Netflix thinks…
…that because I enjoyed “The Great Escape” I will also enjoy “Fiddler On the Roof.”
Netflix is nuts.
Netflix thinks that because I enjoyed “The Adventures of Baron Munchausen” I will also enjoy “The Madness of King George.” In this case, Netflix is probably not nuts, although King George most likely was.
Netlix thinks that because I enjoyed “In Treatment, Season 1” I will also enjoy “In Treatment, Season 2.” This seems eminently sensible. And yet I did not. Go figure.
Makes you wonder. What exactly is the path that takes you from “The Great Escape” to Tevye? That’s worth a post all by itself.
My recommendations are almost as screwy, but that’s because sometimes I pick and sometimes my wife picks. Imagine the results from combining “Eat Pray Love” and “Aliens” or “Amelie” and “The Mysterious Mr Wong.”
I think one season of treatment was probably all you needed.
They’ve been right about some recommendations for me. In any case they’re worth a try.
Netflix thinks? No, it really doesn’t.
Maybe it’s a version of the “Six degrees of Kevin Bacon” game that was popular a few years ago. Try to come up with the connectors that take you from “The Great Escape” to “Fiddler on the Roof”. I suspect it is entirely possible. I, however, have other things to do at the moment.
We watched “Avatar: the Last Airbender” so we’re likely to enjoy… blue’s clues? And Dora? Er, no. Those aren’t even in the same target-age-group.
We enjoyed “Ghost in the Shell” (the anime) so we may enjoy… that animated show with the Mexican kids and the dragons? (Their grandma came to visit from Mexico, from what I remember when watching my sister’s boy– it’s rude to use a generic when you have a specific.)
Me thinks their metric is a bit messed up by a lot of families using the service, and by using over-broad categories. (Although it was oddly accurate to say that we enjoy Man Vs Food so we’d enjoy Stargate: SG1 season 1. Both are very cheerful, not polished-pretty and just have fun…not to be confused with the other Stargates. Bleh.)
Susanamantha: I have it! “Fiddler” is 181 minutes long, and “Escape” is 172. It’s about patience. If you can sit through one, Netflix figures you can sit through the other.
Movies have gotten much longer these days. Those were both VERY long movies for their time.
Watched “White Ribbon”. A good movie that deals with early signs something was going wrong in the psyche of German/Austrian people pre Hitler.
Also “My Brother’s Keeper” is an unusual documentary but very good.
Both are streaming cause i’m the impatient type!
Amazon.de has an interesting recommendation program too. Last year, we bought 2 folding picnic tables, and they keep recommending them to me. How many picnic tables do you need, and who writes these crazy programs? Why not recommend tablecloths that fit or seasonal decorations to use on the table.
Amazon.com does a much better job.
The sort of statistical data mining used by Netflix, Amazon, and other online merchants is a rather inexact science, but they keep using it because it does work a lot of the time. Of course, it’s only as good as the data sets that they are working with. For Netflix, it’s possible that, as Neo postulates, length is one of the variables they look at, along with obvious things such as era, genre, director, and so on, but the more illusive intangibles are not as easy to classify or to categorize, and they can often get things very wrong. Still, it is neat when they are right and you get a movie that you never would have thought to consider otherwise. Two of my favorite discoveries that way have been “Anne of Green Gables” and “Hobson’s Choice,” neither of which I would have considered on my own, and there have been many others I liked, too.
On the other side of the spectrum, though, the two worst suggestions so far have been “Waterland” (dreadful) and “The House of Sand and Fog” (which was almost redeemed by being so bad it was ridiculous).
I just looked up “The Great Escape” and “Fiddler on the Roof” and noticed that Netflix classifies both under the genre “Classics,” among many other terms used for each. So far, then, we’ve got length, genre, era (60s-early 70s), and for some viewers that might be enough for there to be a match.
Pingback:Twitter Trackbacks for neo-neocon » Blog Archive » Netflix thinks… [neoneocon.com] on Topsy.com
running down Netflix??
Hey Gal, this is like trying to put down Monica!
Let the cheap movies roll….
I haven’t used Netflix but I do buy books from Amazon. They periodically send me recommendations which are usually spot on. The trouble is, if I bought every book they recommended, I’d have to buy another house to store them. I’m almost at that point as it is.
