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

101–110 of 437 posts

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

#101

Post by dredmorbius »

NegativeLatency wrote: post #25
dredmorbius wrote: post #9

Earlier quoted context omitted.

That was my experience years ago when I last did so. Logging in offered no visible upsides and multiple downs. I learned not to log in and use search wherever possible. Now I use Invidious (with JS disabled and/or uMatrix blocking YouTube functionality in the browser as an additional nudge), or youtube-dl / ytdl / mpv to play video directly. Usually audio-only for talkpieces. I'll also disable recommendations and com…

I've been using a cron job to watch/download my watch later playlist so I can view the videos on plex at a later time: https://github.com/nburns/utilities/blob/master/youtube.fish... Theres another script that deletes watched videos from the specific youtube plex library every night. It works great as long as you remember to remove the videos from your watched later playlist before watching them on plex. (Have yet to…

The tool that I was in love with for a few months before Google / YouTube effectively killed it off was mpsyt / mps-youtube.

https://github.com/mps-youtube/mps-youtube>

It was (well, still is) a terminal-based YouTube client which provided search, locally-managed playlists, choice of playback mode (video or audio-only), downloads, and a number of other features. YouTube quashed it with an API-key limitations such that it is now effectively unusable.

But for a brief moment, YouTube content was accessible from a terminal window or console, without all the godawful dark patterns of the site itself. That was tremendously useful.

These days I'll snag videos by URL via youtube-dl/ytdl, and/or mpv (which invokes those transparently). Not quite as magickal, but still pretty good.

Hrm... Apparently there's a successor project, yewtube:

https://github.com/mps-youtube/mps-youtube/issues/1191>

https://iamtalhaasghar.github.io/yewtube/#installation>

I'm going to have to check that out...

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

#102

Post by hombre_fatal »

JohnFen wrote: post #49
hombre_fatal wrote: post #11

Earlier quoted context omitted.

Youtube does have a great escape latch though: go to the History tab and delete your viewing history. Your recommendations start from scratch and you’ll get funneled into a potentially new corner of videos. With that in mind, I think Youtube recommendations are quite good. Just do it when your recommendations get stale, monthly or so. I wish other platforms had that like Amazon.

> With that in mind, I think Youtube recommendations are quite good. Having to nuke a valuable reference like your history in order make the recommendations approach something useful is the opposite of "quite good".

Well, I'm one to appreciate what we have available to us especially compared to the alternatives and temper my appreciation accordingly.

I don't know if I use a single service that gives me particularly compelling control over recommendations. Though I also don't find myself deep-diving my history so often that clearing it every few months affects my life, but I suppose I can see how that could be a setback.

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

#103

Post by mdcds »

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…

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)

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

#104

Post by XorNot »

darth_avocado wrote: post #68
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…

The biggest problem with this approach is that it takes a lot of curation to get your recommendation in a decent shape and it takes you one accidental video view to have your feed be full of uninteresting content related to that video. This asymmetry is the root of all the recommendation system problems.

This right here is exactly it: my number one use of incognito mode these days is when clicking on YouTube videos that have been linked from somewhere else where I'm trying to keep their content out of my recommendations.

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

#105

Post by hombre_fatal »

joe_guy wrote: post #16
hombre_fatal wrote: post #11

Earlier quoted context omitted.

Youtube does have a great escape latch though: go to the History tab and delete your viewing history. Your recommendations start from scratch and you’ll get funneled into a potentially new corner of videos. With that in mind, I think Youtube recommendations are quite good. Just do it when your recommendations get stale, monthly or so. I wish other platforms had that like Amazon.

Amazon actually has a pretty detailed page that lets you control what feeds back into your recommendations. https://www.amazon.com/gp/yourstore/iyr/

Thanks, I had no idea. This is great.

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

#106

Post by maroonblazer »

harry8 wrote: post #95
kilroy123 wrote: post #91

Earlier quoted context omitted.

I assume this because I pay and don't see ads.

You are in a minority. Why do you think they built their recommendation system with you in mind at all? You've already paid, you no longer impact the fit of the short term revenue maximization objective function.

I suspect daily/weekly/monthly active users and minutes viewed are a metric they're interested in maximizing. That would be incentive to optimize the recommendation system for paid users too.

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

#107

Post by not1ofU »

I noticed that when youtube implemented the auto-play feature on the homepage, whereby you hover the mouse over a vid, it expands slightly and starts to autoplay the vids, that it actually puts the vids in to your watch history, even if you only watched a few (2-3) seconds. After this i noticed that my recommendations got crappier almost instantly. Basically, I would see something with a 'nice' tumbnail and hover it. Since then, everytime I open youtube, i need to go into the preferences and disable that shit, so not to pollute my history, and thus pollute my recommendations. I clear cookies when browser closes and this setting doesnt get saved, for some reason, google isnt smart enough to save these settings / preferences to my account like all the other setting, oh no, its controlled by a cookie.

This may have already been scailed back abit since I first noticed it, so maybe not 100% accuarate anymore, still pisses me off.

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

#108

Post by ketzo »

skybrian wrote: post #98
ketzo wrote: post #76

Earlier quoted context omitted.

That’s the problem. this player was at the top of the matchmaking ladder that (very small) population had a lot of people who wanted to avoid him he couldn’t get games at his current level, and he couldn’t rise above that level, because there was nothing above that level

Maybe that means he won the game and it's time to do something else until someone with similar skill comes along? Or maybe there should be a handicapping system.

There were players of similar (and higher) skill — but they didn’t want to play with him because he was “annoying” to play against.

Which is a normal part of a competitive multiplayer game, except that people got the option to avoid him (rather than learn to beat him), so many took it!

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

#109

Post by poglet »

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…

Occasionally I have to clear my search history and 'start again'.

Once YouTube think's it knows what I like, I get 1 or 2 categories of videos I might watch and barley anything else related to my subscriptions.

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

#110

Post by amatecha »

proteal wrote: post #46
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…

Overwatch had a feature where you could avoid players, but a top sniper was avoided by enough top players that he could never find a match. The matchmaking system could never find him a match because almost all potential matches had at least one player on each team that had him avoided. The devs eventually reworked that feature so that he could play the game. Sometimes seemingly innocuous features have weird edge cas…

Right, it sounds like Overwatch's implementation was poor. I don't expect to have the player "unmatched" forever, or even completely unmatched for that matter. Just lowered in the "algorithm", probably including a cooldown factor so the "de-rank via reporting" is gone after some time.

Like I said I don't care if matchmaking takes longer (usually it's only ~30 sec). The pool of players is utterly massive, and ranking one player down a bit would make a trivial difference.

Yeah, I could wait a couple minutes and not queue up. But that's making it my problem that I had to deal with a toxic/abusive player. I already did my part and reported them, but now I also have to just not play the game for a while to make sure I don't get matched again? Indeed, my suggested solution of lowering their matchmaking "matchability" to me would benefit the community. Not to mention, the effect could be nullified if I am spamming unfounded reports against players constantly.

I kind of just assume that a competitive multiplayer game under active development for the past 15 years would have functionality like I've mentioned above. Maybe it does, and it just hasn't worked well to prevent me from matching up with toxic players multiple games in a row.

Post reply on HN