OK, listing five movies I love and five I dislike, with they x/10 score I believe they contemporaneously received. 1) The Shawshank Redemption (10/10). Instantly recognized masterpiece. 2) Tommy Boy (8/10). Hysterical, even though retard jokes don't age well out of the 90s. 3) Children of Men (10/10). Cinematic marvel, thought-provoking sci-fi at some of its best. 4) Michael Clayton (9/10). Underrated thriller. 5) The Big Lebowski (10/10). They must have known this would be a cult classic. 6) Enough Said (4/10). I just remember wanting two hours back. 7) Wild (probably got 10/10). I disliked the movie almost as much as I disliked the first 50 pages of the book I read. 8) American Sniper (probably got an 8/10). I was distracted by the raw propaganda of this movie. I've been told that if I hated this movie, I should stay away from Michael Bay's 13 Hours. 9) Ex Machina (8/10). Twice I could not get more than halfway through this movie. 10) .... 2001 A Space Odyssey. (10/10). I know I will be called a stupid, unappreciative millennial. But I spent a month reading Anna Karenina everyday that went by faster than this movie. I can almost feel how wrong I am on this one, but I just can't force myself to like it. Alright, firing up Rotten Tomatoes. 1) The Shawshank Redemption (10/10). 91% 2) Tommy Boy (8/10). 44% (LOL) 3) Children of Men (10/10). 92% 4) Michael Clayton (9/10). 90% 5) The Big Lebowski (10/10). 81% 6) Enough Said (4/10). 96% (Holy Jesus what the fuck?!) 7) Wild (probably got 10/10). 90% 8) American Sniper (probably got an 8/10). 72% 9) Ex Machina (8/10). 93% 10) .... 2001 A Space Odyssey. (10/10). 94% Ok, so I wasn't totally off. I was generally within 10%, except for the times I wasn't. I don't know if my aim is good or not, or if I even want it to be good. But certainly there's an effect of priming: if a movie is rated poorly, I brace myself to enjoy it less, which is a wholly unnecessary waste of 12 bucks. I don't need to give that power away.
Ahh, but what did Netflix think? Because Netflix's algorithm is refined to your inputs. It uses affinity to track how well it thinks you will like something. It's our magic new data driven studio that creates content based on what people really want! And it gave us Orange is the New Black! And House of Cards! And Marco Polo. And that's my point: at the end of the day, the best curve fit is still an approximation of what you'll think, and it doesn't know things like "kleinbl00 hates watching OITNB Season 1 because the sound is shit, and Season 2 because you guys have never asked me how I feel about fat lesbians getting naked together." Some things it doesn't track well. The Napoleon Dynamite Problem has been solved for like eight years now and Netflix still gives four stars to shit that I hate, and two stars to shit that amuses the hell out of me (Netflix doesn't get irony). And never mind the averages. Of your ten movies, you gave perfect scores to half of them. Of those four, three of them effectively got an A-. One of them got a B-. Doesn't mean they're wrong, it means they aren't tracking your tastes. Which is pretty much this discussion in a nutshell. Again - I'll use reviews that illuminate what, exactly, is broken about something. I didn't buy a phone system because a couple reviewers said "yeah, it supposedly does this thing, but we haven't gotten it to work." I got the more expensive version because everyone said "this thing is a total piece'o'cake to set up." It totally isn't, by the way... but tech support is super helpful so I'm happy with my choice. It's all in how you use it.