FlickSurfer – a way to find Netflix gems | Hacker News Reader

FlickSurfer – a way to find Netflix gems

flicksurfer.com

124 pointstrashymctrash56 comments

My experience of UK netflix is that UI is designed expressly to disguise the fact that beyond a thin crust of popular films, there is a desperate lack of content and barely any change week to week.

There is no - "what's new in netflix" - because it would reveal there is only one or two new films a month. Instead you get "popular", "trending", "blockbusters", "highly rated" that let them display the same content in a slightly different order multiple times.

I just feel their only game to overcome the lack of content is to offer you the same film enough times with different forms of social pressure until you finally surrender and watch an Adam Sandler film you have been evading for years...

Not only that, but it's always amusing how the top thrillers overlap with the top action and the top drama films.

I remember when last.fm used to just scrobble (?) my listening habits and offer music suggestions from people who have listened to similar things. One of the key advantages of that approach was that it wasn't constrained by availability of things it wanted to sell me, and the necessity of offering an answer no matter how valid. I'd be interested in using a similar service for films, which could THEN interact with netflix/prime etc and suggest things I might like.

Check out the rather hidden http://movies.netflix.com/WiRecentAdditions

Looks good, but I wish the authors of Netflix apps would include a region selector: there are huge variations in the catalogue depending on where you are.

As a UK resident, this is currently useless to me. It includes stuff that isn't available in the UK (e.g. The X-Files) and excludes stuff that isn't available in the US (e.g., the Shield, the Hobbit, Up). There is overlap, but most of the best content is region-specific.

It's frustrating how much a VPN improves Netflix.

And they've started to lock it down: http://www.reddit.com/r/technology/comments/2r88cu/netflix_c...

No, they really are not.

http://www.engadget.com/2015/01/03/netflix-clamps-down-on-vp...

Netflix tells us that there's been "no change" in the way it handles VPNs, so you shouldn't have to worry about the company getting tough any time soon. With that said, these blocking errors started showing up in the past few weeks, so it's not clear what would have prompted them.

There used to be a site called YMDB, I believe, which let you list your top 20 favorite movies, and then would let you see the lists of, and interact with others who had many of the same movies on their lists.

I was often amazed at how some people who loved many of the same movies I loved also loved some movies I hated. So it wasn't a perfect means of finding enjoyable movies, but it did suggest some interesting possiblities.

Anyone know whatever happened to YMDB and whether there's any other sites like it?

I've been using http://www.criticker.com for a few years and have found some fantastic films. Never heard of YMDB but it seems like the same concept.

Jinni is a pretty good recommendation engine. It's historically pulled from your Netflix ratings and listed where you can watch something on or off line, but that's varied based on APIs. (I think the current status is that they don't have access to any Netflix ratings data anymore.)

Even so, I like a ratings service that's separate from a content platform. I worry about the conflicts of interest Netflix faces (they already tend to underrate movies they only carry in DVD for me).

http://www.jinni.com/

I feel like there's a problem with the data - it shows that, for example, Mystic River has an RT rating of 37% [1], while the actual page has 87% [2], and the IMDB link [3] is broken since the correct URL is [4]

[1] http://imgur.com/M8xYf1G [2] http://www.rottentomatoes.com/m/mystic_river/ [3] http://www.imdb.com/title/tt2427555/ [4] http://www.imdb.com/title/tt0327056/

Edit: After reload I got the correct data. Weird!

Cool. I still don't know of a good website that does this well (RT comes close, except some of their filters are broken).

For some constructive feedback, after using it for a few min, my thoughts were that this would only become useful to me if I had these additional filters:

1. Filter by how recent a movie is (e.g. I may not be in the mood for a sci-fi movie from 1950).

2. Filter by rating (e.g. best RT rating).

3. Exclude certain genres (e.g. not in the mood for Crime).

I'd just love a service that can tell me where something I want to watch is available, particularly among the services to which I'm already subscribed (netflix, prime, directv, hbogo, maxgo, showtimeanytime, etc.).

For instance, I got the urge to watch Funny Farm a couple days ago and couldn't easily check availability without pulling up each app.

I thought the public Netflix API was completely shut down now? Is this using scraping for its content?

I found a comment [1] from the author on a reddit thread [2]. So yes, it seems like the data is already stale and getting staler :(

  [–]WedgeTalon 7 points 5 months ago 
  Interesting, but won't this be a pretty short-lived site since the Netflix API will be shut down November 14? Or are you doing this without using their API?
    [–]n1te1337[S] 2 points 5 months ago* 
    It should work pass that date as the site doesn't rely on the API. I can't promise it though as at the end of the day they own the content no matter how you get it.

1 - http://www.reddit.com/r/movies/comments/2ckxhm/netflix_catal... 2 - http://www.reddit.com/r/movies/comments/2ckxhm/netflix_catal...

Hi, I can confirm that the data is updated daily

I entered John Wayne in the search box and it found nothing. But I know there are at least one or two of his movies on NetFlix.

Yeah that's weird. When you search for "true grit" it finds the movie and John Wayne is correctly listed and one of the cast members. I'll sort it out when I get home, it's probably just a case of adjusting the querying algorithm. Thanks for letting me know.

I searched for [Steve Martin] and my sole result is "iSteve", which, as far as I can tell, doesn't have the token "Martin" in it at all. This is despite five Steve Martin films currently streaming on Netflix US.

It should be fixed, thanks!

It should be fixed now, thanks again!

