Live data from Hacker News

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

engadget.com

401–410 of 437 posts

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

#401
post #396
post #334

Earlier quoted context omitted.

The ballot box as well. Only half joking here.

The flag link for election misinformation....joking

I wasn't making a reference to American election fraud, but rather the notion that even in nominally democratic countries you are often voting for different sides of the same political structure with little fundamental change

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

#402
post #338

Earlier quoted context omitted.

Mine will not give up on family guy. Hundreds of dis likes. Don’t recommend.

"he just keeps watching and engaging with content, let's give him more!" Click triple dot thingy on recommendations and select "not interested"

I do. Constantly. It doesn’t care.

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

#403

Earlier quoted context omitted.

None of this works for me. I get constant recommendations for (what I consider to be) asinine lifestyle videos, and music videos completely unrelated to my interests. It doesn't matter how many times I say I'm not interested in K-pop, it doesn't help.

Dont recommend channel works real well.

Not for me. Just gives me more of the same.

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

#404

Earlier quoted context omitted.

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…

My memory isn't 100% but they essentially said that most players are actually unlikely to be toxic with any regularity. Typically players will have a bad game, and they will lash out. They said that for players that are habitually toxic, a short ban (I think 2 weeks is standard) significantly reduces the chance of reoffending. For players with repeat bans for harassment or aggression they get permanently banned or eventually quit because of frequent longer bans.

Most of their communication should still be online in some form whether on their website or reddit. All this is just from my memory so subject to misremembering.

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

#405
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…

I think that's part of the problem with black box algorithms controlling a large portion of the internet.

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

#406
post #94

Earlier quoted context omitted.

Google have a ton of other ML features about you that they fit to the model and so you have been profiled to be a bit lecherous.[1] Do you feel good about that? Think how it must feel to be profiled to be a crook or a bad credit risk or something else that actually impacts your life and is false. It's a little worrying. [1] Other people like you like this so it is assumed you do also. Without "lecherous" ever being m…

The one nice thing about Google's ML features is that they're so absolutely terrible that what they think means nothing about you at all.

Until it stops you from getting insurance. Or flags you to border control.

But that isn't happening, and won't ever, right? Right?

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

#407

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…

Who decides what is uninteresting?

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

#408
post #337

Earlier quoted context omitted.

Can this not be at least partially solved by relying more heavily on a user's feedback(like/ dislike buttons, numerical rating systems, and any other feedback mechanisms) to better understand preferences of the viewer? Along with simply being more patient and diligent when trying to discern actual user desire? It seems like the problem of asymmetry in recommendation platforms is more a choice from a design perspectiv…

They sell ads, not a good user experience. For them to sell user experience they'd need to have competition breathing down their necks first

Exactly what I'm getting at, well said. Rip when Vimeo being actually competitive.

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

#409

Earlier quoted context omitted.

Search is entirely crippled for new stories now. If I saw some breaking news on Twitter I used to pop over to Youtube to find original or better footage than news services were getting, or to get it faster. Impossible via YT's search now, it's just pages and pages of old-media channels with their thoughts on the footage, always clipped down and talked over. Even Twitter seems to do this now, it's becoming harder to f…

Funny that youtube is basically the opposite of google's search which will often only show you results that are relevant to top/recent stories making it very difficult to get results about anything else. It doesn't even have to be that recent in some cases. Searching for "Sandy Hook Elementary" along with just about any other search term is going to give you a page with results about a shooting that took place a deca…

I use DDG now but I wish the search tools weren't just last week, last month, last year but greater than a week/month/year so I could discard more recent results. It's quite the puzzle to find ways to remove things, as using the -search-term-I-don't-want or +term-I-do-want is increasingly less reliable, in my experience.

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

#410
post #315

Earlier quoted context omitted.

Prepend your search with 'intitle:' and you'll get results with the string in the title.

As a regular user, my expectation is that the very first results from a search would be from matching the title string (possibly after trying to match the unique video id seen in the url), only afterwards would results from keyword matching and other such metadata show up. Maybe I'm out of touch but I don't think this is an unreasonable expectation.

It's not unreasonable, and it's exactly how it used to work until Google started messing around with inserting ads into every aspect of the site (they very much want you to you click on the ad-laden videos you were NOT looking for). They changed the default and relegated it to be a search modifier.
Post reply on HN