Friends without Netflix

Do you have any friends who still haven't tried Netflix? I was astonished to learn that my nurse at the doctor's office still hasn't signed up. I will not rest until she does.

I think it might be ignorance. I have to explain to them how it works. It helps if I show them the envelope, because it's hard for folks to imagine how easy it is. They think they have to get a stamp and address the envelope and buy envelopes, and all that old-school stuff. Even when I tell them it's postage-paid both ways, they are skeptical. They think there's a catch.

I also think it's hard work for some people to find stuff to watch. With TV, it's passive. You watch whatever is on, or you hit one button on the remote. Netflix requires a bit of thought and more effort than switching on the TV. You have to go online, look at pictures, and click a button. If you're really ambitious, you read the description and reviews.

Even with Roku, you have to decide what to watch. The TV doesn't decide for you. People can be paralyzed by choice. Well, Netflix solves that problem by suggesting movies. All you have to do is rate movies. Is that so hard?

I look at some of my Netflix Friends, and I'm frustrated at how few movies they've rated. Or maybe they really haven't seen that many. They must be the people with the social lives and beautiful lawns.

Re-releases on Netflix

I have to check the Netflix New Releases on HackingNetflix.com, because I love rating the ones I've seen. Occasionally, one will show up that I have previously rated! It went out of print, and then came back. Or maybe I saw it in the theatre and rated it before it was available. Or maybe I rented it from Facets or Greencine. It's how I keep track of what I've seen, to improve recommendations.

Netflix recommends...

One of the features I like about Netflix is that it recommends movies you'll love. The more movies you rate, the more accurate these suggestions are. You should browse Netflix and rate every movie you've ever seen, not just the ones you've rented from Netflix. This really makes a difference.

The red star ratings you see on the Netflix site are actually predictions of what you would like, based on other users' opinions. If you rate more movies you have seen, those stars will start to reflect your opinion more than anyone else's.

I have rated 3,246 titles on Netflix.com. It's possible that I've seen more than that, I just can't remember them all. As a result, there are 347 movies that Netflix expects me to love. Does that number seem low to you? Or is it about right?

Back before the Netflix Prize was awarded, I had very few recommendations from Netflix. Less than a hundred. I think my taste in movies is so personal and eccentric, that it's hard to predict what I will love. Even the Netflix algorithm predicts only four stars for "movies you'll love." Shouldn't it predict you'll give FIVE stars to movies you will love? I guess it's because my taste is so idiosyncratic.

LikeMinds alpha

I have rated over 3000 titles on Netflix.com. I use Netflix as my main repository of movie ratings. I like to know how I rated a movie, in case someone asks me. Sometimes I can't remember if I've seen a movie, or if I liked it enough to see it again, so I refer to my Netflix rating. Having to refer to Netflix is inconvenient sometimes. I don't want to have to start over from scratch with ratings on every movie Web site. What if I quit Netflix? What's going to happen to all those ratings? Now I can keep them. The LikeMinds Web app allows you to take your Netflix ratings with you, or move them from one Web site to another. Right now, you can export ratings from Amazon, Blockbuster, Flixster, Imdb, Movielens, Netflix, RottenTomatoes, and Yahoo, and import to Amazon, Blockbuster, Movielens, Netflix, and RottenTomatoes. They will be adding others.

It took about 10 minutes for my 3030 ratings to be imported from Netflix into LikeMinds.com. It took about a minute to download them to a file on my laptop.

It's in alpha at the moment, so they are looking for feedback on problems or features and are quick to respond. Let me know if you use it, and what your experience is.

Netflix recommends I watch Wings of Desire (1987), and predicts that I will give it 3.9 stars, based on my feelings for Nosferatu the Vampyre (with Bruno Ganz), Downfall (a great movie also starring Bruno Ganz) and Big(!). The problem is I've already seen it. I rented it from Netflix in 2006 and rated it 3 stars. I guess it went out of print and then a new edition was made, but the rating I gave it in 2006 didn't transfer from that edition to this. I don't remember exactly why I gave it a lower rating. Should I re-rate it 3 stars, in total disagreement with what Netflix knows about me? Or trust Netflix? I feel bad for disagreeing with the Netflix prediction of my rating, because they spent a million dollars to improve their recommendations.

Two things of note: I love Bruno Ganz, as an actor, and I rate every movie I've seen on Netflix.com.

Freakomendations

Paul Lamere, a researcher at Sun Labs, collects Freakomendations, which are wacky recommendations from Netflix or Amazon. For example, "Because you enjoyed The Killers, Netflix thinks you will like The Looney Tunes Golden Collection".

Netflix recommends I watch ZERO 258 movies

I don't know if it's the new design of the Netflix Web site, but all suggestions have disappeared. I used to have about 200, because I've rated over 2000 titles. Maybe they've been reset. When I rated the last movie I returned, the number jumped to ten.

***UPDATED 3/30/08 I just noticed, all my Netflix suggestions are back!