LAG, from Teyve to “The Great Escape”:
Teyve in “Fiddler on the Roof” was played by Topol; Topol was in “Before Winter Comes” with David Niven; David Miven was in “Eye of the Devil” with Donald Pleasence; Donald Pleasence was in “The Great Escape”. (Also, all three of those actors were in Bond films: Topol was in “For Your Eyes Only”, Pleasence was in “You Only Live Twice”, and Niven was in the first “Casino Royale”.)
If you really want to see some strange stuff, watch something that concerns faith or religion. They seem to think in this area, everything that involves faith or religion, even those movies that bash both, are ones you would want to watch. Really strange things are shown as things you might like to watch. NO, not only would I not care to watch that crap, but it is offensive to even suggest it.
Y’all are not thinking.
this is not hard.
If you watch Aliens and rate it a 4, this is a data point. If this is the only rating you provided, then the following works: of the entire universe Netflix customers who rated Aliens a “4”, the average rating for Tron was a 4.1. This is a conditional. If you rated Aliens a 4 and you rated Tron, then we display the average rating of all of you that rated Tron. This is an easy calculation. You may love or hate Tron, but prior to your viewing and ranking, everyone else who matched your data point on Aliens thought Tron was a 4 and so we think the odds are stong that you will think it is a 3 or more.
As you build the database, you get more information. If you have 100 ratings, then it is more likely that you will have movies you might want to watch be rated, too.
I pretty much ignore the suggestions but I do like NetFlix….been a member for many years.
When in Texas, I can’t watch streaming movies and I miss that convenience. (In Texas, I’m on satellite Internet service.)
Neo, are you saying you didn’t enjoy Fiddler on the Roof?
Even Arafat shed a tear watching that flick ….
There’s an even quicker connection between Great Escape and Fiddler:
Gordon Jackson starred in Great Escape, then starred with Topol in Cast a Giant Shadow..
Mike Lief: didn’t like the movie. Loved the play, which I saw in the original cast production. I generally dislike movie musicals and find them very odd. But that’s because I grew up watching the originals in the theater.
I’m spoiled that way.
Neo, I had the opportunity to see Zero Mostel perform as Tevya back in the mid ’70s, and it was a thrilling performance, one of my favorite theater experiences.
Like you, I think the film version isn’t nearly as good — and Topol offers a very different take on Tevya — but it’s a decent production of the show.
Have to agree with you that people bursting into song makes more “sense” in a stage production, as opposed to the photo-realistic world of the cinema.
FogCity –
No, you’re assuming that it makes that much sense. Most of the recommendations are straight out of the broader category– “classic movie,” “Cartoons and Anime,” “Musical,” “Documentary”– with a few that go off of shared key-words from those who rated it highly.
Rating more movies hasn’t actually improved the recommendations, it’s resulted in more keyword-sharing movies in the same category.
I wish they would use a profiler like Amazon’s– “Z people who liked X liked Y”– it might be more accurate.
The entire point of the post is that the suggestions suck.
Netflix simply looks at who liked the movie you ordered and what else they liked. It is likely more sophisticated than that, but it is just doing stats on what people order and rate. If you start rating its suggestions, it might become more selective. Of course, when sophisticated Opera lover chooses anything with Elina Garanca in it, but also chooses animated stories, thrillers, science fiction spoofs and B&W classics, what’s a poor dumb computer to do? You liked Toy Story so you’ll love Carmen?
Speaking of Zero Mostel, we just threw out a toilet set he had autographed. It said “Zero Mostel shat here” and had a cartoon sketch of himself. He “created” it when he played my father-in-law’s tent theater, loafing through Tevye, and still bringing the house down. We would have Ebayed the seat but it was past even that.
I’ve found Netflix suggestions to be very good when they are 4 stars are better, but I’ve rated a lot of films. I’ll provide more specifics when I have more time, and I hope others will tell how many films they’ve rated.
What didn’t you like about the second season of “In Treatment”? I loved the first season, and I’m considering whether or not to rent the second one.
Liz: I just thought the second season was flat, somehow. I wasn’t as interested in the patients, and Paul seemed more unrelievedly depressed.
Liz: I just thought the second season was flat, somehow. I wasn’t as interested in the patients, and Paul seemed more unrelievedly depressed. That’s not to say it’s not worth renting, if you liked the first one. I plan to rent the third one when it comes out.
The sort of statistical data mining used by Netflix, Amazon, and other online merchants is a rather inexact science, but they keep using it because it does work a lot of the time. Of course, it’s only as good as the data sets that they are working with