blossom Summary
How and why I rate things I like.
Introduction
I really enjoy rating things. I know it’s silly, but there really is a clarifying power behind putting my opinions into words. (You don’t have to agree with them, and I won’t tell you them unless prompted and/or you dig for them in this garden!)
Some places you can find my ratings/reviews:
- Music: music i’m listening to, Rate Your Music, my top 5 albums of 2024
- Books: my bookshelf
- Film: letterboxd
- Restaurants: @bencuan on Beli
Related writing: is there such a thing as bad art?
Criticisms of the 5-star rating scale
Quote
While it was occasionally applied to film criticism as early as the 1950s, the five-star scale wasn’t used to rate hotels until 1979, and it wasn’t widely used to rate books until Amazon introduced user reviews. The five-star scale doesn’t really exist for humans; it exists for data aggregation systems, which is why it did not become standard until the internet era.
—John Green, The Anthropocene Reviewed
Now that we are in the midst of the Internet Era, we’ve become fully star-brained. Any conversation involving where to go for lunch, or which product to buy on Amazon, or dozens of other everyday decision-making inevitably involves leaning on opinions written by strangers online.
This in itself isn’t necessarily a bad thing: some data is better than no data!
The main issue I have with the five-star rating scale is how widely it ranges based on contexts, platforms, and cultures. ‘Four stars’ could range anywhere from amazingly incredible (a restaurant in Japan) to a major red flag (Uber driver rating).
Personal rating scales
contextual 5 star scales
My default 5-star scale skews more critical than most. It roughly breaks down into the following:
- I basically never give 1-star ratings, unless someone/something actively tried to kill me.
- 2 stars is my catch-all for ‘that was really bad and I would not recommend this to anybody.’
- 3 stars is conditionally good. There is probably someone out there I would recommend this to, but it could also be a terrible fit for another.
- 4 stars is very good. I would not hesitate to recommend this to most people.
- 5 stars usually means something was life-changing. I very rarely rate things 5 stars unless they are literally perfect and/or the best of their class.
The one-to-ten scale maps 1-1 onto the 5 star scale: the number of stars is equal to 1/2 of the numerical scale. (1/10 = 0.5 stars, 5/10 = 2.5 stars, and so on).
The Michelin star scale
Whenever I have the choice, I typically use the Michelin star scale for inspiration. The main reason why I like this scale so much is that it’s a recommendation-only scale, i.e. the star rating is far less important than the presence of something on the list itself!
In practice, this breaks down into something like this for me:
- ∅ no stars: YMMV: they range anywhere from ‘just above average’ to ‘absolutely incredible but not for a general audience’. Again, if I did not thoroughly enjoy something I won’t put it on a recommendation list, so zero-star entries are generally great!
one star: fantastic and highly memorable; among the best of its class.
two stars: I recommend this to anybody. This lives rent-free in my head and has influenced me in some significant and repeated way.
three stars: this changed my life and maybe it will change yours too.
Stack ranking
Beli-style stack ranking that assigns scores based on relative positioning (would you rather have X or Y?) works well for lists of things that are similar. However, it becomes reductive very quickly when you need to evaluate something on more than one dimension.