People want a rating when they read a review, some kind of number to summarize whatever more complex thoughts the reviewer might be sharing. Here’s my criteria, based loosely on the Michelin star ratings for restaurants.
I’m generalizing these ratings for any media, but I’m thinking of games as I write this, so I’ll talk about the ratings in terms of games.
No stars: I don’t recommend this game. That doesn’t mean it’s a bad game, or that you wouldn’t enjoy it. I’m looking for amazing game experiences, and most games fall short, while still being perfectly reasonable examples of their genre. Any game that I walk away from and feel no urge to revisit, any game I quit early, any game that makes me wish I were playing a different game; those are no-star games.
One star: I recommend you play this game. It’s good; it has features or content or experiences I found compelling. If you get the chance, give it a try; I think you’ll probably enjoy it.
Two stars: I highly recommend this game. Definitely pick it up the next time you have the opportunity. It is an excellent example of its type of game, or it’s genuinely innovative, or it provided a particularly compelling experience.
Three stars: One of the best games I’ve ever played. You should drop whatever other game you’re playing, and go get this one instead. It defines a new genre of game, offers unparalleled gameplay, or compelling game experiences that left me wanting more. I’m probably still playing this game regularly, and I’m enthusiastically talking about it whenever I get the chance. The game left me with profound or memorable experiences that serve as a standard by which I judge other games.
So why only 0 to 3 stars?
First, I don’t think there’s any value in differentiating levels of ‘bad’. There’s nothing more tepid, limp and useless than a rating of ‘six’ on a ten point rating scale. Is the game above average? What’s ‘average’? 6/10 sounds bad. Is it bad? Is it like, a test in school, where 60% is a D-minus and barely passing?
I have exactly one rating that means ‘bad’. It covers ‘meh’ and ‘this is awful’ and everything in between. I don’t even know how to go about differentiating a review score of, say, 3/10 from 2/10. What are the criteria? What earns that extra point, and since 3/10 is pretty universally accepted as ‘irredeemably awful’, what’s the use of having 2/10 right next door to it?
If a game is uniquely bad, I’ll write that in the review itself.
Second, I don’t stack-rank games I’ve enjoyed. That feels like an elementary school clique: you’re my 2nd best friend, she’s my 3rd best friend, he’s my 4th best friend. How I feel about a game relative to other games I’ve played is largely dependent on which game I’ve played most recently. So there’s no real benefit to distinguishing between a 8/10 and a 9/10 or, god forbid, a 9.5/10. In any of those cases, all I’m really saying is ‘I liked it’, or possibly if I dare to dip down to the 7/10 range, ‘I liked it with reservations.’
So I simplify: ** means ‘I liked it’, and * means ‘I liked it with reservations’. Now I don’t need to be cryptic and mysterious about some pseudo-mathematical criteria I’ve used to construct that 8.125/10 rating, and hope that somehow you’ll distill my secret criteria from the end product, like sieving shit to look for corn.
Third, I don’t ever expect to produce a list of ‘My Top Ten Games of All Time!’ sorted according to some arbitrary metric. So when I say a game is ***, I mean ‘If I were to make a list of the best games I’ve ever played, this would be on that list.’ I mean, I never have a good answer to the question ‘What’s your favorite game?’, so the three-star rating means ‘This might be it.’
To recap: 0 = bad game; * = good game with reservations; ** = good game; *** = amazing game. For more granularity, read the actual text.