gamerankings uses the more reliable sources i've found, and excludes the unreliable (obviously) .. thats why prefer to use that over metacritic, both are fine though for what they do

Metacritic has a superior methodology, there is no doubt about that. Statistically its a vastly superior way to undertake analysis. Quite simply all reviewers are simply not created equal, and shouldn't be considered as such.

Coming from a reasonable statistics background, gamerankings is rather useless. 

Accurate? I think you meant to ask "which do you prefer"...

And for everyone who thinks there is a huge difference, you need to realize that the two sites are not independent, they're associated with each other. But overall Metacritic is a more highly regarded site (it's the one that gets mentioned by industry people).

The reason games like Crash Bandicoot 1 are not on there is because it's a relatively new site that should only be relied upon for recently released games (see their list of top films, which is basically recent movies and a few old movies that got reviewed because they came out on DVD), so games that came out in the last 3-4 years.

i prefer the layout of meta, and how you can scroll down and read the summarys from all the reviews. gamerankings is better for the overall rankings and searching.

Metacritic is the better site, but the rankings in Gamerankings are better.

The main problem I have is they arbitrarily assign a number rating to reviews that don't have a number rating, based on the content of the review. These numbers always seem ridiculous

That comment is pointless unless you offer a site that is better.

 How do you determine which site is superior?  This is where metacritic fails, their method of deciding which sites have greater weight appears to be purely arbitrary, and results in games like Tony hawk ending up in the top ten, when its obviously not that good of a game

 Uhm, weighting data without some sort of reason that is defined by the nature of the data is just inviting innacuracies.  Its one thing to weight test scores which are objectively scored and quite another to weight review scores which are completely subjective.  Put simply, a weighting of a subjective score is another subjective decision and introduces the possibility for further innacuracies.

It would be different if we were interested in collecting the subjective data the reviewers produce but we are actually interested in the quality of the game and substituting the subjective review scores which quite literally represent one person's opinion and not necessarily the game's quality. 

I actually think the GR "weighting" by only counting trustworthy sites is a lot better method.  But it still is open to increased innacuracy due to the weighting.  Sometimes it helps sometimes it hurts.

i prefer gameranking.
i don't like the black design of metacritic.

i like gamerankings because it averages with increment but metacritic updates it faster