And that is the problem I have with metacritic.  If you look at rottentomatoes.com you will see an example of a system that works.  Basically it's a yes/no system that is then averaged into a percentage.  That percentage has real meaning because it means what percentage of people liked it.

I will say that games are more complex, so they should get a different system but at the very least I would suggest an average of scores out of 3:

0 = Stay away.  There is no way anyone could like this.

1 = Not good, but it isn't all bad.  Some people may like it.

2 = Worth playing!  Everyone should enjoy this at least a little.

3 = Great game.  Most should like this.

Then all of the #/3 scores (No decimals) will be averaged into a percentage of what people like.  That way when people see a 75, they will think "Oh well 3 out 4 people like this game, I should see if I am one of those 3!" and they will read the review instead of "75/100?!?! I am not paying $60 for a C!"  and then he/she ignores the game completely.

Multiplayer is incredible. 

I hated when reviewers did somethign like this to Socom. A lot of places gave it bad reviews because the online multiplayer was having problems right at launch. They didn't review the game, they just reviewed the launch. The game was fine a few days after, but the review damage was done. Kinda sucks.

I'm not familiar with the SOCOM situation, but don't you think a developer should be penalized for releasing a buggy, broken, or otherwise incomplete game? If reviewers judge games only by their potential, what incentive is there for studios to release a polished, stable, and complete game?

While I agree that Metacritic is broken, I think in theory it's far superior to Rotten Tomatoes. With the RT system, a game with consistent perfect scores and a game with consistent good scores earn the same "fresh" rating. It's a system without nuance.

With the amount of console and game launches that have problems, I'm starting to understand that launches are just very difficult to pull off.  I can give some slack.  It's just disappointing that they "reviewed" a game by using the launch instability to gloss over the actual game.  I think a few nice people updated their reviews, but I don't think most did.  Sad, because it was a fun game.

The good news is that reviews now have run to give second batch of games 9´s and 10´s.

it depends, sometimes most reviews describes the game very accurate. Others is a total mess.

Call of Duty rairly works well at launch and they NEVER penalize that game...

Except RT also tracks review score averages. There's no reason not to use both methods, so we can clearly see how many positive reviews a game has earned, and how much more positive they actually are.