It's become the common opinion of the game literati that professional reviews are an appendix-like relic of the past, not just useless but potentially destructive. I understand where they're coming from, as the scores can feel meaningless (what does scoring something an 8 versus an 8.2 really mean?), negative press from the power gaming elite has killed off more than one solid title (Paraworld anyone?) and otherwise-meh games have scored highly based solely on the specific experiences of the reviewer (Calling All Cars a 8.5? Really?), and reviewers often concentrate so heavily on the technical aspects of a game that you can barely figure out what genre it fits into, much less if it's any good. These are problems, yes, but I don't think they get to the heart of the underlying issue.
The real problem is that we haven't yet come to a decision as to where reviews should land on the line between objectivity and subjectivity. Should a reviewer be grading a game based on her own experiences (subjective), or what she expects the experiences of most players to be (objective)? For other forms of criticism, the lines are much more well-defined: You don't see a movie critic writing things like "I didn't like this movie at all, but for action fans it'd be high-octane fun." It's understood that their opinions are highly subjective. By the same token, you won't read an electronics review that reads "Everything about this television is perfectly executed, but there was just something about it that I didn't like." A target of anything other than objectivity would be wildly out of place.
Game reviewers, on the other hand, veer wildly back and forth over this line. The notorious Neverwinter Nights 2 review is a great example of how confusing this issue is. He genuinely didn't like the game, so why shouldn't it get a 5/10? Why should his score have to be objective, falling in line with other scores for the game? I'm sure there are plenty of gamers who'd agree with him that the shallow role-playing options are extremely disappointing. From the average gamer's perspective, though, the score is completely misleading. Though it had numerous technical faults that were ironed out in subsequent patches, it's faithful to the highly-regarded Knights of the Old Republic role-playing conventions and hugely fun for those of us (like me) who enjoy having tons of tactical options and character building.
On the other side of the line, you see a lot of games whose review scores fall into a suspiciously narrow range. It's as if the reviewers are basically saying "The average gamer will probably have a lot of fun with [X] but not consider it a new classic of the genre, so I'm going to give it a score in the [7.0-8.7] range to indicate that to them." Where's the guy who's just taken in by the story and swoons, giving it a perfect 10? Where's the girl who's icked out by the character design and spews vitriol all over her 41/100 review? Is the enjoyment/hatred of the game really so universal that everyone agrees, or do they plan their scores to fit in with their expectations of their average readers' experience?
I'm really not sure how things would work in a perfect world. Sixty bucks is a lot to plunk down on a title, so giving a game a high score when there are only a small subset of your readers who're going to love it with the same intensity that you do is dangerously misleading. Going too far in the other direction leads to only the most mainstream titles receiving decent scores, preventing the quirky games that appeal to a very specific type of gamer from getting a second look.
Right now, the industry is clearly working towards greater objectivity in their reviews. As the author of the NWN2 article found out, posting a review that's far out of line with the mainstream opinion is a no-no in the increasingly professional game journalism sphere. While I personally disagree with that review (and, in all honesty, it has problems beyond its subjectivity), I'd like to see more like it.
If a reviewer has strong feelings about a game, they should be able to share them, even if puts them outside the average gamer's experience. If they genuinely dislike a popular game, it should get a low score. If they fall in love with a bizarre indie game that's going to turn off 75% of their readers, they should be able to take that into consideration. Editors should verify that a review is factually correct and well-written, perhaps even adding some balance in extreme cases, but shouldn't kill or materially change a review because some readers are going to strongly disagree with it.
Personally, I like the format that EGM uses in their reviews, where the score is actually a composite of three different reviewers, each with a mini-review all their own. It allows people with strong impressions to share those without dominating the overall score. Sure, it may sometimes drag down a game that'd otherwise receive a strong score or inflate that of a game that most would consider a failure, but that's vastly preferable to seeing an endless list of bland scores from bland reviews that'll inevitably lead to blander games.
