I feel that the ranking system itself has become ineffective, simply because all publications have different internal criteria as to what goes into the score. Some might have graphics as a subscore, while others might not take graphics into the final score at all (or just have a checkbox in regards to passability). Others leave the score up to the whim of the reviewer himself, meaning that the final "score" on the review can either reflect the review as a whole, or smash it with a hammer ("This game sucks. 10/10"). Without a standardized format, rankings by themselves have no merit--and it's hard to standardized an opinion of what is a largely subjective experience. Until then, we're stuck with the many different review systems and their scoring systems (97%, ***/*****, 4/10, etc.) and scoring criteria.
There are other reasons, too--namely spin and buy-outs. Neither is as prevalent (hopefully), but still a contributing factor. Buy-outs should be obvious: if Company X threatens to pull advertising from Magazine Y if Product Z gets a bad score/review, then the odds are that Magazine Y will inflate the score of Product Z and give it a glowing review, even if the product was bland, broken, or otherwise deserving of a low score. Spin, on the other hand, is harder to define, but generally it's bias, both on the parts of the reviewer and the reviewed. If Reviewer Beta is a fan of Genre Delta, and plays a game from Genre Upsilon--a genre he despises--will more than likely give it a lower score, while Reviewer Phi--who loves Genre Upsilon--can either give it a better score due to familiarity with the genre (and forgive its shortcomings), or bash it completely (because it's not like Game ZZ). The flip-side to the other isn't as predominant, as it mainly only applies to games made by smaller companies (i.e., by only one person). This side is willing to take lower scores as something good, and then apply its own unique spin to it to make it seem positive. Hell, I've seen someone say that a score of 2/5 was a great score since it came from someone who didn't really like the game in the first place. And this came from a game made only by one person.
Rankings are becoming obsolete--and with time, it's likely that reviews will as well if more of them begin to succumb to corporate pressure. Hell, one can argue that reviews themselves are already obsolete, seeing as they attempt to quantify a subjective, data-less experience and, at best, merely offer a guess as to whether the consumer reading the review will like the game or not.