Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

YouTube’s ‘dislike’ barely works according to new study on recommendations

engadget.com

Post reply on HN
Print view View original

301–310 of 437 posts

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#301

Post by praestigiare »

john_alan wrote: post #292
qwertox wrote: post #263

Earlier quoted context omitted.

Apparently you cared enough about it to start watching it after reading the title. A video with "Rust" and "Tokio" in the title will probably be unwatchable if it has a like to dislike ratio of 1 to 3, even if the title signals you that you want to see that video. But since there no longer exists this ratio, you'll have to watch it partially only to notice that it is bad. And then dislike it (and then have this disli…

Quoted post unavailable.

I don't know what you think the decision making process was on that, but you're wrong.

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#303

Post by rasz »

daveoc64 wrote: post #131
mikrl wrote: post #80

I didn’t like when they removed the dislike counts, and I still don’t like it. When I look for videos of folks working on Subarus, I don’t want to have to manually scrub through the video to gauge the quality. Before I’d just look at the ratio and if I was feeling generous I’d preview the frames to see if they had what I wanted. Now it takes me at least 2 mins longer to get to the same judgement which begs the questi…

The problem with the dislike button seems to be that nobody agrees on what it should be used for. In previous HN discussions, it's become apparent that some people use it to say "I am not interested in this video". That is of zero use to other users. The video isn't necessarily bad, it's just on a topic you don't care about. If you are instead using it to mean "this video is factually wrong" or "this video is obnoxio…

The _only_ problem with dislike button was that if affected ad buyers (game/movie publishers) and YT corporate - the worst crime of all.

>On December 13, 2018, YouTube Rewind 2018: Everyone Controls Rewind became the most disliked video on the video sharing platform, with 15 million dislikes rapidly surpassed the music video for Justin Bieber's song "Baby"

You can dislike random people, but Susan hearing her pupil star employees content is shit is where you cross the line. Especially when it happens again next year: "YouTube Rewind 2019: For the Record"

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#305

Post by cerved »

TheAceOfHearts wrote: post #32

The trick to improving YouTube recommendations is to regularly prune uninteresting videos and channels. Eventually this should start surfacing videos you're more likely to find interesting. I've actually hit the end of my recommendation stream on the mobile app after marking dozens of videos with Not Interested. Once this happens it starts recommending videos you've already watched or older videos from creators you'v…

TikTok also makes imperfect recommendations, the difference is that's it's trivial to skip a recommendation

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#306

Post by alpaca128 »

TheAceOfHearts wrote: post #32

The trick to improving YouTube recommendations is to regularly prune uninteresting videos and channels. Eventually this should start surfacing videos you're more likely to find interesting. I've actually hit the end of my recommendation stream on the mobile app after marking dozens of videos with Not Interested. Once this happens it starts recommending videos you've already watched or older videos from creators you'v…

I never do this because I have no idea what it'll do. If I tell YT I'm not interested in a video will it hide the whole channel from me? Even secretly unsubscribe me as it randomly did multiple times?

If I had to describe YouTube's recommendation algorithm I'd say "undefined behaviour". It might recommend the same uninteresting video every day, but also hide an interesting, subscribed channel's videos for months even though I liked almost every one of their videos. It might mix videos I already watched once with completely unrelated search results. At this point Google could probably improve their algorithm just by flipping a coin.

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#307

Post by vonwoodson »

They left out the block channel button. Very effective. Seriously, as an avid YouTube consumer (and not a creator) training my algorithm is such a huge part of what my entertainment life is. Definitely have used private mode, on a VPN, with a different browser to watch videos I didn’t want polluting my algorithm. Now that my kid knows how to use YouTube and I let him use my account, my years of work are totally shot… maybe. It’s nice to be able to see and block channels, guide the algorithm away from YTP, and be sure to keep plenty of math and science options on the recommendations. I’ve also noticed that it varies by time of day. I wouldn’t mind more control, but as it is, I learn the system and do have control that I need. I don’t always know what I want, and do like surprises. Like, who knew I’d be into cow hoof trimming? YouTube did.

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#308

Post by blamestross »

"repulsion" in a hyperbolic space (especially a higher dimension one) has basically no power to guide selection. That's what a "dislike" is, a repulsion in a space where there is a lot more room to slide around in than most folks intuition allows for.

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#309

Post by luckylion »

adamdusty wrote: post #79
amatecha wrote: post #19

This reminds me of how games ( cough LeagueOfLegends cough ) match me up with a player that I literally just reported for harassment, griefing etc. in the previous match. Like, there weren't any other possible players to match me with out of the many thousands? I don't mind waiting a few more seconds to match up with someone else! I'm getting a lot of responses as if I expect the player to be totally blocked from mat…

This has been discussed by the community for as long as I've been playing (2012ish), and rito has said they won't do it because it creates a "prisoners island" type situation where all the mean players end up playing together and player retention drops. They argue its much better to attempt to rehabilitate rude players, but this was back when riot Lyte was doing his thing. I don't know if they spoken on the subject s…

That's wild. Their reasoning (or even finding, they should have the data) is that if you isolate the annoying players away from the general population, people overall play less? I'd intuitively expect the opposite, you play less if you constantly have rude players around.

Or is it more of a thing of "60% of our players like to harass people, if we only pair them with other harassers, they don't have any fun and leave"?

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#310

Post by jobigoud »

donio wrote: post #151
mdcds wrote: post #103

Earlier quoted context omitted.

I've been using Return Youtube Dislike addon for Firefox. Idk how accurate it is, but you can gauge the sentiment (mostly positive, mixed, mostly negative)

I believe that uses a separate database that was seeded from scraped youtube data but now it's only updated by those who use the addon. So it would have a much smaller sample size for recent activity.

I'm not sure about this because when going through Youtube shorts I regularly see videos with +100K dislikes, and any "viral" video has dozens of thousands of dislikes. I don't think this many users installed an extension to bring back dislikes and happened to dislike these particular videos.
Post reply on HN