6 ms·

You're basically arguing against every movie rating system ever invented. As are the studios.

by 1_2__4 9y ago

eridius 9y ago
Yep. Systems that boil a movie down to a single number are fundamentally problematic. FiveThirtyEight has an article today about this topic (https://fivethirtyeight.com/features/al-gores-new-movie-exposes-the-big-flaw-in-online-movie-ratings/ https://fivethirtyeight.com/features/al-gores-new-movie-expo...), focusing on polarizing movies like An Inconvenient Sequel which end up with middling scores that don't properly reflect the fact that most ratings are either a 1 or a 10.
- kbenson 9y ago
  Yeah, there are better ways to represent scoring, such as Amazon's histogram of star ratings. Unfortunately, that doesn't help Rotten Tomatoes, because they often don't even have a graduated scale to measure in the source data for critics. When many critics are just recommending or not, their scale is basically 0 or 1. They are extrapolating out of much lossier source data. This is also why I dislike Netflix's rating system. There's two axis by which it's measuring (how good the movie is, and how well it matches your interests), and turning a 4 level rating that already combined those nuances into an 8 level aggregate rating and then altering that score based on similar interests lead to quite a bit of confusion as to why something was rated the way it was. Then they doubled down and made it worse by reducing the source data rating to 2 levels and increasing the aggregated output to 100 levels. :/ The only upside is that they are able to throw some interesting compute power and algorithms and/or machine learning at it to generate some good suggestions occasionally.
- Berobero 9y ago
  The example you cite, however, is an argument for Rotten Tomatoes system; as the FiveThirtyEight article notes, the movie has a "great" 78% among critics. This is due to the RT score being a weight of reviews which have been flattened to thumbsup-thumbsdown.
  - eridius 9y ago
    Not really. It's only a "great" among critics because critics were more favorable than audiences (or rather, the movie was less polarizing for critics). But the criticisms that FiveThirtyEight has are still valid, and you can easily imagine a movie that is polarizing among critics with 50% of them giving a 10 and 50% of them giving a 1.
- Spooky23 9y ago
  Middling scores are not super important. The value is on the very low scores. I see a lot of movies in the theater. Movies like Baywatch and Emoji Movie were awful on any scale, and rotten tomato ratings will keep you away. This whole discussion is a distraction from the larger problem -- studios are slinging out shit sandwiches as feature films. I should be getting a better film experience in a theater than on a made for Netflix movie, but I frequently don't.
- kazagistar 9y ago
  For sure. I find ratings useful when looking for something middling, fun, and fairly entertaining. If I am looking for my new favorite, ratings are near useless. The average of my top 10 movies is likely lower then, say, the average marvel movie. But that doesn't mean it's useless to look at ratings. It's just for deciding between "I guess why not" and "naw, I'd rather play a good game or something".
gambiting 9y ago
Video games have the exact same problem, with metacritic score being a really poor way to measure the quality of a video game. For an example how you can review video games differently look at Eurogamer - they stopped giving scores, instead, the site gives their recommendation - either "Avoid", "Recommended", "Essential" or no recommendation at all. It's a vastly better system than an absolute score out of 5 or 10.

You're basically arguing against every movie rating system ever invented. As are the studios.

by 1_2__4 9y ago

You're basically arguing against every movie rating system ever invented. As are the studios.

Yep. Systems that boil a movie down to a single number are fundamentally problematic. FiveThirtyEight has an article today about this topic (https://fivethirtyeight.com/features/al-gores-new-movie-exposes-the-big-flaw-in-online-movie-ratings/ https://fivethirtyeight.com/features/al-gores-new-movie-expo...), focusing on polarizing movies like An Inconvenient Sequel which end up with middling scores that don't properly reflect the fact that most ratings are either a 1 or a 10.

kbenson 9y ago
Yeah, there are better ways to represent scoring, such as Amazon's histogram of star ratings. Unfortunately, that doesn't help Rotten Tomatoes, because they often don't even have a graduated scale to measure in the source data for critics. When many critics are just recommending or not, their scale is basically 0 or 1. They are extrapolating out of much lossier source data. This is also why I dislike Netflix's rating system. There's two axis by which it's measuring (how good the movie is, and how well it matches your interests), and turning a 4 level rating that already combined those nuances into an 8 level aggregate rating and then altering that score based on similar interests lead to quite a bit of confusion as to why something was rated the way it was. Then they doubled down and made it worse by reducing the source data rating to 2 levels and increasing the aggregated output to 100 levels. :/ The only upside is that they are able to throw some interesting compute power and algorithms and/or machine learning at it to generate some good suggestions occasionally.
Berobero 9y ago
The example you cite, however, is an argument for Rotten Tomatoes system; as the FiveThirtyEight article notes, the movie has a "great" 78% among critics. This is due to the RT score being a weight of reviews which have been flattened to thumbsup-thumbsdown.
- eridius 9y ago
  Not really. It's only a "great" among critics because critics were more favorable than audiences (or rather, the movie was less polarizing for critics). But the criticisms that FiveThirtyEight has are still valid, and you can easily imagine a movie that is polarizing among critics with 50% of them giving a 10 and 50% of them giving a 1.
Spooky23 9y ago
Middling scores are not super important. The value is on the very low scores. I see a lot of movies in the theater. Movies like Baywatch and Emoji Movie were awful on any scale, and rotten tomato ratings will keep you away. This whole discussion is a distraction from the larger problem -- studios are slinging out shit sandwiches as feature films. I should be getting a better film experience in a theater than on a made for Netflix movie, but I frequently don't.
kazagistar 9y ago
For sure. I find ratings useful when looking for something middling, fun, and fairly entertaining. If I am looking for my new favorite, ratings are near useless. The average of my top 10 movies is likely lower then, say, the average marvel movie. But that doesn't mean it's useless to look at ratings. It's just for deciding between "I guess why not" and "naw, I'd rather play a good game or something".

gambiting 9y ago

Video games have the exact same problem, with metacritic score being a really poor way to measure the quality of a video game. For an example how you can review video games differently look at Eurogamer - they stopped giving scores, instead, the site gives their recommendation - either "Avoid", "Recommended", "Essential" or no recommendation at all. It's a vastly better system than an absolute score out of 5 or 10.