Now somebody make a better site than Metacritic, because that thing is terrible

I don't see what's supposed to be so terrible about Metacritic. It's just a site that collects as many reviews as possible and gives out an average. I think what can be terrible is when people decide not to buy a game because the metascore is a little too low, or when developers don't get as much money because their game didn't hit 90, but none of this is Metacritics fault.

Yes, the point is that many people skip games just because it's on the 70s. This 0-10 scale has turned 6-7 games into "bad" and 8 is "okay" apparently. Metacritic is terrible because it uses a MASSIVE 0-100 scale. Plus, Metacritic considers a 74 game as "average". 
Ok, isn't the average a 50? A 74 sounds like something fairly above average. Take something like RottenTomatoes: it gives you a percentage of reviewers that though the movie was good, regardless of the score they gave, and then a little consensus about the film. Of course movie critics are terrible too, but the site is pretty good.

Another thing is that Metacritic has pages for the game on each platform, which is pointless. Here's a pretty good example:

Good:

http://www.metacritic.com/game/playstation-4/lara-croft-and-the-temple-of-osiris

Average:

http://www.metacritic.com/game/xbox-one/lara-croft-and-the-temple-of-osiris

The problem is that for most game-reviewers, 7/10 means average, which is why Metacritic treats 74 as average. On the other hand, movie-reviewers tend to actually use the full scale, which is why a movie that has a rating of 72 is considered good. So, I think if game-reviewers would also use the full scale, Metacritic would reflect that.

I do agree that rotten tomatoes has a better system and that 5/10 should mean average and not bad. I'm not sure about your last point though. Sometimes a game that's good on one system can be much worse on another, because of screen-tearing, more bugs, or stuff like that, and I guess it's good that this system is able to point that out.

Which is why 0-10 ratings are pointless. And honestly, there are barely any games that have any real differences between platforms. Those are few and far between. Off the top of my head, I can think of Bayonetta and FFXIII

I was doing a lot of thinking about game reviews and how to make the perfect review system that would be accurate and transparent and I came up with this. 

Every game starts off with 10 points. And these 10 points are divided into 5 categories.Each category is worth 2 points and is measured in 0.5 point units. 

The 5 categories should be;

1. GAMEPLAY how does the game actually play and how are its game mechanics. Is it fun, does it have depth, is it difficult, does it do anything innovative? These are the things this section should cover and 0.5 points can be knocked off for every thing it gets wrong in the this department. 
2. PRESENTATION graphical quality and artistic design. how good a looker is the game? What resolution does it run at, how is the frame rate? locked or all over the place? 
3. EMOTION this part will cover the games story and music/sound design. A game will only be penalized for trying to do something and failing at it. So a game that doesn't even try to have a story won't get points knocked off for not having one. A game that puts one in but its bad will get points knocked off.
4. QUALITY Is the game well designed? is it broken? Does it feel cheap and rushed? Does it require a day one patch?
5. VALUE How many hours does the game last. The ideal marker should be 20hrs of gameplay for a full 2 points, so as long as a game in normal/ normal + multiplayer mode gives you 20hrs of gameplay it gets full points. 

RolStoppable said:

I am scoring games based on worth buying, not worth playing. After all, a review is meant to be a buyer's guide. Someone who has unlimited funds for gaming doesn't need reviews anyway; they can just buy whatever they think looks interesting without any remorse.

You are right that a game is only worth buying if it's worth playing, but that's not the point here. A game worth playing isn't necessarily worth buying. The most common situation where this is the case is when a game's MSRP is (much) higher than justified.