It seems that, last year there was a disconnect between what people thought the rating system means, and what the rating system should mean if its intended purpose is to show how good a game is.
You see, a rating is a comparison of something to a goal. It is x/y, where x is the thing being reviewed and y is “perfect”, with relation to something. Seriously. That’s as basic as I can make it.
But where it all goes wonky is in the decision of what the game is being compared to. Some strange people tend to rate games based on the games they're succeeding, based on legacy games that they think this "new game in this series" must stand up to; that, to me, seems rather foolish. Sure, you can rate Skyrim based on how it compares to Morrowind, but have you captured how good the damn game is by doing so? Nope! You've very specifically said, "I think this game is yea good, with respect to that really old game I like." That sort of metric isn't a very good one to pick if you want that metric to give people an idea of simply how good a game is; after all, who cares about other games?
One could ask in response to that thinking, "Mark, what the hell do you rate a game in comparison to, if not other games like it, then?" And there are a few answers to that question. Which answer you might pick should depend on the function you're trying to fulfill: are you trying to determine if a particular game is good for yourself, or are you trying to determine the best game for other people? There are other functions you could try to fulfill -- like determining if a particular game is a good game to show to the elderly -- but let's just look at these two, because they're different enough.
If you're trying to figure out if you'll enjoy a game, just figure out the things you enjoy and go with what fits those sordid desires. You should probably judge a game by how much fun you have playing it if the only important thing you're trying to do is play a game that you find genuinely great; who cares what that strumpet Sally down the street thinks about BF3 if you have a whale of a time flying around and screaming pilot gibberish at your screen ("Ace! Niner-Five-Oh! We got Rubidion's Paradox on our 9 o'clock vector, SHIT-BASH, SHIT-BASH, WE ARE ROUTED, AAAH I TOOK AN ARROW TO MY ACHILLES HEEL ABORT)?
Buuuut, if you're trying to tell other people what game they should buy, and your audience is diverse (though uniformly acne-ridden and sexually dolorous), then you should probably judge a game by, well, its aspects. What are its flaws, specifically the ones that aren't only flaws for you (the reviewer), due to opinions that other people may not hold? What is the game trying to be? How does it fail in that regard?
I believe most reviewers try to do that. They try to find a metric to review games by, one that people can look at and say, "So, this title will do X for me." A gamer/customer can then compare one game's value to another, and make the best decision. The only question left is, then, what metrics do people tend to use? Furthermore, can they be successful in picking a good reference to measure against, one that is recognized universally (by their audience)? The question of how "good" the game is, quantified and stuck on a number scale, relies mainly on the person deciding what good is; opinion. A reviewer using a particular metric is, pretty categorically, deciding if a game fulfills that metric, and what, exactly, it means for a game to "do" whatever it takes to get a 10/10 is going to rely entirely on what the reviewer thinks a 10/10 is justified by; to one reviewer, a 8/10 game could legitimately be considered a 9/10 game by another, because it's all a stupid matter of opinion, anyway.
And that's what you get when you go to a reviewer, asking for reviews like a fat wife asking her hubby about jeans: their opinion. Oh wait, a husband can lie, telling his wife what he thinks she wants to hear, in order to make her smile and maybe to get some glorious, blubbery sex later on in the eve. So it is, then, with some reviewers, who don't give you their honest opinions, but instead confirm all your [false] expectations, clandestinely hoping to get some sex out of all their consumers, we can assume. Those reviewers sure fuck all their fans and consumers over, at any rate, like a fat lady in ill-fitting jeans fucks over the eyes of small children. All the people who read that misleading opinion get a shit game, or miss a brilliant title, and then they COMPLAIN. SO MUCH. ALL THE TIME. So, don't do that, reviewers.
Instead, judge a game by your metric, and do it honestly.
So, reviews: what are they? Potential good advice, or fuckery.