
You can turn individual features off, or whitelist YouTubers you like
This is why I use it, also the more people who use it the more accurate it is
What an awful mischaracterisation. While the dislike feature may appeal to misanthropes, it also appeals to the much larger pool of people that are intelligent enough/respectful of their own time to understand the value of the feature in helping to avoid poor quality and misleading content.
It really bears out in the results. For those who recall what old ratios looked like, for sufficiently popular videos, they still hold true with this plugin.
i just think dearrow seems like the better choice for that purpose.
How so? Like/dislike ratios are a very quick and effective means of identifying problematic content. Clickbait titles and thumbnails are another issue - they aren't a reasonable indicator of the quality of the content. It's an emerging trend in an oversaturated environment in which even creators of high quality content feel the need to partake.
but like... the like/dislike buttons are there for the algorithm. they're there to tweak your recommendations. always were. the problem with having the ratio visible is that people will brigade, and that "mainstreams" your recommendations, making them less useful. that's why youtube removed the visible ratios. putting them back, and just for people with the plugin, means your recommendations will start to align only with the people who have the plugin. it literally becomes an echo chamber.
Like I said, many of us use like/dislike ratios to gain insight into quality of content; whether or not that was the intended use of the feature is really irrelevant.
You're massively overstating the incidence of brigading. A tiny fraction of videos have experienced this. The idea that brigading is significantly affecting user recommendations and even causing them to align is hyperbolic nonsense.
well that's good then. personally my experience with the dislike bar is mostly from siIvagunner, where videos with well-used tracks got shitloads of dislikes, more than the views of videos without them. this despite the quality of work put into them. i'm glad to be rid of that.
I have no experience with that channel. What makes you think it is being brigaded? Seems like a very unusual target. Given that it's well-used tracks that appear to be affected, it seems to me that it may just be attracting a greater breadth of viewers - who are generally less familiar with the content and may have different expectations (and would be just as likely to downvote whether visible or not.)
okay, this is going to require some introduction.
back in the olden days of youtube, there was a channel named silvagunner. it uploaded video game music in high quality ripped straight from the game files. this channel is long since gone due to dmca, although it lives on as the lower-profile followup gilvasunner.
during the first channel's reign though, a parody account was created called siIvagunner (that's two i's) which exactly copied the style and text formatting of the original channel, but replaced the tracks with "subtle" changes. this one is one of the more famous examples. if you haven't heard the track in a while you may be lured into a sort of Mandela effect of "maybe it did always sound like this" and then they hit you with the full switch at the midpoint. hilarity ensues. this is, truly, a high-quality video game rip.
siIvagunner is a really popular channel, and is run by committee. they have a vetting team, and scores of people have contributed tracks. like with everything, there have been trends in what content is included in the rips. from absolute dogshit to frankly insane effort, you at least know you're in for a laugh when they upload (which they do a lot; for a while they averaged one track an hour). these trends usually take the form of one track being parts of many rips at once, with different spins.
some people, who i think must have thought that their taste was objective or something, took it upon themselves to police the meme quality of the videos. for a while, it was customary to downvote any rip containing a specific track from the Love Live series, regardless of quality, to simply "warn others". this brigading was so common they even made memes about it.
ok, intro over. fact is, brigading occurred. my point is, this has a snowball effect. we know for a fact that only a small percentage of viewers actually like or dislike, which means that when it's applied as consistently as it was here, great videos are skipped due to some people downvoting out of some sense of "community service". siIva, of course by it's very nature can't make money, but this is in my opinion indicative of a common issue with the platform.
It's honestly relatively representative afaik
Even though it, SponsorBlock etc. could just be neglected by using piped or just federated alternatives.
how can we even know that though? there's no stats anymore.
also, if peertube et al got bigger, surely sponsorblock would be useful there as well? if creators would use those services because people were on them, sponsors would contact those creators because they got the views.
It subjectively feels right, and instances where creators leaked the numbers the RTYTD stats matched pretty closely.
Until peertube grows to that size, it's still a long time. Then it would be useful.