Live data from Hacker News

How useful was the Netflix Prize challenge for Netflix?

quora.com

191–200 of 205 posts

Re: How useful was the Netflix Prize challenge for Netflix?

#191

Earlier quoted context omitted.

Completely agree. I am constantly baffled at the bullshit Netflix suggests me. I just went to the frontpage of my Netflix and...: - "My list" recommends 3 series where I have alread watched the last episode - "Only on Netflix" recommends another two series, I have already wathed. - A section displayed is "Watch together for older kids" - I dont have kids, never watched any kids stuff on my account. - "Documentaries"…

"My list" is not algorithmically generated; it's a list of shows you have chosen to add to it. It would be nice if have you the option to remove a show from it when you finished it but very annoying if it did it automatically behind your back. Although clearly the real problem here is that "my list" isn't a clear enough description to show you manually control it (but I don't have a better suggestion).

Shows need to stay in "my list" in case there's another season later.

Re: How useful was the Netflix Prize challenge for Netflix?

#192
post #190

Earlier quoted context omitted.

The difference now is that, for 2 or 3 dollars, you can individually rent and stream most movies instantly. No need to pay a subscription that requires you to scroll through a small low-quality subset on a irritating interface.

Speaking specifically of the US market (I don't know where you are), the list of available titles for transactional rental at any time is a tiny subset of all movies that exist on digital due to windowing (licensing) restrictions. By far, most movies are not available for rental.

That's fair. Maybe an order of magnitude difference between what's available on say Netflix streaming and what's available for individual rental, and maybe another order of magnitude for all movies? There sure are a lot of movies. I'm not sure where the Netflix DVD rental falls, especially if you account for movies that are technically available but with so few copies that it may take months to actually come to you.

Re: How useful was the Netflix Prize challenge for Netflix?

#193

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

>- If I watched a movie yeah some movies might be more likely to be rewatchable quicker by a larger amount of the population, and some people might be more likely to be able to rewatch movies they like more quickly than a month - and those people might have preferences that indicate their liking to rewatch more often. Thus it might be nice to do AI on this.

Why would you need to be recommended a movie you already watched? You know about it and if you want to watch it again just search for it. Recommendations should be for discovery of new content.

Re: How useful was the Netflix Prize challenge for Netflix?

#194
post #128

I used the same dataset for a small project in my CS master. It was a really fun challenge, and it taught me a lot. Most notably, it taught me that it was incredibly hard to make significant progress past the most simplest and naive approach. That approach was "Take average rating a user gives, take the average rating a movie gets, multiply". (Ratings normalized to be between 0 and 1). Just using this method would gi…

This is an important point about a lot of sophisticated models; you're really fighting for a few percent improvement over simple approaches. Sometimes a basic linear regression will get you 70% there, while a trained neural net will bring that up to... 75%.

A few percent can make a difference, especially in competitive areas; but the biggest win is just getting something in where there was nothing before. It's a bit like optimizing code.

Re: How useful was the Netflix Prize challenge for Netflix?

#195

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

You forgot the "Hidden Gems" section with movies that you had already watched and even rated.

Re: How useful was the Netflix Prize challenge for Netflix?

#196
post #13

Earlier quoted context omitted.

I think Netflix doesn’t have enough content to do this. They have their own originals and just a thin layer of other stuff. So I think any truly personalized content channel would get exhausted quickly. What I’d like to have is just a channel of curated or semi curated movie content that I can leave running or forward through to watch. I recently stayed in a hotel with 6 channels of hbo. It’s kind of refreshing to ha…

…Huh? Netflix has more originals at this point then HBO has across their entire library. That’s ignoring the licensed content Netflix also pays for. I can appreciate that you may not like said content, but they certainly aren’t lacking.

HBO licenses more movies. In my week in my hotel room I haven’t watched a single “hbo original.”

So Netflix does have more original content, I don’t care. I want to watch shows and movies and don’t care what studio produces them.

Can you imagine only listening to Spotify produced music? Or Apple Music or something?

Re: How useful was the Netflix Prize challenge for Netflix?

#197

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

The Netflix catalogue, and the movie rental/streaming ecosystem in general, was completely different when the AI challenge was going on. It was a DVD postal service. Your ratings mattered, and almost any movie you wanted was available. No one was bored scrolling Netflix trying to decide how to kill two hours before bed. They are clearly not using the ratings based automatic recommendations any more. It's not even rel…

> The Netflix catalogue, and the movie rental/streaming ecosystem in general, was completely different when the AI challenge was going on.

With Netflix producing its own content now, and with the cost of acquiring content rights much higher than it used to be (all major streaming platforms want to offer great content), I'm wondering how much the business imperative impacts the recommendations we get -> eg. Netflix giving priority to its own content over licensed shows/movies.

Re: How useful was the Netflix Prize challenge for Netflix?

#198

Earlier quoted context omitted.

I get the value of rewatching stuff, but some people take it to an extreme and I suspect these were the same people whose parents let them watch the same children’s show/movie over and over and over.

I was one of those kids and hate rewatching things now.

I’m glad you got out of it

Re: How useful was the Netflix Prize challenge for Netflix?

#199
post #190

Earlier quoted context omitted.

The difference now is that, for 2 or 3 dollars, you can individually rent and stream most movies instantly. No need to pay a subscription that requires you to scroll through a small low-quality subset on a irritating interface.

Speaking specifically of the US market (I don't know where you are), the list of available titles for transactional rental at any time is a tiny subset of all movies that exist on digital due to windowing (licensing) restrictions. By far, most movies are not available for rental.

It's true. But most movies that people actually want to see, are.

I worked in a video store, back when there were such things, and can attest that the vast majority of people wanted the new thing and ignored the back catalog. It was my job to get them interested in the back catalog. I didn't do very well.

I joined Netflix because they had that back catalog available. But now that I'm old and grumpy, I've seen most of what I want to see in that back catalog, too. There's a ton of stuff in that category of "I'm sure it's great but I just don't want to work that hard". Also... most of that back catalog is crap, just under Sturgeon's Law.

Sadly, Netflix has figured that out, and gotten rid of most of its back catalog of DVDs. I hope the real film buffs have some other place to go get it.

Re: How useful was the Netflix Prize challenge for Netflix?

#200

Earlier quoted context omitted.

The DVD catalogue is still available. They even have an app for impulse adds :) https://dvd.netflix.com/App

Yeah. I finally gave up on it last year not because of the catalogue, but because the turnaround got too slow. Individually renting movies to stream is better all-around.

It's still working out ok for us. We're doing the 2 DVDs at a time, and while I understand netflix always was suspected of "rate-limiting" people who had too much churn, we must not be anywhere near that since turnaround is still just a few of days. We're not impulse watchers so it's fine.

It's still tons cheaper than individual stream rentals, there's more device flexibility, a lot more choice, doesn't require blackbox DRM on the devices (they are probably going to force secure enclaves on linux for their DRM at some point once the kernel patches propagate).

Post reply on HN