Doesn't this go against Netflix's Terms of Use (https://www.netflix.com/TermsOfUse) by scraping for data, if you're not relying on the now shut down public API?

Looks cool, but it's not clear to me how it's better than just browsing Netflix. I clicked on "comedies" and Black Mirror came up, which is not quite what I had in mind...along with Postcards from Buster, which is a children's show.

Everybody but Netflix has a good Netflix browser

Nice work! I did something like this recently in Ionic framework, released it open source and deployed it to the Google Play Store: https://github.com/brentvatne/flix - essentially it allows you to flick through movies as you would people on Tinder, and shows the IMDB rating and a description along with each one. It also allows you to filter by IMDB rating and genre etc.

My frustration with Netflix's current recommendation system is that it includes films that it knows I've already seen or rated.

If you are in a browser, you can fade them out or hide them via browser extensions.

I am building Flix Plus by Lifehacker (http://lifehacker.com/flix-plus-customizes-netflix-to-your-h...) right now which does this within Chrome.

Thanks!

Thanks. Flix Plus looks good, though I use Firefox, not Chrome.

Thanks. Flix Plus looks good, but I use Firefox, not Chrome.

Shame, looks useful but i think this is US Netflix and you cannot filter by country (I'm in Canada)

You really need to filter out the IMDB entries with under 1000 reviews, or you end up with garbage like this among the true masterpieces:

http://www.imdb.com/title/tt3524124/

Love it, Well done. One thing: You can't (officially) get Netflix in Australia - I'm wondering how easy it would be to extend this to all films and then maybe make it call Couchpotato to add them to your Usenet search list?

Couchpotato has a bookmarklet that can run on RT / IMDB pages to add movies easily, it's in the settings pages somewhere.

Been looking for something exactly like this. Great job, and thanks for sharing. A small request: expose the number of votes, especially on IMDB. Without knowing this, the rating doesn't have as much value. Thanks!

Hi, when you click on the movie poster you'll get more detailed information in a modal.

Awesome! Does clicking on "RT" mean 'sort by RT rating'?

I am not the owner, but I think this sorts by the "Rotten Tomatoes" score

RT is the Tomatometer

I am looking for recommendations something like if I like Social Network movie then what are other similar movies that I can watch?

Not sure if there is a site that does this already

Netflix itself will do that if you rate movies.

A third party site would have an incredibly difficult time providing better recommendations than Netflix itself. They offered extremely large cash prizes to anyone who could come up with better results, and the winners were only able to eek out ~10% improvements (which Netflix then adopted). http://www.netflixprize.com/leaderboard

You don't need to rate movies or even have a Netflix account to get this data. For this particular example, go here (found by googling 'the social network netflix'):

http://dvd.netflix.com/Movie/The-Social-Network/70132721

You'll see they suggest {Beyond the Sea, Into the Wild, The Blind Side, Good Will Hunting, and Prefontaine}.

This is the problem I am facing. Their recommendations are not even remotely related with Social Network movie and that is just one example. If they recommended Steve jobs movie or Black hat or Sword fish then it is understandable.

Thanks for your comment though.

Presumably they're not related by movie subject; they're related by viewer preference. That is, Netflix is effectively saying that stereotypical person who watches and enjoys [movie A] will also tend to watch-and-enjoy movies X, Y, and Z.

This is the same algorithm broadcast networks enact when they try to figure out what shows to pick up: "who does our audience consist of? What would they watch if it were offered? Air things like that." This is why, for example, the History channel shows Ancient Aliens documentaries—they have nothing to do with History; they have everything to do with the particular people who tend to watch the History channel.

You can usually figure out pretty explicitly what cluster you're being assumed to be in, and what it's like, by watching the advertising attached to a TV show or movie. Why do they play SUV ads at Disney/Pixar movies? Because the people in the theatre with money to spend tend to be parents-of-five-year-olds, who often are considering purchasing a new car. Thus, if you told Netflix you had watched a Disney/Pixar movie, it would probably recommend you other media that people in your cluster—people who might buy an SUV soon, to put it one way—would watch.

If you want personalized recommendations and not just generic recommendations, you do need to rate movies you've seen. They need to have some input that differentiates you from the average person.

API seems pretty responsive.. what is the technology stack?

I'm assuming it's just plain PHP (X-Powered-By:"PHP/5.5.9-1ubuntu4.2") on Ubuntu with Nginx (Server:"nginx/1.4.6 (Ubuntu)").

I was curious about this too; I didn't see anything from the author on the reddit thread. The client is in AgularJS and uses bootstrap. The headers don't tell us much, but suggest PHP behind nginx... maybe Laravel?

  Connection:keep-alive
  Content-Type:application/hal+json
  Date:Mon, 05 Jan 2015 00:44:29 GMT
  Server:nginx/1.4.6 (Ubuntu)
  Transfer-Encoding:chunked
  X-Powered-By:PHP/5.5.9-1ubuntu4.2

Frontend: angular + bootstrap

Backend: apigility/zf2 + elasticsearch + nginx

Needs a <noscript> tag. I saw a blank white page until I allowed the domain in noscript.

Check out http://flixfindr.com as well.

It supports Amazon Prime and Hulu as well with a better UI.

I like the UI a lot. Content-wise, it lists "Enough" and "Murder by Numbers" as gems. Hmm.

I wish there was a to filter by things like the number of episodes or seasons.

Looks good, Is there something like this for Amazon Prime Video?

Reply on news.ycombinator.com