It's a topic that's been touched on before, but current game review systems have issues. Between The Orange Box and Super Mario Galaxy, I have seen a metric crapton of 10/10's on various websites. Sure they may be good games, I'm not arguing otherwise. The problem is is that 10 seems a little high for what are essentially two sequels to successful franchises.

One of the interesting things about indie music review Pitchfork Media is how little they give out 100's. Extremely solid albums may only wind up in the mid 80's. Not that Pitchfork has found faults with the album, but because 100's are reserved for truly revolutionary works.

So, with that philosophy in mind, here's a different idea of how to rate games:

5 stars: This game is literally genre-defining. It breaks new ground in video games. We should be blown away by the sheer freshness in the face of sequels and clones. Example: Katamari Damacy

4 stars: This game is one of the best in its genre: Great graphics, solid gameplay, and an overall wonderful experience. Innovative, but not revolutionary. Example: Bioshock

3 stars: This game is fun to play. Is it an absolute must play? No, but it's worth your time and money and you won't be disappointed. Example: Jak and Daxter

2 stars: This game is mediocre. While it may have some interesting hooks, ultimately you won't be missing much if you don't play it. Example: Sonic Adventure 2

1 star: Garbage. Shoddy workmanship, uninspiring gameplay, and Bratz licenses will end you up here. You can think of examples.

This would hopefully be more objective, cutting out things like hype from the review. Not saying hyped games are necessarily bad, but they're probably not Top 100 games EVAR material either.