Like/Dislike system - You like a game that's 51%, then you come across your favorite game...
Bad, Maybe, Good system - You give Good to a 67% game, then you come across...
Percent system - You give 100% to a game, then you come across...

You shouldn't mix an absolute choice like favorite / non favorite with the more granular measurement.
And what does incorrect even mean in this context? The correct word would be indiscriminate, but then again you can discriminate in a variety of orthogonal axis.

But whatever.

For 5 star system I think the Amazon categories apply for most: Hate, Dislike, OK, Like, Love.
Maybe if you read them to mean that, you won't consider them to be incorrect ;)

I agreed with you in my post.

I was saying that I wouldn't split ratings percentages evenly like he did because the descriptions of each star category don't align with 20% chunks in my mind. What other percentage could it be, other than a rating?

Again, I was agreeing with you and just providing additional thoughts on why stars work better than percentages in this case...

That's a very good question, and hard to answer. I'm also using percentages in my ratings but I cannot give you a full explanation what they mean. It's very subjective and can be confusing for others, I'm first to admit that. I think scientific mind tends to compare a lot of things to numbers because it's easier for him/her to do so. But it's something which we may call "internal system rating" because it's very difficult to explain to anybody else how it actually works. At least it's so in my case. I also like to compare everything to colors but this is even more confusing so I rarely share those thoughts. For example Age of Wonders is a blue game for me while Giana Sisters is yellow. Does it make any sense to you? ;) I don't think so and yet it works perfectly for me ;)

Color ratings? I love it! You are a genius, my friend. :)

"So West it's East" Virginia.

We are talking about games here right?

That's AWESOME :D!
Would Heroes 3 count as blue or purple? Or something else entirely?

As a side-note, I think I've come up with a solution to this numerical conundrum, although implementation and using such a system would be pretty quaint:
Imagine a 2D board with 5 "helper" zones that can be as small or large as we want. Zones are aligned vertically, with 1 at the bottom and 5 on top. When a person rates a game, it goes into the corresponding zone. If there are already games present in that zone, the person can choose to put it ABOVE (better), BELOW (worse) or ASIDE (comparable) of any title he chooses to. Hell - the zones could even be eliminated at some point or even never be there in the first place.
Given enough people and some clever number-crunching, we could calculate the global, transitive intuitions on what is "better"... except that this doesn't some the problem of incomparable titles, which your coloring methods seems to achieve.
Perhaps different categories should be rated separately, when applicable...? OTOH - we've already seen this approach in game reviews and it doesn't really work all that well :\. The last thing that comes to mind are some "tags", which reminds me of how Gamespot deals with the issue.
Meh. I guess personally solving something that has been troubling people for decades is harder than I have expected...

Right ;-) but the mechanism stays the same (cognitive dissonance).

Quite on the contrary, whenever I spend time on something, I eventually realize how pointless, awful and meaningless it all is, including life itself.
Then again - perhaps I'm in the minority ;P.

No thats a false or premature conclusion or whatever the internet forum term for that is. Maybe the games have a high average because most games here are good.

You remind me of my school sports teacher who only gave me 14 out of 15 in tabletennis because we had some guy in the class who was in the national league or something. Even though I more than perfectly met the requirements of the class (I was in some league,too), I did not get the full rating. And if the world champion had been in the class, I would have gotten 13 and the other guy only 14. Does that seem right to you?

Quite possible. Generally people tend to shift opinions in a way that it goes along with the things they do. That can be quite absurd sometimes (e.g. one treats someone else bad and afterwards he tends to victim blaming).

It's not that easy because I gave you a simplified version with just one simple color and in my mind I'm usually thinking about more complex mixture of colors ;) But if we stick to the simple colors that HoMM 3 is definitely steel blue

EDIT: and it doesn't matter that there are a lot of steel weapons in the game. Alan Wake can be steel blue too!

One should also keep in mind that the average of user ratings tends to be better than it should be for the most part.
Why? You don't rate every game. There are a lot of games you are simple not interested in because you know you won't like them and they aren't even worth the effort to click some kind of dislike button. Therefor the average lacks a lot of low ratings.
It isn't the avarage of all peolpe it's the avarge of those who are interested enough to rate it. Obviously that group of people is more likely to enjoy whatever they are rating otherwise they wouldn't bother.

Yeah, i know the internet is full of people who love hanging around in comminitys just to tell them whatever the community is about sucks. But in the end they are outweighted by those enjoying that stuff.

Good point, I agree, it's also a sample problem of people who are actually rating.