All scores that follow are my opinion.

Army of Two: 71%
CoD4: 88%

In a five star scale, including 0/5, which makes it a six point scale I suppose, Army of Two and CoD4 would both get 4/5. Is that right? Absolutely not. CoD4 is a far superior game and...

Okay, somebody is going to say CoD4 deserves a 95% and therefore a 5/5. Which makes it perfect, or rather as close to perfect as a game can be, given the console and the year of release.

So either way, the five star system loses, in my opinion.

Yahtzee has hated pretty much every game I have ever enjoyed. I watch his reviews for a laugh, not to make any decisions. You shouldn't be using Yahtzee as an argument for a rating system lol.

The thing is that if Yahtzee doesn't like a game, he bashes it. Bashes it hard. Mentions any good points within 20 seconds. If he likes a game, the good points are there throughout.

Subjectivity will exist with or without numbers. Numbers simply help to convey an opinion. Example- Resistance 2:

IGN: The single player is epic in scope and in story, which is practically worth the price of admission by itself. Then you throw in the extremely engaging co-op and competitive play, which expands on replayability in just about every way possible. Toss in a generous number of unlockables, trophies and community support, and you've got a phenomenal experience.

D+Pad: The latter part of Resistance 2 is altogether more successful than its opening, with some diverting puzzles, nods to Valve's storytelling technique and a well-considered balance between checkpoints. But the fact that its levels are so sparse and the production values so surprisingly low suggests this isn't a game in which every action should be savoured, rather one to plough through at full pelt. Look for revolution in Resistance 2 and you'll be found wanting.

Okay, you haven't seen the scores. Who liked the game, and who didn't?

It will not surprise you to know (although you probably know already) that IGN gave the game a 9.5, while D+Pad gave it a 7.

Is one of them wrong? Neither of those reviews will be agreed with by everybody. But, again, does that make them wrong? No, it just means they are not entirely right. Your opinion is just as important as theirs. 

Another thing is that percentages (or my personal favourite variation, the X.Y system that IGN uses) allow for comparisons between games. Not necessarily between, say, Gears of War and Resistance 2, but say you want an FPS on your PS3. You like both alien and human shooters, which of the following will help you more in your decision?

"The single player is epic in scope and in story, which is practically worth the price of admission by itself. Then you throw in the extremely engaging co-op and competitive play, which expands on replayability in just about every way possible. Toss in a generous number of unlockables, trophies and community support, and you've got a phenomenal experience."

and

"This is a solid, confident shooter with plenty to offer the casual and hardcore alike."

OR

9.5

and

9.2?

Holy Hell do you see that thing flying overhead?

It's not Superman! It's the point!

More seriously, though, I think it's pretty easy to tell that I am referring to Yahtzee's method of "scoring", which is to say he has none, not his particular narrative style, which is more comedic than critical. He himself has said that the condensation of complex opinions into absolute values is asinine, and he's right. In spit of anything else he does in his reviews, he is right about that.

I've gone over, several times, why less specific scores are better than more specific ones. In this very topic, even.

http://www.vgchartz.com/forum/post.php?id=1457111

Also here. Specifically tthe points where I refer to the variance in quality between equal points on the scale.