I think the most fundamental problem with games criticism that hones in on what a game isn't -- that focuses on what a game
should be -- isn't that it's inherently bad to write that kind of criticism. I think the bigger issue is really a lack of self-awareness among reviewers who write in this way. As others here have pointed out, a common and intellectually valueless trend in games reviewing is to point out some obvious factual point about the game in question -- say, that it's linear -- and then call it bad for that reason, which in effect is just the reviewer sharing that he or she doesn't care for linear games. The reviewer is free to express his or her opinion, and there are plenty of times when simply offering up a basic impression of something is a good idea, but it feels to me that most of the time it comes down to a lack of self-awareness because a review is being ostensibly presented as intended to accomplish one goal, while its content suggests an entirely different one.
This is what I think the
parody review that WillyFive linked earlier is getting at. It emulates a review posted by a major games website like IGN. A review like that one, that does not analyze the plot in major detail (presumably to avoid spoilers), that is presented to a broad audience, and that gives a numeric score to indicate the quality of the subject is presumably intended to give its readers an idea of whether or not they'd enjoy a game before buying it. Yet the writing and content of the review don't really aim at this goal at all: we instead have the reviewer reading off a wish list of preferences and sighing with noted disappointment whenever some fact about the subject of the review doesn't live up to his or her wishes. Consequently, the review can only achieve its apparent goal of informing potential customers if the reader can be reasonably confident that his or her preferences align well with the reviewer's, which just serves to cut down the portion of the audience that may stand to benefit from the review. Could it have been done better? Sure. The writer could have considered what the audience already knows about the subject and think of a few categories of readers -- say, fans along genre lines -- and then consider the game from their perspectives and write about what they, the readers, might find interesting or disappointing about the subject. The thing is, this requires a degree of self-awareness in the reviewer about what exactly he or she is setting out to accomplish while writing. I notice sometimes that I enter into writing about something with the perception that I basically understand something and that I need to express it in a way that other people will understand, too. I suspect it's a common feeling to come into a discussion feeling that you've got it figured out and that you just need to "tell it like it is" to whatever audience is present. Maybe that's just something that people generally have to improve on when writing critically.
I do want to offer another perspective on comparative criticism, though. Some have expressed that it's either inherently incorrect or invalid to pan something because it's not something that's inherently different -- e.g., this JRPG isn't good because it doesn't have an open world like many western RPGs. I don't think it's quite this simple. For one thing, when a limit is imposed on the types of criticism that are valid, whether as a rule-of-thumb or a hard limit, some type of arguably valid criticism is removed from the space of valid critical arguments. This is where coming at things from a perfectly technically critical mindset, where a work has to be evaluated on the basis of how well it achieves what it sets out to accomplish, can fall flat. If we only consider games from that basis, we lose the ability to ask (in my opinion completely valid) questions such as "Is it a good thing that this game sets out to accomplish x?" For instance, in the example about the difference in world structure between JRPGs and WRPGs, I think it's a completely legitimate question to ask what end developers are trying to achieve by including open worlds and if that's really a valuable thing. If I thought that the goal was creating a more believable world, for instance, I might go on to ask if making a more believable world is a good priority for that kind of game -- and it almost certainly is, but that's just an example of the kinds of questions that lie outside the space of talking about a game solely on its success in doing what it set out to do.
There's also one other thing that got me thinking:
It's a similar thing with, say, movies with certain aesthetic styles or books with certain narrative styles. You may in fact hate those types of narrative styles or those aesthetic choices. That's fine. But that doesn't actually tell the reader anything about how they went wrong in following that aesthetic choice. So you tell the audience that Megaman 9 is an ugly game because it looks like an NES title. Congratulations, you've just informed your audience that the game looks exactly as intended, as a retro love letter to Megaman NES titles.
I agree with this point: saying that Mega Man 9's NES stylings are bad really just tells the reader that a) Mega Man 9 is visually designed to resemble the old NES games, which he probably doesn't need you to tell him, and that b) you don't "like" those graphics for whatever reason. That's perfectly fine, but one of the places where I think games criticism these days falls short is on accepting parts of a game, especially things outside of core gameplay like graphics and music, on the basis of simply whether they are aesthetically pleasing to the reviewer or to the reviewer's audience, rather than whether they serve some purpose besides that. I am of the opinion that a lot of games criticism doesn't live up to its potential and stops short of really meaningful analysis because it is content to give a passing grade to graphics and music and dialogue if they're simply "nice." How many times has a review said something like "the graphics are pleasant to look at and well-drawn and they didn't distract me so the game has good graphics"? I don't think it's impressive for a game to have nice-looking graphics. Relatively low-art-budget mobile games like King's and Rovio's hits look perfectly nice and have arguably appealing art. Instead, I think good analysis should ask how or whether the style of the visuals and the way they are used contributes to the game in a holistic way, whether that be to the tone or the accessibility or the player's ability to react quickly or whatever else. To give a little shameless plug to one of my favorite critics,
Matthewmatosis does a great job of this. Many reviewers looking at Skyward Sword's art would be content to say "it's colorful, it looks nice, it didn't distract me, it was good!" A few might rise above that and point out the visuals' Impressionist roots, and might say "the visuals are Impressionistic and they do a good job of emulating that style, so the visuals are good." But to me, a level of analysis actually worth aspiring to is to take all of the above and ask whether or not an Impressionistic style lends itself to the game's narrative themes or tone or general identity as a Legend of Zelda game and, if so, how.
So, to get back to the idea of whether analyzing games based on what they
are rather than what they
are not is a good rule to follow, I think it can be, and it all comes down to the self-awareness of the reviewer and the the reviewer's understanding of his or her own undertaking. If, in the reviewer's opinion, the game falls flat, then almost by definition it has be because of something the game isn't, because there would presumably be something that, if the game were, the game would not fall flat. For reviewers who have trouble seeing why they're writing, it can be a good rule-of-thumb to keep them on a relatively "objective" track. But for reviewers who have a good understanding of their purposes, I don't think it's a necessary or particularly useful limitation to humor.
Also, let me know if I misinterpreted you at all because I think I went on a serious tangent there D: